Introduction to

Process Engineering
and
Design
Second Edition
 About the Authors
Shuchen B Thakore graduated in Chemical Engineering from Gujarat
University and obtained his ME in Chemical Engineering from M S University,
Vadodara. He is currently Associate Professor of Chemical Engineering at L D
College of Engineering, Ahmedabad, Gujarat. He has been teaching
undergraduate and postgraduate students in Chemical Engineering, for more than
18 years. Besides teaching, he also undertakes consulting assignments in process
and mechanical designs of heat exchangers, distillation columns, storage tanks,
pressure vessels, etc., and environmental audit for various chemical industries.
Prior to joining the academic institute, he worked in the industry for about five
years. He served Alembic Chemical Works, Naveen Heat Exchangers Ltd.,
Gujarat Alcohol and Allied Chemicals Ltd. and Indian Engineering Software Co.
Mr. Thakore has earlier authored a book on Process Equipment Design.
Bharat I Bhatt graduated in Chemical Engineering from the Institute of
Chemical Technology, Mumbai. He is a Consulting Engineer to Chemical
Industries for the last ten years in Ahmedabad. His firm, Aavishkar Consultancy
Services, provides consulting services in debottlenecking in chemical industries,
energy conservation, effluent treatment, etc. He earlier served chemical
industries for nearly 34 years. His industrial experience covered the fields of
project execution, process engineering, debottlenecking, troubleshooting,
production management, management reporting, etc. He spent nearly 18 years
with Indian Farmers Fertiliser Cooperative Limited (IFFCO), Kalol Unit in
various capacities.
Prior to joining the industry, he served at the reputed textile research institute
(ATIRA) for nearly five years where he was involved in industrial research and
consultation to member textile units. He also worked in many fine chemical and
pharmaceutical industries as Vice President (Technical).
Mr. Bhatt has also authored a book titled Stoichiometry, published by McGraw-
Hill Education (India). Till date, five editions are published of which the last
three editions were in SI Units. While first four editions were co-authored by Dr.
S. M. Vora, fifth edition was co-authored by Prof. S. B. Thakore. In addition, Mr.
Bhatt carried out units conversion work from FPS units to SI units of the book
Introduction to Chemical Engineering Thermodynamics, 7th Edition, by Smith,
Van Ness and Abbott, a publication of McGraw-Hill, USA. The other book is,
Design Data Book: Properties of Steam, Selected Refrigerants, n-Hexane and
Brines (SI units).
In 2011, Mr. Bhatt was awarded the UAA Distinguished Alumnus Award.
 Introduction to
Process Engineering
and
Design
Second Edition

Slmchen B Thakore
.uw<ia.. Profr<$Ot"
ll<'",,,,_,,,
of Ch.mical £ "fJi""""i"9
L D CJJlege oIE"fJI" .....I"fJ
..Ihme4ob<l<1 GUj<lrat

Dha ra tl IIha tt
CO"'"Iti"fJ E"fJI,,~r
..,..,.." ~ar eon<ul/on<:v S«vIas
..Ihme4ab<l<1 GUj<lrat

McGraw Hill Edu ca ti on (Ind ia) Private limited

-~~
_ DoN _ _ so .......................... _ c--
"""'.....,..
... ""'"'--
_......,.,
__ ........ _
_ eo,.
r_ _ _-...
 """"_ Hm E_ . _ I......) ....... \0 l""_
,-.0 ... _ """'"'
._".0«<0 ....... 0:. _ _ "... 'W",.
.............. M<G ... H", .....,.....

<:"" ..... 0
,;.,
l<I'~"

................
....
-
,.,..><1"",._ ,,, I......,...., ':-';"""ri.".oo
-..-- .
".

. ._,,_ ...
... ~--- ... ,-~
-Go.,. Hill ,,.,._ ,...., _
_ _ .........., .... . . . . . - .. _ _ w

..............- .............. ,....""'-"""'....-

,..,....~
l "" ~

__
l,'

...... ---.....
_ _ .. ,..-...n.. __ ""... ,..... ,_ .. _ _
'-
.
.....--.
~m-...-.l.,- ..... ~_

_,""'''"
.. ---..... '_'_1.-.. ...,.., ... --.,
!'riM ,:.0;,;",
I SB~ ,Ur "I-9-n-l)<11l-t
I SH~ UO.9-)5· 1)<Ill-X
~
IS"~
.......""...
IU, _9-'~I M l~
IS"~ 11<0> 9-)5· 1MI7'O-l
" - """",,,0 K-.w ......
.......... , ..... .--,~ ... " "'-o ~ V_ .._
..........
~----...:"~ T""' ...."""'.. J..
,, ____ r.-..., ............
A ... _ _ ,, _ _ _
S,_~$....,.

~ ... Gnomor M......--I"....

&Moo_ 'M""'..... , ' ... s.-r.
So-c..,.o..l>0~ "_~"'"
...........
..............
o-nI M . . . ,_,~c........

_.
- .-
......,~"-

...

. _

. .. __
...... _-,
,- ,..... __ .. ..- . ..,""'"'-_ ..
....
-..".. "_... a"'.., _ _ -... ... _,
.......,
"' _ _
,- , ,
- ..... ...
.... _
...
.. _
"' .. ...._._ ..
.. _ _ ...... _ . . , ...-0... .... _ _ _ , _...... _

..... _-_-.. _-_ .._p

... "'......,_
..... -
...
.... _. " _ ..."""H .. ".,,,_

---_ _ ....... _ K _ _ _ _ . . _ ...

 Contents
Foreword
Preface
1. Dimensions and Units
1.1 Introduction
1.2 Fundamental Quantities or Dimensions
1.3 Derived Units
1.4 Recommended Rules for Writing Unit Symbols
1.5 Conversion Factors
 References
2. Stoichiometry
2.1 Introduction
2.2 Basic Chemical Calculations
2.3 Material Balances
2.4 Energy Balances
2.5 Enthalpy Changes Accompanying Chemical Reactions
2.6 Heat Changes in Mixing Processes
 Exercises
 References
3. Process Engineering
3.1 Introduction
3.2 Examples of Alternative Routes in Process Engineering
3.3 Generalized Approach to the Chemical Plant Design
3.4 Preparation of Operating Instructions Manual
3.5 Conductance of Performance Test Runs
3.6 Troubleshooting
3.7 Cost Cutting Measures
3.8 Preparation of Data Sheets of Equipments
3.9 Green Engineering
3.10 Process Intensification
3.11 Conclusion
 Exercises
 References
4. Importance of Process Diagrams in Process Engineering
4.1 Introduction
4.2 Need for PFD and P&ID
4.3 Block Diagrams
4.4 Development and Utility of Process Flow Diagrams
4.5 Development of Piping and Instrumentation Diagrams
4.6 Conclusion
 Exercises
 References
5. Process Design of Piping, Fluid Moving Devices and Flow Meters
5.1 Introduction
5.2 Process Design of Piping
5.3 Process Design of Fluid Moving Devices
5.4 Evaluation of Centrifugal Pump Performance When Handling Viscous
Liquids
5.5 Power Required in Fan, Blower and in Adiabatic Compressor
5.6 Flow Meters
5.7 Process Design of Orifice Meter
5.8 Process Design of Rotameter
5.9 Two-Phase Flow
5.10 Troubleshooting of Fluid Flow Systems
 Exercises
 References
6. Process Design of Heat Exchangers
6.1 Introduction
6.2 Shell and Tube Heat Exchangers
6.3 Various Parts of Shell and Tube Heat Exchangers
6.4 Advantages and Disadvantages of Different Types of Shell and Tube Heat
Exchangers Over Each Other
6.5 Different Types of TEMA Designations
6.6 General Design Method for Shell and Tube Heat Exchangers
6.7 Clasical Method of Heat Transfer and Pressure Drop Calculations for Shell
and Tube Heat Exchangers
6.8 Criteria of Selection between Horizontal Condenser and Vertical Condenser
6.9 Condensation with Non-Condensables in Shell and Tube Heat Exchangers
6.10 Multicomponent Condensation
6.11 Tinker’s Flow Model
6.12 Delaware Method for Design of Baffled Shell and Tube Heat Exchangers
6.13 Process Design of Reboilers and Vaporizers
6.14 Air Cooled Heat Exchangers and Air Heaters
6.15 Plate Heat Exchangers
6.16 Spiral Flow Heat Exchangers
6.17 Brazed Aluminum Plate-Fin Heat Exchanger
6.18 Other Types of Heat Exchange Equipments
6.19 Heat Exchanger Networking for Energy Integration
6.20 Heat Transfer in Solids
 Exercises
 References
7. Process Design of Liquid-Liquid Extractors
7.1 Introduction
7.2 Important Industrial Applications of Liquid-Liquid Extraction
7.3 Theory of Liquid-Liquid Extraction Operations
7.4 Desirable Solvent Properties or Choice of Solvent
7.5 Design of Counter Current Multistage Extractor
7.6 Industrially Important Extractors
7.7 Supercritical Extraction (SCE)
 Exercises
 References
8. Process Design of Distillation Columns
8.1 Introduction
8.2 Criteria of Selection
8.3 Selection of Equipment for Distillation
8.4 Distillation Column Design
8.5 Batch Distillation
8.6 Short Path Distillation
8.7 Reactive and Catalytic Distillation
8.8 Azeotropes and Separation Thereof
8.9 Energy Conservation in Distillation
 Exercises
 References
9. Process Design of Absorbers
9.1 Introduction
9.2 Criteria of Selection
9.3 Design of Packed Tower-Type Absorber
9.4 Absorption with Chemical Reaction
9.5 Process Design of Spray Chamber or Spray Tower-Type Absorber
9.6 Venturi Scrubber
9.7 Process Design of Falling Film Absorber
9.8 Process Design of Sieve Plate Absorber
 Exercises
 References
10. Process Design of Reactors
10.1 Introduction
10.2 Different Types of Reactors
10.3 Batch Reactors
10.4 Continuous Flow Reactors
10.5 Degree of Completion of Reaction
10.6 Mixing for Different Type of Reaction Systems
10.7 Heat Transfer in Batch Reactors
10.8 Bubble Column Reactor
10.9 Thin Film Absorber Cum Reactor
10.10 Catalysis
10.11 Design of Fixed Catalyst Bed Reactors for Gaseous Reactions
 Exercises
 References
11. Process Design of Utility Systems
11.1 Preamble
11.2 Process Heating Systems
11.3 Process Cooling Systems
11.4 Compressed Air Systems
11.5 Inert Gas Systems
11.6 Other Utilities
 Exercises
 References
Appendix 1
Appendix 2
Index
 Foreword
Dr. Jack Welch, CEO of General Electric, built its market cap by more than $450
billion and established himself as the most admired business leader in the world.
He is Doctorate in Chemical Engineering and is proud to be a Chemical
Engineer. In his business autobiography he wrote, “I have always felt that
chemical engineering was one of the best backgrounds for a business career
because both the class-work and required thesis teach you one very important
lesson: There are no finite answers to many questions.”
A chemical process engineer’s job is, therefore, tough. It is hard because he has
to find not only a finite answer but to select the best answer for the situation
from a variety of options. That is why process engineering is interesting and at
the same time intriguing.
A process engineer has to design a plant or a part of it. He may have to design a
unit operation or a unit process or a system. The important aspect is that it must
work and operate reliably and smoothly. The unit is to be designed for a
commercial purpose and, hence, it has to be economical and cost effective. A
good process engineer has to have plant experience and, in addition, a good
understanding of process economics and knowledge of current costs and trends.
Chemical process industry covers a very wide spectrum of industrial products.
Apart from the chemicals they handle, the quantities and sizes also vary. This
requires the process engineer to have flexibility of mind and vision to adopt
itself to any scale. Selection and design of the equipment should therefore be
capable to handle the capacities and instruments to measure the parameters for
large chemical plants to micro units to manufacture enzymes.
Till 1973, fuel and energy costs were a small component of the manufacturing
costs of a chemical product. Since then, the rising fuel prices have made it
mandatory that the plant design and its operation be energy efficient. The
approach to process design has undergone a paradigm change.
Bhopal gas leakage disaster and Chernobyl catastrophe made another impact on
the minds of process engineers and changed the practices followed till then.
Greater responsibility now rests on the process engineers for safer designs of the
plants. HAZAN (Hazard Analysis) and HAZOP (Hazard and Operatability)
studies have become a part of process studies, attaching one more design aspect
in the minds of process engineers. Several lessons regarding storage and
handling were learnt by the process engineer. He has to now respect
environmental and ecological sustainability of the process and its design.
Software for the design of equipment such as heat exchangers, distillation
columns, etc. are available. It has simplified piping design. Drawing and drafting
has become easier, error free and faster. But without understanding the design
principles, explained in this book, one cannot use them efficiently.
Yet, the process flow sheeting or the development of a process scheme still
remains a brain teasing function. It can be explained and perhaps taught but
cannot be standardized. Process engineer has to be innovative and creative.
Preparation of a Process Flow Diagram is still a prominent and challenging
domain of a process engineer.
It is more admirable that Mr. Bharat I. Bhatt and Prof. Shuchen B. Thakore
decided to go deeper into some of these topics and write a book on Process
Engineering and Design. They have done it with success. I had the opportunity
of working with Mr. Bhatt. He is one of the best process engineers I have come
across. The long industrial experience of Mr. Bhatt and extensive teaching
experience of Prof. Thakore have made them eminently suitable to offer their
knowledge to student fraternity and practicing engineers. I have no doubt that
the book will find its place with most academicians involved in teaching the
subject.
The book describes systematic methods for the process design with many
illustrations, drawn from various sections of the industry. They have included
several solved and unsolved problems with sufficient details. Authors’ approach
is user friendly and result oriented. Indepth study of these examples will boost
confidence of the students in solving real life design and operating problems.
Large number of references, cited at the end of each chapter, indicates the
extensive literature study made by the authors. Further, study of these references
will provide more insight to process design aspects.
An exclusive chapter is devoted on development of Process Flow Diagrams
(PFD) and Piping and Instrumentation Diagrams (P&ID) which is justified by
my earlier comments. Sufficient emphasis is placed on ‘green’ engineering.
These aspects have made this book unique.
Many recent topics like pinch technology, energy conservation in distillation,
two-phase flow of fluids, design of loop reactors and many others are to be
appreciated which are hard to find in other books on process design.
I have no doubt that the book will make the library of a student and practicing
engineer richer.
Suhas M Mokashi
Ex-Director of Technology & Marketing
Kaevernar Powergas Pvt. Ltd.
 Preface
Introduction
Chemical process industry is expanding globally to meet new challenges. In this
environment, cost reduction has assumed great importance for the process
engineers. New processes and technologies are knocking the doors of the process
industry to achieve results with stricter environmental regulations. Process
engineers have been engaged in meeting these tasks. With these requirements in
mind, the present edition has been brought out.
New to this Edition
• New chapter on process design of utility systems
• 30 new process flow-charts are added
• 5 new process and instrumentation diagrams are added
• Two-phase flow in pipelines
• Delaware method of design of heat exchanger
• Design of reactive distillation columns
• Short-path distillation units
• Design of catalysis and catalytic reactors
• Excellent pedagogy that includes
 98 Examples
 82 Exercise Questions
 242 Figures
 208 Tables
Salient Features of the Book
• Focus on Process Design leading to understanding of the design of process
plant equipment and systems
• Unit operations to highlight the process design aspects. In each of these
operations, current developments are included which provide energy efficient,
more reliable and safer designs
• Industrial examples for each concept/principle and a large number of solved
problems drawn from the industry to make the book practical and to give
students a good understanding of the design steps and the confidence to solve
problems themselves in real life
• State-of-the-art techniques of supercritical extraction, reactive distillation,
shore path distillation, etc.
About the Book
The text is thoroughly revised and each chapter is refreshed with newer
developments and their design aspects, such as development of piping and
instrumentation diagrams, which will be a welcome feature for newer process
engineers. Design of shell and tube heat exchangers, Delaware method is
considered to offer economic design over the conventional design methods. Its
inclusion in this revised edition will surely make the book richer and will be
helpful to process engineers. Inclusion of new industrial design problems is an
added feature in the book. Chapter 7 on liquid-liquid extractor is rewritten to
include description of various extractors available for selection. In Chapter 8 on
design of distillation column, design of multicomponent batch distillation and
example on extractive distillation are included. Absorption with chemical
reaction is a new topic in Chapter 9. Design of packed tower-type absorber
involving chemical reaction is included with industrial example. Use of falling
film absorber as a reactor is gaining popularity in the industry. Design of such
equipments is covered in Chapter 10.
A new chapter on process design of utility systems (Chapter 11) makes this book
different from other books on the subject. In this chapter, stress is given on role
of a process engineer in selecting and specifying utility requirements correctly
rather than carrying out detailed design of the equipment used in delivering the
utility. Typical P&I diagrams of the utility systems would help process engineer
to develop his own diagrams with confidence.
Green engineering and process intensification have assumed great importance in
process industry. A sincere attempt is made to make aware of these requirements
to the process engineer in his design.
Use of SI units is the highlight of the book. Current developments by
International Bureau of Weights and Measures are well covered in this edition.
Many problems in the book are taken from the industry and have been discussed
from alternative approaches. Advantages and limitations of newer technologies
are spelled out at every stage so that process engineer will use the new ideas
judiciously.
In addition, exercises and references at the end of each chapter, and Appendices
at the end of the book are added advantages to the book.
Web Supplements
For Instructors
• Solutions Manual
• PowerPoint Slides
For Students
• Objective-type Questions with answers
Acknowledgements
We acknowledge the patronage of the faculty of Chemical Engineering,
Industrial Chemistry, Process Engineering, Environmental Engineering,
Biochemical Engineering and others. It was an inspiration for us to write this
revised edition.
We are thankful to the Principal of L D College of Engineering, Ahmedabad for
granting us the permission to revise and publish this edition. We also appreciate
the continued support of our spouses; Dimple Thakore and Minaxi Bhatt without
whom the task of preparing the manuscript could not have been possible.
We are also thankful to the following reviewers for going through our
manuscript and offering valuable suggestions:
A K Sadhukhan    National Institute of Technology, Durgapur
S Srinath       National Institute of Technology, Warangal
P Sivashanmugam   National Institute of Technology, Trichy
Mangalya Kar    MGM College of Engineering & Technology, Kamothe,
Navi Mumbai
We thank Ms Vibha Mahajan and her editorial team at McGraw-Hill Education
(India) for their sincere efforts in bringing out this edition in an attractive format.
Shuchen B Thakore
Bharat I Bhatt
Feedback
Syllabi of various reputed chemical engineering institutes were studied which
are being revised from time to time. Every attempt has been made to address
their valued comments in this edition. Addition of newer topics is commensurate
with revised syllabi.
We are positive that our efforts in making this edition more useful will be
appreciated at large. Nevertheless suggestions and constructive criticism are
most welcome. Readers can write to us at b_bhatt26@hotmail.com and
sbthakore@yahoo.co.in
Shuchen B Thakore
Bharat I Bhatt
Publisher’s Note
We look forward to receiving valuable views, comments and suggestions for
improvements from teachers and students, all of which can be sent to
info.india@mheducation.com, mentioning the title and author’s name on the
subject line.
Piracy related issues can also be reported.
 1
Dimensions and Units
1.1 INTRODUCTION
Knowledge of dimensions and units is imperative in day-to-day dealings. While
dimensions are recognized by our sensory perceptions, their use requires definite
scales of measurement known as units. When a solid is observed, one thinks
about its volumetric and mass dimensions. Similarly, a liquid is recognized by its
volume and density measurements. A gas exerts definite pressure and
temperature which are important for defining its state.
Various quantities, required to identify a substance and its state can be classified
in two categories; namely fundamental and derived. Every quantity can be
expressed as a product of a dimensionless number and a unit where the unit is a
selected reference quantity in terms of which all quantities of the same kind can
be expressed.

1.2 FUNDAMENTAL QUANTITIES OR

DIMENSIONS
Fundamental quantities or dimensions are length, mass, time, thermodynamic
temperature, electric current, amount of substance and luminous intensity. Units,
representing these quantities are known as base units.
There are basically two systems of units; FPS and metric. The former is widely
used in the USA and is also known as US Customary Units. It is based on foot,
pound and second as standard measurements of length, mass and time,
respectively.
Metric (mks) system is based on the unit of length known as metre. The units of
mass and time in this system are kilogram and second. Its important feature is
the decimal expression.
International system of units (or Le Système International d’Unités in French or
simply SI Units) was adopted by the 11th General Conference of Weights and
Measures in 1960. Base units of the MKS system for length, mass and time are
retained in SI. Base units of the other four fundamental quantities are defined in
SI. Table 1.1 lists the fundamental quantities in different units.
The thermodynamic temperature (Kelvin), defined in SI, is accepted as the
absolute temperature in the metric system. For definitions of all the fundamental
quantities, references 1 to 4 may be consulted.
The table of elements based on carbon-12 is updated from time to time by the
International Union of Pure and Applied Chemistry (IUPAC), USA which gives
atomic masses (m) and atomic numbers of the elements. Current listing5 is
available as Atomic Weights of Elements, 2007.
Table 1.1 Fundamental Quantities
System of units with symbolic abbreviation
Symbol Fundamental quantity Dimension
SI Metric fps

l Length Metre (m) Metre (m) Foot (ft) L

m Mass Kilogram (kg) Kilogram (kg) Pound (lb) M
θ Time Second (s) Second (s) Second (s) θ
I Electric current Ampere (A) Ampere (A) Ampere (A) I
T, t or T Temperature Kelvin (K) Celsius (°C) Fahrenheit (°F) T
n Amount of substance Mole (mol) n
Iv Luminous intensity Candela (cd)

In 1971, the concept of the amount of a substance was introduced as a

fundamental quantity. It is defined as the amount of substance of system
containing as many elementary entities as the number of atoms in 0.012
kilogram of carbon-12 element. This unit of quantity was called a mole
(abbreviated as mol). Thus unified scale of mole gives value of the relative
atomic mass (m). When a compound is represented by a chemical formula, then
the total mass computed based on the formula is called molar mass (popularly
known as molecular weight) and is represented by the symbol M.

1.3 DERIVED UNITS

Derived units (including supplementary units) are expressed algebraically and in
terms of base units by means of mathematical signs of multiplication and
division. For example, the SI unit of work or energy is Joule which is defined as
work done when one Newton (N) force acting on a body moves it by a distance
of one metre (see Sec. 1.3.4).
Table 1.2 lists some of the important SI derived units used in Chemical
Engineering with their special names while Table 1.3 lists frequently used
derived units.
Table 1.2 SI Derived Units with Special Names, Including SI Supplementary Units
SI derived unit (Derived quantity) Special Symbol Expressed in terms of SI base units and SI
name derived units
Plane angle radian α, β, γ 1 rad = 1 m/m = 1
Solid angle steradian Ω, ω 1 sr = 1 m2/m2 = 1
Frequency hertz f 1 Hz =1 s–1
Force newton F 1 N = 1 kg · m/s2
Pressure or stress pascal p, σ 1 Pa = 1 N/m2
Energy or Work or Quantity of heat joule E, H 1 J = 1 N · m
Power Radiant flux watt P 1 W = 1 J/s
Electric charge or Quantity of electricity coulomb Q 1 C = 1 A · s
Electric potential or Potential difference, or Tension or volt V 1 V = 1 W/A
Electromotive force
Electric resistance ohm R 1 Ω = 1 V/A
Electric conductance siemens G = 1/R 1 S = 1 Ω–1

Celsius temperature degree t 1°C = 1 K

Celsius1

1Degree Celsius is a special name for the unit Kelvin. For use in stating value of Celsius temperature. t/°C =

T/K – 273.15.
Table 1.3 Additional Frequently Used SI Derived Units
Quantity Symbol SI unit Practical units Notes

Acceleration a m/s2 m/s2, cm/s2

Acceleration of free fall g m/s2 m/s2
Isentropic exponent k k = Cp /Cv (for ideal
gas)
Ambient Pressure pamb N/m2, Pa mbar, bar, Torr

Angular acceleration α rad/s2 rad/s2

Angular velocity ω rad/s rad/s
Area A, As, S, a m2 m2, cm2

Concentration of substance B CB mol/m3 mol/m3, mol/L

Current density, j A/m2 A/m2, A/cm2

Density (mass concentration) ρ kg/m3 kg/m3, g/cm3, kg/L
Relative density/Specific gravity d ρt1/ρt2 — Non-dimensional
quantity
Diametre d, D m cm, mm
Diffusion coefficient D m2/s m2/s, cm2/s
Efficiency η — — Non-dimensional
quantity
Heat (enthalpy) of evaporation λ J/g, J/mol kJ/kg, kJ/kmol
Volumetric flow rate qv m3/h m3/h, m3/s, L/min

Molar flow rate* n. mol/s mol/s, kmol/h

Gas constant (molar) R J/(mol · K) (mbar · L)/(mol · K), (m3 ·
bar)/(kmol · K)

Gas constant of substance B (specific) RB J/(kg · K) (mbar · L)/(kg · K), (m3 · bar)/(kg ·
K)
Heat capacity of a given mass C J /K J/K, kJ/K
Heat capacity, isobaric, molar Cmp J/(mol · K) kJ/(kmol · K)

Heat capacity, isobaric, specific Cp, Cl J/(g · K) kJ/(kg · K)

Heat capacity, isochoric: molar Cmv J/(mol · K) kJ/(kmol · K)

Heat flow rate or heat flux ϕ W, J/s kW, kJ/s, kJ/h

2
Heat transfer coefficient h, U W/(K · m ) W/(K · m2), kcal/(h · m2 · °C)
Height h m m, cm, mm
Kinematic viscosity γ m2/s m2/s, mm2/s, cm2/s, cSt
Kinetic energy Ek J J, kJ

Leak rate (in a vacuum system) qL N · m/s mbar · L/s

Length L m m, cm, mm
Liner expansion coefficient αl m/(m · K), m/(m · K), 1/K

Mass m kg kg, g
Mass flow rate qm . ṁ kg/s kg/s, kg/h, g/s

Mass fraction w, X, Y kg/kg %, ppm ppm = parts per mega

parts
Molality m mol/kg mol /kg —
Molarity M mol/dm3 mol/m3, mol/L, —
Molar mass M kg/kmol kg/kmol, g/mol
Molar volume Vm m3/mol m3/kmol, L/mol
Mole fraction x (in liquid) y (in vapour/ – –
gas)
Number of revolutions n 1/s 1/s, 1/min (or rpm)
Partial pressure of component B pB N/m2, Pa mbar, kPa, Torr

Period (periodic time) θ s s, ms, μs

Permeability (for gases) μ Nm3/(m2 · s · Nm3/(m2 · d · bar)

Pa)
Pumping speed (rate)/ volumetric qv,V̇ m3/s L/s, m3/h, m3/s
flow rate
Quantity (volume) of gas V Nm3 L, Nm3, Sm3
Quantity of heat Q, H, i J J, kJ, kWh, W · s
Radius (also of molecule) r m cm, m, μm
Vapour pressure pv N/m2, Pa mbar, kPa, Pa, Torr

Saturation vapour pressure ps N/m2, Pa mbar, kPa, Pa, Torr

Temperature difference ΔT, Δt K K, °C

Thermal conductivity k W/(K · m) W/(K · m), kW/(m · K), kcal/(h · m
· °C)
Thermal diffusivity a m2/s m2/s
Velocity v m/s m/s, mm/s, km/h
Viscosity (dynamic) μ Pa · s mPa · s, cP
Volume V m3 m3, L, cm3
Volume (specific) v m3/kg m3/kg, cm3/g, m3/kmol
Volume fraction of substance B ϕB m3/m3 L/L, %, ppm ppm = parts per mega
parts
Wavelength λ m nm, Å
Weight (force) G N N, kgf
Specific/absolute humidity H kg /kg kg/kg, g/g
Relative humidity RH — — No dimensions
Saturation ratio φ — — - do -

*In the book, alphabets in italic are also used for molar flow rates.

It will now be appropriate to elaborate a few of the important SI derived units for
better understanding.
1.3.1 Force
According to the Newton’s second law of motion,
F ∝ m × a  (1.1)
or F = K m a  (1.2)
where, K = proportionality constant
Selecting K as unity, the unit newton is defined. Thus, when a force of one
newton (N) is applied to a body having a mass of one kilogram it gives an
acceleration of one m/s2.
Definition of technical unit of force is obtained by selecting K equivalent to 1/gc
where gc is called Newton’s law conversion factor. Its value is 9.806 65 (kg ·
m)/(kgf · s2) and is equal to the acceleration due to gravity (g) at mean sea level.
It may be noted that gc is not equal to g at all locations on the earth.
Kilogram-force (kgf) is the force which when applied to a body having a mass of
one kilogram, gives it an acceleration of 9.806 65 m/s2. It is the base unit in
metric system of units.
1 N = 0.101 972 kgf
or 1 kgf = 9.806 65 N
Although mass is the base unit, it nearly equals weight under gravitational
acceleration (g). This is why ‘mass’ and ‘weight’ terms are commonly used
without differentiation in general practice.
1.3.2 Pressure
Pressure is the force acting on a unit area, exposed to it.
p =   (1.3)

Common units of pressure in SI and metric systems are N/m2 (known as Pascal;
symbol Pa) and kgf/cm2, respectively. Pressure, measured with the help of a
gauge, is called gauge pressure or over pressure and the letter ‘g’ follows the
unit. When (local) atmospheric pressure is added to the gauge pressure, it is
called absolute pressure and the letter ‘a’ follows the unit. Standard atmospheric
pressure (symbol atm) is measured at mean sea level and is assigned the value
101 325 N/m2 or Pa or 1.033 kgf/cm2. Since Pascal is a very small unit of
pressure, bar is chosen as a practical pressure unit in the SI system.
1 bar = 105 Pa = 1.019 716 kgf/cm2 = 0.986 923 atm
1 atm = 1.013 25 bar
Pressure is also expressed in terms of fluid heads, e.g., water column (WC),
liquid column (LC), mercury column (Hg), etc.
Fluid head = Pressure /density = p/ρ  (1.4)
1 atm = 760 mm Hg at 0°C or Torr = 10.33 m WC at 4°C
Vacuum refers to pressure below atmospheric pressure and hence absolute
pressure of a vacuum system is derived by substracting vacuum from
atmospheric pressure.
1.3.3 Volume
Volume is the space occupied by a substance. It is measured in cubic metres and
liters in SI and metric units, respectively. Approximately, 1 m3 ≈ 1000 L ≈ 1 kL.
1.3.4 Work, Energy, Power and Heat
Work is defined as the product of force acting on a body and the distance
traveled by the body.
W = F × d  (1.5)
Units of work (kg · m2)/s2 or N · m or joule (symbol J) and m · kgf in SI and
metric systems, respectively.
Internal energy (Symbol U or E) is a physical entity which is present in a system
in different forms; mechanical, chemical, thermal, etc. Heat is the form of energy
in transit that flows from higher temperature to lower temperature. Total
enthalpy (symbol H ) of a substance is the total energy possessed by it under the
given conditions.
H = U + pV  (1.6)
Unit of energy, heat and enthalpy is joule and calorie (symbol cal) in SI and
metric systems, respectively.
1 calorie (International Steam Tables) = 4.1868 J
It may be noted that unit of work, energy, heat and enthalpy in SI is same, i.e., J
(unlike metric system).
Power P is defined as the work done per unit time
P =   (1.7)

Unit of power is Watt (i.e., J/s) and m · kgf/s in SI and metric systems,
respectively. Practical unit of power in metric system is horsepower.
1 metric system horsepower = 75 (m · kgf)/s = 0.7355 kW
1.3.5 Derived Electrical Units
Electric potential (V ) is defined as
V =   (1.8)

Its unit is volts.

Resistance of the conductor is defined as
R =   (1.9)

Its unit is ohms (Ω). Reciprocal of resistance is called conductance; symbol (G)
and has units siemens (S). Quantity of electricity (symbol Q) is defined as the
product of current and time.
Q = I · θ  (1.10)
Unit of quantity of electricity is coulomb (C).
1 Faraday (F) = 96 485.3415 C/mol

1.4 RECOMMENDED RULES FOR WRITING

UNIT SYMBOLS
Following guidelines may be followed while using SI unit symbols:
1. SI prefixes are given in Table 1.4.
Table 1.4 SI Prefixes
Factor Prefix Symbol Factor Prefix Symbol

1024 yotta Y 10–1 deci d

1021 zetta Z 10–2 centi c

1018 exa E 10–3 milli m

1015 peta P 10–6 micro μ

1012 tera T 10–9 nano n

109 giga G 10–12 pico p

106 mega M 10–15 femto f

103 kilo k 10–18 atto a

 102 hecto h 10–21 zepto z

101 deca da 10–24 yocto y

2. An exponent attached to compound prefix-unit implies that the exponent

refers to the entire compound unit and not just to the base symbol. 1 cm3
means volume of a cube having one cm side.
3. The product of two or more units may be indicated in any one of the following
ways:
Correct: N m, or N · m   Incorrect: Nm
4. A solidus (oblique stroke), a horizontal line or negative powers may be used to
express a derived unit, formed from two others by division, e.g., linear
velocity can be expressed in one of following four ways:
m/s, m · s–1 or m s–1

5. A solidus must not be repeated on the same line unless ambiguity is avoided
by a parenthesis.
Correct          Incorrect
m/s2 or m · s–2 or m s–2     m/s/s
W/(m2 · K) or W m–2 K–1    W/m2 · K or W m–2/K
6. Unit symbols do not change in the plural. For example, 5 metres should be
abbreviated as 5 m and not as 5 ms.
7. Unit symbols are not followed by a full stop (period) except at the end of a
sentence.
Correct: 8 kg   Incorrect: 8 kg.
8. When numerical values fall outside the range of 0.1 to 1000, it is
recommended that the numerals be separated into groups of three with a
space replacing the traditional comma.
Recommended     Not recommended
3 600 or 3600      3,600
19 625 725       19,625,725
0.001 625        0.001,625
0.046 89        0.046,89
In this book, four digits or decimals are grouped.
9. Prefix symbols are printed without any space between the prefix symbol and
the unit symbol.
Correct: 10 kW Incorrect: 10.5 kW
10. Compound prefixed formed by the juxtaposition of two or more SI prefixes
 are not to be used.
Correct: 6 GJ Incorrect: 6 MkJ
11. A prefix should never be used alone.
Correct: 106/m3 Incorrect: M/m3
12. Although kilogram is the base unit in SI units, names of decimal multiples
and submultiples of mass are formed by attaching prefixes to the word
‘gram’.
Correct: 1 mg Incorrect: 1 μkg
13. Good practice recommends selection of a prefix which, whenever possible,
provides a numerical value between 0.1 and 1000. Prefer expression 10.0 kPa
over 0.01 MPa. However, when values are tabulated, they should be
expressed in the same unit multiple even though their numerical value lies
outside 0.1 to 1000 range.
12 × 104 J can be written as: 120 kJ.
1632 W can be written as 1.632 kW.
14. If the magnitude of the number is less than unity, the decimal sign should
preferably be preceded by zero, e.g., .072 should be written as 0.072.
15. The SI prefixed are not to be used with °C or K.
16. In the expression of a quantity, the unit symbol is placed after the numerical
number and a space is left between the two.
Examples 52 kg or 25.4 cm
As an exception, in this book, a space is not left between the numerical value
and the symbol for the value of celsius temperature. Also while expressing
pressure in terms of fluid heads, unit and fluid name are separated.

1.5 CONVERSION FACTORS

Appendix 1 lists the conversion factors in a direct usable form for conversion
from one system of units to another system of unit.

REFERENCES
1. Taylor, B. N. and A. Thompson, The International System of Units (SI),
National Institute of Standards and Technology (NIST) Publication No. 330,
USA, 2008.
2. ISO 1000 : 1992/Amd. 1 : 1998 (E), SI Units and Recommendations for the
Use of their Multiples and Certain other Units (Amendment 1), International
Organization for Standards, Switzerland.
3. Mills, I., T. Cvitas, K. Homann, N. Kallary, and K. Kuchitsu, Quantities, Units
and Symbols in Physical Chemistry, 2nd Ed., IUPAC Chemical Data Series,
Blackwell Science Ltd., UK, 1993.
4. Bhatt, B. I., and S. B. Thakore, Stoichiometry, 5th Ed., McGraw-Hill
Education, New Delhi, 2010.
5. Wieser, M.E., and M.B. Berglund, Pure Appl. Chem., 81(11), 2009, p. 2131–
2156.
 2
Stoichiometry
2.1 INTRODUCTION
Stoichiometry is a branch of chemistry which deals with combining the
proportions of the elements or compounds involved in chemical reactions. It is
based on three concepts; conservation of mass, the relative masses of atoms and
definition of a mole. While the last two concepts were introduced in Chapter 1,
the law of conservation of mass is dealt in this chapter.
In any chemical reaction or in a unit operation, exchange of thermal energy also
takes place. Therefore, heat balance is also important in stoichiometric
calculations. Similar to mass balance, heat balance calculations are based on the
law of conservation of energy (the first law of thermodynamics).

2.2 BASIC CHEMICAL CALCULATIONS

Before attempting mass and energy balance problems, knowledge of basic
principles of chemical engineering calculations is necessary. A brief review of
these calculations is presented as follows.
2.2.1 Atomic Mass and Molar Mass
Elements are basic entities which are combined (or reacted) to form a
compound. As seen in Chapter 1, relative atomic masses for each element are
presented on a unified scale which is known as Atomic Masses (Weights) of
Elements1.
For chemical compounds, a mole is defined as the amount of substance equal to
its formula weight. Molar mass (M) is thus the total mass of elements, contained
in the compound. Thus it is easy to calculate molar mass of a compound,
knowing its chemical formula.
Calculate molar mass of the following:
(a) Sulphurous chloride (SOCl2) and
(b) 2,4,6 – trinitrophenol (also known as picric acid)
Solution
(a) Molar mass of SOCl2 = Atomic mass of sulphur + Atomic mass of oxygen +
 2 × Atomic mass of chlorine
= 32.065 + 15.9994 + (2 × 35.453)
= 118.9704 kg/kmol (exact)
With rounded off values of atomic masses,
molar mass of SOCl2 = 32 + 16 + 2 × 35.5
= 119
(b) Chemical formula of picric acid = C6H2(NO2)3OH
Molar mass of picric acid = 6 × 12.0107 + 3 × 1.007 94 + 3 × 14.0067 + 7 ×
15.9994
= 72.0642 + 3.023 82 + 42.0201 + 111.9958
= 229.103 92 kg/kmol (exact)
≈ 229 (rounded off value)
In this book, rounded off values of atomic masses will be used for convenience.
2.2.2 Mole
In Chapter 1, mole is defined as a base unit in SI. It can be further defined as the
amount of substance containing 6.022 142 × 1023 molecules (also known as
Avogadro’s number, NA). In simple terms, number of moles can be found by
dividing mass of the compound by its molar mass.
Convert the following into moles:
(a) 1000 kg boron trichloride (BCl3)
(b) 1500 kg phenol (C6H5OH)
Solution
(a) Molar mass of BCl3 = 11 + 3 × 35.5 = 117.5 kg/kmol
1000 kg BCl3 ≡ = 8.5106 kmol

(b) Molar mass of C6H5OH = 6 × 12 + 6 × 1 + 16

= 94 kg/kmol
1500 kg C6H5OH ≡ ≡ 15.9574 kmol.

2.2.3 Mass Fraction and Mole Fraction

In a mixture, mass fractions (w) or mole fractions (x, y) are expressed to
represent its composition. Mass fraction is the ratio of the mass of a particular
substance divided by total mass of all the substances. Similarly, mole fraction is
the ratio of the moles of a particular substance divided by total moles of all the
substances. One can be calculated knowing the other.
Aqueous calcium chloride brine contains 35% CaCl2 by mass.
Convert the same in mole %.
Solution
Basis: 100 kg brine
Table 2.1 Composition of Brine
Component Mass kg Molar mass kg/kmol moles kmol mole % (xi )

CaCl2 35 111 0.3153 8.03

H2O 65 18 3.6111 91.97

Total 100 3.9264 100.00

Average molar mass (M) of brine = = 25.47 kg/kmol

Feed gas to ammonia synthesis loop has molar composition

24.75% N2, 74.25% H2, 0.7% CH4 and 0.3% Ar. Calculate its mass composition,
average molar mass and specific gravity (with respect to air).
Solution
Basis: 100 kmol feed gas
Table 2.2 Composition of Feed Gas to Ammonia Synthesis Loop
Component kmol Molar mass kg/kmol Mass kg mass %

N2 24.75 28 693.0 80.14

H2 74.25 2 148.5 17.17

CH4 0.70 16 11.2 1.30

Ar 0.30 40 12.0 1.39

Total 100.00 864.7 100.00

Average molar mass (M) of feed gas = = 8.647 kg/kmol

Similarly, average molar mass of air is be calculated to be 28.9697 in Example

2.10 (rounded off value to be 29).
Specific gravity of feed gas = = 0.2985

While mass fractions are useful in establishing material balances, mole fractions
are logical expressions in chemical reactions.
2.2.4 Density and Specific Gravity
Density of a compound in any state (solid, liquid or gas) is expressed as the ratio
of mass per unit volume. It is normally expressed in g/cm3, kg/L or kg/m3.
Density of iron is 7.75 g/cm3 at 20°C while that of aqueous solution of 24%
BaCl2 at 40°C is 1.243 g/cm3.
Specific gravity (d) of a liquid or solid is a dimensionless number and it is
defined as the ratio of density of solid or liquid at a given temperature to the
density of water at 4°C (which is nearly equal to 1.000 g/cm3). For liquids,
specific gravities are often expressed in terms of certain hydrometers. Among
these, °Twaddell (°Tw) and °Baumé (°Be′ ) scales are well known. In petroleum
industry, hydrometer scale called °API is used which is very useful in finding
certain properties like molar mass, viscosity, heating value, etc. Specific gravity
of 24% aqueous BaCl2 solution at 40°C is 48.6°Tw. Strength of certain important
inorganic solutions such as that of caustic soda solutions, sulphuric acid
solutions, etc., are tabulated on °Tw or °Be′ scale as hydrometers are convenient
for use in industry.
Specific gravity of a gas is the parameter which indicates the heaviness of a gas
as compared to that of air. One definition is the ratio of molar mass of gas to the
molar mass of air. In Example 2.4, specific gravity of feed gas to ammonia
synthesis loop is determined to be 0.2985, indicating that it is much lighter than
air (both at same operating conditions).
Another definition of specific gravity of a gas is the ratio of the density of the
gas at given conditions of pressure and temperature to the density of air at
reference conditions. Although this is a more logical expression, molar mass
ratio is chosen in this book to express heaviness of a gas as compared to the air.
2.2.5 Vapour Pressure and Boiling Point of Liquid
When a liquid is heated, it gets converted to vapour at a given temperature and
pressure. For a pure liquid, since degree of freedom is one, it boils at a constant
temperature for a given pressure. Conversely, for a pure liquid at a given
temperature, its equilibrium pressure is fixed which is known as vapour
pressure. This implies that when both phases exist in equilibrium, there is only
one definite pressure (called vapour pressure) at a given temperature for a single
component.
Boiling point (tB) of a pure liquid at standard atmospheric pressure (1 atm or
101.325 kPa) is called normal boiling point (NBP). At 1 atm, water boils at
100°C, ethyl alcohol boils at 78.4°C, methylene chloride boils at 40.1°C and
refrigerant 134a (tetrafluoroethane) boils at –26.1°C. All are called normal
boiling points.
Relationship of vapour pressure with its saturation temperature is expressed
either empirically or in tabulated form. Variety of equations are known for
calculating vapour pressure of a liquid. Among these, Antoine equation is well
known and is expressed as follows:
log pv = A –   (2.1)

where pv is vapour pressure at temperature T (in K) while A, B and C are

constants. Ref. 1 lists values of A, B and C for several liquids. Validity of
equations for a given liquid is limited in a definite temperature range. Number of
other equations are reported in literature for accurate prediction of vapour
pressure.
Some of the fluids are extensively used in industry over a wide range of pressure
and temperature. Their applications could vary from heating, cooling or
processing. Steam and certain vapour phase heating fluids are used for heating.
Similarly, a variety of refrigerants are used for cooling. n–Hexane is used for
extraction of oil from seeds. Nitrogen, oxygen, etc., are also used as feed in
various chemical reactions. Carbon dioxide, water, propane and butanes are used
in their supercritical states for extraction and for carrying out certain chemical
reactions. Thus, thermo-physical properties of all such fluids are required at
different pressures and temperatures. For this reason, properties of these fluids
are available in tabular form. These tables list temperature and its corresponding
vapour pressure as one property. For example, steam tables (Ref. 2) are widely
used in industry. Data, read from the table, permit quick calculations.
Several other properties of liquids and solutions are of importance in chemical
engineering calculations. References 1 and 3 may be referred for the discussions
on these properties.
2.2.6 Gases
Gaseous state of a pure component is defined by its pressure and temperature.
However, when a mixture of gases is present, its composition is also required to
be specified. In a conventional practice, mole % composition of a gaseous
mixture is specified (as seen in Example 2.4).
Gases behave as ideal in a definite range of pressure and temperature such that
they obey the well-known ideal gas law.
pV = RT  (2.2)
where, p = absolute pressure of the gas, kPa
V = volume occupied by the gas, m3/kmol
R = Universal Gas Constant
= 8.314 472 J/(mol · K) or m3 · kPa/(kmol · K)
T = absolute temperature of the gas, K
When p is chosen as one standard atmosphere (101.325 kPa) and T is chosen as
273.15 K (i.e., 0°C), these are called normal conditions (i.e., NTP).
Molar volume at NTP of an ideal gas is 22.414 m3/kmol.
Conversely an ideal gas is one which follows Eq. (2.2). Most gases under
ordinary circumstances (such as ambient conditions) obey ideal gas law.
However, real gases deviate significantly at high pressures and high
temperatures. Real gas behaviors are described by equations of state (also called
p-V-T relationships). Reference 4 may be referred for more information on
various equations of state. p-V-T relations can be represented on phase diagrams.
p–t diagrams for water and carbon dioxide are given in Figs. 7.28 and 7.29.
Critical conditions of a gas are those beyond which a gas cannot be liquefied.
These conditions are defined by critical temperature, critical pressure and critical
volume. By definition, critical temperature is the maximum temperature at which
a gas can be liquefied. The critical pressure is the saturation (vapour) pressure
corresponding to the critical temperature. The volume, occupied by the fluid at
critical point is the critical volume. At the critical point, densities (or specific
volumes) of gas and liquid (in equilibrium) are equal while latent heat of
evaporation (λ) is zero. Reference 1 lists the critical properties of number of
compounds.
Physical state of a fluid above its critical conditions is known as supercritical
state. In Chapter 7, extraction in supercritical fluids is discussed. Table 7.8 lists
the critical properties of commonly used supercritical fluids.
When a mixture of gases is handled, its composition is specified in mole %. For
an ideal gas mixture,
mole % = volume % = pressure %
Partial pressure (or pure component pressure) of a component in the ideal gas
mixture is defined as total pressure of the system multiplied by the mole fraction
of the component.
pi = p · yi  (2.3)
where, pi = partial pressure, kPa
p = total pressure, kPa
yi = mole fraction of the component in gas phase
For an ideal liquid–vapour mixture, equilibrium at a given temperature is
represented by Raoult’s law.
pi = p · yi = pvi · xi  (2.4)
where, pvi = vapour pressure of ith component at the given temperature, kPa
xi = mole fraction of ith component in liquid phase
Conversely, an ideal liquid mixture is one which follows Raoult’s law. Non-ideal
behavior of the liquid mixture is expressed in different ways for which a
standard text book on thermodynamics may be referred (Ref. 4).
For certain liquid mixtures at a definite composition, liquid and vapour
compositions are same at a given pressure. Composition of such a mixture is
known as azeotropic composition. Ethyl alcohol-water, toluene-water and many
multicomponent mixtures exhibit this property which is advantageously used in
the industry to remove water from organic reaction mass.

2.3 MATERIAL BALANCES

A process represents a series of actions or operations or treatments that result in
an end product. In chemical engineering, physical and/or chemical changes take
place during the operations and the material flows and its inventory in a system
are quantified during these operations which is known as material balance.
A system can be defined as the body or assemblage in which the process is set
out for analysis. When a system is isolated from the surroundings, it is called a
batch system. In other words, there is no transfer of materials from a batch
during the time interval of interest. Contrary to this, in an open system, material
is transferred across the boundaries, i.e., from and to the surroundings. Thus it is
important to define the system being analyzed.
Having defined the system, it is important to fix a basis, e.g., 100 kg feed, 100
kg reactant A, 100 kmol gas mixture, 1000 L liquid, etc. All computations are
performed considering this basis. In a prolonged analysis of the system, it may
become necessary to shift bases but interrelation of all the bases should be clear.
It is thus clear that material balance is a quantitative expression of law of
conservation of matter. It can be expressed as
Input = Output + Accumulation  (2.5)
Equation (2.5) is always true for elements but often not true for compounds
participating in chemical reactions. Because of its fundamental nature, material
balance is one of the most useful tools in system analysis and therefore in
process design.
2.3.1 Material Balances Involving No Chemical
Reactions
In most unit operations, reactions do not take place. Material balances of such
systems are worked out by making balances of a tie material or inert material.
When two or more compounds in the system are simultaneously affected,
material balance equations are solved by satisfying simultaneous equations.
Soybean seeds are batch extracted with n–hexane. Flaked seeds
containing 19% oil, 69% solids and 12% moisture (by mass) are fed to the batch
extractor. At the end of the extraction process, deoiled cake (DOC) is found to
contain 1% oil, 88% solids and rest moisture (by mass). Find the percentage
recovery of oil.
Solution
Basis: 100 kg soybean seeds
Since solids do not take part in the extraction process, its balance can be made.
Solids in the feed = 100 × 0.69 = 69 kg
DOC contains 88% solids.
Quantity of DOC = = 78.41 kg

Oil in seeds = 100 × 0.19 = 19 kg

Oil in DOC = 78.41 × 0.01 = 0.784 kg
Oil recovery = × 100

= 95.87%
n–Butanol-water forms an azeotrope having composition of
57.5% n–butanol (by mass) at 93°C and 1 atm. When azeotropic vapour is
condensed and cooled to 20°C, the mixture separates in two layers. Top layer
and bottom layer consist of 79.93% and 7.81% n–butanol, respectively (by mass)
at 20°C. Calculate the proportions of top and bottom layers.
Solution
Basis: 100 kg azeotropic vapour to condenser
n–Butanol in vapour = 57.5 kg
Water in vapour = 100 – 57.5 = 42.5 kg
In the liquid phase, the mixture separates in two layers. Since both components
are affected, simultaneous equations need to be solved.
Let m1 kg be top layer and m2 kg be the bottom layer.
Overall balance: m1 + m2 = 100
n–butanol balance : 0.7993 m1 + 0.0781 m2 = 57.5
Solving the equations, m1 = 68.9 kg, m2 = 31.1 kg
Proportion of top layer/bottom layer = 68.9/31.1 = 2.215 (mass)
Solving simultaneous equations for a binary system could be easily done
manually. However, when many components are present, use of a mathematical
software such as Mathcad®, will be advantageous.
2.3.2 Material Balances Involving Chemical Reactions
Material balance calculations of processes involving chemical reactions follow
the same law of conservation of mass. However, moles are important rather than
mass in chemical reactions. For material balance calculations, conversion and
yield are required.
Conversion is defined as the fraction of the reactant in the feed which is
converted to product(s). Conversion is normally defined on the basis of the
limiting reactant which is present in the smallest stoichiometric amount. Yield is
defined by the following equation:
Yield =   (2.6)

where, stoichiometric factor (vi) = stoichiometric requirement of the specific

reactant (in moles) per mole of the
desired product
Selectivity is the ratio of amount of limiting reactant that reacts to give the
desired product to the amount that reacts to give undesirable product(s).
One more term used in reaction engineering is the excess reactant which is
defined as the reactant present in excess of stoichiometric (theoretical)
requirement of the limiting reactant on mole basis.
While conversion is related to the degree of completion of reaction, yield and
selectivity are related to the degree to which the desired reaction proceeds.
Study of oxidation reaction to produce nitrogen monoxide can be useful to
understand various terms.
Following are the chief competing reactions taking place in the reactor in a nitric
acid plant:
4 NH3 + 5 O2 → 4 NO + 6 H2O(a)
4 NH3 + 3 O2 → 2 N2 + 6 H2O(b)
4 NH3 + 4 O2 → 2 N2O + 6 H2O(c)
Above reactions are carried out over a precious metal catalyst at high
temperature. Aim is to maximize production of NO. Ammonia and oxygen (in
the form of air) are the reactants. At the exit of the reactor, ammonia is
completely consumed and hence it is the limiting reactant. Oxygen is fed in
excess over the stoichiometric requirement (5/4 times ammonia feed rate) and is
termed as excess reactant. In this case, conversion is 100% based on ammonia.
Stoichiometric factor (vi) for reaction (a) is one, i.e., 1 mole ammonia is required
per mole of NO. Typical yield of NO is found to be 0.93–0.98 depending on
operating temperature and catalyst activity. Take for example, yields of NO, N2
and N2O as 0.95, 0.02 and 0.03, respectively.
NH3 consumed for desired product = 0.95 kmol
NH3 consumed for undesired products = 0.02 × 2 + 0.03 × 2
= 0.1 kmol
Selectivity of the catalyst = 0.95/0.1 = 9.5
Combustion of a fuel is another oxidation reaction in which oxygen (i.e., air) is
fed in excess to achieve complete combustion. Fuel is thus the limiting reactant.
Oxygen is supplied in excess over the stoichiometric requirement in the form of
air. Knowing the flue gas and fuel analyses, it is possible to calculate the excess
air.
Flue gas mixture from a waste incinerator is analysed and found
to contain 12% CO2, 5.5% O2 and rest N2 (by mole) on dry basis. Assuming
complete combustion, calculate the excess air and C : H mass ratio in fuel.
Solution
Basis: 100 kmol dry flue gases
Table 2.3 Flue Gas Analysis on Dry Basis
Component kmol O2 accounted kmol

CO2 12 12
 O2 5.5 5.5

N2 82.5 –

Total 100.0 17.5

Assume negligible nitrogen in the fuel (i.e., waste).

O2 supplied in the form of air = × 82.5 = 21.93 kmol

O2 consumed in reaction with hydrogen = 21.93 – 17.5

= 4.43 kmol
Hydrogen reacted = 4.43 × 2
= 8.86 kmol
Carbon in waste = 12 × 12 = 144 kg
Hydrogen in waste = 8.86 × 2 = 17.72 kg
C : H ratio in waste =

= 8.13:1 by mass
Total oxygen consumed for combustion (or stoichiometric oxygen requirement)
= 12 + 4.43
= 16.43 kmol
Excess oxygen or air = (5.5/16.43) 100 = 33.48%
2.3.3 Recycling and Bypassing Operations
Recycle stream is a process stream from exit or downstream from a unit back to
the inlet or upstream of the same unit. Nature is a classical example in which
nitrogen and oxygen are recycled. Animals inhale oxygen and exhale carbon
dioxide. The latter is used by plants for manufacture of food in presence of
sunlight. In this photosynthesis reaction, oxygen is released. In unit operations
and in chemical reactions, recycling operations are performed for a variety of
reasons. One of the chief reasons of recycling is to maximize the utilization of
reactant(s) in the reaction. Hydrogen is used in excess in many reactions. After
the reaction, excess hydrogen is separated and recycled from the separation unit
back to the reaction unit. For recovery of volatile organic compounds (VOC)
from vent gases, carbon adsorption units are commonly employed. Nitrogen is
used for regeneration of the carbon bed from which VOC is separated. After
separation of most VOC, nitrogen is recycled.
Other reasons for recycle are utilization of heat from the outgoing stream, to
improve the performance of an equipment, to improve selectivity of a product, to
improve safety of the process, to minimize waste generation, etc.
Chemical engineers are familiar with reflux in a distillation column. It is
basically a recycle stream to maintain specific quality of the top product. In
Chapter 8, number of examples are given for calculating reflux ratio which
represent material balance in a unit operation.
In a loop reactor, fraction of product stream (after cooling) is recycled back to
the reactor to control the exhothermicity of reaction.

2.4 ENERGY BALANCES

It is known that most physical and chemical changes are accompanied by
enthalpy changes. It is therefore necessary to study the enthalpy balances.
Energy exists in different forms such as potential energy, kinetic energy, thermal
energy, electrical energy, etc. In SI units, the unit of all types of energy is Joule
(represented as J). Conversion of one form of energy into another is feasible.
However, total energy of the universe is constant.
In this section, discussions will be limited to thermal energy and the terms
energy, enthaply and heat are essentially used for thermal energy. While enthalpy
of a substance is associated with its state (i.e., pressure and temperature), the
term heat is used for the enthalpy in transit. In physical and chemical changes,
flow rate of heat (in W in SI units) is of interest to process engineers.
2.4.1 Sensible Heat Changes
Sensible heat changes in solid, liquid or gas means change in temperature of the
substance without phase change. Increase in temperature is termed as heating
while a decrease is termed as cooling. For evaluation of the heat change, mass,
temperatures and heat capacity are required.
For any substance, heat capacity is defined as the heat required to effect the
temperature change by 1°C of 1 kg substance. Thus, units of heat capacity are
kJ/(kg · °C). If the heat capacity of one kmol substance is specified, it is known
as molar heat capacity and its units are kJ/(kmol · °C).
For gases, heat capacity depends on the route, followed for changing the
temperature. One route is known as constant pressure path while another route is
constant volume path. Majority of the processes follow constant pressure route
and the heat capacity for the route is denoted by Cmp and has units kJ/(kmol ·
°C). This quantity is dependent on pressure and temperature. At 1 atm pressure,
it is denoted by symbol C° mp. A number of equations are developed, offering a
relationship of C° mp with temperature at 101.325 kPa. Among these, polynomial
forms are used extensively.
C° mp = a + bT + cT 2  (2.7)
or C° mp = a + bT + cT2 + dT 3  (2.8)
where, a, b, c and d are constants (separate values for both equations) and T is
temperature in K.
Values of C° mp, derived from the equations, are accurate at 1 atm but they can
be safely used for subatmospheric pressure and up to about 10 bar a.
Correlations, graph and tables are available in literature for prediction of Cmp at
high pressures (Ref. 1, 3, 4).
For a mixture of ideal gases, sensible heat changes are calculated by calculating
heat capacity of the mixture, using the following equation:
C° mp mix = Σ yi · C°mpi
where, C ° mp mix = heat capacity of gas mixture, kJ/(kmol · °C)
yi = mole fraction of ith component
C°mpi = heat capacity of ith component of gas mixture, kJ/(kmol · °C)
This equation is strictly valid for an ideal gas mixture but can be used for most
practical purpose up to moderate pressures.
In case of liquids, heat capacities at constant pressure and at constant volume are
nearly equal. Also variation of the heat capacity with pressure is insignificant for
most liquids. Symbol for heat capacity of liquid is Cl. Polynomial equations are
also developed for correlating Cl with T. For aqueous solutions, liquid mixtures,
petroleum fluids, etc., graphical and tabulated data are available in literature.
Heat capacity of an immiscible liquid mixture is given by
Clm mix = Σ xi · Clm i  (2.9)
where, Clm mix = heat capacity of liquid mixture, kJ/(kmol · °C)
xi = mole fraction of ith component
Clm i = heat capacity of ith component of liquid mixture, kJ/(kmol · °C)
Heat capacity of solids (Cs) are lower than that of liquids and increase with the
increase in the temperature. However, variation of Cs with respect to T is
generally insignificant.
2.4.2 Heat Changes During Phase Change
A study of p–t diagram (Fig. 7.28 and Fig. 7.29) reveals interesting facts about
phase changes. Above triple point, there are two phase changes. When a solid is
melted above its freezing point, latent enthalpy (or heat) of fusion is to be
supplied. Ice melts at 0.01°C at standard atmospheric pressure (101.325 kPa).
Latent enthalpy of fusion at this temperature is 339.4 kJ/kg. In general, melting
point of a pure fluid increases with increase in pressure while latent enthalpy of
fusion decreases.
When heat is supplied to convert liquid to vapour at a constant temperature, it is
called that latent enthalpy or latent heat of vaporization (λ). Water evaporates at
100°C at 101.325 kPa when heat equivalent to 2256.9 kJ/kg is supplied.
Saturation temperature of a pure fluid increases with increase in pressure while
latent enthalpy of vaporization decreases and becomes zero at critical point.
Below triple point, a fluid sublimes from solid to vapour. Heat, supplied to
change the solid phase to vapour phase of a pure substance, is called the latent
enthalpy (heat) of sublimation. Sublimation temperature and latent enthalpy of
sublimation decreases with decrease in pressure below triple point. Carbon
dioxide sublimes at –78.46°C at 101.325 kPa and latent heat of sublimation is
571.0 kJ/kg. At triple point, latent heat of sublimation can be considered a total
of latent heat of fusion and latent heat of vaporization. Triple point of carbon
dioxide is –56.56°C at 5.18 bar. At this temperature, latent heat of fusion is
197.22 kJ/kg and latent heat of vaporization is 350.38 kJ/kg. Thus latent heat of
sublimation at this temperature is 547.60 kJ/kg.
Above critical point, again sublimation of solid to vapour is experienced as there
is no liquid phase. Critical temperature and critical pressure of water are
374.15°C and 221.2 bar, respectively. Water sublimes at –5°C at 600 bar and
latent heat of sublimation is 319.92 kJ/kg at this temperature.
Various correlations are available in literature for prediction of latent heat of
fusion and vaporization. However, latent heat of vaporization (λ) is most useful
to a process engineer. Tables of thermodynamic properties (Ref. 2) are available
in literature for large number of compounds. At standard atmospheric pressure,
Riedel equation is found to give latent heat of vaporization within 5% accuracy:
=   (2.10)

where, λ = latent heat of vaporization at TB, kJ/kmol

pc = critical pressure, kPa
TB = normal boiling point, K
TBr = reduced temperature at TB = TB /Tc
R = 8.314 472 kJ/(kmol · K)
Knowing latent heat of vaporization at one temperature, it can be calculated at
another temperature, using the following Watson equation:

  (2.11)

where, λ1 and λ2 refer to latent heat of vaporization at temperatures T1 and T2 (in

K) respectively, Tc is critical temperature (in K) and n is characteristic constant,
having a value of 0.38. Watson equation is simple and reliable in most cases.
For heat changes calculations during phase change, latent heat of appropriate
phase change is to be taken into account.

2.5 ENTHALPY CHANGES ACCOMPANYING

CHEMICAL REACTIONS
When any chemical reaction takes place, enthalpy changes are encountered. This
is known as heat of reaction. For combustion reactions, it is termed as heat of
combustion. For calculation of heat of reaction, heat of formation data are
required.
Standard heat of formation is defined as the isothermal enthalpy change in a
synthesis reaction from the elements in their standard states. Standard heat of
formation, denoted by Δ H° f of large number of compounds, is available in
literature1 in tabular form. Standard states of compounds are pure crystalline
solid at 25°C (298.15 K) and for pure liquid and pure gas are 1 bar a and 25°C
(298.15 K). Since combustion reactions are easy for experiments, standard heat
of combustion (Δ H°c) data so obtained are used to calculate the standard heat of
formation.
Standard heat of formation data are obtained for isothermal conditions. It varies
with temperature. Empirical relation of ΔHf with T is a polynominal equation,
similar to that for heat capacity of gas [Eq. (2.7) or Eq. (2.8)].
Consider a chemical reaction between A and B to produce C and D.
Stoichiometric relation can be written as
nA A + nB B = nC C + nD D
where, ni = number of moles of ith component. Standard heat of reaction (Δ H ° r
) for the above reaction can be calculated by the following equation:
ΔH ° r = Σ (ni · ΔH °fi)Products – Σ (ni · ΔH°fi)Reactants
where, ΔH ° f i = Standard heat of formation of ith component
Since ΔH °f i will be at 25°C, ΔH ° r will be also at 25°C. However, if ΔHf i at
some another temperature is taken (but all at one specific temperature), ΔHr will
correspond to that temperature.
Enthalpy is known to be a state function. Hence, change in enthalpy taking place
while going from one state to another state (either in physical operations or in
chemical reactions) is independent of the pathway. This principal is known as
Hess’s Law.
If heat of reaction is positive, the reaction is said to be endothermic while if the
heat of reaction is negative, the reaction is termed exothermic.
Standard heat of formation of liquid oleic acid (C18H34O2) is –
802.9 kJ/mol at 25°C. Calculate its gross calorific value at 25°C.
Solution
Basis: 1 mole liquid oleic acid
Gross calorific value or higher heating value is the heat liberated when a mole of
the combustible compound is burnt with stoichiometric amount of pure oxygen
at 25°C, product carbon dioxide is in gaseous form while product water is in
liquid form. Combustion reaction of oleic acid can be written as
C18H34O2(l) + 25.5 O2(g) = 18 CO2(g) + 17 H2O(l)
ΔH° f i –802.9 0 18 × (–393.51) 17 × (–285.83)
Standard heat of formation data for CO2 and H2O are taken from Ref. 1.
ΔH°r = 18 (–393.51) + 17 (–285.83) – (–802.9)
= –11 139.4 kJ/mol of liquid oleic acid
Literature reports ΔHr at 20°C for liquid oleic acid as –11 154 kJ/mol.
Natural gas having following molar composition is considered
as a raw material for manufacture of ammonia.
CH4: 78.5%, C2H6: 8.2%, C3H8: 4.1%, i-C4H10: 0.6%, n-C4H10: 0.4% and CO2:
8.2%
Hydrogen is produced by reforming of hydrocarbons with steam as shown
below.
CnH2n+2 + 2n H2O = n CO2 + (3n + 1) H2
Dry air, having molar composition 78% N2, 21% O2 and 1% Ar, is used for
supplying nitrogen. Oxygen of air will react with H2 as under.
2 H2 + O2 = 2 H2O
(a) Assume 100% conversions of all reactions, 100% removal of ammonia from
the reaction mass and venting of Ar to atmosphere. Calculate stoichiometric
requirement of natural gas per tonne of ammonia.
(b) Considering net calorific values, of hydrocarbons, calculate specific energy
requirement per tonne of ammonia.
Solution
Basis: 1000 kg ammonia production
Molar mass of ammonia = 17.0305
Molar production of ammonia = 1000/17.0305
= 58.7182 kmol
Reaction of ammonia manufacture:
3 H2 + N2 = 2 NH3
Stoichiometric H2 requirement = (3/2) 58.7182
= 88.0773 kmol
Stoichiometric N2 requirement = (1/2) 58.7182
= 29.3591 kmol
Supply of air brings 21 O2 kmol of 78 kmol N2.
O2 equivalent to N2 requirement = (21/78) 29.3591
= 7.9044 kmol
H2 consumed by O2 = 2 × 7.9044
= 15.8088 kmol
Total H2 required = 88.0773 + 15.8088
= 103.8861 kmol
H2 produced by reforming reaction of natural gas
= 0.785 × 4 + 0.082 × 7 + 0.041 × 10 + 0.01 × 13
= 4.254 kmol/kmol
Natural gas requirement for 1000 kg ammonia
= 103.8861/4.254
= 24.4208 kmol
≡ 547.37 Nm3Ans. (a)
Table 2.4 Composition of Natural Gas
 Component Molar mass (M) kmol NCV* (ΔH°c) kJ/mol ni · NCVi kJ
ni

CH4 16.0425 0.785 802.62 630.057

C2H6 30.0690 0.082 1428.64 117.148

C3H8 44.0956 0.041 2043.11 83.768

i-C4H10 58.1222 0.006 2648.12 15.889

n-C4H10 58.1222 0.004 2657.32 10.629

CO2 44.0095 0.082 – –

Total 1.000 857.491

*Ref. 1.
NCV of natural gas = 857.491 kJ/mol
Specific energy requirement = 857.491 × 1000 × 24.4208/106
= 20.941 GJ/t ammoniaAns. (b)
In the manufacture of nitric acid, ammonia is oxidized to nitric
oxide on a Pt/Rh gauge (as catalyst) as per the following reactions:
4 NH3(g) + 5 O2(g) = 4 NO(g) + 6 H2O(g) (a)
4 NH3(g) + 3 O2(g) = 2 N2(g) + 6 H2O(g) (b)
4 NH3(g) + 4 O2(g) = 2 N2O(g) + 6 H2O(g) (c)
Liquid ammonia at atmospheric pressure and –33.33°C is pumped to 9.0 bar a. It
is first vaporized with cooling water and subsequently superheated to desired
temperature using low pressure saturated steam at 5 bar a. Oxygen in the form of
air is supplied. Air is compressed to 9.3 bar a and 130°C in a two–stage turbo
compressor, driven by a steam turbine. Further it is heated to 260°C with the
reactor exit stream. Superheated ammonia at the rate of 3266 Nm3/h and hot air
at the rate of 29 394 Nm3/h are fed to the reactor. Ambient air has 0.016 kg
moisture/kg dry air. In the reactor, ammonia is completely consumed. Yields of
NO, N2 and N2O are 95%, 2% and 3%, based on ammonia conversion,
respectively. Reactor exit gas stream temperature is observed to be 915°C at 7.5
bar g. Carry out the material and energy balances of the reactor system and
preheating equipments.
The process flow diagram is given as Fig. 2.1.
Solution
Basis: Ammonia flow rate to reactor, qam = 3266 Nm3/h
Molar flow rate of ammonia, qam = 3266/22.414
= 145.71 kmol/h
Volumetric flow rate of moist air, qa = 29 394 Nm3/h
Molar flow rate of moist air, qa = 29 394/22.414
= 1311.44 kmol/h
Assume dry air molar composition to be 21% O2, 1% Ar and 78% N2.

Fig. 2.1 Manufacture of Nitric Acid

Average molar mass of dry air = 0.21 × 31.9988 + 0.01 × 39.948 + 0.78 ×
28.0134
= 28.9697 (referred in Example 2.4)
Moisture in dry air = 0.016 kg/kg dry air
= 0.016 × 28.9697/18.0153
= 0.0257 kmol/kmol dry air
Moist air = 1.0257 kmol/kmol dry air
Moisture, entering with air = 1311.44 × 0.0257/1.0257
= 32.86 kmol/h
Dry air flow rate to reactor = 1311.44 – 32.86
= 1278.58 kmol/h
O2 supply to reactor = 1278.58 × 0.21 = 268.50 kmol/h
Argon supply to reactor = 1278.58 × 0.01 = 12.79 kmol/h
N2 supply to reactor = 1278.58 – 268.50 – 12.79
= 997.29 kmol/h
Table 2.5 Thermodynamic Data1
Heat capacity equation constants
Component State Molar mass kg/kmol ΔH° f kJ/mol
ai bi × 103 ci × 106 di × 109

NH3 gas 17.0305 –45.94 25.6503 33.4806 0.3518 –3.0832

O2 gas 31.9988 0 26.0257 11.7551 –2.3426 –0.5623

N2 gas 28.0134 0 29.5909 –5.141 13.1829 –4.968

Ar gas 39.948 0 20.7723 – – –

NO gas 30.0061 90.25 29.4867 –2.0524 11.3379 –4.8195
N2O gas 44.0128 82.05 23.2082 63.6695 –42.3402 10.4061

NO2 gas 46.0055 33.18 23.5804 53.5944 –31.4901 6.5394

H2O gas 18.0153 –241.82 32.4921 0.0796 13.2107 –4.5474

H2O liquid 18.0153 –285.83

HNO3 liquid 63.0128 –174.10

Conversion = 100% of NH3 and Yield of NO = 95%

NH3 converted to NO = 145.71 × 0.95
= 138.42 kmol/h
NO produced = 138.42 kmol/h
NH3 converted N2 = 145.71 × 0.02
= 2.91 kmol/h
N2 produced = 2.91 × 2/4
= 1.457 kmol/h
NH3 converted to N2O = 145.71 – 138.42 – 2.91
= 4.38 kmol/h
N2O produced = 4.38 × 2/4
= 2.19 kmol/h
Total O2 consumed = 138.42 (5/4) + 2.91 (3/4) + 4.38 (4/4)
= 173.025 + 2.183 + 4.38
= 179.588 kmol/h
O2 in reactor exit stream = 268.50 – 179.588
= 88.912 kmol/h
Total H2O produced = 145.71 × 6/4
= 218.565 kmol/h
Total H2O in reactor exit stream = 218.565 + 32.86
= 251.425 kmol/h
Total N2 in reactor exit stream = 1.457 + 997.29
= 998.747 kmol/h
Table 2.6 Reactor Ingoing Stream
Heat capacity equation constants
kmol/h mole %
Component kg/h m.i n.i · bi n.i · ci n.i · di
n.i (wet) n.i · ai
× 103 × 106 × 109

NH3 145.71 10.00 2 481.51 3 737.5 4878.5 51.3 –449.3

N2 997.29 68.44 27 937.75 29 510.7 –5127.1 13 147.2 –4954.5

O2 268.50 18.43 8 591.68 6 987.9 3156.2 –629.0 –151.0

Ar 12.79 0.87 510.93 265.7 – – –

H2O 32.86 2.26 591.98 1 067.7 2.6 434.1 –149.4

Total 1457.15 100.00 40 113.85 41 569.5 2910.2 13 003.6 –5704.2

Average molar mass of reactor ingoing stream

= 40 113.85//1457.15 = 27.53 kg/kmol
Table 2.7 Reactor Exit Stream
Heat capacity equation constants
Component kmol/h n.i mole % (wet) kg/h m.i n.i · bi n.i · ci × n.i · di
n.i · ai
× 103 × 106 × 109

N2 998.747 66.92 27 978.30 29 553.8 –5134.6 13 166.4 –4961.8

O2 88.912 5.96 2 845.08 2 314.0 1045.2 –208.3 –50.0

Ar 12.790 0.85 510.93 265.7 – – –

NO 138.420 9.27 4 153.44 4 081.5 –284.1 1569.4 –667.1
N2O 2.190 0.15 96.39 50.8 139.4 –92.7 22.8

H2O 251.425 16.85 4 529.50 8 169.3 20.0 3321.5 –1143.4

 Total 1492.484 100.00 40 113.64 44 435.1 –4214.1 17 756.3 –6799.5

Average molar mass of reactor exit stream

= 40 113.64/1492.484 = 26.88 kg/kmol
It may be noted that mass flow rates of ingoing and outcoming streams are equal
which should be checked during every material balance calculations. However,
molar flow rates are different.
Energy Balance
For carrying out energy balance calculations, reference (or base) temperature
need to be fixed. Since standard heat of formation data are based on 25°C,
selection of 25°C as the reference temperature is logical.
T0 = 25 + 273.15 = 298.15 K
Reactor exit gas temperature is desired at 915°C (1188.15 K).
Heat flow rate of reactor exit stream

ϕ8 = (44 435·1 – 4214·1 × 10–3 T + 17 756·3 × 10–6 T 2 – 6799.5 × 10–9 T 3)

dT
= 43 156 523 kJ/h ≡ 11 987.92 kW
Standard Heats of Reactions
For reaction (a):
ΔH°r1 = [4(90.25) + 6(–241.82)] – [4(–45.94) + 5(0)]
= 361 – 1450.92 + 183.76
= –906.16 kJ/4 mol NH3 reacted
≡ –226.54 kJ/mol NH3 reacted (exothermic)
For reaction (b):
ΔH°r2 = [2(0) + 6 (–241.82)] – [4 (–45.94) + 3(0)]
= –1450.92 + 183.76
= –1267.16 kJ/4 mol NH3 reacted
≡ –316.79 kJ/mol NH3 reacted (exothermic)
For reaction (c):
ΔH°r3 = [2 (82.05) + 6 (–241.82)] – [4 (–45.94) + 4(0)]
= 164.1 – 1450.92 + 183.76
= –1103.06 kJ/4 mol NH3 reacted
≡ –275.765 kJ/mol NH3 reacted (exothermic)
Total heat liberated during reaction,
ϕ7 = (226.54 × 138.42 + 316.79 × 2.91 + 275.765 × 4.38) 1000
= 33 487 377 kJ/h ≡ 9302.05 kW
Heat input to reactor (i.e., of the reactants’ stream)
ϕ6 = 43 156 523 – 33 487 377
= 9 669 146 kJ/h
≡ 2685.87 kW
Let temperature of reactants’ stream be T4.

ϕ6 = (41 569.5 + 2910.2 × 10–3 T + 13 003.6 × 10–6 T 2 – 5704.2 × 10–9 T 3)

dT

= 9669 146 kJ/h

Solving by trial and error or by Mathcad, T4 = 515.22 K or t4 = 242.07°C
Among the two reactants, air is introduced at 260°C or 533.15 K (T3).
Heat input of air,
Heat capacity equation constants for air are drawn from Table 2.6.

ϕ5 = (37 832 – 1968.3 × 10–3 T + 12 952.3 × 10–6 T 2 – 5254.9 × 10–9 T 3)

dT
= 9142 364 kJ/h ≡ 2539.55 kW
Heat input by superheated ammonia,
ϕ4 = 9669 146 – 9142 364
= 526 782 kJ/h ≡ 146.33 kW
If temperature of superheated ammonia is T2,

ϕ4 = (3737.5 + 4878.5 × 10–3 T + 51.3 × 10–6 T2 – 449.3 × 10–9 T3) dT

= 526 782
Solving by Mathcad,
T2 = 395.41 K or t2 = 122.26°C
Ammonia is fed at –33.33°C (saturation temperature at 101.325 kPa) as liquid.
Assume that in the evaporator, liquid ammonia is vaporized and heated at 9.0 bar
a to its saturation temperature of 21.5°C with cooling water.
T1 = 21.5 + 273.15 = 294.65 K
From ammonia property tables (Ref. 2):
Enthalpy of liquid ammonia, h = 49.08 kJ/kg at –33.33°C
Enthalpy of vapour ammonia at 9 bar a and 21.5°C, H1 = 1480.78 kJ/kg
Note that the values of both the above enthalpies are based on reference
temperature of 0°C in ammonia tables which is different from original reference
temperature of 25°C. Nevertheless, since difference in two enthalpy values only
is required, it does not matter.
Heat duty of ammonia evaporator,
ϕ1 = 145.71 × 17.0305 (1480.78 – 49.08)
= 3552 784 kJ/h
≡ 986.88 kW
Cooling water at 32°C is used to evaporate liquid ammonia. Assume a drop in
cooling water temperature by 5°C.
Cooling water flow, ṁCW =

= 169 714 kg/h

≡ 169.7 m3/h
Enthalpy of ammonia gas at 9.0 bar a and 122.26°C, H2 = 1739.55 kJ/kg
Heat duty of superheater, ϕ2 = 145.71 × 17.0305 (1739.55 – 1480.18)
= 643 630 kJ/h
≡ 178.79 kW
If heat capacity data (at 1 atm) for ammonia, are used for calculations of ϕ2,

ϕ2 = (3737.5 + 4878.5 × 10–3 T + 51.3 × 10–6 T 2 – 449.3 × 10–9 T 3) dT

= 544 914 kJ/h

≡ 151.37 kW
Error = × 100

= 15.34% – significant
Error indicates justification of tabular data, particularly at high pressures.
Saturated steam at 5 bar a is used for superheating ammonia vapour from 21.5°C
to 122.26°C.
Latent heat of vaporization of steam at 5 bar a, λ = 2107.4 kJ/kg (Ref. 2)
Steam consumption in superheater,
ṁs = = 305.4 kg/h

Heat duty of air preheater,

ϕ3 = (37 832 – 1968.3 × 10–3 T + 12 952.3 × 10–6 T 2 – 5254.9 × 10–9 T 3)

dT

= 5098 328 kJ/h ≡ 1366.16 kW

2.6 HEAT CHANGES IN MIXING PROCESSES

When a solid or gas is dissolved in a solvent, heat change takes place. This is
known as heat of solution. Also when two liquids are mixed, a heat change takes
place which is termed as heat of mixing. These heat changes are measured at
constant temperature, say at 25°C (298.15 K) and at standard atmospheric
pressure. Units of the heat changes are kJ/mol solute, kJ/kg solution, etc.,
depending on user’s convenience.
When solids like caustic soda, caustic potash, sodium carbonate, etc., are
dissolved in water, heat is evolved. Similarly, when acids are mixed (e.g., a
mixture of HNO3 + H2SO4 + H2O), heat is released. Dissolution of ammonia or
formaldehyde in water is exothermic. These heat changes should be considered
while making heat balance calculations.
Heat of mixing data are extensively covered in many handbooks. Heat of
formation data for large number of compounds and their solutions are published
by National Institute of Standards and Technology, USA which permits heat of
mixing calculations. Reference 1 may be consulted for more discussions on the
subject.
Following example will illustrate the method of calculations relating to heat of
absorption in gas–liquid system.
Refer to Example 2.10. Reactor exit stream is cooled in various
heat exchangers to 40°C and introduced in an absorber along with the
condensate (formed in the coolers), operating at 1.5 bar g. Secondary air at the
rate of 8145 Nm3/h is also introduced at the bottom of the absorber at 40°C
which converts NO to NO2. Fresh demineralized water is introduced from the
top at 25°C which scrubs NO2 to produce HNO3. Absorption of NO2 in water is
exothermic and therefore cooling coils are provided in the absorber. Tail gas,
leaving the absorber at 0.1 bar g and 50°C, is found to contain 0.2% NO (on dry
basis). Aqueous 58% acid (by mass) leaves absorber at 50°C from the bottom.
Make material and energy balance of the absorber. Find cooling water
requirement for the absorption system assuming availability of cooling water at
32°C and a rise of 5°C.
Solution
Basis: Same as that of Example 2.10.
Pressure of gas at the inlet of absorber, = 1.5 bar g
= 251.325 kPa a
Vapour pressure of water at 40°C = 7.375 kPa
Refer to Table 2.6.
Moisture in ingoing gas mixture =

= 37.519 kmol/h at 40°C

≡ 675.92 kg/h
Liquid water entering absorber = 251.425 – 37.519
= 213.906 kmol/h at 40°C
≡ 3853.58 kg/h
Pressure of outgoing mixture = 0.1 bar g
= 111.325 kPa a
Vapour pressure of water at 50°C = 12.335 kPa
Water content of tail gas =

= 0.1246

Secondary air flow = 8145 Nm3/h

Molar flow rate of wet air = 8145/22.414
= 363.4 kmol/h
Moisture, entering with air = 363.4 × 0.0257/1.0257
= 9.11 kmol/h
Dry secondary air flow rate = 363.4 – 9.11
= 354.29 kmol/h
O2 entering the absorber = 354.29 × 0.21
= 74.40 kmol/h
Ar entering the absorber = 354.29 × 0.01
= 3.54 kmol/h
N2 entering the absorber = 354.29 – 74.40 – 3.54
= 276.35 kmol/h
Table 2.8 Gas Stream Entering Absorber
Heat capacity equation constants
Component kmol/h n.i mole % (wet) kg/h m.i
n.i · ai n.i · bi × 103 n.i · ci × 106 n.i · di × 109

N2 1275.097 77.66 35 719.80 37 731.3 –6555.3 16 809.5 –6334.7

O2 163.312 9.95 5 225.79 4 250.3 1919.7 –382.6 –91.8
Ar 16.330 0.99 652.35 339.2
NO 138.420 8.43 4 153.44 4 081.5 –284.1 1569.4 –667.1
N2O 2.190 0.13 96.39 50.8 139.4 –92.7 22.8
H2O 46.629 2.84 840.04 1515.1 3.7 616.0 –212.0

Total 1641.978 100.00 46 687.81 47 968.2 –4776.6 18 519.61 –7282.8

Average molar mass of gas mixture, entering absorber = = 28.43

Reaction
2 NO(g) + 1.5 O2(g) + H2O(l) = 2 HNO3(l)(d)
Let n.1 kmol/h NO is reacted by the above reaction.
O2 consumed = 1.5 n.1/2
= 0.75 n.1 kmol/h
NO in tail gas = 138.42 – n.1 kmol/h
Dry tail gas = 1641.978 – n.1 – 0.75 n.1 – 46.629
= 1595.349 – 1.75 n.1 kmol/h
Tail gas contains 0.2% NO on dry basis.
= 0.2

n.1 = 135.70 kmol/h

Dry tail gas = 1595.349 – 1.75 × 135.70
= 1357.874 kmol/h
Moisture in tail gas = 0.1246 × 1357.874
= 169.191 kmol/h
 Table 2.9 Absorber Exit (Tail Gas) Stream
kmol/h mole % kg/h Heat capacity equation constants
Component n.i m.i
(dry) n.i · ai n.i · bi × 103 n.i · ci × 106 n.i · di × 109

N2 1275.097 93.90 35 719.80 37 731.3 –6555.3 16 809.5 –6334.7

O2 61.537 4.54 1 969.11 1 602.9 724.0 –144.3 –34.6
Ar 16.330 1.20 652.35 339.2 — — —
NO 2.720 0.20 81.62 82.3 –5.7 31.6 –13.4
N2O 2.190 0.16 96.39 50.8 139.4 –92.7 22.8
H2O 169.191 — 3 048.03 5647.3 13.8 2296.1 –790.4

Total 1527.065 100.00 41 567.30 45 453.8 –5683.8 18 900.2 –7150.3

Average molar mass of exit gas stream = 41 567.3/1527.065 = 27.22 kg/kmol

Reference temperature, T0 = 298.15 K
Heat input to absorber (of gas stream)

ϕ9 = (47 968.2 – 4776.6 × 10–3 T + 18 519.6 × 10–6 T 2 – 7282.8 × 10–9 T 3)

dT
= 720 459 kJ/h
≡ 200.13 kW
Heat in tail gas stream

ϕ10 = (45 453.8 – 5683.8 × 10–3 T + 18 900.2 × 10–6 T 2 – 7150.3 × 10–9 T 3)

dT

= 1132 458 kJ/h

≡ 314.57 kW
Heat of reaction of Reaction (d):
ΔH°r = 2 (–174.10) – [2 (90.25) + 1.5(0) + (–285.83)]
= –348.2 – 180.5 + 285.83
= –242.87 kJ/2 mol NO reacted
≡ –121.435 kJ/mol NO reacted (exothermic)
Total heat generated during reaction:
ϕ11 = 121.435 × 1000 × 135.7
= 16 478 730 kJ/h
≡ 4577.42 kW
Water evaporated in absorber = 169.191 – 46.629
= 122.562 kmol/h
≡ 2208 kg/h
Assume average temperature of 45°C in the absorber.
λ at 45°C = 2394.9 kJ/kg (Ref. 2)
Heat absorbed during evaporation,
ϕ12 = 2208 × 2394.9
= 5287 939 kJ/h ≡ 1468.9 kW
Total HNO3 produced = 135.7 kmol/h
= 8549.1 kg/h (100%)
Strength of aqueous acid at the bottom of the absorber = 58%
Aqueous acid produced =

= 14 739.8 kg/h
Water in the acid = 14 739.8 – 8549.1
= 6190.7 kg/h
Fresh demineralized water input
= 6190.7 + (135.7/2) 18.0153 – (3853.58 + 840.04 – 3048.03)
= 5767.4 kg/h
≡ 5.76 m3/h
Since demineralized water is added at 25°C, its heat input is zero.
ϕ13 = 0
When HNO3 dissolves in water, heat change takes place.
ΔH° f of 58% HNO3 solution = –197.0 kJ/mol HNO3
ΔH°f of 100% HNO3(l) = –174.2 kJ/mol HNO3
Heat change = –197.0 – (–174.2)
= –22.8 kJ/mol HNO3 (exothermic)
Heat evolved by dissolution of HNO3 in water at 25°C:
ϕ14 = 22.8 × 1000 × 135.7
= 3093 960 kJ/h
≡ 859.43 kW
Heat capacity of 58% HNO3 = 2.64 kJ/(kg · K)
Enthalpy of liquid 58% HNO3 over 298.15 K
= 2.64 (313.15 – 298.15)
= 39.6 kJ/kg at 313.15 K
Heat of 58% acid at 40°C over 25°C
ϕ15 = 8549.1 × 39.6
= 338 544 kJ/h
≡ 94.04 kW
Heat of secondary air at 40°C

ϕ16 = (37 832 – 1968.3 × 10–3 T + 129 523 × 10–6 T 2

– 5254.9 × 10–9 T 3) dT
= = 159 153 kJ/h

≡ 44.21 kW
Enthalpy of condensate, entering absorber:
ϕ17 = 3853.58 (40 – 25) 4.1868
= 242 013 kJ/h
≡ 67.23 kW
Heat to be removed by cooling water
= ϕ9 + ϕ11 + ϕ13 + ϕ14 + ϕ16 + ϕ17 – ϕ10 – ϕ12 – ϕ15
= 200.13 + 4577.42 + 0 + 859.43 + 44.21 + 67.23 – 314.57 – 1468.9 – 94.04
= 3964.95 kW
Cooling water is supplied to internal coils at 32°C and leaves it at 37°C.
Required cooling water flow rate, ṁCW =

= 681 849 kg/h

≡ 681.8 m3/h
An example relating to absorption of formaldehyde in water to produce 37%
aqueous solution is solved as Example 9.6 in Chapter 9. Both these examples
refer to absorption with heat of absorption.

EXERCISES
2.1 A cylinder is filled with 6 Nm3 hydrogen. Its pressure and temperature are
150 bar g and 35°C. If the actual specific volume of hydrogen at these
conditions is 0.1847 m3/kmol, what is the geometric capacity of the cylinder?
Instead of using actual specific volume data, if ideal gas law is used for
calculations, what will be the % error?
2.2 The chloride of some element contains 90.8% chlorine (by mass). What is
 the formula of the chloride compound?
2.3 A mixture, amounting to 18.3 kg of a trivalent metal and its oxide, is
dissolved in 32.0 kg sulphuric acid having 98% strength. The mixture reacts
with stoichiometric amount of sulphuric acid and forms metal sulphate. There
is no excess sulphuric acid leftover. Calculate the composition of the mixture.
2.4 Complete combustion of an alkane produces water vapour, equivalent to the
oxygen consumed (on mole basis). Find the alkane.
2.5 Nitration product of a hydroxybenzene [C6H(6–a)(OH)a] contains 49% oxygen
by mass. Determine the structure of the compound.
2.6 Reduction of nitrobenzene is to be carried out in a batch reactor. At the end
of the reaction, aniline is steam distilled and collected in a receiver at 40°C.
Distilled mass is separated in two layers. Top layer is decanted in another
vessel and measured to be 910.2 kg. Bottom layer is measured to be 1089.8
kg. It is known from the literature5 that top aniline rich layer contains 94.3%
aniline while the bottom aqueous layer contains 3.83% aniline by mass at
40°C.
(a) Calculate total aniline distilled.
(b) Assume that all aniline, produced during reaction, is distilled out. If 1200 kg
nitrobenzene was charged to the reactor and its conversion is 100%, what is
the yield of aniline?
2.7 A wet scrubber is deployed to capture suspended particles from a gas stream.
For this purpose water is circulated in the scrubber in countercurrent direction
to gas stream which scrubs the suspended particles. A finite purge stream is
maintained to control accumulation of solids in the scrubber.
A wet scrubber is to be designed for treating 50 000 Am3*/h at near room
temperature with a particulate loading of 1830 mg/Am3. Assume that the
scrubber particulate removal efficiency of 95% and the maximum suspended
solids level, desirable in the scrubber, is 2% by mass. Use a liquid to gas ratio
of 1.1 L/Am3. Calculate the liquid purge rate and required recirculation pump
flow rate for the scrubber system. Assume specific gravity of the circulating
solution to be 1.02.
2.8 Natural gas, specified in Example 2.9, is considered as a raw material for the
manufacture of methanol. H2 and CO2 produced by reforming with steam are
reacted at specified conditions in a converter.
CO2 + 3 H2 = CH3OH + H2O
(a) Consider 100% conversion of all reactions and calculate theoretical
requirement of natural gas per tonne methanol.
(b) Considering net calorific values of hydrocarbons, calculate specific energy
requirement per tonne of methanol.
2.9 Refined soybean oil is split with water at 30 bar g and 240°C in a high
pressure autoclave. Composition of the soybean oil is given in Table 2.10.
Table 2.10 Composition of Soybean Oil
Fatty acid Formula mass % Molar mass

Palmitic acid C16H32O2 11.1 256.42

Stearic acid C18H36O2 3.8 284.48

Oleic acid C18H34O2 27.3 282.46

Linoleic acid C18H32O2 51.1 280.45

Linolenic acid C18H30O2 6.7 278.43

Assume that all fatty acids are present as triglycerides and refined oil does not have any other impurity.
(a) Calculate theoretical requirement of soybean oil for production of 100 kg
fatty acids.
(b) Calculate theoretical water requirement for splitting.
(c) If the mixed feed to the autoclave contains 0.35 kg water per kg soybean oil,
what will be the excess water fed (as percentage of theoretical required).
(d) Splitted dry mass contains fatty acids and glycerine. It is analysed by testing
with aqueous KOH solution. Acid value of the dry mass is defined as mg
KOH required for neutralization of 1 g mass product. Calculate theoretical
acid value of the dry product, assuming 100% splitting.
2.10 A medium oil length alkyd resin is prepared by processing a mixture of 300
kg soybean oil, 200 kg phthalic anhydride [C6H4(CO)2O] and glycerol
[(CH2OH)2CHOH] together. Basically these are estrification reactions.
Assume soybean oil composition as that given in Exercise 2.9. The process
stipulates 10% excess of total hydroxyl (–OH) equivalents (eB) over total acid
(–COOH) equivalents (eA). Calculate the required quantity of glycerol to be
used in the mixture.
2.11 Ethylene chlorohydrin and sodium bicarbonate are reacted in a continuous
stirred tank reactor. Reaction takes place as given in the following equation at
110 kPa a and 82°C:
CH2OH CH2OH
+ NaHCO3 = + NaCl + CO2
CH2Cl CH2OH
Aqueous solution of 30% of ethylene chlorohydrin and aqueous solution of 15%
NaHCO3 (both by mass) are fed to the reactor in equimolar proportion.
 Conversion of the reactants is 95% while the yield of ethylene glycol is
100%. Calculate the mass % and mole % compositions of the product stream.
Carbon dioxide, produced during the reaction, is vented out at 110 kPa a and
82°C. Calculate the concentrations of carbon dioxide, water vapour, ethylene
chlorohydrine and ethylene glycol in the vent gas stream. Assume ideal
behavior of the liquid mixture.
Data:
(a) Vapour pressures of water, ethylene chlorohydrine and ethylene glycol at
82°C are 51.329 kPa, 17.065 kPa and 0.79 kPa, respectively.
(b) Lowering of vapour pressure of water due to the presence of inorganic salts
(NaCl + NaHCO3) having less than 10% concentration (by mass) in water
at 82°C is 2.62 kPa.
2.12 Effluents generate at various locations in a composite textile unit. Reject
water from reverse osmosis plant for production of water for boiler feed (BF)
make-up purpose, boiler and cooling tower blowdowns, dye-house effluents,
fabric washing liquor, etc., collect in a pit. Raw water from an underground
source and combined effluent analyses are given in Table 2.11.
Table 2.11 Analysis of Raw Water and Combined Effluent
Parameter Raw water Combined effluent

pH 7.8–7.9 7.5–7.9
Dissolved solids (DS), mg/L 1250 12 600
Total hardness as CaCO3, mg/L 230 500
Permanent hardness as CaCO3, mg/L Nil Nil
Calcium hardness as CaCO3, mg/L 170 375
Total alkalinity as CaCO3, mg/L 310 1190
Chlorides as Cl, mg/L 320 6260
Sulphates as SO4, mg/L 156 210
Silica as SiO2, mg/L (max.) 3 30
Suspended solids, mg/L (max.) Nil 100
Chemical oxygen demand (COD), mg/L (max.)-dichromate value Nil 800
Biological oxygen demand (BOD), mg/L (max.) Nil 75
Oil and grease, mg/L (max.) Nil 3
Colour None Dark

At first, the textile unit adopted primary and secondary treatments for removal of oil, grease, suspended
matter (< 30 mg/L), COD (< 250 mg/L), BOD (< 30 mg/L) and colour. Effluents with nearly 13 000
mg/L DS after these treatments used to go to solar evaporation lagoons. Capacity of the plant was 100
m3/d.
Shortage of water of underground source over a period forced the textile unit to
consider integrated additional treatments for recovery of water for reuse. The
 unit requires about 75 m3/d of raw water for BF Make-up purpose for passing
through the reverse osmosis (RO) plant.
In the newly designed treatment facility (see Fig. 2.2) effluent from the existing
treatment plant is first passed through an ultrafilter in which multicapillary
hollow fiber membrane is used. This filter acts as a polishing filter in which
suspended solids, organic impurities, colloidal silica, etc., are removed. Clear
effluents with nearly same composition (in terms of inorganic compounds)
and 12 250 mg/L DS is passed through two stage RO units. In the first stage,
spirally wound membranes for brackish water are housed. Effluents enter at
25 bar g to this unit. Recovery of 62% is achieved as permeate which contains
520 mg/L DS and 275 mg/L chlorides (as Cl). Reject from the first unit is
pumped to 50 bar g and passed through the second stage of the RO unit. This
unit houses spirally wound membranes, designed for high solids brackish
water desalination. Recovery of 28% is achieved from the second stage as
permeate which contains 960 mg/L DS and 486 mg/L Cl. Permeates from
both the stages are collected, mixed and passed through the existing reverse
osmosis plant at 10 bar g to produce BF make-up water.
 Fig. 2.2 Proposed Effluent Treatment Facility for a Composite Textile Unit

Reject from the second stage RO plant is pumped again to 7.5 bar g and passed
through a nanofilter. In this unit, special spiral wound membranes are housed
which selectively removes chlorides in the permeate. Recovery from the
nanofilter is 90% in which 95% of the chlorides, present in the feed, are
 dissolved. Reject from the nanofilter is sent to solar evaporation lagoons.
(a) Calculate the total permeates, available from two stages of RO system. Check
whether this quantity is adequate for production of BF make-up water.
(b) Brine (i.e., permeate with high chlorides from the nanofilter) is reused for
production of 12% NaCl solution for dyehouse. Calculate the flow rate and
concentration of brine.
(c) Calculate the reduction in final effluent, discharged to solar evaporation
lagoons.
Hint : Assume chlorides as NaCl at all points in the treatment plant.

2.13 Ethyl benzene is an important intermediate for the manufacture of styrene.

This intermediate can be manufactured by reacting benzene with ethylene
over a catalyst in vapour phase. Patented Mobil–Badger process uses Zeolite
as catalyst. Chief reaction can be written as under.
C6H6 + CH2 = CH2 = C6H5CH2CH3
Benzene Ethylene Ethyl benzene
Alkylation reaction does not stop at monoethyl benzene but a part of ethyl
benzene is further alkylated to diethyl benzene, triethyl benzene, etc.,
collectively referred to as polyethyl benzenes.
C6H5CH2CH3 + CH2 = CH2 = CH3CH2C6H4CH2CH3
Ethyl benzene Ethylene Diethyl benzene
A high benzene to ethylene ratio in the feed gives a low polyethyl benzenes to
ethyl benzene ratio in the product stream. Excess benzene is recovered from
the product gas stream and recycled to the feed.
While pure ethylene stream can be used as a feed, dilute gas stream, such as
treated off gas mixture from Fluid Catalytic Cracking (FCC) unit of a
petroleum refinery with low ethylene concentration, can also be used as the
feed.
Figure 2.3 is a simplified flow diagram of the process. Fresh benzene is
vaporized and mixed with recycled benzene vapour. Ethylene in the form of
offgas mixture from an FCC unit of refinery, having molar composition of
CH4: 37%, C2H6: 19%, C2H4: 19%, H2: 9%, N2: 13% and CO: 3%, is
compressed to 12 bar g and fed to the reactor after heating in a heater, using
molten salt.
Reactor has three 3 catalyst beds and operates at 12 bar a. Offgases are heated
and sent to first bed. An overall molar ratio of 10 : 1 of benzene to ethylene is
used for the reaction. Of the total benzene, 40% quantity is fed to the first bed
at 350°C. Assume 50% conversion of ethylene in the first bed with 90% yield
to ethyl benzene. Offgas temperature is so controlled that the product gas
mixture from the first bed is maintained at 420°C. A definite quantity of
benzene at 300°C is mixed into the production gas mixture from the first bed.
In the second bed, additional 25% conversion of ethylene (on total feed basis)
is achieved with 90% yield to ethyl benzene. Temperature of product gas
from the second bed is maintained at 400°C. The product gas mixture from
the second bed is quenced with balance quantity of benzene at 300°C. In the
third bed, rest 25% conversion takes place with 90% yield to ethyl benzene.
All polybenzenes may be assumed as diethyl benzene for calculation purpose.
 Fig. 2.3 Manufacture of Ethyl Benzene

Table 2.12 Data on Gaseous Components1

Gaseous heat capacity equation constants at 1 bar
Component Formula ΔH° f , (gas) kJ/mol at 1 bar & 25°C
ai bi × 103 ci × 106 di × 109

Benzene C6H6 82.93 –37.99 490.42 –321.39 79.36

 Methane CH4 –74.52 19.25 52.11 11.97 –11.32

Ethane C2H6 –83.82 5.41 178.09 –67.37 8.71

Ethylene C2H4 52.50 4.13 155.02 –81.55 16.98

Hydrogen H2 0 28.61 1.02 –0.15 0.77

Nitrogen N2 0 29.59 –5.14 13.18 –4.97

Carbon monoxide CO –110.53 29.12 0.25 –2.07 5.90

Ethyl benzene C8H10 29.92 –36.72 671.12 –422.02 101.15

Diethyl benzene C10H14 22.27 –37.41 867.09 –556.0 141.10

Calculate: (a) Required temperature of the offgas mixture, entering the first bed.
(b) Benzene fed (in % of total) to the second bed as quench, and (c)
Temperature of product gas stream, leaving third bed.
Note: Assume that enthalpy of formation and heat capacity data of 1 bar are valid at 12 bar a.

REFERENCES
1. Bhatt, B. I., and S. B. Thakore, Stoichiometry, 5th Ed., McGraw-Hill
Education, New Delhi, 2010.
2. Bhatt, B. I., Design Databook; Properties of Steam, Selected Refrigrants, n-
Hexane and Brines, CBS Publishers & Distributors, New Delhi, 2006.
3. Himmelblaue, D. M., Basic Principles and Calculations in Chemical
Engineering, 5th Ed., Prentice-Hall of India Pvt. Ltd., New Delhi, 1994.
4. Smith, J. M., H. C. Van Ness, M. M. Abbott, and B. I. Bhatt, Introduction to
Chemical Engineering Thermodynamics, 7th Ed., McGraw-Hill Education,
New Delhi, 2010.
5. International Critical Tables, Vol. 3, McGraw-Hill, USA, 1928, p. 389.
6. Chopey, N. P., Handbook of Chemical Engineering Calculations, 2nd Ed.,
McGraw-Hill, Inc., USA, 1994.
7. Kapoor, K. L., Numerical Problems in Chemistry, McGraw-Hill Education,
New Delhi, 1991.
8. Reklaitis, G. V., and D. R. Schneider, Introduction to Material and Energy
Balances, John Wiley & Sons, USA, 1983.
9. Hougen, R. A., K. M. Watson, and R. A. Ragatz, Chemical Process
Principles, Part–I, CBS Publishers & Distributors, New Delhi, 1995.
10. Fedler, R. M., and R. W. Rousseau, Elementary Principles of Chemical
Processes, 2nd Ed., John Wiley & Sons, USA, 1986.
*Am3 means volume of gas at actual flowing conditions.
 3
Process Engineering
3.1 INTRODUCTION
A process is simply a method by which products can be manufactured from the
raw materials. Selection of process steps for the manufacture of the product to
the required specifications is the job of a process engineer. There could be
number of methods to manufacture the same product. The best method for
manufacturing a product has neither been discovered nor can it be said that a job
can be performed in only one way. The process engineer faces alternative
solutions to manufacturing problems. This possibility comes out of a
combination of factors such as information gained by him from the past
processes, imagination, his own knowledge of known processes, etc. However,
economy of manufacture is a factor that frequently decides the best route to be
followed. Thus, it is safe to say that innovation, imagination and economy are
the key factors in process engineering.

3.2 EXAMPLES OF ALTERNATIVE ROUTES IN

PROCESS ENGINEERING
Process engineering and design can be explained best with the following
illustrations:
1. Consider treatment of water for various end uses. End use of treated water is a
chief governing factor in deciding the process steps to be selected. Production
of drinking water demands different steps as compared to production of high
quality water for boiler feed (BF) make-up purpose.
Drinking water requirements can be met by different water treatment steps. If
water contains less dissolved solids, say less than 500 mg/L, clarification and
disinfection by chlorination could be satisfactory. However, if raw water has
high dissolved solids, dual filtration followed by desalination by reverse
osmosis and pH adjustment would be necessary. These are relatively simple
process steps.
In case of BF make-up water requirements for high pressure boilers, number of
routes of the process are possible. Clarification, activated carbon treatment
and demineralization by ion exchange is one route. However, this route may
not be economical if raw water has high dissolved solids (say more than 1500
 mg/L). In the latter case pre-treatment, reverse osmosis treatment, followed
by ion exchange treatment would become necessary. The final process route
to be followed will depend on raw water quality, operating pressure of the
boiler, effluent disposal problem, etc.
2. In another case, consider the classic example of production of anhydrous
(100%) ethanol. Ethanol produced by fermentation of molasses can be first
distilled to produce azeotropic mixture (containing 95.6% alcohol by mass)
which can be further dehydrated to 100% by use of an entrainer such as
benzene in azeotropic distillation. New development of pervaporation process
can be used in place of azeotropic distillation. Similarly, drying of azeotropic
mixture by using molecular sieve beds can be also employed (Refer to Table
8.65). Innovations thus have a strong impact on process design alternatives.
3. Partially hydrogenated edible oils are extensively used for variety of
applications. It is normally manufactured by hydrogenation of an edible oil in
presence of a nickel catalyst under pressure (say at 2 bar g) and at 130 to
160°C in an autoclave. In a batch process, the reactor can have an anchor
agitator with convolute at the top of the liquid level for proper mixing of
hydrogen with the oil. This age old technology is replaced by installing a gas
induced agitator (Refer to Exercise 10.2) which can result in a better product
in a shorter period. This innovation has several advantages over the
conventional design. Further research has demonstrated that use of carbon
dioxide along with hydrogen at high pressure results in a product with
stringent specifications (e.g., with less trans fatty acids).
4. Modern ammonia plants are large size plants (1800 t/d or more) which are
continuous and reliable. These are quite compact than those originally
developed to operate at very high pressures. Also, newer plants are safe and
highly energy efficient, proving the economy of the scale.
Several examples can be cited where technological developments have totally
changed the manufacturing methods. This can be attributed to newer
technologies, efficient equipment designs, use of alternate raw materials, etc.
Process engineer’s job is, therefore, to evaluate various alternatives available to
him and decide over the best method to be adopted. Batch type hydrogenation
plants are still preferred over continuous one due to many reasons such as
relatively small capacities, different feedstocks for different batches, safety in
handling, etc. Design of ammonia plants are currently more guided by available
raw material and specific energy consumption. Thus it cannot be said that the
best method for manufacturing a product can be performed in only one way. The
possibility of alternate process operations comes as combination of information
gained from the past, imagination, and the process engineer’s knowledge of the
known processes.

3.3 GENERALIZED APPROACH TO THE

CHEMICAL PLANT DESIGN
Design of chemical plant involves chemical technology, unit operations,
instrumentation/controls, use of specialized equipments, safety considerations,
materials of construction, etc. Chemical technology is the branch of chemical
engineering dealing with chemical changes and physical methods involved in
industrial process. It is necessary to supplement the knowledge of unit operations
with considerations of chemical technology in order to obtain a concept of the
complete plant.
From the above discussions, it is clear that before attempting process
engineering design it is necessary to fix the product specifications. Capacity of
the plant can then be decided based on demand and/or investment. Evaluation of
available technologies will have a say on investment. In order to arrive at
consumption figures of raw materials and energy, preliminary stoichiometric
calculations would be necessary. This would give some rough idea of cost of
production due to inputs which normally accounts for higher proportion in the
cost.
Following sequence might be followed in implementation of the project right
from conception of the idea to construction and operation of the plant:
1. Process Research 4. Process Development
2. Preliminary Process Design 5. Final Process Design
3. Preliminary Process Evaluation 6. Final Process Evaluation
3.3.1 Process Research and Pilot Studies
Chemically feasible process is fundamentally characterized by process research.
It is an accepted fact that process research is a basic necessity of the process
industry. Orderly research and development has become an integral part of all
leading industrial organizations and is synonymous with the progress. Basic
operating conditions are explored and sufficient information is obtained from
process research for preparing a preliminary material and energy balance flow
sheet.
In many instances, development of a pilot plant is a necessity, particularly where
laboratory scale data are not sufficient to scale-up or safety aspects are not well
known. It is better to commit blunders on a small scale and make profits on a
large scale. Such pilot plants can result in eliminating production of the off-spec
product, good yields, reduced cost of manufacturing, etc., by fixing best
operating conditions, installation of corrosion resistant plants, deployment of
best equipment, reduced recycles due to better yields and so on.
Need for a pilot plant is justified in any manufacturing (Refs 1, 2, 3 and 4). Take
for example, manufacture of bulk drugs. Manufacture of antibiotics involve
reactions (of exothermic nature) at very low temperatures; at times at cryogenic
temperatures. Use of ice or dry ice suffice at laboratory scale but at production
level it is necessary to deploy industrial refrigeration which can be simulated in a
pilot reactor. Design of a reactor changes significantly with industrial
refrigeration in which possibly use of liquid nitrogen may be necessary.
Similarly, solvent to reactant ratio used at laboratory scale may not be
economically viable at the plant scale and it may be necessary to reduce the ratio
to an optimum by experiments on a pilot scale.
As seen in the earlier section, hydrogenation of edible oil is carried out in a
slurry comprising a catalyst. Selection of agitator is crucial. Pilot studies are
valuable in the development of an agitator ideally suited for the reaction.
Similarly, ozonation is a dangerous reaction and is very much dependent on
intensive diffusion of gas in the liquid. Use of a static mixer with liquid
circulation and simultaneous injection of ozone in it could prove to be
advantageous. However, such designs should be supported by pilot studies for its
efficacy and control of reaction. Design of a seal for agitator can become an
important design factor when dealing with high pressure gas reactions or with
dangerous gases such as hydrogen, chlorine, ozone, etc. Hydrogenation of
glucose to sorbitol is carried out in presence of slurry catalyst at quite high
pressure (> 50 bar). Design of an agitator for such an autoclave and its sealing
arrangement are challenges for the design engineers. Viscosity of reactant mass,
operating pressure and temperature, suspended solids concentration,
heterogeneous reaction mixture, etc., can have a strong bearing on reactor design
and process research with a pilot plant could therefore be fully justified.
Pilot scale studies may be warranted for other unit operations. Its requirement
can be decided based on the data available in literature or from the lab scale
experiments. For example, if sufficient vapour–liquid equilibrium data are
available for separation of a liquid mixture, a distillation equipment can be
designed with a degree of certainty and pilot study may not be required.
Similarly, mutual solubility data, if available, extraction equipment design can be
carried out with known principles of unit operations. In the event of a specific
design of the equipment, pilot studies are recommended. For example, for
carrying out high vacuum distillation of a heat sensitive product, specialized
equipment such as short path distillation or a wiped film evapourator is
recommended. Manufacturers of such equipments provide pilot study on a small
scale equipment with reasonable scale up confidence.
Pilot scale study can provide more reliable flow sheet than the one obtained from
lab scale study. Versatile equipments which can be procured off the shelf are
available for pilot studies. Alternately process engineers can design a tailor made
pilot scale equipment for the study.
Pilot plant can be used for ongoing improvements of the existing process or
development of a next generation process. One of the objectives of such a study
is substitution of a raw material. Improvement in product quality, reduction in
process costs, safety and minimizing the pollution effects are other objectives of
the pilot study.
Introduction of a new product adds a dimension of risk relative to the existing
product. Manufacturing a small volume of the new product in a pilot plant can
be valuable in supplying it for commercial testing to a customer whose
evaluation is crucial before introducing the product on a commercial scale. In
many instances, when the customer demands some product specifications
change, it could be achieved by process research on a pilot plant.
It can, therefore, be said that a pilot plant is an intermediate sized plant,
modeling the process of interest, constructed and operated so as to test the model
in a convenient, less expensive and safe manner.
Results obtained in the pilot plant builds confidence in management and help
them in taking commercial, technical and financial decisions, thereby focusing
their attention and allocating the resources accordingly.
For the design of pilot plants, useful guidelines can be obtained from the book
by Jorden4. Several examples are drawn from industry in the book to develop the
tailor-made pilot plants.
Cost of a pilot plant could be significant but risks are relatively low and payoffs
could be substantial.
Another important consideration in pilot plant operation is the selection of
manpower. A project leader (a chemical engineer), familiar with the chemical
process and the equipments, is chosen for the job. He should be assisted by one
or more chemists, depending on the size and complexicity of the pilot plant.
Careful observations during the operation and interpretation of data are the skills
desired from the personnel. Effective communication among the staff is crucial.
Cost of such talented staff would be high but is rewarding. Motivation to the
staff enhances self satisfaction and increases dedication towards the work.
Logging of data can range from visual examinations to instrument readings.
Relatively inexpensive data acquisition system based on electronic instruments
is available which can be installed and connected to a computer to record the
data, generate trend analysis, know critical events, find temperature or pressure
spikes, etc.
Instrumentation needs for the pilot plant depends on its complexity. A reasonably
straight forward process may not justify complex instrumentation/control system
but a high pressure and hazardous pilot plant will call for an expensive
instrumentation package. Take, for example, a pilot plant for the manufacture of
alkyd resins or antibiotics require normal instrumentation. However pilot plants
for hydrogenation, ozonation, supercritical extraction with carbon dioxide,
carbonylation with carbon monoxide, etc., may call for full automatic operation
with data logging systems.
Pilot plant can be used as a training tool to plant operating personnel. It is well
known that aeroplane pilots are trained on a simulator. Pilot plant can be
considered as a simple version of ‘simulator’ of the main plant.
Project leader will be able to prepare a preliminary operating manual of the main
plant based on his study of the pilot plant. Pre-startup testing, startup, normal
operation and shutdown, emergency shutdowns, etc., can be well visualized by
the project leader. The manual can be subsequently updated by the plant
personnel based on their operating experience.
A process flow diagram can be offered by a team leader for further process
design based on pilot plant studies.
3.3.2 Process Design and Development
After carrying out process research and pilot plant study, the process details are
now available to the process engineer which can be scaled up with sufficient
confidence level to a commercial plant. Scale-up from a pilot plant to
manufacturing scale plant cannot be simply carried out by geometric scaling.
Use of dimensionless numbers is recommended in scale-up for process designs.
Zlokarnik5 has given large number of interesting design problems of stirrers,
absorber, solids processing, etc., in his book on scale-up. The book primarily
deals with the scale-up methods involving geometric similarity principles and
use of various dimensional numbers. In addition, it also describes the selection
of process parameters that are to be critically studied during the experiments,
decision on size of model experiment, model to be adopted for scale-up, etc. The
book will generate a confidence to a process engineer in his design. Immediately
the process engineer is not concerned with detailed design of pieces of
equipment but is concerned with the overall process covering following major
items:
1. Preparation of a block diagram for presentation to the management for
preliminary feasibility of the project.
2. Carryout overall and individual material balances showing the yield and
quality of all materials to and from all pieces of equipment.
3. Heat balances over each piece of equipment.
4. Prepare a Process Flow Diagram (PFD) based on data in Steps (2) and (3).
5. Establish flow quantities and suction and discharge conditions of all rotating
equipments.
6. Size reactors and storage tanks.
7. Separation process calculations with optimum operating conditions to effect
the desired separation.
8. Estimate utility requirements such as steam, cooling water, refrigeration,
electricity, thermic fluid, fuel, etc.
9. Prepare a Piping and Instrumentation Diagram (P&ID) so that requirement of
instrumentation/control system is assessed.
10. Develop a tentative layout of the plant that should take into considerations of
safety requirements.
11. Estimate preliminary capital investment and production cost.
Preliminary process design is necessary to find economic viability of the project
and focus on important practical aspects. Take for example carbonylation of
ethanol to propionic acid. This is high pressure and high temperature catalytic
reaction of ethanol and carbon monoxide. Toxic nature of carbon monoxide
dictates prominently in deciding over the safety requirements and practical
operational aspects over the economic viability of the project.
Preliminary design will show the preferred process and whether further research
and development (R&D) work is necessary. This additional R&D efforts will be
specifically targetted to obtain missing information. Many a times physical
properties of intermediates, corrosion data, heat of reaction, etc., are not
available in literature and may have to be obtained by R&D efforts.
Process engineering calculations do not end with stoichiometric calculations.
Process design of equipments for unit operations and unit processes require
detailed literature survey to select best equipment suited for the service. While
design and selection methods for many such equipments are discussed in
Chapters 5 to 11, property determination of pure fluids and their mixtures is a
task by itself. It is recommended that serious attempts be made to find properties
of pure fluids that are reported in literature. Only when measured values could
not be found in literature, estimation methods may be considered. Reed et al7 has
given extensive treatment to the subject of estimation of properties. For example,
for pure fluids, group contribution methods are developed for a number of
property parameters. Take a case of heat of formation data. While accurate data
are published for large number of compounds by NIST, AIChE, etc., there are
many compounds for which data are not available. Group contribution data by
Reed et al. are quite useful for such compounds. Properties of mixture of fluids
(such as density, viscosity, surface tension, etc.) are needed in most process
design calculations. While some data are available for specific mixtures in
literature, estimation of these properties by various correlations is carried out.
Process engineer must ensure that correlations, selected for calculations of the
mixture properties, are reliable. In chapters to follow on process design of unit
operations and reactors in the book, number of such correlations are used for
calculation of properties of mixture of fluids.
Duration of completion of project can be predicted based on the preliminary
design. Based on this information, marketing strategy of the product can be
developed.
Development of PFD and P&ID are dealt in detail in Chapter 4 with specific
examples.
After completion of the preliminary process design and its evaluation, detailed
process design work is taken up. For the detailed design, variety of ‘tools’ are
used in calculations. Estimation of properties by group contribution methods,
literature search for different types of reactors, correlations for heat transfer,
pressure drop and mass transfer calculations, mixing, etc., are the tools available
to a process engineer. Many a times empirical sizing of equipment is carried out
based on experience or literature survey. Numerous process design examples are
worked out in Chapters 5 to 10 relating to fluid flow, heat transfer, separation
processes and reaction kinetics.
Instrument engineer is simultaneously consulted in selecting the final control
strategy. Suggestions/comments of the project leader of the pilot plant are quite
valuable on the preliminary PFD and are incorporated as far as possible.
On completion of all the above steps, final PFD and P&ID are prepared by the
process engineer and handed over to the project manager with equipment
datasheets (containing process specifications) for further execution of the
project.
During project execution stage also, role of the process engineer cannot be
undermined. He will have an advisory role to ensure that process will function
smoothly as anticipated during design stage.
3.3.3 Batch vs Continuous Operations
When designing processes or facilities for some applications, process engineer
must approach the task differently than he would for a project in other chemical
process industries. Bulk drugs, pharmaceutical intermediates, dyes
intermediates, fine chemicals, etc., are the products in which batch operations
dominate. Take the pharmaceutical industry sector. In this sector, lots of
manufactured product are required to be segregated. In this way waste is
minimized and regulations could be easily conformed to. Secondly, most
pharmaceutical processes deal with relatively smaller amounts of material than
with typical chemical processing operations. For this reason, batch processing
equipment is often the preferred choice. Many a times, an equivalent continuous
based unit operation that is small enough to accommodate the volume of
material to be processed is not available. Further, most pharmaceutical processes
come out of the lab where operations are predominately in batches. These batch-
type processing techniques migrate to the larger production scale.
Although one would not propose changing batch processes to continuous process
for small volumes, many operations could be made continuous. A continuous
unit operation can be used to process a tank, full of material. For example, a
continuous filtration method, such as a rotary drum filter can be used instead of
batch method such as a plate and frame filter.
Another example of making continuous unit operations work involves two-phase
extraction process. The traditional batch approach is to use a simple mixer-settler
that mixes two phases together in a tank before they are allowed to separate. In a
continuous analog, a small extraction column could be used in which two phases
flow past one another continuously to facilitate phase contact. This continuous
operation can be performed in a much smaller equipment. It also provides
numerous ‘stages’ of phase contact resulting in more efficient mass transfer
operation. Similar examples can be cited for distillation, fermentation,
precipitation, crystallization, drying, etc.
Conversion from batch processing to continuous processes would thus allow
smaller equipment with lower capital cost. Continuous equipment could be
designed to be enclosed and cleaned easily, allowing the use of less costly
production space. Many of these continuous equipment would operate at ‘steady
state’ and it would be easy to automate them, reducing labour cost.
Batch operations and the equipment that facilitate them are by no means
obsolete. Continuous equipment by its nature must be designed around a
standard process that will be repeated for a significant period. For products that
are made infrequently it is impractical to design continuous operations to support
them. Processes where the input material varies over time also exist. These
applications require an adjustment to the processing techniques to accommodate
new variables. In many such cases only batch operations are flexible enough to
handle these situations. Further some processes handle materials that are so
valuable or hazardous that the transients that might occur in a continuous process
are unacceptable. In these cases, only batch processes offer the control and
containment, required to make them successful.
Continuous extraction of oil from oil-rich cake with n-hexane is practiced in
industries on a large scale. However, when residual oil is to be extracted from
spent bleaching earth before its disposal, a batch operation is normally
performed as its quantity is small.
Bleaching of edible and non-edible oils was performed traditionally in batch
equipment. Development of continuous bleachers give better quality of bleached
oil with less bleaching earth/activated carbon and eliminate need for
intermediate storage of bleached oil before its passage to continuous deodorizer.
Also, the size of a continuous bleacher is much smaller than a batch bleacher.
Many a times, solvent to reactants ratio is kept high (may be in excess of 15) in a
batch reactor based on lab studies. Many of such reactions could be relatively
fast and endothermic. At the end of the reaction, solvent is required to be
distilled out. Such a process lends itself for conversion to continuous operation.
A mixed stream of solvent and reactants (possibly with much lower ratio) can be
heated in a standard shell and tube heat exchanger, passed through a mixing
equipment (such as a static mixer) and sent to a reactive distillation column
where sufficient reaction time can be provided and simultaneously solvent is
distilled out. Such a conversion can result in substantial energy savings apart
from reduced time cycle of production.
Large capacity chemical plants are continuous in nature. Manufacture of
ammonia, methanol, refining of petroleum products, acetic acid, carbon black,
formaldehyde and many others are the examples where continuous plants are
designed and operated. Consistent quality, optimum energy input, lower cost of
installation of plant per unit (say kg) of product, etc., are inherent benefits of
continuous plants.
The process engineer must evaluate both the operations before deciding the best
operation, suited for the product. As explained in previous paragraphs, batch
operations should be restricted to a minimum so that economy of scale, a feature
of continuous operations, can be availed to maximum degree.
Reference 6 deals exclusively on batch process design.

3.4 PREPARATION OF OPERATING

INSTRUCTIONS MANUAL
One of the less appreciated tasks of a process engineer is preparation of an
Operating Instructions Manual (also known as Operation Manual) — OIM of the
plant. When a calculator is purchased, a leaflet is provided along with the
calculator which describes its functioning. In a similar fashion, OIM is prepared
for safe, smooth and efficient operation of the plant. Well-prepared OIM serves
as a tutor to the operating personnel. It is also a valuable reference tool in day-to-
day operation.
Safe operation of the plant is the prime objective of preparation of the manual.
Therefore, safe operating limits are prescribed clearly for each equipment and
process. Dangers, if any, of crossing the critical parameters must be underlined.
Start-up and shut-down procedures are outlined in the manual. These include
purging, slow pressure build-up, well-defined heating procedure, etc.
Normal operating procedures are covered for achieving optimum performance of
the plant. These procedures cover desired operating parameters’ range,
generation of minimum effluent quantity, optimum utility consumptions, etc.
Emergency handling procedures are well covered in the manual. Power failure,
leakages, utility failure, fire, etc., are the emergencies which must be well
thought of and a plan of each emergency must be described.
PFDs and P&IDs are included in OIM. Wherever required, additional sketches
are included to clear the operational aspects. Calibration charts of tanks, reactors,
dosing systems, etc., are included in OIM. For example, capacity of a screw
feeder with its speed is an essential requirement of an operator.
Process datasheets and manufacturers’ datasheets are normally compiled as
Equipment Datasheets and it is a separate manual by itself. Instrument
specifications are also included in this manual.
Useful guidelines are drawn from the pilot plant operation for the preparation of
OIM. Equipment and instrument supplier catalogues provide their own operating
instructions. OIM should cover comprehensively operating instructions of the
manufacturers.
In many instances, standby pumps, filters, etc., are provided. Periodic operation
of standby equipments is to be included in OIM. Similarly, periodic calibration
of instruments is needed to ensure that they are functioning in order. For
example, in case of pressure gauges, zero error is checked periodically. In case
of temperature measuring instruments, calibration with a standard instrument is
recommended periodically.
Preparation of OIM is normally undertaken by the process engineer during
project execution. Draft copy of OIM is circulated to operating personnel and
project leader of pilot plant. Suggestions received from them should be critically
reviewed and if found useful should be included in the final OIM.
Over a period of time, OIM needs upgradation as more experience is gained in
actual operation. Also the changes in OIM should cover the modifications that
are carried out over the years. Thus upgradation is a never-ending process.
Preparation of log sheets for manual logging of data is normally the job of
operating personnel. Nevertheless, the process engineer can review the draft of
the log sheets and may offer suggestions. For medium scale and small scale
plants, these are also known as batch sheets.
The process engineer should study the log sheets periodically and offer
suggestions to the operating personnel. If necessary, OIM may require
upgradation on account of review of the log sheets.

3.5 CONDUCTANCE OF PERFORMANCE TEST

RUNS
On installation of the plant, process engineer is required to conduct a
performance test of an equipment or entire plant. When a centrifugal pump is
commissioned, its capacity, head and power consumption are checked with the
datasheet. When a water treatment plant; say a reverse osmosis plant, is
commissioned, its performance is checked for permeate quality (vis-a-vis feed
water quality), permeate/feed ratio and power consumption. In the same manner,
performance of cooling tower is checked for achieving anticipated cooling water
temperature for the prevalent wet bulb temperature of ambient air and for its
power consumption.
In large chemical plants, performance guarantees are well specified in the
contract. These include product quality specifications, raw materials
consumptions and utility consumptions. All these parameters are averaged over a
specified continuous run of the plant, say for 120 hours or so. Average values are
compared with those guaranteed.
Conductance of performance test run requires specialized skills. It is carried out
by the process engineer in coordination with the plant personnel and contractor
personnel. Before starting the test run, instruments are calibrated. Any specific
deviation from normality is documented in the log sheet.
Test run report comprises log sheets, observation sheets, analytical reports and
summary of findings. It is used as Acceptance Report for the equipment or the
plant.

3.6 TROUBLESHOOTING
Troubleshooting is yet another task of the process engineer. During
commissioning or normal operation, many operational problems are
experienced. Some of them are related to the process. Process engineer is
required to study the problem, discuss with operating personnel and present a
solution.
Take an example in which cooling water cools a fluid. If cooling is less effective
than design, reasons for the same are to be determined. Less flow of cooling
water or high supply cooling water temperature or inadequate heat transfer
surface could be the problem for a new system. Material and energy balance
across the cooler will reveal the problem and corrective action can be taken.
During the course of operation, if the cooler performance deteriorates, fouling of
cooling surface could be anticipated.
Take another example of a separator in which organic liquid is separated from
water. While studying the performance of the separator, check moisture content
of the organic layer and chemical oxygen demand (COD) of the water layer. If
the results are not satisfactory in terms of separation, check operating
temperature of the separator or its design.
In a distillation column, if the distillate quality is not maintained, one can check
the reflux ratio and other parameters.

3.7 COST CUTTING MEASURES

In a normal running plant, process engineer has to consider cost cutting
measures in the plant. This is a never ending exercise as there exists an
opportunity for improvement in each plant. For finding the opportunity, indepth
study of the plant is recommended.
Consider a chilling plant in which brine at –10°C is produced using refrigerant
R-22. It is a known fact that in this temperature range Coefficient of
Performance (COP) of ammonia refrigeration cycle is better than that of R-22
cycle. Thus replacement of the refrigeration skid of R-22 system by ammonia
system will consume less power. However, fresh investment for the ammonia
refrigeration system will be needed. A cost benefit analysis will have to be
carried out by process engineer.
Take another case of a plant in which hot oil at 150–175°C is used to heat
reaction mass and carry out the reaction at around 105°C. Alternative system
with saturated steam at 3.5 bar g could be equally or more effective. Both
economical and practical aspects need be examined before any change in the
plant is considered.
In large chemical plants such as ammonia plants many opportunities exist for
cost reduction. With methane and ambient air as feedstocks at 25°C, theoretical
energy consumption is 20.618 GJ/t ammonia. Actual ammonia plants consume
energy equivalent to 27 to 42 GJ/t ammonia, depending on the feedstock and
technology. This indicates a potential for reduction in energy consumption and
thereby in cost of production. Similar opportunities can be found in each plant.
Newer technologies and innovations can contribute significantly to cost
reduction. Process engineering plays a vital role in their implementation.

3.8 PREPARATION OF DATA SHEETS OF

EQUIPMENTS
Ordering of equipments starts as soon as the process design of equipments is
completed. Inquiries are sent to prospective vendors for getting their technical
and commercial offers. For this reason, it is necessary that adequate technical
data are furnished to the vendors in the technical data sheets. Most engineering
firms have their own standard format for data sheets which are filled by process
engineers. This important task is not to be undermined as incomplete data may
invite incomplete or at times wrong offer from vendors. As an example,
specifications of a heat exchanger can be filled in TEMA data sheet. Figures
10.7 and 10.8 are sample TEMA data sheets. It will be easy for vendors to
submit their correct technical offer based on such data sheets.

3.9 GREEN ENGINEERING

A set of 12 guiding principles of ‘green chemistry’ were established by Dr. P.
Anastas and J. C. Warner in 1998 for creating safer chemicals and designing for
degradation to save the environment. Further to these guidelines, a
multidisciplinary group of engineering disciplines met in May 2003 in Florida,
USA to draft the principle of Green Engineering. The conference focused on the
definition of the scope and tone of green engineering relative to all engineering
disciplines; including chemical, mechanical, civil, electrical and environmental.
The objective of green engineering is to transform existing engineering
disciplines and practices to those that promote sustainability. It incorporates the
development and implementation of technologically and economically viable
products, processes and systems to promote human welfare while protecting
human health and elevating the protection of the biosphere as a criterion in
engineering solutions. To implement these solutions, engineers are advised to
use the following principles8:
1. Engineer processes and products holistically, use systems analysis and
integrate environmental impact assessment tools.
2. Conserve and improve natural ecosystems while protecting human health and
well-being.
3. Use life-cycle thinking in all engineering activities.
4. Ensure that all material and energy inputs and outputs are as inherently safe
and benign as possible.
5. Minimize depletion of natural resources.
6. Strive to prevent waste.
7. Develop and apply engineering solutions while being cognizant of local
geography, aspirations and cultures.
8. Create engineering solutions beyond current or dominant technologies;
improve, innovate and invent (technologies) to achieve sustainability.
9. Actively engage communities and stakeholders in development of engineering
solutions.
Process engineer can play a vital role in achieving the objectives of green
engineering. At every stage of process engineering, the engineer must consider
and review whether the design adheres to the principles of green engineering. A
few examples will demonstrate implementation of principles of green
engineering in practice.
1. In solvent extraction plants, n-hexane is used to extract oil from oil-rich cake.
It is known that n-hexane is not totally safe for human beings. New
supercritical extraction process based on carbon dioxide as an extracting
agent is found to be eco-friendly, safe for human beings and also gives good
quality oil.
Water can also be used as supercritical fluid to produce biocrude from organic
waste. In this application, water is a safe extracting agent while it helps in
safe ‘disposal’ of organic (including hazardous) waste.
2. Hydrogenated edible oils are used extensively in bakery and other food
products. Presence of trans-fatty acids in the hydrogenated fat is known to be
unfit for human consumption. Technological developments should be
targetted to produce the fat with low trans-fatty acids. One such development
is hydrogenation in presence of carbon dioxide at low temperature. Another
development relates to the use of noble metal catalyst (instead of nickel) for
hydrogenation.
3. Methyl tertiary butyl ether (MTBE) is known an oxygenating product for
improvement of octane number of the gasoline. Research has revealed the
health hazards of MTBE when autovehicles emit gases to atmosphere.
Addition of anhydrous ethanol (a renewable product) is said to improve
octane number without creating health hazards.
4. Trans esterification of fatty acids with methanol can produce biodiesel which
is considered an eco-friendly fuel for DG sets or for autovehicles. Fatty acids
can be obtained from edible and non-edible oils which are renewable
resources.
5. Ozone depleting substances are threats to human beings. Montreal protocol,
1987 bans the use of certain chlorofluoro hydrocarbons which are established
to have very high ozone depleting potential, apart from their low
degradability. Classical example is dichlorodifluoromethane (CFC-12) which
is banned for use as a refrigerant, aerosole agent, etc. Engineers are advised to
use ammonia (R-717) or chlorodifluoromethane (CFC-22) in place of CFC-
12. Eventually CFC-22 is also to be phased out.
6. In a classical Bechamp process, nitro-compound is converted to a
corresponding amine by using iron power and water mixture. Reaction takes
long time. At the end of the reaction, total mass is required to be filtered out
of the iron sludge (fine in nature). Later the mass is separated in two layers.
Water, so separated, has high Chemical Oxygen Demand (COD) and pose an
effluent problem.
In new developed processes, hydrogen is used in presence of a catalyst under
pressure. Use of gas-induced agitator, nickel or noble metal catalysts, etc., are
the developments which result in reduced time of reaction and also do not
generate effluents. Thus, it is an eco-friendly process.
7. While methanol is used as a solvent in many chemical plants, it is toxic to
 human beings. Its use is considered unsafe in pharmaceutical and food
processing. Its higher homolog ethyl alcohol is safe for use in the industry as
a solvent.

3.10 PROCESS INTENSIFICATION

Process intensification can be defined as improving productivity of a system by
changing operational parameters and/or by carrying out desirable modifications.
In a process plant, improvement in productivity means product quality
improvement, yield/production rate increase, reduction in energy consumption,
reduction in other inputs (such as raw materials), safer operation, or reduction in
pollution load of the plant. Process intensification differs from troubleshooting
(Sec. 3.6) as the latter mainly focuses on solving the problems, faced by
operating personnel. Cost cutting measures (Sec. 3.7) can be considered as one
of the objectives of process intensification.
There is a caveat to the definition of process intensification. Development of a
new chemical route or a change in composition of a catalyst, however, do not
qualify as process intensification.
Process engineer can contribute significantly in process intensification; either at
the process design stage or in the running plant. Scientific approach to the design
and critical analysis and review of the operating data are the keys to process
intensification.
At the process design stage, a number of options are available to the process
engineer. Consider design of a multiple-effect evaporator system. Apart from the
process parameters; such as concentration of ingoing/outcoming solution,
heating medium and its conditions (e.g., pressure and temperature), etc., process
engineer can design the evaporators with equal or unequal heat transfer area
(HTA) in the stages. Optimization techniques such as Evolutionary Operation
(EVOP) or dynamic programming can be used to find HTA for both the
alternatives, thereby cost of the installed systems for both the alternatives can be
worked out. Optimum design of the system can then be selected.
A cascading type steam balance offers an interesting example of utility system
optimization as a part of process intensification. Modern heavy chemical plants
such as ammonia plant, methanol plant, ethylene manufacturing, petroleum
refinery, etc., have cascading steam systems which can be optimized to consume
minimum fuel in the boiler and also reduce load on cooling tower.
In a simpler case, selection of an agitator in a reactor can result in substantial
energy saving-with improved agitation.
Linear programming is another tool which helps in optimizing the proportions of
components in a reaction mixture. Consider manufacture of alkyd resins by
reaction of polyols (such as glycerol) with fatty acids by azeotropic distillation
method. Fatty acids could be either distilled (pure) fatty acid or in the form of an
oil. Similarly, a variety of alcohols such as glycerol, sorbitol, etc., are used.
Theoretical calculations can be made with the help of linear programming to
achieve desired properties of the product with best possible yield which can later
be confirmed in laboratory trials.
Innovations and adoption of newer technologies are integral to process
intensification. Adoption of a standard operating practice in one industry into
another industry should be considered for a similar situation. A fluid bed cooler
is used in large urea plant to produce urea of uniform (narrow range particle size
distribution) size prills at low temperature so that urea prills do not agglomerate
to lumps on storage in a silo which produce dust on breaking them. Similar
problems are also experienced in a carbon black plant. Adoption of the fluid bed
cooler should be considered in a carbon black plant.
Absorption of hydrogen chloride in water is highly exothermic and is efficiently
carried out in a Falling Film Absorber, made from graphite blocks. Such an
absorption system is found more effective than a packed bed absorber. Such type
of an absorber can well be considered for absorption cum reaction of sulphur
trioxide (gas) in organic compounds such as linear alkyl benzene, methyl ester of
saturated fatty acids, etc.
Use of a static mixer in mixing an organic compound with a mixed acid (HNO3
+ H2SO4) to form nitro compound is found to be very effective in controlling
side reactions with efficient temperature control in a batch reactor.
Several examples can thus be cited to demonstrate that process intensification at
the design stage or in an operating plant can result in a plant with lower
operating cost, possibly lower initial cost, ease of operation and safer operation.
In an operating plant, critical review of operational parameters can result in
substantial savings. Use of statistical methods are invaluable in this exercise. In
fact, ISO 9001 gives stress on statistical methods for improvement in quality
control which is advised as a continuous improvement process. An article by
Bader9 gives a systematic method for quality assurance by statistical means.
In a similar manner, it is advised that variation in process parameters and its
effect on plant performance be studied which can reflect the best and safer
operating conditions.
Modifications in the plant can be aimed at reduction in energy consumption or
reduction in raw materials consumptions. In ammonia plants, purge gas from
synthesis loop was being vented or used a fuel in the furnace in many grass root
plants. Adoption of purge gas recovery plant to recover and recycle hydrogen
from the purge gas could result in increase in production by about 5% and
simultaneously reduce energy consumption of the entire production by about
2%.
In a bulk drug plant, some reactions are carried out at very low temperatures
(around –80°C). Injection of liquid nitrogen is quite common for achieving such
low temperature in the industry. This method is rather inefficient and expensive.
Instead, a cascading type mechanical refrigeration system with R-23 and R-
22/R-717 in tandem can be considered for adoption to achieve the same
temperature at much lower investment and operating cost. Also the space
requirement is less for the refrigeration system. Adoption of cascading
refrigeration system may bring down the refrigeration cost by about 50% as
compared to liquid nitrogen injection system for a given plant.
Use of aqueous calcium chloride brine was common in refrigeration. However,
numerous operating problems were experienced with it such as slimy spills,
chocking of pipelines due to freezing, difficulty in maintaining brine
concentration, etc. Aqueous monoethylene glycol (MEG) is found safe and
convenient substitute to aqueous calcium chloride brine, thereby design of brine
circulation system is simplified. Use of aqueous methanol brine is made for low
temperature applications (–20 to –50°C) but the brine solution is flammable in
nature and requires specific care in design. Thus adoption of new systems is a
part of process intensification.
In large chemical plants, the network of heat exchangers can be reworked in line
with pinch technology (Sec. 6.19). Such an exercise can result in substantial
energy savings.

3.11 CONCLUSION
Process engineering is the heart of any process design. While adequate attention
at the design stage can pay rich dividends, adoption of process engineering
principles in a running plant is always rewarding.

EXERCISES
3.1 Give brief answers of the following:
(a) Name ten biofuels that are used for steam generation.
(b) Differentiate petroleum based lube oils and eco-friendly lube oils.
(c) An innovative hydrogenation process for edible oil is to be developed which
is based on supercritical principle. Carbon dioxide and propane are
considered as alternative supercritical fluids (SCFs) for the purpose. Study
the advantages and disadvantages of each for the purpose.
(d) Name five ‘green’ solvents for use in chemical synthesis.
(e) An aqueous iso–propyl alcohol mixture, containing 50% alcohol (by mass),
is to be separated to a stream of 100% alcohol. Suggest process steps with
reasonings.
(f) Nitrogen stream at atmospheric pressure and dew point of –60°C is used for
deriming a cold box of air liquefaction plant. If argon is used instead of
nitrogen, what should be the dew point on equivalent moisture content
basis?
(g) For relatively small capacity storage of liquefied gases (such as ammonia,
LPG, etc.), a sphere construction is selected. Why?
(h) As a rule of thumb for most of the distillation cases, an increase in column
pressure means: (a) decrease or (b) increase in the column capacity and (i)
better or (ii) work fractionation, given other parameters.
(i) Opening stock of a fuel oil took was calculated as 600 kL × SG of 0.94 = 564
t, and the closing stock as 550 kL × SG of 0.94 = 517 t in a day, i.e.,
transfer of 47 t to the boiler. Temperature variations were ignored thinking
that the variation in SG is about 1% and should, for all practical purposes,
be acceptable. In reality, the temperatures of the fuel oil at the time of
measurements of opening and closing stocks were such that corresponding
SG were 0.94 and 0.95, respectively. It really meant an error of about as
much as (i) 5.5, (ii) 3.5, (iii) 1.5, (iv) 0.15 or (v) 0.05 t.
(j) Name 2 alternative refrigerants which can be used in place of R-12 (an ozone
depleting compound).
3.2 Chlorination of an organic compound is to be carried out with chlorine.
Hydrogen chloride gas with some chlorine are vented out from a batch reactor
at near atmospheric pressure. Consider the various options, available for
trapping hydrogen chloride gas from the reactor and develop process flow
diagrams for each of them.
3.3 Study a batch reactor operation. Write start-up, normal operating and
shutdown procedures for the reactor, including safety requirements. Also
prepare a log sheet for the batch operation with a view to collect consolidated
information for review and comments.
3.4 In a chemical plant, a cooling tower of 1000 kW capacity is operated. An
operator reports in summer that the production is limited due to cooling
tower. How will you check the efficiency of the cooling tower?
3.5 In a batch reactor, reaction is carried out under reflux conditions for long
hours with a condenser using cooling water at 32°C. Solvent used in the
vessel is methanol. Plant data indicate substantial loss of methanol to
atmosphere during the reaction. Analyze the situation and suggest remedial
measures for loss prevention with reasonings.
3.6 A pharmaceutical plant, manufacturing bulk drugs, uses acetone for final
purification of a product. After filtration, the wet cake is dried in a fluid bed
drier in which nitrogen is used as a drying medium. The offgases from the
dryer is saturated nitrogen with vapour of water and acetone at near
atmospheric pressure and 75 to 90°C. Local pollution control authority
requires that the emissions from the drier must contain acetone below 200
ppm. Suggest a process scheme such that the final gas mixture meets the local
pollution authority’s requirement with reasons for selection of the scheme.

REFERENCES
1. Jones, J., W. Asher, J. Bomben, D. Bomberger, C. Marynowski, R. Murray, R.
Phillips, D. Robers, K. Semaru, R. Swidler, C. Witham, and J. Zuegel, Chem.
Engg., 100 (4), April 1993, p. 138.
2. Lowenstein, T. G., Chem Engg., 92 (25), December 9, 1985, p. 62.
3. Frye, A. L., Ind. Engg. Chem., 50 (4), 1958, p. 577.
4. Jordan, D. G., Chemical Pilot Plant Practice, Interscience Publishers, Inc.,
USA, 1995.
5. Zlokarmic, M., Scale-up in Chemical Engineering, 2nd Ed., Wiley-VCH
Verlag GmbH, Germany, 2006.
6. Henry H. C. and J. B. Gilbert, Ind. Eng. Chem. Process Des. Dev., 12 (3),
1973, p. 328.
7. Reed, R. C., J. M. Prausnitz, and B. E. Poling, The Properties of Gases and
Liquids, 4th Ed., McGraw-Hill, USA, 1987.
8. AIChExtra, Chem. Engg. Progr., 99 (9), 2003, p. 78.
9. Bader, M.E., Chem. Engg., 87 (11), June 2, 1980, p. 123.
10. Sharrat, P. N., Ed., Handbook of Batch Process Design, Kluwer Academics,
USA, 1999.
11. Bhatt, B. I., and S. B. Thakore, Stoichiometry, 5th Ed., McGraw-Hill
 Education, New Delhi, 2010.
12. Kletz, T., Process Plants: A Handbook for Inherently Safer Design, Taylor
and Francis, USA, 1998.
13. Ludwig, E. E., Applied Process Design for Chemical and Petrochemical
Plants, Vol. 1–3, 3rd Ed., Gulf Publishing Co., USA, 1999–2001.
14. Elvers, B., S. Hawkins, and G. Schulz, Ed., Ullmann’s Encyclopedia of
Industrial Chemistry, 5th Ed., Vol. B1 to B4, VCH Verlagsgesellschaft,
Germany, 1988–1992.
15. McKetta, J. J., and A. W. Cunningham, Encyclopedia of Chemical
Processing and Design, Vol. 1 to 68, Marcel Dekker, Inc., USA, 1976–1999.
16. Evans, L. Jr., Equipment Design Handbook for Refineries and Chemical
Plants, Vol. 1 and 2, Gulf Publishing Co., USA, 1971.
 4
Importance of Process Diagrams in
Process Engineering
4.1 INTRODUCTION
Engineering Diagrams are vital for understanding of a process to chemical
engineers, of equipments to mechanical engineers, of electrical circuits to
electrical engineers, and so on. These diagrams are prepared in many different
ways to suit the specific requirements. Process Flow Diagrams (PFD) and Piping
and Instrumentation Diagrams (P&ID) are basic requirements of process
engineering and are prepared before attempting any detailing. Steps followed in
preparation of these diagrams and interpretation of the same are covered in this
chapter.

4.2 NEED FOR PFD AND P&ID

When any chemical manufacturing project is conceived, immediate information
is sought on the requirements of raw materials and utilities. This information can
be calculated using the Laws of Conservation of Mass and Energy, covered in
the previous chapter on Stoichiometry. When these data are presented in a
sequential manner with symbolic representation of equipments, it is known as
PROCESS FLOW DIAGRAM (PFD) or PROCESS FLOW SHEET. It can be
said that PFD is a schematic illustration of the system. It shows relationships
between the major components in the system. It also tabulates the process design
values for the components.
PIPING & INSTRUMENTATION DIAGRAM (P&ID) is different from PFD.
While PFD is a quantitative representation of the process, sequential
presentation of flow, piping with instrumentation and controls that are required
for smooth operation of the process are given on a P&I diagram. It can therefore
be said that P&ID is a schematic illustration of functional relationship of piping,
instrumentation, and system equipment components. It shows all the piping
including the physical sequence of branches, fittings, equipments,
instrumentation and control interlocks.
PFD is a useful diagram in effective communication of design information. It
helps an operator in adjusting his operating parameters, a shift supervisor in
checking the plant operation, management in discussions across the table and
project engineers in comparison and evaluation of different processes and their
quantitative requirements.
Modern plants have extensive instrumentation and controls such that a minimum
attention is required of an operator. Apart from controlling the plant variables,
these plants give better productivity in terms of yield and quality. Also, less
manpower is required to operate the plant with increased safety.
Process engineers are involved in preparation of both these types of diagrams.

4.3 BLOCK DIAGRAMS

Simplest flowsheets are block diagrams. Process engineer begins the process
design with a block flow diagram in which only the feed and product streams are
identified. Such input-output diagrams are not very detailed and as such are most
useful in early stages of process development.
Figure 4.1 is a block diagram of the process to manufacture carbon dioxide from
flue gas of power generation plant using electrochemical membrane separation
technique.
 Fig. 4.1 Block Diagram for Manufacture of 200 t/d CO2 by Electrochemical Membrane Process1
(Reproduced with the Permission of American Institute of Chemical Engineers, USA)

Absorption of carbon dioxide from the flue gas mixture using an aqueous amine
solution or caustic potash solution is difficult because of presence of oxygen.
Electrodialysis cell has an edge over the absorption technology in which
presence of oxygen is not a hindrance to the recovery of carbon dioxide. The
block diagram is useful in understanding overall process and also gives rough
quantitative picture of raw materials and utility consumptions. These data can
help in working out preliminary variable cost.
The processes, described in the block diagram, are then broken down into basic
functional elements such as reaction and separation sections. Process engineer
can also identify the recycle streams and additional unit operations to achieve the
desired operating conditions. Identification of basic elements help in performing
material and energy balance calculations which in turn help in preparation of the
PFD.

4.4 DEVELOPMENT AND UTILITY OF PROCESS

FLOW DIAGRAMS
Any product or process development starts with stoichiometric calculations. At
first, required capacity of the final product is fixed. The product can normally be
manufactured by inputting different raw materials and following different
process routes. For example, manufacture of ammonia is possible with different
hydrocarbon feed stocks such as natural gas, naphtha, fuel oil, coal, etc. Process
route for each feed stock is different. For example, reforming is a preferred route
for natural gas or naphtha while partial oxidation is the route followed for fuel
oil or coal. Nitrogen for the ammonia manufacture is invariably obtained from
air but it may be fed as air or nitrogen is separated from air by liquefaction and it
is fed to the process. PFD (Process Flow Diagram) for each continuous process
is different.
Another example is reduction of nitro aromatic compounds to corresponding
amines. In a classical Bechamp process, nitro compound is mixed with water and
fine iron powder in a batch reactor. The reaction mixture is moderately heated in
presence of a catalyst to initiate the reaction (basically in an acidic medium).
Hydrogen is generated in situ and reduces the nitro compound to corresponding
amine. Reaction is exothermic and temperature is controlled by cooling.
Reaction takes long time to complete. Product mixture is separated in organic
and aqueous layers. While aqueous layer is drained out, organic layer is
siphoned and purified by filtration and distillation under vacuum.
In newer version of the process, hydrogen gas is bubbled in the organic liquid
under pressure. Nickel or noble metal-based catalyst is used. Gas dispersion
through the liquid is improved by different methods such as use of static mixer,
use of gas induced agitator, liquid jet ejector system, etc. Such a system is quite
efficient and the reaction can be completed in relatively short period. Both the
process routes are used in industry. PFDs for both the routes are different.
Above examples involve chemical reactions. However, there are PFDs which
involve unit operations but no chemical reactions. For example, separation of
components of natural gas in a distillation column by cryogenic means, removal
of acid gas by absorption–desorption, removal of Volatile Organic Compounds
(VOC) by adsorption–desorption, etc., do not involve a chemical reaction. Each
PFD is, therefore, to be treated on its own requirements.
A Process Flow Diagram generally includes following information:
1. Flow rate of each stream in case of continuous process or quantity of each
reactant in case of a batch process.
2. Composition of streams.
3. Operating conditions of each stream such as pressure, temperature,
concentrations, etc.
4. Heat added or removed in a particular equipment.
5. Flows and design operating conditions of utilities such as steam, cooling
water, brine, hot oil, chilled water, thermic fluid, etc.
6. Major equipment symbols, names and identification.
7. Any specific information which is useful in understanding the process. For
example, symbolic presentation of a hazard, safety precautions, sequence of
flow, etc.
PFD of a laboratory scale process could be somewhat different from that of plant
scale PFD. In laboratory, more emphasis is given on safety aspects. Also some
utilities could be different from those used on a plant scale. For example, in
laboratory use of ice and dry ice are common for achieving required chilling. On
a plant scale, chilled water or chilled brine at different temperatures are used.
Similarly, once through raw water system is common for cooling in condensers
in the laboratory while closed circuit of cooling water is a practice in the
industry. Batch distillation is commonly employed in the laboratory while in
plants, batch or continuous distillation columns are employed with specific
controls to achieve desired purity of the distilled materials.
Process Flow Diagram of a laboratory experimental unit could be simple and
different from that of a pilot plant. Thodos et al2 have described a PFD for the
study of reaction kinetics to synthesize ethyl chloride from ethylene and
hydrogen chloride gas. This is a gas phase endothermic catalytic reaction,
studied in the pressure range of 8–11.5 bar a and temperature range of 150°C–
300°C. A 100 mm steel pipe reactor with electric heating element, wrapped over
it, was used. Dry ice trap, following the reactor, removed ethyl chloride while
excess HCl was scrubbed in a water trap. Thus a continuous experimental set-up
provided useful data for determining reaction rate equation.
When the above process is to be studied on a pilot plant, the reactor could still be
a pipe (bigger in size) with jacket in which thermic fluid (hot oil) circulation is
provided. For separation of ethyl chloride, heat exchanger with chilled brine can
be considered. Scrubbing of unreacted HCl can be performed in an efficient
packed tower or in a falling film absorber.
Prugh3 has developed a laboratory scale PFD for the manufacture of phenylethyl
alcohol by reacting ethylene oxide and benzene in presence of aluminum
chloride catalyst. Such a PFD is helpful in transferring information from the
research chemist to the process engineer, engaged in designing a pilot plant or in
developing a PFD for a commercial plant. Sufficient stress is laid on safe scale-
up. Effectiveness of the research efforts is thereby improved in designing a plant.
In the laboratory, many hazardous reactions are carried out such as chlorination,
hydrogenation, ozonation, carbonylation, etc. Some of them are not only
explosive but also prone to health hazards. During scale-up of such processes,
these aspects need special attention so that the plant is safe in operation and
environmental pollution problem is not created. For example, venting of
hydrogen containing gas mixture must ensure that the gas mixture is outside
flammability limits. Also a right type of flame arrestor is to be installed on the
vent. Similarly chlorine to be vented is to be scrubbed in aqueous caustic
solution. These details form a part of process (although not a part of reaction)
and therefore PFD must include these equipments.
PFD of a commercial plant covers material and energy balance details
extensively. In PFDs of batch processes, quantities of inputs and outputs are
specified with normal operating conditions. For a continuous process plant, a
good practice is to give process stream flow rates and compositions in a tabular
form at the bottom of PFD. Reference 4 gives several PFDs for batch as well as
continuous processes.
Development of PFD of a process plant starts with conceptual design. At first the
process flow sequence is established with the help of symbols of equipments or
blocks. While symbols of equipments give better idea of the type of equipment
to be deployed, block diagrams are satisfactory to start with. End product and
capacity of production are defined which are entered at the end of PFD. Then an
overall theoretical material balance is established to find the theoretical raw
materials’ requirements. These quantities are entered at appropriate places in
PFD where the specific raw materials are to be introduced. Thus a preliminary
PFD is prepared. Now start establishing material balance and energy balance
across each equipment taking into conversion limitations, yield, selectivity,
catalyst activity, equipment limitations, etc. It is well known that conversions are
dictated by kinetic considerations and hence, it is not possible to achieve 100%
conversion every time. These calculations are performed for each equipment
with actual conversions and results are entered in the PFD. With these revised
data, PFD becomes more like a working PFD. To this, add utilities and safety
aspects. Final PFD is now ready for presentation.
Material and energy balance calculations for the manufacture of formaldehyde
from methanol based on Formox process are well covered in Ref. 4. Using this
information, its PFD is developed for a plant having methanol feed rate of 4000
kg/h (Fig. 4.2). In this diagram, process flow rates and operating conditions are
given along the process lines. Heat duties of various equipments are also
specified near the respective equipment. Stream compositions are given in
tabular form at the bottom of the PFD.
The above PFD is quite useful to a process engineer in detail design of the
equipments, selection of right equipment for the duty, pipe sizing, selection of
right instrumentation and control system, etc. Further, PFDs, submitted by
different vendors at the quotation stage, help the project engineer in comparison
and evaluation of alternative process routes. Competitive processes can be
evaluated from different angles such as ease of operation, minimum energy
requirement, less hazardous operation, turndown capability of the plant, etc.
Such an evaluation becomes useful in discussions with the management across
the table thereby the technologically best plant can be selected.
PFD of an operating plant provides design data at 100% plant load. Such a
diagram helps the operator in adjusting his process parameters and the shift
supervisor in checking/controlling the plant operation to optimize the
production. Manufacture of formaldehyde is a catalytic process. Comparison of
actual operating data with the design conditions (such as those given is Fig. 4.2)
can provide a trend analysis thereby a decision can be taken on when to change
the catalyst. However, if the supervisor finds that productivity of the plant is not
commensurate with the anticipated catalyst life, he would look for other factors
like poisoning of catalyst, fouling of coolers, etc., responsible for the same and
corrective actions can be taken.
In Fig. 4.3, PFD for the production of liquid oxygen from air at low pressure (10
bar g) is given. In this process, only physical separation takes place unlike
chemical reaction, described in Fig. 4.2. This PFD was developed to compare the
low pressure process with the high pressure (135–140 bar g) process. It does not
give the process flows and stream compositions and hence, it is not directly
useful to plant operating personnel. Nevertheless it is useful to the project
engineer in preliminary evaluation.

Fig. 4.2 Process Flow Diagram for Manufacture of Formaldehyde

Fig. 4.3 Process Flow Diagram for Liquefying Oxygen at Low Pressure (Linde Process)5 (Reproduced with
the Permission of McGraw Chemical Engineering by Access Intelligence, USA)
 Fig. 4.4 Process Flow Diagram for NGL Recovery from Natural Gas by Expander Process (Reproduced
with the Permission of the McGraw-Hill, Inc., USA)

Figure 4.4 is the PFD for separation of ethane and heavier components from
natural gas. In this process also only physical separation takes place but
operating conditions of all streams are specified. Flow rates of major streams
and their compositions are given in a tabular form at the bottom of the PFD. An
interesting feature of the PFD is incorporation of two heat pumps which will
result in substantial energy saving. Also an expander is provided to generate
power when gas pressure is reduced. External refrigeration is provided at
sufficiently high temperature level so that a standard refrigeration plant with R-
22 can be used and cryogenic liquids are not required. Development of PFD thus
provides an opportunity to a process engineer for energy conservation.
PFD of a utility system is somewhat different from the PFD of the process plant.
Steam generation, steam distribution and utilization, cooling water circuit,
thermic fluid circuit, instrument/plant air system, nitrogen (inert) gas system,
refrigeration system, etc., fall under this category. Figure 4.5 is a typical cascade
steam balance of a large chemical plant. This is also a PFD of the steam system.
Development of an efficient steam system can pay rich dividends in terms of
optimum energy input.
In Chapter 11, detailed discussion on utilities is given with different P&I
diagrams and flow diagrams.
In Ref. 4, several steam balances are worked out for the same plant and
corresponding PFDs are presented to exemplify the alternatives available to a
process engineer during the design stage.
Similar PFDs can be developed for other utility systems. Liu et al8 have
presented PFDs for raw water in a chemical plant. Reduction in raw water
consumption, as much as 50%, could be achieved by optimizing such a PFD.
Cooling water is yet another utility for which an efficient distribution network
can be developed. In Fig. 9.24, cooling water network is demonstrated.
Process flowsheeting can be carried out with the help of a computer. Steady state
material and energy balance calculations can be performed on a spreadsheet or
with the help of a mathematical software which also can generate block
diagrams. Specialized software such as Microsoft Visio, Aspen Plus, etc., are
available for the preparations of PFDs. References 9 to 11 are useful in
developing computerized flow sheets.

Note: All flows are in kg/h.

Fig. 4.5 Steam Balance in a Large Chemical Plant7 (Reproduced with the Permission from Chemical
 Engineering by Access Intelligence, USA)

4.5 DEVELOPMENT OF PIPING AND

INSTRUMENTATION DIAGRAMS
Having understood the requirements of PFD, it is rather easy to understand the
requirements of P&ID. Emphasis is changed from process flows to process
control. Controls are not possible without measurements of variables; such as
pressure, temperature, concentration, etc., for which instruments are required. A
diagram incorporating piping, instruments and controls is known as P&ID.
In simple plants minimum instruments are installed. They are just adequate for
manual control. However, it is needless to emphasize that installation of
adequate instruments justify the cost in terms of productivity and safety. Many
batch processes are being controlled in manual manner. But as the scale of plant
is getting bigger and bigger, role of instrumentation and control (I&C) is
becoming important. This I&C hardware not only results in smoother and
efficient plant operation but also gives safer operation and high onstream factor.
Modern chemical plants incorporate sophisticated I&C, thereby centralized
operation from a control room is possible. PLC (Programmable Logic Control)
and DCS (Distributed Control System) have become common terminology in
I&C. Touch screen operation of the panel gives a wealth of information to the
operator at his finger tips.
In addition to I&C, P&ID also incorporates other details such as pipeline details,
location of valves and fittings, safety valves’ locations, etc. In major chemical
engineering projects, piping and fittings are classified to cover specific range of
pressure and temperature for each fluid. This classification lists material of
construction, thickness, types of fittings required, etc., for piping. This
classification is mentioned on each pipe symbol on P&ID in addition to the size
of the pipe. Many times, fluid flowing in the pipe, is indicated by symbolic
letters, e.g., NG for natural gas, S for steam, etc. This helps project engineer to
specify the piping material with fittings for procurement with ease.
Apart from instrumentation and control details, many other details are also
covered on P&ID. Cold and hot insulations are marked. At times elevation is
mentioned when relative height of a specific equipment is important. For
example, barometric condenser is installed at 10.5 m height from hotwell water
level as standard atmospheric pressure equals 10.33 meters of water column.
Similarly, when a pump takes suction from a vacuum system, NPSH
requirements are to be met which means positive suction is required depending
on the vacuum. Deaerator for boiler feed water (BFW) is installed at a height to
provide sufficient NPSH to the BFW pump. In many instances, minimum height
of the stack, is specified by the local pollution control authority.
When complete draining of a vessel is desired, slope of specific pipeline is
desired which can be indicated on P&ID in a symbolic manner. Man-entry and
maintenance requirements are kept in mind while preparing P&ID. For this
purpose, purging of vessels with inert gas/compressed air is necessary although
this requirement is not indicated on PFD. Location of sample points and their
arrangements are also shown on P&ID.
Symbolic presentation on the diagram should be easily understood by the
engineer. There are standard symbols for such presentation. All engineering
companies have their own standard symbols which are supplied along with
P&ID.
However, standards are prevalent for the graphical symbols. Among them,
following are prominently followed by many:
Indian Standards:
IS: 3232 Graphical Symbols for Process Flow Diagrams
IS: 9446 Graphical Symbols for Pipelines and Instrumentation Diagrams
US Standards:
ANSI Y32.2.3 Graphical Symbols for Pipe Fittings, Valves and Piping
ANSI Y32.2.11 Graphical Symbols for Process Flow Diagrams
ISA 5.5 Graphical Symbols for Process Displays
British Standards:
BS: 1646 1 to 4 Symbolic Representation for Process Measurement, Control
Functions and Instrumentation
German Standards:
DIN 19227 P1 to P3 Graphical Symbols and Identifying Letters for Process
Measurement and Control Functions
Figures 4.6(a) and (b) give recommended symbols for equipments while Figs
4.7(a) and (b) give recommended symbols for piping, instruments and controls
which will be generally followed in this book.
Fig. 4.6(a) Symbols for Equipments
Fig. 4.6(b) Symbols for Equipments
Fig. 4.7(a) Symbols for Piping
Fig. 4.7(b) Symbols for Instrumentation and Controls
 Fig. 4.7(c) Symbols for Instrumentation and Controls (Contd.)

Bartow et al.12 have presented a P&ID for a pilot plant to carry out research and
development work relating to manufacture active pharmaceutical ingredients
(API) in a bulk drug plant. Different types of reactions can be carried out in such
a pilot plant with heating and cooling, varying form cryogenic conditions (up to
–100°C) to hot conditions (up to +250°C). Process engineer can develop P&ID
for such an universal facility.
4.5.1 Development of Simplified P&I Diagrams
With the use of guidelines and symbols, given in previous paragraphs, P&I
diagrams for various systems can be developed with ease.
Consider the unit operation of distillation. Prime objectives of instrumentation
and control (I&C) system of this unit operation are to maintain purities of top
and bottom products. There could be variety of disturbances in the system and
I&C should be developed to take care of these disturbances. Change in feed flow
rate and/or its composition, change in thermal condition of the feed, changes in
heating/cooling medium conditions, changes in ambient conditions, etc., are
some disturbances that are to be taken care and purities must be maintained.
For simplicity, a binary distillation system will be considered. Cooling water is
considered as a cooling medium while steam at suitable pressure is considered
for heating in the reboiler. Purity of the top product is generally controlled by
controlling the top temperature and varying the reflux ratio while purity of the
bottom product is chiefly controlled by maintaining a specified temperature of
the solution in the reboiler. Figure 4.8(a) gives a typical P&I diagram for a
system in which distillation is carried under vacuum in a packed column. In this
case, top product purity is of prime importance. Temperature of vapour, leaving
the column, is of prime importance. Reflux ratio is adjusted to control this
temperature. While temperature of liquid in the reboiler dictates the quality of
the bottom product, differential pressure across the column is controlled near
design value by energy input to the reboiler.
Where bottom product quality is of significant importance, a scheme shown in
Fig. 4.8(b) could be considered. It may be noted that the temperature of liquid in
the reboiler is of prime consideration. Hence, a casade controller with reboiler
liquid temperature as a master parameter and differential pressure across the
column as the slave parameter could prove useful. In this case, distillation is
carried out under pressure in a plate column. In a binary distillation,
temperatures of vapour, leaving top or liquid leaving bottom are nearly constant
for specified purities. In addition to the temperature control, it is desired that
pressure drop across the column is also controlled to avoid excessive boiling in
the reboiler. Selection of plate or packed column is process designer engineer’s
choice but control philosophy remains nearly unaffected by the column internals.
However, it is possible that a sharp (or significant) change in temperature at a
particular plate in plate column or at a particular point in packed column can
take place with change in the feed composition or its thermal condition. This
location can be adjudged by tray to tray calculations and additional measurement
control of temperature at this location can help in controlling the purities of both
products during changes in feed conditions. In Fig. 4.8(a), it is considered that
sharp change in temperature is expected in rectification section while in Fig.
4.8(b), the sharp change in temperature in the stripping section is visualized. P&I
diagram, shown in Fig. 4.8(a) relates to a vacuum system and hence, elevation
marking was considered necessary on it to ensure that equipments’ placement
and piping details can be worked out accordingly.
Considered another system in which a continuous stirred tank reactor (CSTR) is
used for carrying out an exothermic reaction. For initiation of the reaction,
initially heat transfer fluid (hot oil) from a thermo pack is used to supply heat to
the reactor. Once the reaction has picked up and exothermic heat of reaction is to
be removed, cooled thermic fluid can be circulated in the jacket. This scheme is
shown in Fig. 4.9(a). In an alternate arrangement, a CSTR with limpet coils can
be considered. One coil can be used for heating with hot thermic fluid while
another coil can be used for cooling with cooling water. Note that curved baffles
are provided in the jacket so that thermic fluid flows in a guided manner and
channeling of it can be avoided.
 Fig. 4.8(a) P&I Diagram for High Purity Distillate

Figure 4.9(b) is the P&I diagram for this arrangement. Latter case can be
considered when the thermic fluid is simultaneously used in different
equipments for heating and cooling
of thermic fluid. However, it may be a waste of energy in the thermo pack. If
CSTR has only one jacket or one limpet coil, heat removal can be considered by
circulation of liquid in an external heat exchanger with the help of a pump and
providing sufficient liquid head for hot reactants to overcome required net
positive suction head (NPSH)R. Figure 4.9(c) gives the details of such a system.
In these three P&I diagrams, different type of agitators are shown but the control
philosophy is unaffected by these internals.
Fig. 4.8(b) P&I Diagram for High Purity Bottom Product
Fig. 4.9(a) Use of Thermic Fluid for Heating/Cooling in CSTR with Jacket
Fig. 4.9(b) Use of Thermic Fluid for Heating and of Cooling Water for Cooling in CSTR with Limpet Coils

Fig. 4.9(c) Use of Thermic Fluid for Heating/Cooling in CSTR with Thermopack and External Cooler
Fig. 4.9 P&I Diagrams for Temperature Control in CSTR

Development of the above simplified P&I diagrams with reasoning may prove
useful in learning development of P&I diagrams for entire plants involving
several equipments and having number of utilities.
 Fig. 4.10 Piping and Instrumentation Diagram for Manufacture of Formaldehyde

Figure 4.10 is P&ID for formaldehyde manufacturing, described in Fig. 4.2 with
recycle option. It is developed to ensure safe and smooth operation of the plant
in an efficient manner.
Types of control system, shown in Fig. 4.10 can vary from one process engineer
to another. For example, air flow to the reactor can be controlled by a control
valve on feed air line (A-3-750-CS1). However, in Fig. 4.10, vent controls the
feed air flow. Such a control system is found more reliable and accurate in
practice. Also, discharge pressure of the blower could be lower in the suggested
scheme. Flow of demineralized (DM) water to the absorber may not be a critical
requirement. Manual control could therefore be adequate and costly I&C system
for its flow control may not be warranted. Thus, preparation of P&ID provides
an opportunity to the process engineer to demonstrate his skills. Further, the
diagram has to take care of start-up requirements, emergency response, purging,
etc.
In certain plants, universal facilities are required for the manufacture or
processing of similar products. For example, manufacture of different amino and
alkyd resins call for such a universal facility. Lanson13 has presented a P&ID for
the resin manufacturing either by azeotropic distillation process or by vacuum
process. An edible oil refinery in which edible oils of different seeds can be
refined also represents a similar case.
Several softwares are available for preparation of P&ID also. Microsoft Visio,
Aspen Plus, Smart Sketch, etc. can be used to reduce the time in such
preparation.

4.6 CONCLUSION
PFD and P&ID are indispensable as transmitters of information, essential in the
design stages. They can also be considered as the instruments of communication.
In the final analysis, process diagrams communicate an astonishing amount of
engineering information about many things to many groups for a variety of uses.
It is not surprising that preparations of PFD and P&ID have assumed great
importance in chemical process industry. Process engineers can produce
innovative designs which can be well presented by these diagrams. Time spent
on preparation can easily result in cost savings, efficient plant layout, and safer
plant operation.

EXERCISES
4.1 Prepare P&I diagrams for following cases with safety measures:
(a) Hydrogen from a bullet at 10 bar g and 40°C is let down to 2 bar g for
feeding to edible hydrogenation reactor.
(b) Carbon dioxide is to be cooled from 110°C to 40°C using cooling water in an
intercooler at 68 bar a. Cooling water is supplied at 32°C. Critical pressure
and critical temperature of carbon dioxide are 73.77 bar and 31°C,
respectively.
(c) A heat sensitive product is concentrated in a wiped film evaporator (refer to
Chapter 8) at absolute pressure of 1 × 10–3 Pa and 180°C. Ultra high
vacuum is achieved with the help of dry vacuum pump and three-stage
ejector system (with barometric condenser) in series. A cold trap is
provided with brine/steam utilities for knocking off organic vapours, being
carried to the vacuum system. Temperature in the evaporator is controlled
by thermic fluid, entering at 240°C in the jacket and leaving at 220°C from
the jacket. Adequate instrumentation and controls are to be provided for
smooth operation.
(d) Visit a nearby liquefied gas (such as ammonia, carbon dioxide, LPG, etc.)
storage facility, study the same and prepare a P&I diagram of the system,
including storage tank and holding refrigeration system.
4.2 A pilot plant reactor (SS 316L) of 250 L geometric capacity is to be designed
 for conducting sponsored research at any temperature, ranging from –80°C to
+220°C. For this purpose, it is desired to use a specialty synthetic (silicone
polymer based) heat transfer medium, such as Syltherm XLT® of Dow
Chemical Co., USA. Typical properties of Syltherm XLT are as under.
Chemical: Dimethyl polysiloxane
Average molar mass: 317 kg/kmol
Flash point: 54°C
Auto ignition temperature: 350°C
Freezing point: Less than –110°C
Density @ 25°C: 0.852 kg/L
Gross heat of combustion at 25°C = 32 800 kJ/kg
The reactor will be suitable for maximum operating pressure of 25 bar g. It will
be provided with a turbine agitator with a variable frequency (speed) drive for
speed control in the range, 80–200 rpm. A flameproof electric motor of 3 kW
capacity, operating on 415 ± 10% V, 3 phase, 50 ± 2 Hz power will be used as
a drive. Agitator will be sealed with a mechanical seal with cooling
arrangement with cooling water circulation.
The reactor will also be suitable for operation under vacuum up to 700 Torr.
Reflux arrangement with two condensers; one using cooling water at 32°C
and another vent condenser, using brine at –10°C, will be provided.
Heat transfer medium (HTM) will be circulated in the jacket with the help of a
gear pump. A degassing tank of 100 L capacity will be provided at 2 m
elevation relative to the outlet nozzle of the reactor.
(a) HTM will be cooled from –40°C to –100°C with liquid nitrogen in a welded
plate heat exchanger (PHE).
(b) HTM will be cooled with the help of refrigerant R–22 in the temperature
range –40°C and +40°C in the welded PHE.
(c) HTM will be cooled with cooling water at 32°C in the temperature range
+40°C and +80°C in a shell and tube heat exchanger.
(d) HTM will be heated up to +160°C with saturated steam at 7 bar g in a shell
and tube heat exchanger.
(e) HTM will be heated beyond +160°C up to +220°C with the help of thermic
fluid, supplied at +280°C in the shell and tube heat exchanger.
Heat transfer duty will be 5 kW for each case.
Based on above information, develop a P&I diagram of the above system.
4.3 Prepare a PFD for the manufacture of nitric acid based on stoichiometric
calculations, given in Examples 2.10 and 2.11.
4.4 Prepare a P&ID for VOC emission control scheme, developed for Exercise
3.6.

REFERENCES
1. Winnick, J., Chem. Engg. Prog. 86 (1), 1990, p. 41.
2. Thodos, G., and L. F. Stutzman, Ind. Engg. Chem., 50 (3), 1958, p. 413.
3. Prugh, R. W., Chem. Engg. Prog., 70 (11), 1974, p. 72.
4. Bhatt, B. I., and S. B. Thakore, Stoichiometry, 5th Ed., McGraw Hill
Education, New Delhi, 2010.
5. Guccione, E., Chem. Engg., 72 (4), 1965, p. 144.
6. Johnson, J. E., and D. J. Morgan, Chem. Engg., 92 (14). 1985, p. 72.
7. Slack, J. B., Chem. Engg., 79 (20), 1972, p. 107.
8. Liu, Y. A., B. Lucas, and J. Mann, Chem. Engg., 111 (1), Jan. 2004, p. 30.
9. Westerburg, A.W., H. P. Hutchison, R. L. Motard, and P. Winter, Process
Flowsheeting, Cambridge University Press, UK, 1980.
10. Turner, J., Chem. Engg. Progr., 88 (8), 1992, p. 59.
11. Julian, F.M., Chem. Engg. Progr., 81 (9), 1985, p. 35.
12. Bartow, M., and E. Spark, Chem. Engg., 111 (1), Jan. 2004, p. 49.
13. Swaraj Paul, Surface Coating; Science & Technology, 2nd Ed., John Wiley &
Sons, USA, 1985, p. 127.
 5
Process Design of Piping, Fluid
Moving Devices and Flow Meters
5.1 INTRODUCTION
A fluid is a substance which when in static equilibrium cannot sustain tangential
or shear forces. This property of action under stress distinguishes it from the
‘solid’ state of the matter. Fluids are commonly divided into two substances;
liquids and gases. A liquid occupies a definite volume, independent of the
dimensions of its container, while a gas tends to expand to fill any container in
which it is contained. A vapour differs from a gas by being readily condensable
to a liquid on cooling.
Strictly speaking all fluids are compressible. However, gases are frequently
regarded as compressible while liquids are as incompressible as the latter
undergoes small changes in the properties under compression. Ambient air is
treated as incompressible fluid as its density changes insignificantly with minor
changes in atmospheric pressure and temperature. However, compressed air is
treated as a compressible fluid.
In a pump, a fan or a compressor, mechanical work is transformed into fluid
energy. A turbine or an expander is a machine for transforming fluid energy into
mechanical work at the rotating shaft. The word ‘pump’ is normally used for a
machine handling liquid. In a fan, the density change is so small, due to
relatively low pressure difference, that the gas is regarded as incompressible for
evaluation. In a compressor, the density change is appreciable.
This chapter deals with fluids, its transportation through various moving devices
and various flow measuring instruments.

5.2 PROCESS DESIGN OF PIPING

Design of piping system incorporates all major disciplines of engineering like
chemical engineering, mechanical engineering, civil engineering,
instrumentation and control engineering, electrical engineering, etc.
Process design of pipe is a balance between size or diameter of pipe and pressure
drop in the pipe. For a given flow rate of fluid if larger size of pipe is selected
then it gives lesser pressure drop. Larger size of pipe increases the fixed cost of
pipe and lesser pressure drop means lesser power consumption or lesser
operating cost of pipe. Lesser pressure drop in pipe may reduce the size of flow
moving device like pump and thereby it may reduce the fixed cost of pump.
Thus capital cost of pipe increases with diameter, whereas pumping or
compression cost decrease with increasing diameter. Ideally one should find the
pipe size which gives the sum of amortized capital cost plus operating cost a
minimum or one should find the optimum pipe size.
5.2.1 Optimum Pipe Size
Several equations and nomographs are available in literature to estimate the
optimum pipe diameter. But these equations or nomographs are not reliable as
optimum pipe diameter depends on the current cost of material, cost of power,
rate of interest at particular place and at particular time. It changes with place
and time. For example, optimum pipe diameter equation, derived in USA, cannot
be used to find the same in India. One such equation1 derived for carbon steel
pipe and turbulent flow of an incompressible fluid is
dopt = 293 ṁ0.53 ρ–0.37  (5.1)
where, dopt = Optimum pipe size, mm
ṁ = Flow rate, kg/s
ρ = Density of fluid, kg/m3
This equation was developed by the Institution of Chemical Engineers, UK
based on cost data of 1987, updated to mid 1992.
5.2.2 Recommended Fluid Velocities in Pipe
The velocities, recommended in Table 5.1, are only for approximate calculations
of pipe diameter which can be used for the quick calculation of short distance
pipe line or for estimating pipe size as a starting point for pressure drop
calculations.
Table 5.1 Recommended Fluid Velocities
Fluid Service Recommended velocity, m/s
Water Pump suction line 0.3 to 1.5
Pump discharge line 2 to 3
Average service 1 to 2.5
Gravity flow 0.5 to 1
Steam 0 to 2 atm g, saturated 20 to 30
2 to 10 atm g, saturated 30 to 50
Superheated below 10 atm g 20 to 50
Superheated above 10 atm g 30 to 75
 Vacuum lines 100 to 125
Air 0 to 2 atm g 20
> 2 atm g 30
Ammonia/ Liquid 1.8
refrigerant Gas 30
Organic liquids and oils 1.8 to 2
Natural gas 25 to 35
Chlorine Liquid 1.5
Gas 10 to 25
Hydrochloric acid Liquid (aqueous) 1.5
Gas 10
Inorganic liquids 1.2 to 1.8
Gas and vapours 15 to 30

5.2.3 Pressure Drop in Pipe

Relation between pressure drop and pipe diameter is given by Fanning or Darcy
equation. It is derived for steady flow in uniform circular pipes running full of
liquid under isothermal conditions.

  (5.2)

For SI units, gc = Newton’s law conversion factor = 1

where, Δp = Pressure drop, Pa
L = Length of pipe, m
ṁ = Mass flow rate of fluid, kg/s
ρ = Density of fluid, kg/m3
Di = Pipe inside diameter, m
v = Velocity of fluid, m/s
f = Fanning friction factor
The Fanning friction factor is a function of the Reynolds number (Re) and the
roughness of inside surface (ε).
Values of ε for various materials are given in Table 5.2.
Table 5.2 Values of Surface Roughness for Various Materials
Material Surface roughness ε, mm

Commercial steel or Wrought iron 0.045 72

Galvanized iron 0.152
Cast iron 0.259
Concrete 0.305 – 3.05
Riveted steel 0.914 – 9.14
 Brass, Lead, Glass, Cement, and Bituminus Linings 0.001 524

A plot of Fanning friction factor as a function of Reynolds number (Re) and

relative roughness, ε/D, is given as Fig. 5.1.

Fig. 5.1 Fanning Friction Factors2

(Reproduced with the Permission of McGraw-Hill Book Co., USA)

For turbulent flow in commercial steel pipe, f = 0.0394 Re–0.16  (5.3a)

Above equation gives the value f on higher (safer) side. One of the more
accurate relationships between f and Re for turbulent flow (Ref.: 3)

  (5.3b)

where, Re = (Reynolds number)  (5.4)

μ = Viscosity of fluid, Pa · s or kg/(m · s) or N · s/m2

From Eqs.   (5.2),   (5.3a) and   (5.4),
If G in kg/s, μ in kg/(m · s) or in N · s/m2, Di in m, ρ in kg/m3, L in m, then Δp is
in Pa or in N/m2.
Normally pipe diameters are expressed in mm. If Di is in mm then Δp in Pa is
given by

  (5.5a)

If Δp is in kPa or in kN/m2, then

  (5.5b)

Estimate the optimum pipe diameter for a flow of dry chlorine

gas of 10 000 kg/h at 6 atm a and 20oC through carbon steel pipe using Eq. (5.1).
Solution
For carbon steel pipe, optimum pipe diameter
dopt = 293 ṁ0.53 ρ–0.37  (5.1)
ṁ = 10 000 kg/h = 2.7778 kg/s

ρ = 17.71 kg/m3
dopt = 293(2.7778)0.53(17.71)–0.37
= 173.855 mm
Use next heigher size of standard pipe, i.e., 200 mm (8 in) NB standard pipe.
5.2.4 Standard Pipes
Standard pipes are available from 3 mm (1/8 in) to 600 mm (24 in) size.
Standard pipes are specified with three different diameters, namely inside
diameter, outside diameter and nominal diameter. For standard pipes having
diameter more than 300 mm (12 in), nominal diameters are equal to the actual
outside diameter but for smaller pipe there is no relation between nominal
diameter and inside diameter or outside diameter.
The wall thickness of standard pipe is indicated by the schedule number as per
the US standards.
Schedule Number, SCH =   (5.6)

where, psf = Safe working pressure, N/mm2

σs = Safe allowable stress, N/mm2
Standard pipes are available with variety of schedule numbers; 5, 10, 20, 30, 40,
60, 80, 100, 120, 140, 160. With carbon steel pipe, schedule 40 up to 150 mm (6
in) and with stainless steel pipe, schedule 10 are used for general purposes.
Thickness of standard pipe increases with increases in schedule number.
Seamless standard pipes do not have welding joint. Fabricated pipes are
fabricated by rolling the plates followed by joining the ends of plate by welding
which are also known as Electric Resistance Welded (ERW) pipes.
In matric system (including those specified in Indian Standards), thickness of a
pipe is specified in mm.
Thickness of pipe, subjected to internal pressure, is determined by equation.
  (5.7)

where, t = Thickness of pipe, mm

p = Internal design pressure, N/m2
ri = Inside radius of pipe, mm
ro = Outside radius of pipe, mm
σ = Allowable stress of pipe material at design temperature, N/m2
E = Joint efficiency; for seamless standard pipe, E = 1
CA = Corrosion allowance, mm
5.2.5 Pressure Drop in Fittings and Valves
In addition to pipes, piping system contains fittings and valves. These fittings
and valves offer the additional frictional loss or additional pressure drop. This
additional frictional loss of a fitting or of a valve is expressed either as an
equivalent straight pipe length (Le) or as a number of velocity heads (K ), lost in
pipe of same size and of same material.
Equivalent Length of Pipe (Le) for Fittings and Valves Equivalent length of a
valve or of a fitting is the length of straight pipe of same size creating the same
friction loss as the fitting or the valve in consideration. Nomographs for finding
equivalent length of pipe for various fittings and valves are given in various
books (Ref. 4, 5 and 6).
Often, Le is expressed in terms of inside diameter of pipe. Then Di where,
Di = Inside diameter of pipe.
Values of Le/Di for few valves and fittings are given in Table 5.3.
Table 5.3 Equivalent Lengths of Selected Fittings and Valves

Valve or fitting

Gate valve (fully open) 7 to 10

Gate valve (3/4 closed) 800 to 1100
Gate valve (1/2 closed) 190 to 290
Globe valve (fully open) 330 to 480
Angle valve (fully open) 165 to 220
Plug valve (fully open) 18
90° elbows (standard radius) 30
45° elbows (long radius) 5.8
45° elbows (short radius) 8.0
Return bend (medium radius) 39 to 56
Coupling or union Negligible
Tee, straight through 22

Another way of calculating pressure drop through the fittings and valves is the
use of factor K.
“Number of velocity heads (K) lost in pipe” for fittings or valves is defined by
equation

= K  (5.8)

or   (5.9)

where, ΔF = Additional frictional loss, J/kg or N · m/kg

Δp = Additional pressure drop, N/m2
v = Average fluid velocity through the pipe of same size as valve or fitting, m/s
Values of K for normally used fittings and valves are given in Table 5.4.
Table 5.4 Equivalent Velocity Heads of Fittings and Valves
Type of fitting or valve Equivalent number of velocity heads (K ) (applicable only for turbulent flow)
Gate valve (open) 0.17
Globle valve, bevel seat (open) 6.0
Plug valve (open) 0.4
Diaphragm valve (open) 2.3
Angle valve (open) 2.0
Check valve (swing) 2.0
Foot valve 15
Coupling 0.04
90o elbows (standard) 0.75
90° elbows (long radius) 0.45
45o elbows (long radius) 0.2
45o elbows standard 0.35
90° bend 0.75
180° bend (closed return) 1.50
Tee straight through 0.4
Butterfly valve (α = 5°) 0.24
Check valve (swing type) 2.0

Simpson7 has given values of K for variety of other fittings, such as miter bends,
different types of reducers, different bends, etc.
Carbon dioxide is to be conveyed from the top of the stripper of
ammonia plant to urea plant. Calculate the pipe size required based on following
data:
Flow rate of CO2 = 1000 t/day
Total length of pipe = 800 m
Available pressure at inlet of pipe = 24 kPa g
Discharge pressure of CO2 from pipe required = atmospheric
No. of 90° elbows in pipe line = 8
No. of butterfly valve = 1
No. of flow nozzle = 1
Temperature of gas = 60°C
Viscosity of CO2 gas = 0.016 mPa · s or cP
Solution
Density of carbon dioxide (CO2) gas at the discharge of pipe:
ρ =

p = 1 atm, Molar mass of CO2, M = 44

R =

ρ =

Let the velocity of gas through pipe, v = 20 m/s

Di = 0.677 m
Selecting fabricated pipe having inside diameter, Di = 677 mm
Reynolds number

= 1361 955 > 10 000 (i.e., turbulent flow)

Pressure drop in straight steel pipe for turbulent flow
= 4.07 × 1010ṁ1.84μ0.16Di–4.84ρ–1  (5.5)

= 4.07 × 1010(11.574)1.84(0.016 × 10–3)0.16(677)–4.84(1.6094)–1

= 7.8046 × 10–3 kPa/m ≡ 7.8046 Pa/m
Pressure drop in butterfly valve:

= 77.25 Pa
Pressure drop in 90° elbow (standard):

= 241.41 Pa
Pressure drop in the flow nozzle is considered negligible.
Total pressure drop in pipe line
Δpt = 7.8046 × 800 + 77.25 + 8 × 241.41 = 8252 Pa ≡ 8.25 kPa
Available pressure is 24 kPa. Hence, to decrease the fixed cost, smaller pipe can
be considered. Similar calculations for smaller pipe sizes give the following
results:
Table 5.5 Comparison of Pressure Drop Calculations
Nominal size Schedule No. ID of pipe, mm Δp, kPa

677 mm — 677 8.25

600 mm (24 in) 20 590.55 15.558
500 mm (20 in) 20 488.95 37.525

Select 600 mm, SCH-20 pipe size for the given duty.
Calculate the pipe size based on following data. Fluid flowing
through pipe is carbon monoxide. Discharge pressure of carbon monoxide
required from the pipe is atmospheric.
Available pressure at inlet of pipe = 50 kPa g
Length of pipe = 4 km
Flow rate of CO = 1500 kg/h
Temperature of gas = 50°C
No. of gate valves in pipeline = 2
No. of 45° elbows = 3
No. of 90° elbows = 6
Viscosity of CO = 0.018 mPa · s or cP
Solution
Density of carbon monoxide (CO) gas:
Molar mass of CO, M = 28

= 1.0558 kg/m3
Pressure drop in straight pipe for steel material and for turbulent region:
  (5.5)

Let v = 5 m/s which implies Di = 0.317 m.

For the first trial calculations let the nominal size be 12 in NB (300 mm), SCH-
40 pipe having inside Di = 304.8 mm.

= 1.2724 × 10–3 kPa/m ≡ 1.2724 Pa/m

Reynolds numbers, Re =

= 96 697 > 10 000, i.e., CO is flowing in turbulent region

Velocity of gas, v = = 5.4 m/s

For gate valve,

Similarly for 45° elbows, ΔpE2 =

= 5.388 N/m2 per 45° elbow

For 90° elbows, ΔpE3 =

= 11.545 N/m2 per 90° elbow

Total pressure drop in pipe line
Δpt = 1.2724 × 4000 + 2.617 × 2 + 5.388 × 3 + 11.545 × 6
Δpt = 5180.27 Pa ≡ 5.18 kPa ≡ 0.0511 atm
For 300 mm NB SCH-40 pipe, pressure drop in piping is very low. Hence, to
reduce the fixed capital investment, smaller pipe diameter can be considered.
Similar calculations of smaller pipe diameters gives the following results:
Table 5.6 Comparison of Pressure Drop Calculations
Nominal size Schedule No. ID of pipe, mm Δpt, kPa Remarks

100 mm (4 in) SCH-40 102.26 1012.38 Very high

150 mm (6 in) SCH-40 154.05 139.75 High
200 mm (8 in) SCH-40 202.72 37.10 Reasonable
300 mm (12 in) SCH-40 304.80 5.18 Very low
Considering fixed capital investment, operating cost and safety margin in design,
200 mm (8 in) SCH-40 pipe can be selected for the given application. For this
size of pipe, velocity of gas through pipe is equal to 12.23 m/s which is
acceptable (refer Table 5.1).
100 000 kg/h of water is to be transferred from canal to
reservoir by gravity. Maximum height of water level above the discharge of pipe
in reservoir will be 2 m. Difference between minimum level of water in canal
and maximum level of water in reservoir is 6 m.
Length of pipe = 3000 m
Equivalent length of pipe for fittings and valves = 200 m
Maximum temperature of water = 40°C
Density of water at 40°C = 993 kg/m3
Viscosity of water at 40°C = 0.67 mPa · s or cP
Surface roughness for carbon steel, ε = 0.0457 mm
Surface roughness for concrete, ε = 1.2 mm
Material of pipe is carbon steel. Determine the suitable pipe size. If material of
pipe is concrete, will there be any change in pipe size required?
Solution
Maximum allowable pressure drop in pipe = 6 m WC
Δpmax = 6 m WC = 58.84 kPa = 0.58 atm
Let the velocity of water through pipe v = 0.5 m/s (For first trial calculations)
v = 0.5 m/s =

Di = 0.2669 m ≡ 266.9 mm
Select 300 mm NB, SCH-40 carbon steel pipe.
ID of pipe, Di = 304.8 mm, μ = 0.67 × 10–3 kg/(m · s)

Reynolds number, Re = = 173 188

From Fig. 5.1, f = 0.0042

Pressure drop in pipe, Δp =   (5.2)
Δp =

Δp = 12 871.2 Pa ≡ 12.871 kPa < Δpmax

For the next smaller size of pipe, i.e., for 250 mm NB, SCH-40 pipe, Di = 254.5
mm, Re = 207 417.3
= 1.7956 × 10–4, f = 0.0041 and Δp = 30.959 kPa < Δpmax

Select 250 mm NB SCH-40 carbon steel pipe.

If pipe material is concrete, then for the same size
and Re = 207 417.3

From Fig. 5.1, f = 0.0077

Pressure drop Δp = 58.142 kPa ≈ Δpmax
Thus calculated pressure drop is nearly equal to the maximum allowable
pressure drop. For safer design, next higher size of concrete pipe, i.e., 300 mm
(12 in) pipe, should be used.
An organic liquid is discharged at the rate of 5000 kg/h from a
reactor to a storage tank at 50°C as shown in Fig. 5.2. Reactor is under pressure
at 600 kPa g. Density of the organic liquid is 930 kg/m3 and viscosity is 0.91
mPa · s or cP. Assume negligible flashing of the organic liquid across the control
valve.

Fig. 5.2 Liquid Flow System

Piping system details:

Linear length of straight pipe = 50 m
No. of 90° elbows of standard radius = 6 (All are not shown in Fig. 5.2.)
No. of Tees = 2
Pressure drop in orifice meter = 40 kPa
No. of gate valves = 4
No. of globe valve = 1
No. of flow control valve = 1
Determine the pipe size. Assume it to be uniform throughout. Also find the
residual pressure drop that must be taken by the flow control valve.
Solution
Flow rate of liquid, ṁ = 5000 kg/h = 1.4 kg/s
Volumetric flow rate of liquid:

= 1.4934 × 10–3 m3/s

Based on Table 5.1, select velocity of liquid to be 1.8 m/s.
Inside diameter of pipe required based on selected velocity

Selected 40 mm (1.5 in) NB, SCH-40 pipe having Di = 40.894 mm

Reynolds number, Re =

(i.e., turbulent flow)

For the turbulent flow in commercial steel pipe
  (5.5)

= 0.413 65 kPa/m
Linear length of pipe = 50 m
Table 5.7 Equivalent Lengths of Fittings and Valves
Type No. Pipe ID, mm Le, m (From Table 5.3)
 90° Elbows 6 40.894 6(30Di) = 7.36

Tees 2 40.894 2(22Di) = 1.8

Gate valves 4 40.894 4(10Di) = 1.636

Total 10.796

Total pressure drop

Δpt = 0.413 65 (50 + 10.796) + Pressure drop in orifice meter
= 0.413 65 (50 + 10.796) + 40 = 65.15 kPa
Pressure balance:
Pressure in reactor – Δpt – Δpcontrol valve = Pressure in storage tank
(600 + 101.325) – 65.15 – Δpc = 101.325
Δpc = 534.85 kPa ≈ 5.35 bar
This residual pressure drop should be offered by the control valve. Specifications
of the control valve should be prepared accordingly.

5.3 PROCESS DESIGN OF FLUID MOVING

DEVICES
It includes process design of pump, fan, blower, and compressor.
5.3.1 Process Design of Pump
Pump is a device used in flow system of liquid to increase the mechanical energy
of the flowing liquid. Pumps are broadly divided into two types; (i) centrifugal
pumps and (ii) positive displacement pumps. The following terminologies are
used with pump:
1. Capacity of pump: Flow rate (qv) of fluid created by the pump is known as
capacity of pump. In SI units, capacity is expressed in m3/h or L/s.
2. Total Dynamic Head: Total dynamic head H of a pump is the difference
between total discharge head hd and total suction head hs.
H = hd – hs of Liquid Column (LC)   (5.10)
Before installation of pump it is possible to estimate the total discharge head hd
by using following equation:
hd = hsd + hfd LC   (5.11)
where, hsd = Static discharge head = p′± Z′   (5.12)
p′ = Absolute pressure over free surface of liquid in receiver
Z′ = Vertical distance between free surface of liquid in receiver and centerline of
pump placed horizontally (For vertical pump Z′ is distance between free
surface of liquid and eye of suction of impeller).
hsd = p′ + Z′, if pump is below the level of free surface of liquid in receiver and
hsd = p′ – Z′, if it is placed above the free surface of liquid in receiver.
hfd = Discharge friction head = frictional loss in discharge line.
During operation, total discharge head hd can be determined by following
equation:
hd = hgd + atm pressure + hvd   (5.13)
where, hgd = Discharge gauge pressure measured by pressure gauge. If pressure
is below atmospheric, vacuum gauge reading is used for hgd in Eq. (5.13)
but with negative sign.
hvd = Velocity head at the point of gauge attachment in discharge line
Before installation of pump, total suction head can be estimated by following
equation:
hs = hss – hfs LC  (5.14)
where, hss = Static suction head = p ± Z
p = Absolute pressure over free surface of liquid at source
Z = Vertical distance between free surface of liquid at source and centerline of
horizontal pump (or suction eye impeller for vertical pump)
hss = p + Z, if pump is going to be installed below the free surface of liquid
hss = p – Z, if pump is going to be installed above the free surface of liquid
hfs = Friction loss in suction line
In existing installation, total suction head hs can be determined by following
equation:
hs = atm pressure + hgs + hvs LC   (5.15)
where, hgs = suction gauge pressure,
hvs = velocity head at the point of gage attachment.
In the above equations, all pressure heads are in terms of liquid heads (LC),
flowing in the system.
3. Net Positive Suction Head (NPSH): Net positive suction head is the excess
of sum of velocity head and pressure head in suction line over the vapour
pressure head of liquid. When pump installation is designed, available net
positive suction head (NPSH)A must be equal to or greater than net positive
suction head required; (NPSH)R.
(NPSH)R is normally specified by the pump supplier, while based on the
installation of pump, (NPSH)A should be calculated and specified by the process
engineer. Theoretically, (NPSH)A should be greater than zero to avoid cavitation.
(NPSH)R depends on properties of liquid, the total liquid head, pump speed,
capacity and impeller design. Practical curves of (NPSH)R vs capacity and speed
of pump are supplied by the pump manufacturer. Figures 5.3 and 5.4 can be used
as a guideline to find the value of (NPSH)R for centrifugal pump handling water
at temperatures below 100oC and above 100°C, respectively.

Fig. 5.3 Net Positive Suction Head for High Pressure Centrifugal Hot-Water Pumps8
(Reproduced with the Permission of the Hydraulic Institute (USA)

When (NPSH)A is less than (NPSH)R, cavitation can take place and bubbles of
vapour form in the suction line. Eventually these bubbles collapse inside the
casing of the pump when pressure is exerted on them by the impeller of the
pump. Such collapse of bubbles can cause severe erosion and damage to the
pump. It may form minor cavities on the inside surface of the casing and of the
impeller. Hence, this phenomenon is called cavitation.
When a pump installation is being designed, the available net positive suction
head (NPSH)A can be calculated by the following equation:
(NPSH)A = hss – hfs – pv   (5.16)
where, hss = Static suction head, m LC = p ± Z
hfs = Friction loss in suction line, m of liquid column (LC)
pv = Vapour pressure of liquid at suction temperature expressed in m of liquid
column (LC)
For existing installation, (NPSH)A can be determined by following equation:
(NPSH)A = atm pressure + hgs – pv + hvs   (5.17)
where, hgs = suction gauge pressure, m of liquid column (LC)
pv = vapour pressure of liquid at suction temperature expressed in m of liquid
column (LC)

Fig. 5.4 Temperature Correction Chart for Net Positive Suction Head Requirement for Centrifugal Hot
Water Pumps8
(Reproduced with the Permission of the Hydraulic Institute, USA)
hvs = Velocity head at the point of gauge attachment, m of liquid column (LC)
As a general guide (NPSH)A should preferably be above 3 m for pump capacities
up to a flow rate of 100 m3/h, and 6 m above this capacity.
For a given system, if (NPSH)A is less than (NPSH)R, following remedial
measures are recommended:
(a) Change the location of the pump to improve (NPSH)A. In other words,
positive suction head (hss) may be increased.
(b) Provide jacketed cooling in the suction line to decrease the vapour pressure
(pv) of the liquid.
(c) Reduce operating speed of the pump; thereby specific speed of the pump is
reduced and its (NPSH)R is less.

5.3.2 NPSH Requirement for Liquids Saturated with

Dissolved Gases
In many situations, the liquid to be pumped is saturated with gases, which have
definite solubilities in the liquid. When a suction system for such a liquid is to be
designed for a centrifugal pump, (NPSH)A calculations differ from Eqs.   
(5.16) or   (5.17).
Pumping of cooling water (saturated with air), pumping of condensate from a
knock-out drum of a compressor, pumping of solution from an absorber, etc., can
be cited as examples of such situations where the liquids are saturated with
gases.
Dissolved gases start desorbing when the pump is started and suction is
generated at the pump eyes. Normally, a pump can tolerate 2 to 3% flashed gases
at the pump eye without encountering cavitations. If the design of the suction
system is made to restrict about 2.5% flash, it is considered safe for the pump
operation.
In Eq. (5.16) suction source pressure is the system pressure (p) minus the vapour
pressure (pv) for a normal liquid having practically no dissolved gases. For a
liquid saturated with dissolved gases, pv is replaced by pva which is called
artificial liquid vapour pressure. For evaluation of pva, following procedure is
recommended:
1. Calculate molar mass of the gas mixture, dissolved in the liquid.
2. Calculate mass fraction (wo) of the dissolved gas mixture.
3. Calculate psuedo-critical properties of the dissolved gas mixture, if system
 pressure is high.
4. Calculate specific volume of the dissolved gas mixture (VGa) at the operating
conditions.
Steps (1) to (4) can be avoided if the solubility of the gas mixture in the liquid
(such as air in water) is known from the literature.
5. Calculate volume fraction of the dissolved gas (GVP) in a hypotheoretical
gas-liquid mixture. Consider one unit mass of the liquid in which the gas
mixture is dissolved. If GVP is less than or equal to 2.5%, Eq. (5.16) can be
safely used to calculate (NPSH)A using vapour pressure (pv) of the liquid at
the operating temperature. If higher than calculate volume fraction (a) of
flashed gas mixture (as pressure is lowered) over the liquid, saturated with the
dissolved gas mixture, using the following equation9:

a =   (5.18)

where, p = liquid pressure at pump eye, kPa

pv = vapour pressure of liquid at the operating temperature, kPa
p0 = system pressure, kPa
VGa = specific volume of the dissolved gas mixture, m3/kg [see Step (4) above]
VL = specific volume of the liquid at the operating conditions, m3/kg
Equation (5.18) assumes that the dissolved gas mixture follows ideal gas law,
Dalton’s law and Henry’s law.
6. Calculate a for different value of p. Draw a graph of a vs p. Read p
corresponding to a = 0.025 which is called pva. Alternately by trial and error,
calculate pva, for a = 0.025.
7. Use Eq. (5.16) and insert pva in place of pv and calculate (NPSH)A. The above
procedure is well exemplified in Example 5.10.
5.3.3 Power Required for Pumping
Power required for pumping an incompressible fluid is given by equation:
  (5.19)
where, P = Power required, kW
H = Total dynamic head, m of liquid column
qv = Capacity, m3/h
ρ = Fluid density, kg/m3
η = Efficiency of pump
A centrifugal pump draws benzene from an overhead tank.
Operating vacuum in the tank is 700 Torr. Vertical distance between free surface
of liquid in tank and centerline of pump is 12 m. Maximum operating
temperature is 50°C. Vapour pressure of benzene at 50°C is 280 Torr. Density of
benzene at 50°C is 870 kg/m3. Frictional loss in suction line of pump is 1 m of
benzene column. Calculate the (NPSH)A of centrifugal pump.
Solution
(NPSH)A = hss – hfs – pv  (5.16)
hss = p + Z
p = 760 – 700 = 60 Torr a
1 atm = 10.33 m WC
p =

= 0.937 m LC
LC

(NPSH)A = p + Z – hfs – pv = 0.937 + 12 – 1 – 4.37 = 7.567 m LC

A centrifugal pump is drawing methanol from an underground
tank. Operating pressure in the underground tank is 1 atm. Vertical distance
between centreline of pump and free surface of liquid in the tank is 3 m.
Maximum operating temperature is 50o C. Vapour pressure and density of
methanol at 50o C are 400 Torr and 785 kg/m3, respectively. Friction loss in
suction line is 1 m of liquid column. Calculate the (NPSH)A of centrifugal pump.
Solution
(NPSH)A = hss – hfs – pv = p – Z – hfs – pv  (5.16)
p =

pv =

(NPSH)A = 13.16 – 3 – 1 – 6.925 = 2.235 m LC

Note: Effect of vapour pressure (pv) of the fluid is significant in calculation of
(NPSH)A. Selection of the pump is critical. As a general guideline (NPSH)A
should be greater than 3 m for a centrifugal pump.
Installation of pump is to be decided for the a pump which will
handle hot water at a temperature of 176.7°C and a suction pressure of 1.034
MPa g. Capacity of pump will be 10 m3/h. It is planned to use a centrifugal
pump with single suction for this applications. Speed of its impeller will be 2900
rpm.
Find the minimum distance that must be kept between free surface of liquid at
source and centreline of horizontal centrifugal pump.
Solution
(NPSH)A ≥ (NPSH)R
From Figs 5.3 and 5.4, data for centrifugal pump having a speed of 3550 rpm are
given. Assume these data to be applicable for 2900 rpm (a safer design). For
3550 rpm, 10 m3/h, 176.7°C and p < 35 bar:
(NPSH)R = 3.048 + 3.6576 m (additional)
= 6.7 m
(NPSH)A ≥ 6.7 m
p ± Z – hfs – pv ≥ 6.7 m
p = 1.034 MPa g = 1.034 × 106 Pa g
1 atm = 101.325 kPa
p = (1.034 × 106 + 101.325 × 103) Pa a
p = 1.135 325 × 106 Pa a
Vapour pressure of hot water at 176.7°C = 0.928 MPa
pv = 0.928 × 106 Pa
Density of water at 176.7°C, ρ = 890 kg/m3
p – pv = 0.207 325 × 106 =

= = 23.746 m WC

Select the pipe size and piping system such that hfs ≤ 1 m.
–Z – 1 + 23.746 ≥ 6.7 m
(if pump is installed above the free surface of liquid at source)
or Z ≤ 16.046 m
Pump can be installed maximum up to an elevation 16.046 m above the free
surface of liquid at source to avoid cavitation. Pump can be installed at any
distance, if it is to be installed below the free surface of liquid at source.
In the previous example, if hot water is saturated at 176.7°C
temperature at source, what will be the minimum height that must be kept
between free surface of liquid at source and centreline of selected pump? Type,
speed and capacity of pump are same as that for previous example.
Solution
For the single suction centrifugal pump having speed 2900 rpm and capacity 10
m3/h,
(NPSH)R = 6.7 m
(NPSH)A ≥ (NPSH)R or (NPSH)A ≥ 6.7 m
Here hot water is saturated at 176.7° C at source.
p = pv
In this case pump must be installed below the free surface of liquid at source.
(NPSH)A = p + Z – hfs – pv = Z – hfs ≥ 6.7 m
Let the velocity of water in suction line = 1 m/s
qv = v × (π/4)Di2 = 10 m3/h ≡ 2.7778 × 10–3 m3/s

Di = = 0.059 47 m

From the table of standard pipe select 62.5 mm NB SCH-40 steel pipe having Di
= 62.71 mm
Re =
ρ of water at 176.7°C = 890 kg/m3
Velocity based on Di = 62.71 mm, v = 0.9 m/s
Viscosity of water at 176.7°C and at saturation pressure, 9.08 atm a must be
determined. Effect of pressure on viscosity of liquid is insignificant up to about
50 atm.
For reduced temperature Tr < 0.75, relation between viscosity of liquid and
temperature is given by equation
ln μw = , T in K  (5.20)

where, C = 17.71 – 0.19 TB, TB is normal boiling point (i.e., at 1 atm) in K

For water, C = 17.71 – 0.19 (100 + 273) = –53.16
From Appendix 14 of Ref. 10:
μw at 149°C = 0.185 cP, μw at 115.6°C = 0.242 cP or mPa · s

ln (0.185) =

ln (0.242) =

A + 0.0027 112 B = –1.6874, A + 0.002 981 B = –1.4188

Solving simultaneous equations,
A = –4.3865, B = 995.55
ln μw = – 4.3865

μw = 0.1532 cP or mPa · s at 176.7°C

μw = 0.1493 mPa · s at 180°C from JSME Steam Tables – a check

Re = = 327 876.7 (i.e. turbulent)

For turbulent flow through commercial steel pipe

= 4.07 × 1010 ṁ1.84 μ0.16 ρ–1  (5.5)

qv = 10 m3/h = 2.7778 × 10–3 m3/s

ṁ = mass flow rate = qv × ρ = 2.7778 × 10–3 × 890 = 2.4722 kg/s
= 4.07 × 1010 × (2.4722)1.84 × (0.1532 × 10–3)0.16 × (62.71)–4.84 × (890)–1
 = 0.1186 kPa/m

For 6.7 m elevation,

Let L = actual length of pipe = 7 m, including horizontal length for pump
suction.
Assuming three 90° elbows and one globe valve in the suction line.
Table 5.8 Equivalent Lengths of Fitting and Valve
Valve or Fitting No. Le

90° elbows (standard) 3 3 (30 Di) = 5.64 m

Globe valve (open) 1 350 Di = 21.95 m

L ≅ 7 + 5.64 + 21.95 ≅ 34.59 m

Δp = 4.102 kPa = Liquid Column (i.e., WC at 176.7°C)

hfs = Δp = = 0.47 m of WC

Z ≥ (6.7 + 0.47) m or Z ≥ 7.17 m

Minimum elevation required between free surface of water at source and
centreline of pump = 7.17 m
Carbon dioxide gas (> 99.5% pure) from ammonia plant at
1.01 atm a and 40°C is washed with cooling water in a packed column before
compression in urea plant. Cooling water from bottom of the column is pumped
back to cooling tower. Solubility of carbon dioxide in water is 0.0973 kg/100 kg
water at 1 atm and 40°C. Find the volume fraction of carbon dioxide over
cooling water. Also calculate artificial liquid vapour pressure.
Solution
Molar mass of carbon dioxide, M = 44.0095
From JSME Stream Tables, vapour pressure of water at 40°C,
pv = 0.073 75 bar
= 0.0728 atm
Mass fraction of carbon dioxide in the liquid
w0 =
= 9.8177 × 10–4
VGa =   (5.18)

= 5.676 × 10–4 m3/kg H2O

From JSME Steam Tables
VL = 1.0078 m3/kg H2O at 40°C

GVP =

= 0.0563%
Thus, volume per cent of carbon dioxide is only 0.0563% (< 2.5%) over cooling
water and there is no danger of pump cavitation.
po = 1.01 atm

a =   (5.18)

By trial and error,

For p = 0.145 atm, a = 0.0247
Hence, pva = 0.145 atm
Thus, if vapour pressure of water (0.0728 atm) is used instead of pva (i.e., 0.145
atm), (NPSH)A would have been erroneous by 0.75 m H2O.
Note: Above calculations indicate that at higher temperature and lower suction
pressure, pva could have significantly affected (NPSH)A.
Benzene at 37.8°C is pumped through the system at a rate of
9.09 m3/h with the help of a centrifugal pump. The reservoir is at atmospheric
pressure. Pressure at the end of discharge line is 345 kPa g. The discharge head
is 3.05 m and the pump suction head is 1.22 m above the level of liquid in
reservoir. The friction loss in suction line is 3.45 kPa and that in the discharge
line is 37.9 kPa. The mechanical efficiency of the pump (η) is 0.6. The density of
benzene is 865 kg/m3 and its vapour pressure at 37.8°C is 26.2 kPa.
Calculate (a) (NPSH)A and (b) power required by centrifugal pump.
Solution
(a) (NPSH)A = hss – hfs – pv = p – Z – hfs – pv

p = 1 atm = = 11.94 m LC

Z = 1.22 m (suction)

hfs = 3.45 kPa = = 0.4066 m LC

pv = 26.2 kPa ≡ 3.088 m LC

(NPSH)A = 11.94 – 1.22 – 0.4066 – 3.088 = 7.2254 m of LC
(b) Power required by centrifugal pump

  (5.19)

p′ = Pressure at discharge = 345 + 101.325 = 446.325 kPa a

Z′ = 3.05 – 1.22 = 1.83 m

hsd = 52.6 + 1.83 = 54.43 m LC

hfd = 37.9 kPa = = 4.47 m LC

hd = 58.9 m LC
hs = hss – hfs = p – Z – hfs = 11.94 – 1.22 – 0.4066 = 10.31 m
H = 58.9 – 10.31 = 48.59 m
P = = 1.735 kW

It is proposed to pump 10 000 kg/h of saturated toluene at

114°C and 1.1 atm a from the reboiler of a distillation tower to a second
distillation unit without cooling. If the friction loss in the line between the
reboiler and pump is 7 kPa and the density of toluene is 866 kg/m3, how much
liquid level in the reboiler is to be maintained to give a net positive suction head
of 2.5 m? Calculate the power required to drive the pump if the pump is to
elevate the toluene to 10 m to a second unit at atmospheric pressure. Assume
friction loss in the discharge line to be 35 kPa. Pump efficiency is 62%.
Solution
In this case, the pump is installed below the liquid level.
(NPSH)A = p + Z – hfs – pv = 2.5 m
Liquid in reboiler is saturated liquid at reboiler conditions. Therefore, p = pv
Z – hfs = 2.5 m
hfs =

= 0.824 m LC
Z = 2.5 + 0.824 = 3.324 m LC
Pump must be installed at a level 3.324 m below the free surface of liquid in
reboiler.
Power required to drive the pump
  (5.19)

p′ =

Z′ = 10 m
hsd = 10 + 11.927 = 21.927 m

hfd = 35 kPa = = 4.12 m LC

hd = 21.927 + 4.12 = 26.047 m

Total suction head hs = hss – hfs = p + Z – hfs  (5.14)

p = 1.1 atm =

hs = 13.12 + 3.324 – 0.824 = 15.62 m LC

H = hd – hs = 26.047 – 15.62 = 10.427 m LC
qv = = 11.5473 m3/h

P = = 0.4582 kW

A centrifugal pump is drawing water from an overhead tank,

exposed to atmosphere. Vertical distance between free surface of liquid in the
tank and centerline of pump is 10 m. Capacity of centrifugal pump is 10 000
kg/h. Maximum operating temperature is 50°C. Vapour pressure of water at
50°C is 92.51 Torr. Total length of suction pipe 10.5 m, having two 90° elbows.
Determine the size of suction pipe and (NPSH)A of centrifugal pump. Material
of pipe is carbon steel. Density of water = 1000 kg/m3, Viscosity of water =
0.558 mPa · s (or cP).
Solution
In Table 5.1, velocity of water in suction line (v) is recommended in the range of
0.3 to 1.5 m/s.
Let velocity of water in suction line, v = 1 m/s

Di = 0.0595 m
Select 62.5 mm NB SCH-40 pipe, Di = 62.7 mm

Actual velocity, v = 0.9 m/s through the selected pipe.

Re =

= 101 129 (i.e., turbulent flow)

Surface roughness of carbon steel, ε = 0.0457 mm

From Fig. 5.1, f = 0.005

  (5.2)
= 129 Pa/m
For 90° elbows, additional pressure drop
  (5.9)

per elbow

Total pressure drop in suction line

Δpt = 129 × 10.5 + 303.75 × 2
Δpt = 1962 Pa ≡ 1.962 kPa ≡ ≡ 0.2 m WC

For 62.5 mm NB SCH-40 pipe,

(NPSH)A = hss – hfs – pv = p + Z – hfs – pv  (5.16)

(NPSH)A = 10.33 + 10 – 0.2 – 1.26 = 18.87 m WC

A centrifugal pump is used to transfer hot water at 110°C from
a deaerator to boiler as shown in Fig. 5.5. Hot water is saturated liquid in the
deaerator at 110°C. Capacity of pump is 16.5 m3/h and its efficiency is 60%.
Speed of the impeller is 2900 rpm. Pipe line to the boiler contains one control
valve and five 90° bends. Its straight length is 25 m.
Calculate:
(a) Recommended size of suction line
(b) Minimum vertical distance Z required between lowest liquid level in the
deaerator and centerline of pump and
(c) Recommended size of pipe line to boiler.
 Fig. 5.5 Typical Boiler Feed Water System

Solution
(NPSH)A ≥ (NPSH)R
For the single suction centrifugal pump having speed 2900 rpm and capacity
16.5 m3/h,
rpm × = constant

qv1 = 11 m3/h
Operating pressure, p = 0.147 MPa g (From JSME Steam Tables corresponding
to the 110°C saturation temperature.)
p = 0.147 MPa g = 248.325 kPa = 2.48 bar < 35 bar
From Figs 5.3 and 5.4, (NPSH)R = 3.048 + 0.4572 = 3.5052 m, say 3.5 m WC
(NPSH)A = p + Z – hfs – pv ≥ 3.5 m
In the deaerator, water is saturated, hence p = pv
Z – hfs ≥ 3.5 m
Let hfs = 1 m (assumed) Z ≥ 4.5 m
Zmin = 4.5 m, Suggested Z = 5 mAns. (b)
Size of suction line
Let velocity of hot water in suction line = 1 m/s (Table 5.1)

Di = 0.0764 m
Select 75 mm NB, SCH-40 MS pipe
Di = 77.92 mm

Reynolds number, Re =

Velocity of water based on 77.92 mm diameter, v = 0.961 m/s

Viscosity water at 110°C, μ = 0.256 mPa . s (or cP), ρ at 110°C = 950.12 kg/m3

Surface roughness of steel pipe = 0.045 72 mm (Table 5.2)

From Fig. 5.1, f = 0.0045, ṁ = 16.5 × × 950.12 = 4.3547 kg/s

  (5.2)

= 101.38 Pa/m
For 90° elbows, additional pressure drop

per elbow

Total pressure drop in suction line:

Δpt = 101.38 × 5 + 1 × 329.05 = 835.95 Pa
= = 0.09 m WC

Corrected value of Zmin = 3.59 m, say 3.6 m

The recommended size of suction line is 75 mm NB, SCH-40 pipe and minimum
vertical distance Zmin required between lowest liquid level in deaerator and
centerline of pump is 3.6 m. Ans. (a) and (b)
Size of discharge line to boiler:
Let the velocity of water in discharge line, v = 2 m/s

Di = 0.054 m
Select 50 mm (2 in) NB, SCH-40 MS pipe for which Di = 52.5 mm
Velocity of water based on 52.5 mm inside diameter v = 2.12 m/s

= 413 079.5
Surface roughness of steel pipe ε = 0.045 72 mm

From Fig. 5.1, f = 0.0047

ṁ = 4.3547 kg/s
For the turbulent flow through steel pipe,

= 762.6 Pa/m
Table 5.9 Equivalent Velocity Heads of Fittings and Valves
Fittings/valves No. (ni) Ki

Gate valve 1 0.17

Control valve 1 6
90° Bends 5 0.75
Check valve 1 2

Total pressure drop in discharge line

= 19 065 + 25 450.5 = 44 515.5 Pa ≡ 44.515 kPa

This value of pressure drop is within permissible limit.
Hence size of discharge line to boiler is recommended to be 50 mm NB, SCH-
40.
Crude dichlorobenzene is pumped from a storage tank to a
distillation column. The tank is blanketed with nitrogen and the pressure above
the liquid surface is held constant at 0.1 bar g. The minimum depth of liquid in
the tank is 1 m.
The distillation column operates at a pressure of 500 Torr a. The feed point to the
column is 12 m above the base of the tank. The tank and column are connected
by a 50 mm internal diameter commercial steel pipe, 200 m long. The pipe,
running from the tank to the column, contains the following valves and fittings:
20 No. of standard radius 90° elbows, two gate valves (fully open), an orifice
plate and a flow control valve.
If the maximum flow rate required is 20 000 kg/h, calculate the power required
by pump. Pump efficiency is 70%. Pressure drops across the control valve is 0.5
bar. Density of the dichlorobenzene is 1300 kg/m3 and viscosity is 1.4 m Pa · s
(or cP).
Flow meter/Valve/Elbows Equivalent number of velocity heads, K
(i) Standard radius 90o Elbows 0.75
(ii) Gate valve (Fully open) 0.17
(iii) Orifice meter 10
Solution
Power required by pump is calculated by following equation:
  (5.19)

qv =

η = 0.7
H = hd – hs = (p′ + Z′ + hfd) – (p + Z – hfs)
p′ = Operating pressure in distillation column
= 500 Torr a = 66 661.184 Pa
To convert p′ in m of liquid column,
p′ =

Z′ = 12 m
p = 0.1 bar g
or p = 10 000 + 101 325 = 111 325 N/m2 a
=

Z = 1 m
hfs = Friction loss in suction line ≅ 0 m LC (i.e., negligible)
hfd = Total friction loss in discharge line in m of liquid column
hfd =

where, Δp = Total pressure drop in discharge line

Velocity of dichlorobenzene in discharge line

(i.e., turbulent)

For the turbulent flow of fluid through commercial steel pipe

= 1.535 73 kPa/m

+ Δp across 90° elbows, gate valves and orifice meter

= 1.535 73 × 103 × 200 + 0.5 × 105 + (20 × 0.75 + 2 × 0.17 + 1 × 10) ×

= 435 171.63 Pa ≡ 435.172 kPa

H =   (5.227 + 12 + 34.123) – (8.7293 + 1 – 0)

= 41.62 m LC
P =

= 3.24 kW

5.4 EVALUATION OF CENTRIFUGAL PUMP

PERFORMANCE WHEN HANDLING VISCOUS
LIQUIDS
When centrifugal pumps, designed and tested for handling water are used for
viscous liquids, significant increase in power consumption and decrease in head
and capacity occur.
Figures 5.6 and 5.7 provide means of determining the performance of a
conventional centrifugal pump handling a viscous liquid when its performance
on water is known. The same can also be used as an aid in selecting a pump for a
given application.
Fig. 5.6 Viscosity Correction Chart8 (Reproduced with the Permission of the Hydraulic Institute, USA)
Fig. 5.7 Performance Correction Chart8 (Reproduced with the Permission of the Hydraulic Institute, USA)

Limitations on use of viscous liquid performance correction chart (Figs 5.6 and
5.7):
1. Extrapolation is not recommended.
2. Use only for pumps of conventional hydraulic design in the normal operating
range with open or closed impellers. Do not use for mixed flow or axial flow
pumps.
3. Use only for Newtonian liquids. Do not use for gels or slurries or other non-
Newtonian liquids.
 Performance of a centrifugal pump on water is known. It can
deliver 9.5 m3/h of water at 50 m total head of water with 80% efficiency. By
using Figs 5.6 and 5.7 predict the performance of this pump if it is used to
deliver lubricating oil with 43 × 10–3 m2/s (or 43 cSt) viscosity at 40°C.
Determine the percentage change in Head, Capacity and Power required by the
pump. Density of lubricating oil at 40°C is 869 kg/m3 at the operating
temperature.
Solution
Power required by centrifugal pump for delivering 9.5 m3/h of water at 50 m
total head with 80% efficiency
  (5.19)

(Density of water at 40oC, ρ = 992.215 kg/m3 from Table 3.28 of Ref. 2)

Enter the graph (Fig. 5.6, viscosity correction chart) with 9.5 m3/h capacity, go
up to 50 m head, over to 43 cSt viscosity, and then up to the correction factors.
CQ = 0.9
CH = 1
CE = 0.71
Capacity of centrifugal pump when it handles the lubricating oil
Qoil = CQQW = 0.9 × 9.5 = 8.55 m3/h
Head developed by pump
Hoil = CH × HW = 1 × 50 = 50 m
Efficiency of pump Eoil = CE × EW = 0.71 × 0.8 = 0.568
Note that there is a significant drop (29%) in the efficiency.
Power required by pump when it handles the lubricating oil,

% decrease in capacity of pump = × 100 = 10%

% decrease in head developed of pump = 0% (i.e., no change)

% increase in power required by pump = × 100 = 11.02%

Figure 5.8 shows the performance curve of a centrifugal pump

operated on water. If the same pump is used for handling oil with a specific
gravity of 0.9 and a viscosity of 1000 SSU (220 cSt) at the pumping temperature,
develop the performance curve of the same pump for oil.
 Fig. 5.8 Sample Performance Chart

Solution
On the performance curve for water, locate the best efficiency point. In this
example, from Fig. 5.8 (from the efficiency vs capacity curve for water), it is 170
m3/h. For water with 170 m3/h capacity, efficiency of pump is maximum (86%).
Tabulate capacity, head and efficiency for 0.6 × 170, 0.8 × 170 and 1.2 × 170
m3/h flow rates. By using Fig. 5.8, for 170 m3/h, 31 m of head is developed by
the pump when it handles the water (from head vs capacity curve for water).
Enter the graph (Fig. 5.7, viscosity correction chart) with 170 m3/h capacity, go
up to 31 m head, over to 1000 SSU (220 cSt) viscosity and then up to correction
factors.
CE = 0.645
CQ = 0.95
CH = 0.895 (for 1.2 × QN)
CH = 0.92 (for 1 × QN)
CH = 0.94 (for 0.8 × QN)
CH = 0.96 (for 0.6 × QN)
Sample calculations for 0.6 × QNW capacity:
QNW = 170 m3/h
0.6 × QNW = 102 m3/h = QW
HW = 34.62 m (from head vs capacity curve for water in Fig. 5.8)
ηW = 0.76 (from efficiency vs capacity curve for water in Fig. 5.8)
Viscosity of oil, μ = 220 cSt
CQ = 0.95 Qoil = CQ · QW = 0.95 × 102 = 96.9 m3/h
CH = 0.96 Hoil = CH · HW = 0.96 × 34.62 = 33.235 m
CE = 0.645 ηoil = CE · ηW = 0.645 × 0.76 = 0.4902
Here again, note the drop (24%) in efficiency.
Power required by pump when it handles the oil

Calculations for obtaining the performance curve for oil given in Table 5.10.
Same data are presented in Fig. 5.8.
Table 5.10 Revised Performance of Centrifugal Pump for Handling Oil
Capacity 0.6 × QNW 0.8 × QNW 1.0 × QNW 1.2 × QNW

Qw m3/h 102 136 170 204

HW m 34.62 33.21 21 27.3

ηW 0.76 0.84 0.86 0.85

CQ 0.95 0.95 0.95 0.95

CH 0.96 0.94 0.92 0.895

CE 0.645 0.645 0.645 0.645

Qoil = CQ × QW 96.9 129.2 161.5 193.8

Hoil = CH × HW 33.325 31.217 28.52 24.434

ηoil = ηW × CE 0.4902 0.5418 0.5547 0.5483

P, kW 16.11 18.26 20.363 21.18

5.5 POWER REQUIRED IN FAN, BLOWER AND

IN ADIABATIC COMPRESSOR
Fans, blowers and compressors are used to increase the mechanical energy of
gases. Discharge pressure of a fan is less than 3.45 kPa g (350 mm WC). If
discharge pressure required is higher than 3.45 kPa g then a blower is used.
Maximum discharge pressure of blower is 1 atm g. If the discharge pressure
required is more than 2 atm a then a compressor is used. Discharge pressure of
compressor ranges from 2 atm a to thousands of atmospheres.
5.5.1 Power Required by Fan
Fans are of two types: (i) centrifugal fans and (ii) axial flow fans. Efficiency of a
fan ranges from 40 to 70%, depending on its specific speed. Power input to a fan
is given by the equation
P0 = 2.72 × 10–5 qv pd /η  (5.21)
where, P0 = Power required by fan, kW
qv = Capacity of fan, m3/h
pd = Discharge pressure of fan, cm WC (gauge pressure)
= Sum of pressure and velocity head, developed by fan, cm WC
η = Efficiency of fan
5.5.2 Power Required in Blower and in Adiabatic
Compressor
Blowers are always operated in an adiabatic manner. Many compressors are also
operated in adiabatic manner (Example: air compressors up to 10 bar g). In
adiabatic compressor, jacketed cooling is not provided. While in other types;
polytropic and isothermal compressors, jacketed cooling is provided around the
compressor section. In isothermal compressor, inlet temperature of gas is equal
to outlet temperature of gas. If both temperatures are not same even after
providing the jacket cooling, it is called polytropic compressor.
For adiabatic compression

  (5.22)

where, k = Cp/Cv
= ratio of specific heat at constant pressure to specific heat at constant volume
p2, p1 = outlet/inlet pressure of gas, kPa
V2, V1 = outlet/inlet volume of gas, m3/kmol
T2, T1 = outlet/inlet temperatures of gas, K
Power required in single stage blowers or in single stage adiabatic compressor,
   (5.23)

where, P0 = Power required, kW

p1, p2 = Absolute inlet pressure, kPa
qv1 = Volumetric flow rate of gas based on inlet condition, m3/h
η = Efficiency of blower
Discharge temperature of gas from the blower or single stage adiabatic
compressor,

  (5.24)

Value of k ranges from 1.39 to 1.41 for air and perfect diatomic gases. For k =
1.395, value of bracketted expression in Eq. (5.23) is known as X factor

  (5.25)

where, k = 1.395 for air or a diatomic gas

r = Compression ratio
Power required for blower and adiabatic compressor
P0 = 9.81 × 10–4qv1 p1 X/η kW  (5.26)
Other gases are having value of k in the range of 1 to 1.4. Power required by
blower or adiabatic compressor, for the gases having value of k other than k =
1.395, is given by the following equation:
P0 = 6.37 × 10–4 qv1p1 kW  (5.27)

where, d = 2.292 [(k – 1)/k]

Values of and d can be obtained from Fig. 5.9 for the different values of k
and for different values of r.
For a multistage compressor with N stages of compression in which
1. adiabatic compression takes place in each stage,
2. near equal work done in each stage (i.e., equal compression ratio in each
 stage; r) and
3. intercooling is provided between each two stages of compression to same inlet
gas temperature (T1),
Power requirement can be calculated by the following equation:

kW  (5.28)

For the multistage compressor, discharge temperature from the last stage can be
predicted by the following equation:
T2 = T1 K  (5.29)

where, N = number of stages

 Fig. 5.9 (XG/X) vs k and d vs k (Ref. 2)

(Reproduced with the Permission of McGraw-Hill Education, USA)

Find out the power required for a turbo blower for the
following duty:
Fluid = Atmospheric air
Capacity = 1000 Nm3/h
Discharge pressure = 2 atm a
Efficiency of turbo blower, η = 0.7
Also find the discharge temperature of air.
Solution
Power required for turbo blower

  (5.23)

For air, k = 1.395

qv1 = Volumetric flow rate based on inlet condition in m3/h

= 33.4 kW
Discharge temperature of air

  (5.24)

Discharge temperature of air, t2 = 120°C

A three stage reciprocating compressor is used to compress
306 Sm3/h of methane from 0.95 atm a to 61.3 atm a. The inlet temperature is
26.7° C.
Specific heat ratio of methane, k = 1.31
Calculate (a) Power required for compression, if mechanical efficiency is 80
percent and
(b) Discharge temperature of gas after 1st stage
Solution
For a multistage compressor, total power required is minimum if each stage
consumes the same power or same compression ratio in each stage.
For same compression ratio in each stage,
Compression ratio in each stage r =

Power required for each stage

  (5.23)

k = 1.31, n = 0.8
qv1 = Volumetric flow rate based on inlet condition m3/h

= 19.45 kW
If power loss in allied piping and intercoolers is neglected, then total power
required for three stages,
P′0 = 3 × 19.45 = 58.35 kW
(b) Discharge temperature after first stage

T2 = (273 + 26.7)(4.01)(0.31/1.31)
T2 = 416.3 K, t2 = 143.3°C
or
for k = 1.31 and r = 4.01, from Fig. 5.9:
d = 0.54, (XG/X) = 0.8

XG = 0.8 × 0.4818 = 0.3854

Power required for each stage
  (5.27)
Total power required P′0 = 3 × 19.34 = 58.03 kW
(Neglecting pressure loss in the intercoolers and allied piping)

5.6 FLOW METERS

A variety of flow meters are in use in chemical process industry. One category is
the traditional differential pressure (DP) type volumetric flow meters. In many
applications, the volumetric flow rate is of direct interest to the user. Because of
its accuracy, simplicity and relatively lower cost, these flow meters are popular
in the chemical industry. By multiplying the flow rate with the actual density,
mass flow rate can be obtained. Pitot tube, orifice meter, venturi meter and flow
nozzles fall under this category. Among these, orifice meter is by far the most
popular in the industry. Figure 5.10 shows these meters schematically.
Less commonly used flow meter is vortex meter which works on Von Karman
effect. Flowing fluid separates on two sides of shedder bar face. Vortices form
behind the face and cause alternating differential pressures (DP) around the back
of the shedder.
The frequency of the alternating vortex development is linearly proportional to
the volumetric flow rate. Although vortex meters are as reliable and as accurate
as other DP instruments, they can be used for only clean liquids and are
recommended for turbulent flows.
 (a) Various Tap Locations for Orifice Meter

(b) Pipe Diameters from Inlet Face of Orifice Plate

 (Extracted from the ASME Flow Meter Computation Handbook, 1961)

(c) Orifice Meter with Vena Contracta Formation

Fig. 5.10 Orifice Meter

Another popular category is of variable area flow meters which are also known
as rotameters. Figure 5.11 shows two different types of rotameters. While DP
type instruments can be installed in horizontal position, rotameters are installed
in vertical position. Bypass rotameters are installed in horizontal position which
is a combination of an orifice and a rotameter. Bypass rotameters are installed on
larger size pipelines and occupy quite less space. Near linear scale of the meter is
an advantage over logarithmic scale for the DP meters. However, rotameters
with glass tubes are relatively less accurate and are mainly used for low pressure
applications.
Disadvantage of low pressure applications due to glass tubes has led to the
development of magnetic rotameters in which a magnetic float is used in a
metallic tapered tube and its accurate position is monitored outside the tube.
Float’s movement is calibrated on a dial-type indicator. Both DP type and
magnetic rotameter can be connected to a microprocessor-based control unit for
totalizing and/or controlling purpose.
Positive displacement-type flow meters incorporate a pair of gears or a turbine
which rotate on their axes and the revolutions are measured. Since volume
displaced per revolution is nearly constant, volumetric flow is calculated by
multiplying two parameters. Traditional water meters are of gear type. Turbine
flow meters are available for number of liquids such as ammonia, chlorine, etc.,
which flow under relatively high pressure. These instruments are sufficiently
accurate and are used for custody transfer.
Weirs are popularly used for volumetric flow measurements in open canals and
alike situations. Figure 5.12 shows different types of weirs in use.

(a) Standard Rotameter

 (b) Magnetic Rotameter

Fig. 5.11 Rotameters

Fig. 5.12 Flow Measurement with Weirs

Another category relates to ultrasonic flow meters. Its underlying principle is

measurement of the effect a flowing fluid has on an ultrasound signal either
transmitted through or reflected from the fluid. These instruments are largely of
clamp-on designs and offer simplicity of installation. These meters suffer from a
reputation for unreliability and poor long-term stability, compared with DP type
meters. However, newer designs are being developed for high pressure and
temperature fluids, including gases. Designs are available when a single
instrument can be clamped on a range (though limited) of pipes and volumetric
flows can be checked periodically. This type of meter is quite handy to a process
engineer in checking plant performance from time to time.
Electromagnetic flow meters (also known as magmeters) work on Faraday’s law
of conductance which states that when a conductive fluid flows through a
magnetic field, generated perpendicular to the flow, the voltage induced in the
fluid is linearly proportional to the volumetric flow rate. Thus, this type of
instrument operates only with electrically conductive liquid. Non-invasive nature
of ultrasonic flow meter is also the feature of magmeter. Both permit
unobstructed flow through the measuring element. Both are clamp on meters
whose transducers are strapped to the pipeline. Unlike ultrasonic meters,
magmeters have a reputation of accuracy.
True mass flow meter based on Coriolis principle was introduced in 1977 by
Mirco Motion, Inc., USA, a division of Emerson Process Management. The
Coriolis principle, discovered by Gustave Coriolis in 1935 states that fluid
flowing through a tube, vibrating at its natural frequency, generates a force that
slightly distorts the tube. This extremely small distortion, picked up by sensors,
is directly proportional to the mass flow rate (Newton’s second law of motion)
and is independent of any other property of the fluid; say pressure, temperature,
density or viscosity. Figure 5.13 schematically represents the mass flow meter.

Fig. 5.13 Coriolis Mass Flow Meter (Courtesy: Micro Motion, Inc., USA)
In a single mass flow meter, four parameters of the fluid can be measured at a
time; mass flow, volumetric flow, temperature, and density. This instrument
offers high accuracy and can be used for any fluid; clean or dirty, transparent or
opaque, conductive or non-conductive. Originally developed instrument required
extra space for installation but recently developed meter has a straight tube
configuration. Among different types of flow meters, Coriolis device is the
costliest.
In the forthcoming sections, process designs of orifice meter and rotameter are
discussed. Reader is advised to refer standard handbooks on
instrumentation/flow meters for design of other flow meters.

(a) Flow Nozzle with Differential Gauge

(b) Venturi Meter

Fig. 5.14 Differential Pressure Type Flow Meters
5.7 PROCESS DESIGN OF ORIFICE METER
Orifice meter is a widely used flow meter in chemical industry, compared to
venturi meter and rotameter.
Advantages of orifice meter are:
1. Fixed cost is less
2. Easy to fabricate and install
3. Occupies less space as compared to the venturi meter
4. Provides more flexibility. Orifice plate can be easily replaced.
Disadvantages of orifice meter is higher power consumption and hence operating
cost of orifice meter is higher than the same of venturi meter and of rotameter.
In orifice meter, as shown in Fig. 5.10, a square edged or sharp edged orifice
plate is mounted between two flanges at the flanged joint. When fluid flows
through the orifice it forms free flowing jet. This free flowing jet first contracts
and then expands. Minimum flow area achieved by free flowing jet is known as
Vena Contracta.
Mass flow rate through orifice is given by the following equation:

ṁ = Co Y Ao   (5.30)

where, ṁ = Mass flow rate of fluid, kg/s

A0 = Area of orifice = (π/4)d20, m2
p1, p2 = Pressure at upstream and downstream static pressure taps, respectively,
Pa
ρ = Density of fluid, kg/m3
β =

Y = Expansion factor
= 1 for liquids
= 1 – [((1 – r)/k)(0.41 + 0.35 β 4)] for gases  (5.31)
where, r = , ratio of downstream to upstream pressure

k = , specific heat ratio

Co = Coefficient of orifice = f (Reo, β, Location of taps)

where, Reo = Reynold’s number for the fluid flowing through orifice

Reo =   (5.32)

where, uo = Velocity of fluid through orifice, m/s

μ = Viscosity of fluid, kg/(m · s)
Equation of orifice meter (Eq. 5.30) is empirical in nature and it requires
experimental support.
In addition to Reo and β, Co also depends on the location of pressure taps.
Location of pressure taps on upstream side and down stream side are also
standardized as discussed in Fig. 5.10(a).
There are total five standard locations of pressure taps:
1. Corner taps: Static holes made in upstream and downstream flange. They are
very close to the orifice plate. With corner taps, it is possible to drill both
static holes in the orifice plate itself. Then entire orifice meter can be easily
inserted in any flanged joint without drilling the holes in pipe or flanges.
2. Flange taps: Static holes made at a distance 25.4 mm on upstream side and
25.4 mm on downstream side.
3. Radius taps: Static holes located at a distance one pipe diameter on upstream
side and 1/2 pipe diameter on downstream side. Radius taps are the best from
practical standpoint of view as it gives reasonably good pressure difference,
compared to other taps except vena contracta taps. Higher pressure difference
means more accurate measurement of flow rate.
4. Vena contracta taps: Upstream static hole is 1/2 to 2 times pipe diameter
from the plate. Downstream tap is located at the position of minimum
pressure. Vena contracta taps give the maximum pressure difference for a
given flow rate. But it is not suitable, if orifice size is changed from time to
time.
5. Pipe taps: Static holes are located at 2.5 times pipe diameter upstream side
and 8 times pipe diameter on down stream side. This means fluid is flowing
normally on both sides without being affected by turbulence, created by the
orifice plate.
For Reo > 30 000,
Co = Between 0.595 to 0.62 for vena contracta taps
Co = Between 0.595 to 0.8 for radius taps
Co = 0.62 for corner taps
Relation between discharge coefficient Co, β and ReD is given by Stolz equation
(ISO: 5167).
Co = 0.5959 + 0.0312 β 2.1 – 0.184 β 8 + 0.0029 β 2.5 (106/ReD)0.75
+ 0.09 L1 β 4 (1 – β 4)-1 – 0.0337 L2 β 3   (5.33)
where, ReD = Reynolds number based on internal diameter of pipe D

where, do = diameter of orifice

β = do/D
l1 = Distance of the upstream tapping from the upstream face of orifice plate,
mm
l2 = Distance of the downstream tapping from the downstream face of the orifice
plate, mm
Design an orifice meter based on the following data:
Name of fluid = water
Flow rate = 100 000 kg/h
Inside diameter of pipe = 154 mm (6 in, SCH-40 pipe)
Operating temperature = 32oC
Density of water at 32oC = 995.026 kg/m3
Viscosity of water at 32°C = 0.765 mPa · s or cP
Manometer fluid = Mercury
Density of Mercury at 32oC = 13 516.47 kg/m3
Solution
Inside diameter of pipe D = 154 mm
Let β = 0.5, i.e., do/D = 0.5
Diameter of orifice, do = 77 mm
ReD = Reynolds number based on inside diameter of pipe

= 300 210.2
Use square edged circular orifice plate with radius taps.
Radius taps are best from practical stand point. In radius taps, downstream tap is
located approximately at the mean location of vena contracta (1/2 times pipe
diameter) and also upstream tap is sufficiently far so that it is not affected by
distortion of flow in the vicinity of orifice.
Based on Stolz’s equation,

  (5.33)

Co = 0.5959 + 7.2776 × 10–3 – 7.1875 × 10–4 + 1.264 × 10–3 + 6 × 10–3

– 4.2125 × 10–3
= 0.6055
Mass flow rate of water

  (5.30)

Y = expansion factor for liquid = 1

Δp = 45 722.597 Pa ≡ 45.722 kPa ≡ 4.661 m WC

Δp = Rm

45 722.597 = Rm × (13 516.47 – 995.026)

Rm = Manometer reading = 0.3722 m Hg

≡ 372.2 mm Hg
In actual industrial practice, DP transmitter is used. Select the transmitter with
550 mm WC range so that the normal reading is at about 85% of the range.
Similarly, one can find different values of Rm for the different variation in flow
rate or in other words calibration of orifice meter can be carried out.
Design an orifice meter based on the following data:
Name of fluid = Chlorine gas
Flow rate = 1500 Nm3/h
Operating pressure = 1.2 atm a
Operating temperature = 30°C
Viscosity of chlorine gas at 30°C = 0.0145 mPa · s or cP
Inside diameter of pipe = 154 mm (6 in. SCH-40)
Specific heat ratio for Cl2 gas, k = 1.355
Solution
Let β = 0.5
= 0.5 or do = 77 mm

ReD =

Density of Cl2 gas at normal condition

ρ =

ṁ = 1500 × 2.902 = 4353 kg/h

ReD =

= 689 457.6
If corner taps are made then l1 = l2 = 0
Coefficient of orifice meter can be determined by Stolz’s equation.

+ 0.09L1β4(1 – β4)–1 – 0.0337L2β3  (5.33)

= 0.5959 + 7.2776 × 10–3 – 7.1875 × 10–4 + 6.7755 × 10–4
= 0.6031
Expansion factor
Y = 1 – [((1 – r)/k)(0.41 + 0.35β4)]
= 1 – [((1 – r)/1.355) × 0.4319]
For the first trial calculation let r = 0.8, Y = 0.9362
Mass flow rate of chlorine gas
   (5.30)

Density of chlorine gas at operating condition:

ρ =

Δp = 28 946.4 Pa
≡ 28.946 kPa
p1 = 1.2 atm, p2 = p1 – Δp = 1.2 × 101.325 – 28.946
p2 = 92.644 kPa
r =

Δp ∝

Δp2 =

r = = 0.7557 ≡ 0.762 (close enough to the previous value of r)

Δp2 = 29.7 kPa ≡ 3.028 m WC (final)

Similarly, for the different variation in the flow rate Δp can be determined. In
this case also, a DP transmitter having 3500 mm WC range can be selected.
Normal reading will be at about 87% of the range.

5.8 PROCESS DESIGN OF ROTAMETER

Rotameter is used for the flow measurement of liquids and gases. It is a variable
area flow meter in which pressure drop is nearly constant and the area through
which the fluid flows varies with the flow rate. It consists of tapered tube with
the smallest diameter at the bottom. The tube contains a freely moving float
which rests on a stop at the base of the tube. When the fluid is flowing the float
rises until its weight is balanced by the upthrust of the fluid, its position then
indicates the rate of flow. Float is not floating but is completely submerged in
the fluid and its density is greater than the density of fluid. The tube is marked in
divisions and the reading of the meter is obtained from the scale reading at the
reading edge of the float which is taken at the largest cross section of the float. It
is always mounted in the perfect vertical position. Rotameters are designed to
measure the flow rate of fluid in certain range like 5 m3/h to 50 m3/h, 10 m3/h to
100 m3/h, 3 Lpm to 30 Lpm, etc.
For a linearly tapered tube with a diameter at the bottom about the float
diameter, flow is almost a linear function of the height or position of float. As a
tube material glass is the first choice. For high temperatures or pressures or for
other conditions where glass is impracticable, metal tubes are used. In this case
since, in metal tube, the float is invisible, means must be provided for either
indicating or transmitting the meter reading.
Mass flow rate through rotameter is given by the equation

qm = CD A2   (5.34)

where, qm = Mass flow rate, kg/s

CD = Discharge coefficient depends on the shape of the float and the Reynolds
number though annulus. It can be obtained from literature such as Fig.
6.22 of Ref. 11. For the float having the shape same as shown in Fig. 5.15,
value of CD is almost constant for the value of Re above 500 and it is
0.775.
A2 = Cross-sectional area of annulus between float and tube, m2
A2 = A1 – Af
A1 = Cross-sectional area of tube at the float position, m2
Af = Maximum cross-sectional area of the float in a horizontal plane, m2
g = Acceleration of gravity = 9.81 m/s2 at mean sea level
Vf = Volume of float, m3
ρf = Density of float, kg/m3
ρ = Density of fluid, kg/m3
Determine the minimum and maximum flow rates of water at
40°C which can be measured by rotameter of the given dimensions. Dimensions
of tube and float of rotameter are given in Fig. 5.15.
Viscosity of water of 40°C = 0.7 cP or mPa · s
Density of water at 40°C = 992.2 kg/m3
Density of stainless steel = 8000 kg/m3
Float is made from stainless steel plate of 2 mm thickness. Calibrate the
rotameter or find the different values of flow rates for the different float
positions.

Fig. 5.15 Typical Details of Float and Tube

Solution
Volume of float

Vf =

Vf = 3.338 × 10–6 m3
Mass of float
mf =

= 0.023 25 kg
Density of float, ρf = kg/m3 < 8000 kg/m3

Maximum flow rate at top:

Mass flow rate through Rotameter is given by the equation

  (5.34)

For the first trial calculation, let discharge coefficient, CD = 0.75.

A1 = Cross-sectional area of tube
= (40)2 = 1256.6 mm2 ≡ 1256.6 × 10–6 m2

Af = Maximum cross-sectional area of the float in a horizontal plane, m2

= (25)2 = 490.87 mm2 ≡ 490.87 × 10–6 m2

A2 = Cross-sectional area of annulus between float and tube, m2

A2 = A1 – Af = 765.767 × 10–6 m2, = 0.6094

qm = 0.75 × 0.765 767 × 10–3

= 0.6441 kg/s
≡ 2318.76 kg/h ≡ 2.337 m3/h (taking ρ = 0.9922 kg/L for water)
Mass velocity through annulus
G = = 841.12 kg/(m2 · s)

Equivalent diameter
de =
= = 0.015 m

Re = = 18 024 > 500

From Fig. 6.22 of Ref. 11:

CD = 0.775
qm = × 0.644 = 0.6655 kg/s

≡ 2395.7 kg/h ≡ 2.414 m3/h (corrected)

First mark for indicating flow rate is at 27 mm diameter of tube.
A1 = (27)2 = 572.55 mm2 ≡ 572.55 × 10–6 m2

A2 = A1 – Af = 81.685 × 10–6 m2

= 0.142 67

Let CD = 0.775 (for first trial)

qm = 0.775 × 81.685 ×

= 0.056 873 kg/s ≡ 204.7 kg/h ≡ 0.2063 m3/h

G = = 696.25 kg/(m2 · s)

de = = 0.002 m

Re = = 1989.3

For Re = 1989.3, CD = 0.775 (from Fig. 6.22 of Ref. 11). Hence, no change in qm
or qv. Based on similar calculations, Table 5.11 is prepared.
Table 5.11 Flow Rate Calculation for Rotameter
Tapered tube A1 A2 A2/A1 qv m3/h
diameter, mm mm2 mm2
 25 490.87 — .— 0

27 572.55 81.685 0.142 67 0.2063

30 706.85 215.98 0.3056 0.5672

32 804.25 313.38 0.3897 0.85

34 907.92 417.05 0.4593 1.174

36 1017.876 527 0.5178 1.54

38 1134.11 643.24 0.5672 1.9528

40 1256.6 765.73 0.6094 2.4148

Let angle made by tapered tube with vertical plane is 2 degrees.

Let h = Float position in mm from the bottom-most point (i.e., from the point
where diameter of tube is 25 mm)
For qv = 2.4146 m3/h, inside diameter of tapered tube = 40 mm
Value of h at 40 mm diameter

h = = 214.77 mm

Similarly float position h can be determined for the different flow rates.
Table 5.12 Float Position Calculations for Rotameter
Diameter of tapered tube, mm qv, m3/h h, float position from bottom, mm

25 0 0

27 0.2063 28.64

30 0.5672 71.59

32 0.85 100.23

34 1.174 128.86

36 1.54 157.5

38 1.9528 186.135

40 2.4148 214.77

Draw the graph of h vs qv.

Refer Fig. 5.16.
 Fig. 5.16 Calibration of Rotameter

In Example 5.22, a rotameter is calibrated to measure the flow

rate of water at 40°C. If the same rotameter is used for measuring the flow rates
of the following fluids then what is the change in calibration?
(a) Monoethylene glycol (MEG) brine containing 60% MEG by mass12
Operating temperature: –25°C
Operating pressure: 3 bar g
Density: 1098 kg/m3
Viscosity: 40 mPa · s (or cP)
(b) Ethyl alcohol
Operating temperature: 35oC
Pressure: 3 bar g
Density: 781 kg/m3
Viscosity: 0.9 mPa · s (or cP)
Solution
At the topmost position of the float,
h = 214.77 mm, diameter of tapered tube = 40 mm,
A1 = 1256.6 mm2, A2 = 765.73 mm2, = 0.6094

For this position of float, flow rate of MEG:

qm = CD A2   (5.34)

For the first trial calculations, let CD = 0.775

qm = 0.775 × 765.73 × 10–6

= 0.6939 kg/s
or qv = × 3600 = 2.275 m3/h

Mass velocity, G = = 906.194 kg/(m2 · s)

Re = = 339.82

Value of CD = 0.77 (Fig. 6.22 of Ref. 11)

qm = × 0.6939 = 0.6894 kg/s

or qv = 2.26 m3/h
For the same position of float, flow rate of ethyl alcohol:
Let CD = 0.775 (for the first trial calculations)

qm = 0.775 × 765.73 × 10–6

= 0.6 kg/s
qv = × 3600 = 2.77 m3/h
Mass velocity, G = = 783.57 kg/(m2 · s)

Re = = 13 059.5

CD = 0.775 (From Fig. 6.22 of Ref.: 11)

No change in the values of qm and qv.
At diameter of tapered tube = 27 mm,
Float position, h = 28.64 mm,
A1 = 572.55 mm2, A2 = 81.685 mm2, A2/A1 = 0.142 67
Flow rate of MEG: For the first trial, let CD = 0.77

qm = 0.77 × 81.685 × 10–6

qm = 0.059 kg/s

qv = × 3600 = 0.193 44 m3/h

G = = 722.287 kg/(m2 · s)

Re = = 36.114

CD = 0.7 (from Fig. 6.22, Ref. 11)

qm = × 0.059 = 0.0536 kg/s

qv = 0.1757 m3/h
For the same position of float, flow rate of ethyl alcohol:
Let CD = 0.775 (for the first trial calculation)

qm = 0.775 × 81.685 × 10–6

qm = 0.051 34 kg/s
qv = 0.2367 m3/h
G = = 628.5 kg/(m2 · s)

Re = = 1396.67

CD = 0.775 (From Fig. 6.22 of Ref. 11)

Hence qm and qv remain unchanged.
By similar calculations, following table is prepared.
Table 5.13 Flow Rates of Different Fluids for the Same Position of Float
Float position h, mm Diameter of tapered tube D, mm qv m3/h

Water MEG solution Ethyl alcohol

0 25 0 0 0
28.64 27 0.2063 0.1757 0.2367
71.59 30 0.5672 0.5274 0.65
100.23 32 0.85 0.8 0.976
128.86 34 1.106 1.348
1.174
157.5 36 1.54 1.45 1.766
186.135 38 1.9528 1.84 2.24
214.77 40 2.4148 2.26 2.77

Following table is made from the graph (Fig. 5.16).

Table 5.14 Calibration for Rotameter
Float Position, h mm
qv, m3/h
Water MEG Solution Ethyl Alcohol

0.1 14.31 15.94 11.828

0.2 27.8 31.02 23.66
0.4 52.67 55.8 45.224
0.6 75.06 79.22 66.28
0.8 95.21 100.00 84.77
1.0 113.81 118.94 102.08
1.2 131.0 137.37 117.47
1.4 146.93 155.47 132.73
1.6 161.89 170.05 147.63
1.8 175.96 183.46 161.88
2.0 189.19 196.67 172.9
2.2 201.81 209.83 183.93
2.4 213.91 — 194.78

Comments: It can be seen from Table 5.13 that rotameter calibration is

essentially immune from the viscosity effect (say up to about 100 mPa · s or
cP). However, calibration is significantly affected by the density of the fluid
which changes drag force.
In actual practice, rotameters are tested for actual water capacities on a test
bench to confirm the calibration.

5.9 TWO-PHASE FLOW

In many applications, two-phase flow is experienced in industry. If pipe sizing is
not done properly, it can result in serious problems. A few such problems are
illustrated as follows:
1. When a saturated liquid under pressure is flashed through a control valve,
two-phase flow will be realized in the downstream piping. Pressure drop in
the two-phase flow could be much higher (several fold in some cases) in the
pipe. This will result in much higher back pressure on the control valve,
restricting its capacity. Apart from excessive pressure drop, there could be
erosion in the pipeline.
2. When vapour and liquid flow in upward direction and the line is sized for low
pressure drop, it may cause slug flow, i.e., slugs of liquid flow in vapour. This
can result in pressure pulsation and severe vibration. This phenomenon is
commonly observed and is known as water hammer (refer to Sec. 5.9.2) in
steam lines, particularly during cold start-up when steam condenses during
travel and water slugs flow with steam. Water hammers can break open the
flanges or damage the supports. Adequate draining facility is provided on
long steam pipelines to avoid such water hammers.
3. If cooling water is flowing downward at temperature in excess of 40oC from a
height of 10.5 m or more to an atmospheric pond, vibrations could result due
to vapour formation at the top and two phase formation in downflow pipe. A
provision of appropriate size orifice at the base of the pipe can solve the
problem.
A process engineer should visualize such a problem and design the piping
suitably so that both the phases flow smoothly without creating any problem.
Estimation of pressure drop in an isothermal two-phase system can be done with
empirical relations, developed by Martinelli’s and Dukler’s correlations.
Simpson7 has dealt at a length on both the correlations (refer to Sec. 5.9.1)
A common problem of two-phase flow is experienced in steam condensate lines.
When steam is used for heating in a heat exchanger, its condensate is discharged
through a steam trap. While the heat exchanger operates under pressure, the trap
discharges condensate to low pressure. In most cases, the condensate line opens
into an atmospheric storage tank while in a few instances, condensate is
conveyed to a deaerator, operating under positive pressure. In both cases,
flashing of condensate takes place which calls for larger size pipe. If not
designed properly, water hammer could be anticipated and also the back pressure
on the trap could be much higher than anticipated. Steam traps are sized and
rated for a specific pressure drop and condensate flow discharge rate. If the
pressure drop is less than design, trap’s capacity will be reduced and condensate
accumulation is expected in the heat exchanger, thereby heat transfer will be
affected.
Empirical correlations could be used for calculating pressure drop in steam
condensate pipes. However, pipe sizing could be done with the help of Fig. 5.17
which is based on approximate pressure drop of 0.1 kPa/m (or 1 mbar/m) length
of pipe. This represents only friction loss. To this, head loss and equipment
pressure (such as that of deaerator) are to be added to arrive at a back pressure of
the trap.
Saturated steam at 10 bar g is used in a heat exchanger at the
rate of 2000 kg/h. Condensate is discharged through a trap and taken to a
deaerator, located at 10 m above the heat exchanger. Pipe length, including
equivalent length of fittings, is calculated to be 60 m. Deaerator operates at 0.4
bar g. Size the condensate pipe and calculate the back pressure on the trap.
Solution
Assume back pressure of 2 bar on the trap.
Steam at 10 bar g has saturation temperature of 184.13°C and condensate has
specific enthalpy of 781.6 kJ/kg (from Steam Tables). At 2 bar g, specific
enthalpy of condensate and flash vapours will be 562.2 kJ/kg and 2725.5 kJ/kg,
respectively. If ṁ is the flash steam rate after the trap, enthalpy balance can be
written as
2000 × 781.6 = ṁ × 2725.5 + (2000 – ṁ) 562.2
ṁ = 202.8 kg/h
This represents 10.14% flashing. Specific volume of flash steam at 2 bar g and
133.7°C is 0.603 m3/kg.
Thus volumetric flow rate of flash steam will be 122.29 m3/h.
From Fig. 5.17, pipe size for 25 m/s flash steam velocity is read between 40 and
50 mm NB. Select 50 mm NB condensate pipe, downstream of trap which could
be of 25 mm size.
Fig. 5.17 Steam Condensate Line Sizing Chart (Reproduced with the Permission of Spirax-Sarco Ltd., UK)

Friction loss in pipe = 0.1 × 60 = 6 kPa ≡ 0.06 bar

Deaerator pressure = 0.4 bar
Elevation of deaerator = 10 m = 0.98 bar
Total back pressure = 0.06 + 0.4 + 0.98 = 1.44 bar
Calculated back pressure is less than the assumed back pressure. Condensate
pipe design should therefore be satisfactory.
Refrigerants are another class of fluids which produce flash vapours on pressure
reduction of saturated liquids. Quantity of flash vapour normally varies from 15
to 40%, depending on the refrigerant and the pressure difference across the
expansion valve. To avoid any problem of high pressure drop, the expansion
valve is normally situated close to the evapourator.
Charts, such as Fig. 5.17, are not readily available in literature for other fluids.
Hence, empirical equations for two phase flow are to be used for pressure drop
estimation.
5.9.1 Generalized Approach to Two-Phase Flow
For the cocurrent two-phase flow in a horizontal pipe, Lockhart and Martinelli
correlations13 are used for pressure drop calculations. These can be written as
shown below.
  (5.35a)

  (5.35b)

where (Δp/L)l or (Δp/L)g can be calculated using the Fanning equation [Eq.
(5.2)] by assuming that each phase is flowing alone in the pipe. Raynolds
number (Re) and pressure drop (Δp) are determined based on superficial velocity
of either phase.
Yl = f1(X) and Yg = f2(X)   (5.36)

where, X =   (5.37)

Based on foregoing equations, it can be deduced that

Yg = X2 Yl  (5.38)
Yg and Yl can be obtained from Fig. 5.18 in which 3 separate curves are
shown for Yg vs X and Yl vs X. These three curves represent the following cases:
1. Flows of liquid and gas are in turbulent regions.
2. Flow of one phase is turbulent while another phase is laminar, i.e., viscous.
3. Flows of both phases are in laminar regions.
If Re ≤ 2000, flow is considered laminar and if Re > 2000, flow is considered
turbulent. These correlations can be applied to pipe diameters up to 100 mm (4
in) NB. Based on actual measurements, pressure drops, predicted by the above
correlations are high for stratified, wavy and slug flows and low for laminar
flow.

Fig. 5.18 Parameters for Pressure Drop in liquid-gas flow through horizontal pipes. [Reproduced with the
permission of McGraw-Hill Book Company, [USA]

5.9.2 Water Hammer

It is a hydraulic shock which occurs in pipe whenever there is a sudden change
in the velocity of liquid, flowing in the pipe. Conditions that can result in water
hammer can be as given below.
1. Rapid closing of a valve. A control valve may shut instantly (may be in 1 or 2
seconds).
2. Start/stop or an abrupt change in pump speed.
In water hammer, kinetic energy of moving liquid is transformed into pressure
energy due to sudden stoppage or abrupt change in velocity. Such a pressure rise
in piping often results in severe mechanical damage.
If water is flowing through a pipeline then maximum pressure developed by
water hammer is given by the following equation:
  (5.39)

where,
hwh = Maximum pressure developed by water hammer, m LC
Vw = Reduction in velocity, m/s
g = Gravitation acceleration = 9.81 m/s2
aw = Velocity of propagation of elastic vibration in the discharge pipe, m/s

=   (5.40)

R = Ratio of inside pipe diameter-to-wall thickness

Ka = Ratio of elastic modulus of water to the pipe material
Table 5.15 Values of Ka
Pipe material Ka
Steel 0.01
Copper 0.017
Alluminum 0.03

Time interval (θs) required for the pressure wave to travel back and forth in the
pipe is
θs =   (5.41)

where, θs = time, s
L = Length of pipe, m
When the actual shut-off time for the valve (or pump stoppage) is shorter than θs,
then maximum pressure equal to hwh will be exerted on the closed system or
pipeline which can cause water hammer.
A 200 mm (8 in) NB SCH-40 carbon steel pipe is used for
transferring water from reservoir to a user plant. Capacity of pump is 250 m3/h.
Length of discharge pipe from pump to control valve (including equivalent
length of fittings) is 800 m. Calculate the maximum pressure, developed in pipe,
when the control valve is suddenly closed.
Solution
Thickness of 200 mm NB SCH-40 pipe, t = 8.18 mm
Inside diameter of pipe = 202.72 mm
R =

Vw = = 2.153 m/s

aw =

hws =

≡ 19.3757 atm
Total pressure in piping:
p = 19.3757 atm + discharge pressure of pump
Time interval for pressure wave travel:
θs =

If actual shutoff time of the control valve is less than 1.75 s then water hammer
pressure can be equal to p.

5.10 TROUBLESHOOTING OF FLUID FLOW

SYSTEMS
Utilities such as cooling water, steam, refrigeration, etc., are important to process
industry. Design of utility network (Chapter 11) should therefore be carried out
carefully so that no user equipment starves of any utility. This can also result in
energy saving.
Apart from utilities, piping of process fluids also require careful design. Piping
should neither offer excessive pressure drop nor should be overdesigned as
initial cost could be a deterrent.
In many instances, problems might arise in actual operation due to various
reasons. Plant might be operated at a higher capacity thereby pumping capacity
might become limiting, higher pressure drop might be experienced in pipe line,
etc. In another example, fluid distribution in manifold piping might not be
satisfactory. In yet another instance, a new boiler of higher capacity is added and
steam from the new boiler is carried in the existing pipelines. In all these
situations, process engineer is required to evaluate the bottlenecks and carry out
troubleshooting exercise.
Take the instance of installation of new higher capacity boiler. If pressure drop
calculations reveal high pressure drop in the existing steam pipeline, process
engineer can consider laying of another parallel pipeline from the boiler to
nearest steam consuming equipment. Alternately, steam at a higher pressure is
conveyed from the boiler if the existing pipeline rating permits its operation at
higher pressure.
Instrument air manifold becomes a bottleneck in many situations as more
instruments are added over a period of operation. Also, fluctuations in supply
instrument air pressure could be a problem. After careful evaluation, installation
of an air receiver of adequate size near the plant may prove valuable.
For pumping system, upgradation of existing centrifugal pumps offers many
options. While evaluating the options, safety, reliability, ease of maintenance and
cost benefit analysis must be considered. Chaware et al.14 have dealt in detail the
upgrading options of a centrifugal pump. Impeller replacement, operation of two
pumps in parallel, installation of a bigger pump, change of a type of a pump,
etc., are the options that can be considered. In small and medium scale industry,
suction pipe of the centrifugal pump is observed smaller or equal to suction
nozzle size. If suction pipe size is selected one size higher than the nozzle size,
pump capacity is observed to be in the design range. Process engineer can
contribute significantly in troubleshooting.

EXERCISES
5.1 Answer the following:
(a) At 7 bar g, the leakage of steam through 3 mm, 6 mm and 12.5 mm diameter
holes amounts to 34, 230 and 540 kg/h, respectively. One of the leakage
figures is a typographical error. Which figure? Why?
(b) A two-stage reciprocating compressor is operating at 0.7 bar a inlet and 6.2
 bar a outlet pressure. If the pressure drop across the intercooler is 0.14 bar,
the pressure at the inlet of second stage will be approximately; (a) 3.38 bar
a, (b) 2.69 bar a, or (c) 2.0 bar a.
(c) If volumetric flow rate and head of a centrifugal pump are proportional to
speed raised to power 0.8 and 1.5, respectively (for a given specific speed),
then power would be proportional to speed raised to power: (a) 3, (b) 2.3,
(c) 2.0, (d) 1.5 or (e) 1.7.
(d) Pressure drop in a 150 mm NB diameter pipe in a well-designed distribution
system is 0.07 bar due to water flowing at a certain rate. What will be the
pressure drop in 80 mm NB pipe for the same length and same flow rate?
(a) 4.4 bar, (b) 1.62 bar, (c) 1.1 bar, (d) 0.55 bar, (e) 0.276 bar or (e) 0.138
bar.
(e) An orifice type flow meter, installed on a liquid stream is coupled to a
differential pressure (DP) transmitter and a flow recorder. Normal reading
is 5 on a 0–10 square-root scale and 25 on a 0–100 linear scale. If the range
of the DP transmitter is doubled, the reading will be
(i) 2.5 on square-root scale and 6.3 on linear scale.
(ii) 2.25 on square-root scale and 12.5 on linear scale or
(iii) 7.1 on square-root scale and 50 on linear scale.
(f) For a centrifugal pump, power formula can be written as P = 80 K (d)3
where, P = power in hp
K = dimensionless pump constant
d = impeller diameter, in
To write this formula in metric units, P should be in kW and d should be in cm.
What will be the constant (instead of 80)?
5.2 A saturated gas mixture, obtained from a coal liquefaction plant after the acid
removal section, has composition of 38.2% H2, 12.3% CO, 18% CO2, 31.0%
CH4 and 0.5% N2 (by mole) on dry basis. It is available at 120 kPa a and
60°C. It is cooled to 40°C in a heat exchanger and sent to suction drum of a
compressor. Condensate from the drum is to be continuously pumped out
with the help of a centrifugal pump. Calculate the volume fraction of the gas
mixture over the condensate and the artificial vapour pressure using ideal gas
low.
Table 5.16 Solubility of Gases9 in Water at 1 atm and 40°C
Gas Solubility, kg/100 kg water

H2 0.000 1384

CO 0.002 075
 CO2 0.0973

CH4 0.001 586

N2 0.001 391

5.3 Natural gas is to be delivered through 0.6 m diameter steel pipe. Straight
length of pipe is 3 km. Equivalent length of fittings and valves of this pipe
line is 300 m. Desired flow rate of natural gas is 180 000 Nm3/h. For the
purpose, a centrifugal compressor is installed at the Gas Gathering Station
(GGS). Natural gas is to be delivered at the plant at 313 K and 170 kPa a and
it leaves the compressor at 313 K. What pressure must be developed at the
compressor discharge in order to achieve this flow rate? Average molar mass
of natural gas is 17.4 and its viscosity at 313 K is 0.01 cP.
5.4 It is proposed to pump 5000 kg/h of acetic acid at 129.5°C and 1.4 atm a
pressure from the reboiler of a distillation column to a hold up tank without
cooling. If the friction loss in the line between the reboiler and pump is 8 kPa
and the density of acetic acid is 1048 kg/m3, how far above the pump must the
liquid level in the reboiler be maintained to give a net positive suction head of
2.5 m? Friction loss in discharge line is 40 kPa. Maximum liquid level of
acetic acid above the discharge point is 5 m. Pump efficiency is 60%.
Calculate the power required to drive the pump.
5.5 Performance of centrifugal pump on water is known. It can deliver 100 m3/h
of water at 80 m total head of water with 65% efficiency. By using Fig. 5.7,
predict the performance of this pump, if it is used to deliver the oil of
viscosity 176 cSt (or 800 SSU) at the pumping temperature. Determine the
head, capacity and power required by this pump when it is used to deliver oil.
Density of oil at pumping temperature is 800 kg/m3.
5.6 Hexane at 37.8°C is pumped through the system at a rate of 9.09 m3/h. The
tank is at atmospheric pressure. Pressure at the end of discharge line is 345
kPa g. The discharge head is 3.05 m and the suction lift is 1.22 m above the
level of liquid in the tank. The friction loss in suction line is 3.45 kPa and that
in the discharge line is 37.9 kPa. The mechanical efficiency of the pump is
0.6. The density of hexane is 659 kg/m3 and its vapour pressure at 37.8°C is
33.71 kPa.
Calculate (a) (NPSH)A, and (b) power required by centrifugal pump.
5.7 In a carbon dioxide liquefaction facility to liquefy 25 000 kg/day, it is
condensed at –20°C and stored in a bullet. Before liquefaction, gas is
compressed from atmospheric pressure and 35°C to 20 bar a, dried and sent to
a chiller. Reciprocating compressors are selected for the duty. One engineer
 decides to use a two stage compressor while another decides to use a three
stage compressor. Considering 75% overall mechanical efficiency of either
compressor and equal compression ratio in each stage, calculate the power
requirement in each case. Also calculate temperature of the gas at the
discharge of each stage.
Data: Specific heat ratio, k for CO2 = 1.3
5.8 Design an orifice meter based on the following data:
Name of fluid = Natural gas
Flow rate = 1000 Nm3/h
Operating pressure = 1.35 atm a
Operating temperature = 38°C
Average molar mass of natural gas = 17.2
Viscosity of natural gas = 0.01 cP at 38°C
Inside diameter of pipe = 128.2 mm (5 in, SCH-40)
Specific heat ratio for natural gas (k) = 1.3
5.9 Design a venturi meter based on the data given in Example 5.21.
Mass flow rate through venturi meter is calculated by the following equation:
  (5.42)

where, Cv = Venturi coefficient = 0.98

At = Throat area of venturi =
Dt = Diameter of throat, m
β = Ratio of diameter of throat to diameter of pipe
gc = 1
Δp = Pressure drop measured by manometer, Pa
ρ = Density of flowing fluid, kg/m3
Angle of upstream cone of venturi meter is 30° and the same of downstream
cone is 10°. Take β = 0.5. Draw a sketch of suggested venturi.
Determine: (a) Throat diameter of venturi, (b) Pressure drop and manometer
reading, and (c) Lengths of upstream and downstream cones.
5.10 Determine the minimum and maximum flow rates of air at 40°C and at 1.5
atm a, which can be measured by the rotameter of the dimension of tube and
float of rotameter, given in Fig. 5.15, except material of float is
polypropylene.
Viscosity of air at 40°C = 1850 × 10–5 cP
Density of polypropylene = 905 kg/m3
Thickness of float = 2 mm
Calibrate the rotameter or find the different values of flow rates for the different
float positions.
5.11 In Example 5.22, a rotameter is calibrated to measure the flow rate of water
at 40°C. If the same rotameter is used for measuring flow rates of glycerine at
30°C then what is the change in calibration?
Viscosity of glycerine at 30°C = 400 cP
Density of glycerine at 30°C = 1.25 kg/L
5.12 In a single stage ammonia-water absorption refrigeration plant, pure
saturated liquid ammonia is flashed from 15.55 bar a to 2.90 bar a through a
throttling valve. On flashing, 18.3% of liquid ammonia is converted to
vapour. If the distance between the throttling valve and evapourator (chiller)
is 2 m, determine the pressure drop for two-phase flow in the horizontal line.
Flow of saturated liquid ammonia (at upstream) = 0.1935 kg/s
Inside diameter of the pipe = 100 mm
Temperature of liquid-vapour mixture after flashing = –10°C.
At –10°C, density of liquid ammonia is 639 kg/m3 and viscosity values of liquid
and vapour ammonia are 0.15 cP and 0.0095 cP, respectively.

REFERENCES
1. Sinnott, R. K., Coulson & Richardson’s Chemical Engineering, Vol. 6, Asian
Book Private Ltd., Revised 2nd Ed., New Delhi, 1998.
2. Perry, R. H., and Green D, Perry’s Chemical Engineer’s Handbook, 6th Ed.,
McGraw-Hill Education, USA, 1984.
3. Techo, R., Tinker, R. R., and R. E. James, J. Appl. Mech., 32, 1965, p. 443.
4. Ludwig, E. E., Applied Process Design for Chemical and Petrochemical
Plants, Gulf Publishing Company, Vol. 1, 3rd Ed., USA, 1999.
5. Flow of Fluids Through Valves, Fittings and Pipe, Crane Co., New York,
USA, 1977.
6. Brown G. G., Unit Operations, John Wiley and Sons, Inc, USA, 1950.
7. Simpson, L. L., Chem. Engg., 75 (13), June 17, 1968, p. 190.
8. Hydraulic Institute Standards, 13th Ed., Hydraulic Institute, USA.
9. Tsai, M. J., Chem. Engg., 89 (15), July 26, 1982, p. 65.
10. McCabe, W. L., J. C. Smith, and P. Harriott, Unit Operations of Chemical
Engineering, 7th Ed., McGraw-Hill Education, USA, 2005.
11. Coulson, J. M., and J. F. Richardson, Chemical Engineering, Vol. 1, Butter
Worth-Heinemann, 6th Ed., UK, 1999.
12. Carrier System Design Manual: Refrigerants, Brines, Oils by Carrier
Corporation, USA, 1969.
13. Lockhart, R. W., and R. C. Martinelli, Chem, Engg. Progress, 45 (1), 39,
1949.
14. Chaware, D. K. Shukla, and R. B. Swamy, Hydrocarbon Processing, 84 (5),
2005, p. 47.
 6
Process Design of Heat Exchangers
6.1 INTRODUCTION
Second Law of Thermodynamics states that heat flows from higher temperature
to lower temperature. All chemical reactions and unit operations involve heat
transfer. Solids, liquids and gases require heating or cooling during various
operations that are performed to get the desired product. This transfer can be
performed in a variety of equipments called heat exchangers. Each of the
equipments offer advantages and have limitations for a given service. This
chapter is devoted to selection and process design of various types of heat
exchangers and their integration in the process.
Shell- and tube-type heat exchangers are most common in the process industry.
Plate heat exchangers, spiral heat exchangers, brazed finned-type heat
exchangers, etc., are also used for specific applications. It is intended to cover all
these type of heat exchangers in this chapter.

6.2 SHELL AND TUBE HEAT EXCHANGERS

Shell and tube heat exchangers are the most widely used equipments in chemical
industry. This is because segmented baffles make such a heat exchanger sturdy
enough for industrial use as longitudinal tube bundle when inserted in the shell
gets support. They are mostly used as heat transfer equipments but in a few cases
they are also used as Reactors, Falling Film Absorbers, etc. Sizes of various
parts of shell and tube heat exchangers like shell, tubes, tierods are standardized.
Standards developed by Tubular Exchanger Manufacturers Association, USA
(TEMA) are universally used for design of shell and tube heat exchangers.
Equivalent Indian Standard is IS:4503. TEMA or IS:4503 specify standard sizes
of shell, tubes, tierods, etc., and also maximum allowable baffle spacing,
minimum tube sheet thickness, baffle thickness, numbers of tierods required, etc.
For mechanical design and fabrication, in addition to TEMA standard, ASME
Code Section VIII DIV. I or IS:2825 is used. In TEMA standard, shell and tube
heat exchangers are classified in three categories:
1. Class R covers heat exchangers which are used for severe duties in petroleum
and related industries. Also it covers heat exchangers that are going to handle
toxic gas, highly flammable fluid or hazardous fluid.
2. Class B covers the heat exchangers which are used in chemical process
industries not involving severe duties.

(a) Internal Floating Head Heat Exchanger with Backing Ring, Type AES

(b) Fixed Tube Sheet Heat Exchanger, Type BEM

(c) Outside-Packed Floating Head Heat Exchanger, Type AEP

 (d) U-Tube Heat Exchanger, Type CFU

(e) Kettle-type Floating Head Reboiler, Type AKT

(f) Exchanger with Packed Floating Tube sheet and Lantern Ring, Type AJW
Nomenclature of Parts of Various Shell and Tube Heat Exchangers
1. Stationary Head—Channel
2. Stationary Head—Bonnet
3. Stationary Head Flange—Channel or Bonnet
4. Channel Cover
5. Stationary Head Nozzle
6. Stationary Tube sheet
 7. Tubes
8. Shell
9. Shell Cover
10. Shell Flange—Stationary Head End
11. Shell Flange—Rear Head End
12. Shell Nozzle
13. Shell Cover Flange
14. Expansion Joint
15. Floating Tube sheet
16. Floating Head Cover
17. Floating Head Flange
18. Floating Head Split Backing Ring
19. Split Shear Ring
20. Slip-on Backing Flange
21. Floating Head Cover—External
22. Floating Tube sheet Skirt
23. Packing Box Flange
24. Packing
25. Packing Gland
26. Lantern Ring
27. Tie rods and Spacers
28. Transverse Baffles or Support Plates
29. Impingement Plate
30. Longitudinal Baffle
31. Pass Partition
32. Vent Connection
33. Drain Connection
34. Instrument Connection
35. Support Saddle
36. Lifting Lug
37. Support Bracket
38. Weir
39. Liquid Level Connection
Fig. 6.1 Different Types of Shell and Tube Heat Exchangers1 (Fig. N-2 of TEMA 8th Edition), Reproduced
with the Permission of the Tubular Exchanger Manufacturers Association Inc., USA

3. Class C covers the heat exchangers which are used in commercial and in less
important process applications. Example: Heat exchangers used for recovery
of energy from an effluent stream.
Most popular and reliable softwares used for the design of shell and tube heat
exchangers are as follows:
1. HTRI: Heat Transfer Research Inc., USA
2. HTFS: Heat Transfer and Fluid Flow Services, UK
3. BJAC: USA based company
4. HEI: Heat Exchange Institute, USA.
Design methods and equations used by these softwares are not available in open
literature. For the design of shell and tube heat exchanger involving fluid
without phase change, methods used by these softwares are based on Tinker’s
flow model (Sec. 6.11).

6.3 VARIOUS PARTS OF SHELL AND TUBE HEAT

EXCHANGERS
6.3.1 Shell
Shell is the costliest part of the heat exchanger. Cost of shell and tube heat
exchanger sensitively changes with change in the diameter of shell. As per the
TEMA standard, shell size ranges from 6 in (152 mm) to 60 in (1520 mm).
Standard pipes are available up to 24 in size (600 mm NB). If shell size is
greater than 24 in, it is fabricated by rolling a plate.
Shell diameter depends on tube bundle diameter. For fixed tube sheet shell and
tube heat exchanger, the gap between shell and tube bundle is minimum, ranging
from 10 to 20 mm. For pull through floating head heat exchanger, it is
maximum, ranging from 90 to 100 mm.
Required shell diameter is determined by one of the following three methods:
1. Based on the actual tube sheet layout drawing: All latest computer
softwares available for the design of heat exchangers use this method to find
shell inside diameter. Popular computer program, written in Autolips, is used to
draw the tube sheet layout. To draw the tube sheet layout or to find the tube
bundle diameter, total six parameters are fixed: (i) Outside diameter (OD) of
tube, (ii) Numbers of tube side passes, (iii) Tube pitch (center-to-center distance
between adjacent tubes), (iv) Tube arrangement (triangular, square, etc.), (v)
Total number of tubes, and (vi) Type of shell and tube heat exchanger (fixed tube
sheet, U-tube or floating head).
2. Use of standard tables: Based on actual tube sheet layout, standard tables are
prepared. These tables give the maximum number of tubes that can be
accommodated in various standard sizes of shell ranging from 6 in (150 mm NB)
to 120 in (3000 mm NB), for the different numbers of tube side passes, different
values of tube pitch and size and for the different arrangements of tubes. Table
6.1 gives the data on tube sheet layouts.
Table 6.1 Tube Sheet Tube Hole Count2,3

Table 6.1 (a) 5/8 in OD tubes on 13/16 in square pitch

Shell ID TEMA P or S Number of passes TEMA U Number of passes
 mm in 1 2 4 6 2 4 6

203 8 55 48 34 24 52 40 32
254 10 88 78 62 56 90 80 74
305 12 140 138 112 100 140 128 108
337 13 ¼ 178 172 146 136 180 164 148
387 15 ¼ 245 232 208 192 246 232 216

438 17 ¼ 320 308 274 260 330 312 292

489 19 ¼ 405 392 352 336 420 388 368
540 21 ¼ 502 484 442 424 510 488 460
591 23 ¼ 610 584 536 508 626 596 562
635 25 700 676 618 600 728 692 644

686 27 843 812 742 716 856 816 780

737 29 970 942 868 840 998 956 920
787 31 1127 1096 1014 984 1148 1108 1060
838 33 1288 1250 1172 1148 1318 1268 1222
889 35 1479 1438 1330 1308 1492 1436 1388

940 37 1647 1604 1520 1480 1684 1620 1568

991 39 1840 1794 1700 1664 1882 1816 1754
1067 42 2157 2112 2004 1968 2196 2136 2068
1143 45 2511 2458 2326 2288 2530 2464 2402
1219 48 2865 2808 2686 2656 2908 2832 2764

1372 54 3656 3600 3462 3404 3712 3624 3556

1524 60 4538 4472 4310 4256 4608 4508 4426

Table 6.1 (b) 3/4 in OD tubes on 15/16-in triangular pitch

Shell ID TEMA L or M Number of passes TEMA P or S Number of passes TEMA U Number of passes

mm in 1 2 4 6 1 2 4 6 2 4 6

114 4.5 14 14 — — — — — — — — —
152 6 30 30 24 22 — — — — — — —
203 8 64 48 34 24 34 32 16 18 32 24 24
254 10 85 72 52 50 60 62 52 44 64 52 52

305 12 122 114 94 96 109 98 78 68 98 88 78

337 13 ¼ 151 142 124 112 126 120 106 100 126 116 108
387 15 ¼ 204 192 166 168 183 168 146 136 180 160 148
438 17 ¼ 264 254 228 220 237 228 202 192 238 224 204
489 19 ¼ 332 326 290 280 297 286 258 248 298 280 262
540 21 ¼ 417 396 364 348 372 356 324 316 370 352 334
591 23 ¼ 495 478 430 420 450 430 392 376 456 428 408
635 25 579 554 512 488 518 498 456 444 534 500 474

686 27 676 648 602 584 618 602 548 532 628 600 570
737 29 785 762 704 688 729 708 650 624 736 696 668
 787 31 909 878 814 792 843 812 744 732 846 812 780
838 33 1035 1002 944 920 962 934 868 840 978 928 904
889 35 1164 1132 1062 1036 1090 1064 990 972 1100 1060 1008

940 37 1304 1270 1200 1168 1233 1196 1132 1100 1238 1200 1152
991 39 1460 1422 1338 1320 1365 1346 1266 1244 1390 1336 1290
1067 42 1703 1664 1578 1552 1611 1580 1498 1464 1632 1568 1524
1143 45 1960 1918 1830 1800 1875 1834 1736 1708 1882 1820 1770
1219 48 2242 2196 2106 2060 2132 2100 1998 1964 2152 2092 2044

1372 54 2861 2804 2682 2660 2730 2684 2574 2536 2748 2680 2628
1524 60 3527 3476 3360 3300 3395 3346 3228 3196 3420 3340 3286
1676 66 4292 4228 4088 4044
1829 72 5116 5044 4902 4868
1981 78 6034 5964 5786 5740

2134 84 7005 6934 6766 6680

2286 90 8093 7998 7832 7708
2438 96 9203 9114 8896 8844
2743 108 11696 11618 11336 11268
3048 120 14459 14378 14080 13984

Table 6.1 (c) 3/4 in OD tubes on 1 in square pitch

Shell ID TEMA P or S Number of passes TEMA U Number of passes

mm in 1 2 4 6 2 4 6

152 6 12 12 — — 14 — —
203 8 28 26 16 12 28 24 12
254 10 52 48 44 24 52 44 32
305 12 80 76 66 56 78 72 70
337 13 ¼ 104 90 70 80 96 92 90
387 15 ¼ 136 128 128 114 136 132 120

438 17 ¼ 181 174 154 160 176 176 160

489 19 ¼ 222 220 204 198 224 224 224
540 21 ¼ 289 272 262 260 284 280 274
591 23 ¼ 345 332 310 308 348 336 328
635 25 398 386 366 344 408 392 378

686 27 477 456 432 424 480 468 460

737 29 554 532 510 496 562 548 530
787 31 637 624 588 576 648 636 620
838 33 730 712 682 668 748 728 718
889 35 828 812 780 760 848 820 816

940 37 937 918 882 872 952 932 918

991 39 1048 1028 996 972 1056 1044 1020
1067 42 1224 1200 1170 1140 1244 1224 1212
 1143 45 1421 1394 1350 1336 1436 1408 1398
1219 48 1628 1598 1548 1536 1640 1628 1602

1372 54 2096 2048 2010 1992 2108 2084 2068

1524 60 2585 2552 2512 2476 2614 2584 2558

Table 6.1 (d) 3/4 in OD tubes on 1 in triangular pitch (most popular arrangement)
Shell ID TEMA L or M Number of passes TEMA P or S Number of passes TEMA U Number of passes
mm in 1 2 4 6 1 2 4 6 2 4 6

114 4.5 14 14 — — — — — — — — —
152 6 26 26 24 22 14 14 — — 14 12 10
203 8 42 40 26 24 31 26 16 12 32 24 24
254 10 73 66 52 44 56 48 42 40 52 48 40

305 12 109 102 88 80 88 78 62 68 84 76 74

337 13 ¼ 136 128 112 102 121 106 94 88 110 100 98
387 15 ¼ 183 172 146 148 159 148 132 132 152 140 136
438 17 ¼ 237 228 208 192 208 198 182 180 206 188 182
489 19 ¼ 295 282 258 248 258 250 228 220 226 248 234
540 21 ¼ 361 346 318 320 320 314 290 276 330 316 296
591 23 ¼ 438 416 382 372 400 384 352 336 400 384 356
635 25 507 486 448 440 450 442 400 392 472 440 424

686 27 592 574 536 516 543 530 488 468 554 528 502
737 29 692 668 632 604 645 618 574 556 648 616 588
787 31 796 774 732 708 741 716 666 648 744 716 688
838 33 909 886 836 812 843 826 760 740 852 816 788
889 35 1023 1002 942 920 950 930 878 856 974 932 908
940 37 1155 1124 1058 1032 1070 1052 992 968 1092 1056 1008
991 39 1277 1254 1194 1164 1209 1184 1122 1096 1224 1180 1146
1067 42 1503 1466 1404 1372 1409 1378 1314 1296 1434 1388 1350
1143 45 1726 1690 1622 1588 1635 1608 1536 1504 1652 1604 1560
1219 48 1964 1936 1870 1828 1887 1842 1768 1740 1894 1844 1794
1372 54 2519 2466 2380 2352 2399 2366 2270 2244 2426 2368 2326
1524 60 3095 3058 2954 2928 2981 2940 2932 2800 3006 2944 2884
1676 66 3769 3722 3618 3576
1829 72 4502 4448 4324 4280
1981 78 5309 5252 5126 5068

2134 84 6162 6108 5954 5900

2286 90 7103 7040 6898 6800
2438 96 8093 8026 7848 7796
2743 108 10260 10206 9992 9940
3048 120 12731 12648 12450 12336

Table 6.1 (e) 1 in OD tubes on 11/4 in square pitch

 Shell ID TEMA P or S Number of passes TEMA U Number of passes

mm in 1 2 4 6 2 4 6

203 8 17 12 8 12 14 8 6
254 10 30 30 16 18 30 24 12
305 12 52 48 42 24 44 40 32
337 13 ¼ 61 56 52 50 60 48 44
387 15 ¼ 85 78 62 64 80 72 74

438 17 ¼ 108 108 104 96 104 100 100

489 19 ¼ 144 136 130 114 132 132 120
540 21 ¼ 173 166 154 156 172 168 148
591 23 ¼ 217 208 194 192 212 204 198
635 25 252 240 230 212 244 240 230
686 27 296 280 270 260 290 284 274
737 29 345 336 310 314 340 336 328
787 31 402 390 366 368 400 384 372
838 33 461 452 432 420 456 444 440
889 35 520 514 494 484 518 504 502
940 37 588 572 562 548 584 576 566
991 39 661 640 624 620 664 644 640
1067 42 776 756 738 724 764 748 750
1143 45 900 882 862 844 902 880 862
1219 48 1029 1016 984 972 1028 1008 1004
1372 54 1310 1296 1268 1256 1320 1296 1284
1524 60 1641 1624 1598 1576 1634 1616 1614

Table 6.1 (f) 1 in OD tubes on 11/4 in triangular pitch

Shell ID TEMA L or M Number of passes TEMA P or S Number of passes TEMA U Number of passes
mm in 1 2 4 6 1 2 4 6 2 4 6

203 8 27 26 8 12 18 14 8 12 14 12 6
254 10 42 40 34 24 33 28 16 18 28 24 24
305 12 64 66 52 44 51 48 42 44 52 40 40
337 13 ¼ 81 74 62 56 73 68 52 44 64 56 52
387 15 ¼ 106 106 88 92 93 90 78 76 90 80 78

438 17 ¼ 147 134 124 114 126 122 112 102 122 112 102
489 19 ¼ 183 176 150 152 159 152 132 136 152 140 136
540 21 ¼ 226 220 204 186 202 192 182 172 196 180 176
591 23 ¼ 268 262 236 228 249 238 216 212 242 224 216
635 25 316 302 274 272 291 278 250 240 286 264 246

686 27 375 360 336 324 345 330 298 288 340 320 300
737
 29 430 416 390 380 400 388 356 348 400 380 352
787 31 495 482 452 448 459 450 414 400 456 436 414
838 33 579 554 520 504 526 514 584 464 526 504 486
889 35 645 622 586 576 596 584 548 536 596 572 548

940 37 729 712 662 648 672 668 626 608 668 636 614
991 39 808 792 744 732 756 736 704 692 748 728 700
1067 42 947 918 874 868 890 878 834 808 890 856 830
1143 45 1095 1068 1022 1000 1035 1008 966 948 1028 992 972
1219 48 1241 1220 1176 1148 1181 1162 1118 1092 1180 1136 1100
1372 54 1577 1572 1510 1480 1520 1492 1436 1416 1508 1468 1442
1524 60 1964 1940 1882 1832 1884 1858 1800 1764 1886 1840 1794
1676 66 2390 2362 2282 2260
1829 72 2861 2828 2746 2708
1981 78 3368 3324 3236 3216

2134 84 3920 3882 3784 3736

2286 90 4499 4456 4370 4328
2438 96 5144 5104 4986 4936
2743 108 6546 6494 6360 6300
3048 120 8117 8038 7870 7812

(Reproduced with the Permission of McGraw-Hill Education, USA)

3. Use of approximate equations: Approximate equations are available in

literature to find the shell diameter or to find the tube bundle diameter. One of
such equations3 is given as follows:

Db = do   (6.1)

where, DB = Tube bundle diameter, mm

do = Tube OD, mm
Nt = Total number of tubes
k1 and n1 are constants, values of which depend on ratio ,

arrangement of tubes and number of tube side passes.

Use Tables 6.2 and 6.3 for finding constants of Eq. (6.1).
Table 6.2 Constants for Eq. (6.1)
A. For Pt /do = 1.25, Triangular Pitch

No. of tube side passes 1 2 4 6 8

k1 0.319 0.249 0.175 0.0743 0.0365

 n1 2.142 2.207 2.285 2.499 2.675

Table 6.3 Constants for Eq. (6.1)

B. For Pt /do = 1.25, Square Pitch

No. of tube side passes 1 2 4 6 8

k1 0.215 0.156 0.158 0.0402 0.0331

n1 2.207 2.291 2.263 2.617 2.643

Equation (6.1) can be used for fixed tube sheet type and floating head shell and
tube-type heat exchangers. For U-tube heat exchangers, following equation can
be used:

  (6.2)

where,   (6.3)

Nt = Number of tube holes on tube sheet

After finding tube bundle diameter Db, shell ID (Inside diameter); Di can be
determined by following equation:
Di = Db + C   (6.4)
where, C = Diametric clearance between shell ID and Db
C = 10 to 20 mm for fixed tube sheet and U-tube
C = 50 to 80 mm for split-ring floating head
C = 90 to 100 mm for pull through floating head
6.3.2 Shell Side Pass Partition Plate
Single pass shell is used in the most of the cases. Two pass shell is rarely used
and is recommended where shell and tube temperature difference is
unfavourable for the single shell side pass. For such cases, normally two or more
smaller size 1-1 heat exchangers, connected in series, are recommended. Shell
side pass partition plate is not provided to improve shell side heat transfer
coefficient but it is provided to avoid the unfavourable temperature difference or
to avoid the cross of temperatures between hot fluid and cold fluid.
Hot oil is to be cooled from 45°C to 37°C by cooling water.
Cooling water will enter at 32°C and leave at 40°C. Select the suitable shell and
tube heat exchanger.
Solution
Minimum driving force required for heat transfer in shell and tube heat
exchanger is 3 to 5°C. Hence, for this case 1-1 shell and tube heat exchanger can
be selected. But it will provide poor heat transfer coefficient and may require
large heat transfer area. 1-2 shell and tube heat exchanger cannot be selected for
this case because of the occurrence of hypothetical temperature cross between
hot and cold fluids. 2-2 shell and tube heat exchanger can be selected. 2-2 heat
exchanger will provide higher heat transfer coefficient than 1-1 heat exchanger.
Hence, either select 2-2 heat exchanger or provide two identical heat exchangers
with smaller diameters, connected in series.
6.3.3 Baffles
There are two functions of baffles:
1. Baffles are used in the shell to direct the fluid stream across the tubes to
increase the velocity of shell side flow and thereby to improve the shell side heat
transfer coefficient. In other words, baffles are used in shell to increase the
turbulence in shell side fluid.
Note: This function is useful only if there is no phase change in shell side fluid.

2. Baffles indirectly support the tubes and thereby reduce the vibrations in tubes.
If shell side liquid velocity is higher; say more than 3 m/s, vibration analysis
calculations should be carried out to check whether baffle spacing is sufficient or
not. Similarly, for very high velocity of gas or vapour and also for the higher
baffle spacing (higher than shell ID), vibration analysis calculation must be
carried out to check the baffle spacing. Vibration analysis calculations are given
in TEMA standard.
Different types of baffles are used in shell and tube heat exchangers; (i)
Segmental baffle, (ii) Nest baffle, (iii) Segmental and Strip baffle, (iv) Disk and
Doughnut baffle, (v) Orifice baffle, (vi) Dam baffle, etc. Figure 6.2 shows
various types of baffles.
Most widely used type of baffle is segmental baffle. Other types of baffles like
nest baffle, segmental and strip baffle and disk and doughnut baffle provide less
pressure drop for the same baffle spacing but provide lower heat transfer
coefficients as compared to segmental baffle. Situations when other types of
baffles might be used could be as given below.
1. For shell and tube heat exchangers shell side pressure drop controls the overall
design. For example, intercoolers of compressors and heat exchangers used in
very high vacuum system. In intercoolers of compressors maximum allowable
shell side pressure drop may be as low as 3.45 kPa. For such cases different
types of shells are also used. Split flow (G shell) or divided flow (J shell) designs
provide low shell side pressure drop compared to commonly used single pass (E
shell) as shown in Fig. 6.3.

Fig. 6.2 Different Types of Baffles used in Shell and Tube Heat Exchangers2 (Reproduced with the
Permission of McGraw-Hill Education, USA)

2. For shell and tube heat exchanger in which boiling or condensation is carried
out on shell side. In such a case baffles are required only to reduce the vibrations
in tubes.
For other type of baffles (other than segmental baffle) correlations for finding
heat transfer coefficient and pressure drop are not easily available in open
literature.
Helical baffles are also developed by some heat exchanger manufacturers. These
are claimed to substantially alter the shell side flow pattern by inducing a
swirling pattern with a velocity component parallel to the tubes. Optimum helix
angle of 40° is recommended. This arrangement is expected to improve the heat
transfer and reduce the pressure drop. However, its design procedure is
proprietary.

Fig. 6.3 (a) One Pass Shell (E Shell) (b) Split Flow (G Shell) (c) Divided Flow ( J Shell)

Segmental baffle is made by cutting the circular metal disk. Segmental baffles
are specified in terms of % baffle cut. Fractional baffle cut is denoted by (x).
% Baffle cut = (h/Db) × 100 = x × 100  (6.5)
where, h = Height of segment removed
Db = Diameter of circular metal disk from which segment is removed
= ID of shell – Diametric clearance (C)
% Baffle cut ranges from 15 to 45%. If there is no phase change on shell side
fluid, then decrease in % baffle cut increases shell side heat transfer coefficient.
Decrease in window area results in increase in window velocity but at the
expense of pressure drop. Baffle cut equal to 20 to 25% is found to be common
in use.
All calculations in the book cover only segmental baffles.
Baffle outside diameter is always less than shell inside diameter. Certain
clearance between baffle OD and shell ID is provided to facilitate removal and
insertion of tube bundle for maintenance. Recommended diametric clearance
between shell ID and baffle OD is as follows.
Table 6.4 Clearance Between Shell and Segmental Baffle1
Shell ID Diametric clearance, C
Shell ID – Baffle OD = (Di – DB)

Standard pipe shell 6 in to 24 in (150 mm to 600 mm) 1.6 mm

Plate shell 6 in to 25 in (150 mm to 625 mm) 3.2 mm
Plate shell 27 in to 42 in (675 mm to 1050 mm) 4.8 mm

If shell is fabricated from a plate then it may not be a perfect cylinder. Hence, for
plate shell, more clearance is a permitted. For heat exchanger without phase
change, increase in clearance between shell ID and baffle OD could mean
decrease in shell side heat transfer coefficient due to leakage. Fraction of shell
side fluid flowing through this clearance is not utilized for heat transfer. Hence
this clearance, if it is more than recommended value, is undesirable.
In kettle-type reboiler, full baffle (with 0% baffle cut) is used, as in this type of
heat exchanger extra space is available for the flow of shell side liquid.
In case of horizontal inclined condenser, segmental baffles with horizontal cut
should not be used as they create resistance to the flow of condensate. Hence, in
this case there are two options; (i) Use segmental baffles with vertical cut (Fig.
6.2 (h)) or (ii) Use segmental baffles with horizontal cut but their base must be
trimmed (Fig. 6.2(g)).
Tube holes are made in baffle. Diameter of tube hole in baffle is kept higher than
tube OD. Normally this diametric clearance is less than 0.8 mm.
Thickness of baffle depends on size of shell and spacing between baffles. For the
given size of shell, thickness of baffle increases with increase in baffle spacing.
It can be determined from the table given in TEMA standard. As per Kern5,
baffle spacing Bs ranges from 0.2 to 1 times shell ID. Lesser baffle spacing gives
higher heat transfer coefficient but at the expense of higher pressure drop.
Optimum baffle spacing is in between 0.3 to 0.5 times shell inside diameter.
6.3.4 Tube
Tube size range from 1/4 in   (6.35 mm) to 2.5 in (63.5 mm) in shell and tube
heat exchanger. Data for standard tubes are given in TEMA standard and in Ref.
2. For the standard tubes, its size is equal to outer diameter of tube. Thickness of
standard tubes are expressed in BWG (Birmingham Wire Gauge). Increase in the
value of BWG means decrease in tube thickness. For no phase change heat
exchangers and for condensers, 3/4 in (19.05 mm) OD tube is widely used in
practice. While for reboiler 1 in (25.4 mm) OD tube size is common. Tubes are
available in standard lengths like 6 ft (1.83 m), 8 ft (2.44 m), 12 ft (3.66 m), 16 ft
(4.88 m) and 6 m.
6.3.5 Tube Side Pass Partition Plate
Tube side passes are provided to decrease the tube side flow area and to increase
tube side fluid velocity thereby to improve the tube side heat transfer coefficient
at the expense of pressure drop. This is true only if there is no phase change on
tube side. Hence, more number of tube side passes are recommended only if
there is no change in the phase of tube side fluid.
For example, at the design stage, if the number of tube side passes are increased
from one to two, then for the given volumetric flow rate, flow area becomes half
and velocity becomes double. Since, tube side heat transfer coefficient, hi ∝ ut 0.8
(where ut is tube side fluid velocity), on increasing number of tube side passes
from 1 to 2, hi nearly becomes 1.74 times. But, Δpt ∝ ut2.8, so the pressure drop
increases by 6.96 times. Increase in hi means decrease in heat transfer area
required and decrease in fixed cost. Increase in Δpt means increase in power
required for pumping the tube side fluid and increase in operating cost. Hence,
ideally optimum number of tube side passes must be decided. Many a times,
tube side velocity in a single pass could be calculated very low (say 0.3 to 0.5
m/s). Under such circumstances, two passes or four passes could be beneficial. It
is recommended that ut > 1 m/s. This is essential if tubeside fluid is cooling
water; otherwise rate of fouling will be higher.
Tube side passes are very common and are advantageously used for improving
tube side heat transfer coefficient. These passes can be achieved in many ways
by locating partition plates in channel covers. Figure 6.4 gives different designs
for achieving desired tube passes.

Fig. 6.4 Nozzle Orientation Designs for Tube Side Passes4

6.3.6 Tie Rods

Baffles are supported by tie rods (Fig. 6.5). Tie rods are made from solid metal
bar. Normally four or more tie rods are required to support the baffles. Diameter
of tie rod is less than the diameter of tube. Diameter and number of tie rods
required for given shell diameter are specified by TEMA standard and IS:4503.

Fig. 6.5 Baffle Spacer Detail

6.3.7 Spacers
Spacers (Fig. 6.5) are used to maintain the space between baffles. Spacers are the
pieces of pipes or in the most of the cases they are the pieces of extra available
tubes. Spacers are passed over the tie rods and because of them baffles do not
slide over tie rods under the effect of the force of fluid. Hence, spacers fix the
location of baffles and maintain the space between them. Length of spacer is
equal to space between the baffles.
6.3.8 Tube Sheet
Tubes and one end of tierods are attached to tube sheet (also called tube plate).
Hence, entire load of tube bundle is transferred to one or two tube sheets. In U-
tube shell and tube heat exchanger (as shown in Fig. 6.1(d), type CFU or Fig.
6.1(e), type AKT) only one tube sheet is used. While in fixed tube sheet shell
and tube heat exchanger, two tube sheets are used. One surface of tube sheet is
exposed to tube side fluid and other surface is exposed to shell side fluid. This
point is very important in the selection of material for tube sheet and also in
determining tube sheet thickness.
In majority cases tube to tube sheet joints are two types; (a) Expanded joint, and
(b) Welded joint as shown in Fig. 6.6.

Fig. 6.6 Tube to Tube Sheet Joint5

In expanded type joint, tube holes are drilled in a tube sheet with a slightly
greater diameter than the tube OD. Two or more grooves are cut in the wall of
each hole. The tube is placed inside the tube hole and a tube roller is inserted
into the end of the tube. Roller is slightly tapered. On application of the roller,
tube expands and tube material flows into grooves and forms an extremely tight
seal. Welding joint is used only for the cases where leakage of fluid can be
disastrous.
Detailed method for calculating tube sheet thickness is given in TEMA and in
IS: 4503. Minimum required tube sheet thickness is also specified in the same.
Simplified equation is also available for calculating tube sheet thickness.

  (6.6)

where, t = Effective tube sheet thickness (minimum), mm

F = constant (1 for fixed tube sheet and 1.25 for U-tube)
p = Design pressure, kPa
f = Allowable stress of tube sheet material at design temperature, kPa
G = Mean diameter of gasket, mm
Then thickness of tube sheet T is given by equation
T ≥ (t + depth of pass partition groove + tube side corrosion allowance + shell
side corrosion allowance)
Also T ≥ Minimum thickness specified in the standard
6.3.9 Sealing Strip
It is a shell side component. Sealing strips are attached on the inside surface of
shell as shown in Fig. 6.7 throughout the length of shell.

Fig. 6.7 Sealing Strips

There are two functions of sealing strips:

1. Sealing strips reduce the amount of bypass stream of shell side fluid flowing
through the clearance between shell inside diameter and tube bundle diameter
and thereby improve the shell side heat transfer coefficient. (This is valid only
if there is no phase change of shell side fluid).
2. Sealing strips also make the removal of tube bundle from the shell easy. Hence,
they are also known as sliding strips.
In Kern’s method (Ref. 5), it is assumed that entire shell side fluid is flowing
across the tube bundle and between the baffles. Actually, shell side fluid is
flowing in various ways. In all latest methods of process design, shell side fluid
flow is divided in five streams. Among these streams, one of the streams is
flowing through the clearance between shell ID and tube bundle. This clearance
is significant in case of pull through floating head heat exchanger (about 90 to
100 mm). In addition, this clearance provides low pressure drop path for shell
side fluid. Hence, actual fraction of shell side fluid bypassed through this
clearance is considerably higher. Sealing strip partially blocks the gap between
bundle and shell and thereby reduces the amount of this bypass stream
considerably. In case of pull through floating head heat exchanger, all latest
computer programs, like HTRI, suggest the use of sealing strip. Also in such
programs shell side heat transfer coefficient is a function of dimension and
numbers of sealing strips provided. Kern’s method is not recommended for use
to predict heat transfer coefficient and pressure drop for pull through floating
head heat exchanger.
6.3.10 Expansion Joint
Expansion joint is attached to shell wall. In this case, shell is made from two
pipe pieces. Two pipe pieces are joined together by an expansion joint as shown
in Fig. 6.8.
 Fig. 6.8 Expansion Joints2 (Reproduced with the Permission of McGraw-Hill Education, USA)

Expansion joint is used in fixed tube heat exchanger to permit the differential
thermal expansion or contraction between shell and tubes which otherwise is not
permitted by fixed tube sheet heat exchanger. Differential thermal expansion
between shell and tube is significant if there is a large temperature difference
between temperature of shell material and temperature of tube material during
operation or if tube material and shell material are different. For the given case
of fixed tube sheet heat exchanger, whether an expansion joint is required or not
can be determined by calculations given in TEMA. If expansion joint is not
provided in fixed tube sheet heat exchanger, then fixed tube sheet does not
permit the unequal expansion or contraction between shell and tube and it can
result in the development of thermal stress across the tube sheet. If this thermal
stress is higher than permissible value, then it may develop a crack in tube sheet,
and can result in leakage at tube to tube sheet joint. Other options, available to
avoid the development of thermal stress in the tube sheet, are use of either U-
tube heat exchanger or floating head heat exchanger.

6.4 ADVANTAGES AND DISADVANTAGES OF

DIFFERENT TYPES OF SHELL AND TUBE HEAT
EXCHANGERS OVER EACH OTHER
6.4.1 Advantages of U-tube and Floating Head Heat
Exchangers (Figs. 6.1(a) and 6.1(d)) Over Fixed Tube
Sheet Heat Exchangers
1. Maintenance and cleaning of tube bundle is easier. In U-tube and floating
head heat exchanger, entire tube bundle can be easily taken out for cleaning
and maintenance. While the same is not possible in fixed tube sheet heat
exchanger.
2. U-tube and floating head heat exchangers permit the differential thermal
expansion or contraction between shell and tube. Therefore, use of expansion
joint is not required in U-tube and floating head heat exchangers.
6.4.2 Disadvantages of U-tube and Floating Head Heat
Exchangers Over Fixed Tube Sheet
1. For a given heat duty, U-tube and floating head heat exchangers are costlier
than fixed tube sheet heat exchangers because of the following reasons:
(a) For the same number of tubes, tube arrangement and tube pitch, shell sizes
required by U-tube and floating head heat exchangers are higher than the
same required by fixed tube sheet heat exchangers. In U-tube heat
exchanger, minimum bend radius of U-tube is 1.5do. Hence, in central
portion of tube bundle, tubes cannot be provided. In floating head heat
exchanger more clearance between shell ID and tube bundle diameter (50
to 100 mm) is required.
(b) In case of cooling or heating of shell side fluid (no phase change) heat
transfer coefficients obtained in U-tube and floating head heat exchangers
for the given heat duty are less than the same obtained in fixed tube sheet
heat exchangers as bypassing area is higher in case of U-tube or floating
 head as compared to fixed tube sheet heat exchanger.
6.4.3 Advantages and Disadvantages of U-tube Heat
Exchanger Over Floating Head Heat Exchanger
Advantage Fixed cost of U-tube heat exchanger is always less than the same of
floating head heat exchanger for the given duty.
Disadvantages
1. In U-tube, tube material becomes weak in bending portion, hence use of U-
tube heat exchanger is not recommended for severe conditions. While for the
same conditions floating head heat exchanger can be used.
2. In U-tube heat exchanger number of tube side passes are fixed (two). It
reduces the flexibility in design calculations.
3. U-tube heat exchanger is not recommended for tube side dirty fluid. As it is
difficult to remove dirt from U-tube. While floating head heat exchanger can
be used where both; shell side fluid and tube side fluid are relatively dirty.
6.4.4 Floating Head Heat Exchangers
There are two types of floating head heat exchangers.
1. Split-ring floating head heat exchanger
2. Pull through floating head heat exchanger
Split-ring floating head heat exchanger is widely used in the chemical process
industries (Fig. 6.1(a) with designation S of Fig. 6.9). With this type, split
backing ring is provided. Advantages and disadvantages of this heat exchanger
over pull through floating head (see designation T of Fig. 6.9) heat exchanger are
as follows:
Advantage Compared to pull through floating head heat exchanger, less diameter
is required for the shell to accommodate the same number of tubes.
Disadvantages More maintenance time is required. In maintenance, shell cover
is removed first, then the split backing ring, then floating head cover and finally
tube bundle is removed from the stationary end.
Pull through floating head heat exchanger is the costliest type of shell and tube
heat exchanger because it requires largest shell diameter for the given numbers
of tubes. With this heat exchanger, it is possible to remove the entire tube bundle
including floating head assembly from the stationary end.
6.5 DIFFERENT TYPES OF TEMA
DESIGNATIONS
In TEMA, standard different types of shell and tube heat exchangers are
designated by three letters. First letter indicates type of front head, second letter
indicates type of shell and third indicates type of rear head. Different types of
TEMA designations are shown in Fig. 6.9.
For example, BEM shell and tube type heat exchanger means bonnet type (B
type) front head, single pass (E type) shell and stationary head with fixed tube
sheet (M type) rear head. BFL type shell and tube heat exchanger means bonnet-
type front head, two pass shell with longitudinal baffle (F type shell) and
stationary head with fixed tube sheet and removable flat cover (L type rear
head).

6.6 GENERAL DESIGN METHOD FOR SHELL

AND TUBE HEAT EXCHANGERS
6.6.1 Calculation of Heat Duty
1. For cooling or heating or for no phase change, heat duty is calculated by the
equation
ϕ = ṁCp Δt  (6.7)
where, ϕ = Heat duty required, kW
ṁ = Mass flow rate of fluid, kg/s
Cp = Specific heat of fluid, kJ/(kg · ° C)
Δt = Temperature difference to be effected, °C
 Fig. 6.9 TEMA-type Designations for Shell and Tube Heat Exchangers1. (Fig. No. N-1.2 of TEMA 8th
Edition)
(Reproduced with the Permission of Tubular Exchanger Manufacturers’ Association, Inc. USA.)

2. For condensation with subcooling

ϕ = ṁλ + ṁCL Δt  (6.8)
where, ṁ = Mass flow rate of vapour, kg/s
CL = Specific heat of condensate, kJ/(kg · °C)
λ = Latent heat of vaporization, kJ/kg
3. For reboilers
ϕB = ṁvλ × 1.05 (considering 5% heat loss)  (6.9)
where, ṁv = Vaporization rate, kg/s
λ = Latent heat of vaporization, kJ/kg
Actually heat duty of the reboiler is determined based on energy balance around
distillation column.
6.6.2 Selection of Cooling Medium or Heating
Medium
Selection of cooling or heating medium mainly depends on the temperature
(inlet temperature) at which cooling medium or heating medium is required.
Other factors like heat transfer coefficient provided by cooling medium or
heating medium, cost of the medium, etc., are also considered in the final
selection. Tables 6.5 and 6.6 can be used as a guideline for the selection of
heating medium and cooling medium respectively.
Table 6.5 Commonly Used Heating Fluids
Heating medium Recommended temperature range

Saturated steam 100 to 180°C

Oil (Thermic Fluid) 180 to 300°C
Dowtherm E 180 to 260°C
Dowtherm A 300 to 400°C
Molten salt 400 to 590°C
Na-K alloys 590 to 750°C
Flue gas or Hot air 750 to 1100°C
Electric heating > 1100°C

Table 6.6 Commonly Used Cooling Medium

Cooling medium Operating range, °C Recommended to use for temperature

1. Brines –68 to 5 < 8°C

Examples:
Ethanol water solution –5 to 5 < 8°C
Methanol water solution –40 to –1 < 8°C
Ethylene glycol solution –34 to 5 < 8°C
Aqueous solution of calcium chloride –20 to 0 < 8°C
Aqueous solution of sodium chloride –9 to 5 < 8°C
Methylene chloride –68 to –37 < –33°C
 Trichloroethylene –68 to –37 < –33°C
Trichlorofluoromethane –68 to –37 < –33°C
2. Chilled water 5 to 12 Below ambient
3. Cooling water 32 to 60 ambient to 100°C
4. Oil –1 to 300 < 300°C
5. Air Atmospheric temperature > 60°C

Heating mediums If temperature of heating medium is required in between 100°

to 180°C, saturated steam is preferably used as heating medium. Saturated steam
condenses in dropwise manner, hence it provides very high heat transfer
coefficient [about 6000 W/(m2 · °C)]. Superheated steam is not generally
selected as heating medium for indirect (through the wall) heat transfer because
desuperheating zone provides very low heat transfer coefficient [about 100-150
W/(m2 · °C)].
If the temperature of heating medium required is greater than 180°C, then use of
saturated steam as a heating medium is not economical. From 180°C to 300°C,
hot oil (thermic fluid) is preferred as heating medium. Hot oil system is easily
available at comparatively low cost. Hot oil provides low heat transfer
coefficient and also properties of oil are uncertain. Dowtherm A® is an organic
fluid of high heat stability, an eutectic mixture containing 73.5 percent diphenyl
oxide and 26.5 per cent diphenyl by mass. Dowtherm A is used as a heating
medium in both phases; as a liquid as well as saturated vapour. Dowtherm E is
specially processed orthodichlorobenzene. For the precise application on a large
scale, Dowtherm E can be considered as an alternate to hot oil. Molten salts are a
molten mixtures of inorganic salts; one of which is an eutectic, consisting 40%
NaNO2, 7% NaNO3 and 53% KNO3. Sodium potassium alloys; 56% Na–44% K
and 22% Na–78% K, are also used as heating mediums.
Cooling mediums In refrigeration terminology a brine is any liquid cooled by a
refrigerant and circulated as a heat transfer fluid. Operating temperature range is
–68°C to 5°C. For any fluid to be cooled below 8°C (minimum possible
temperature of chilled water, 5°C + minimum driving force required for heat
transfer in shell and tube heat exchanger, 3°C), brines are used as cooling
medium. Brine may be
1. an aqueous solutions of inorganic salts, e.g., sodium chloride solution, calcium
chloride solution, etc.
2. aqueous solutions of organic compounds such as alcohols or glycols, e.g.,
 ethanol, methanol, ethylene glycol, propylene glycol, etc.
3. chlorinated or fluorinated hydrocarbons, e.g., methylene chloride, trichloro-
ethylene, trichlorofluoromethane, halogenated hydrocarbons as refrigerants,
etc.
Outlet temperature of chilled water from chilled water plant ranges from 5 to
15°C. While its return temperature is kept from 7°C to 12°C higher than inlet
temperature. When any fluid is to be cooled below the temperature equal to
design temperature of cooling water from cooling tower + minimum driving
force required for heat transfer in shell and tube heat exchanger (3 to 5°C),
chilled water is used as cooling medium.
Cooling water is the most widely used cooling medium at near ambient
temperature.
Design temperature of cooling water from cooling tower = Design wet bulb
temperature + Approach to the wet bulb temperature (ranges from 4 to 6°C)
Cooling water can be used as cooling medium even though inlet temperature of
hot fluid is more than 100°C. But in this case extra care must be taken as at any
time cooling water can be converted into steam. In case of exothermic reaction,
oil (thermic fluid) is used for dual purpose. Initially to achieve the reaction
temperature, oil is used as heating medium. Thereafter as the reaction proceeds,
the same oil is used as cooling medium. For this, one heater and one cooling
water cooler (or economizer) are provided in oil cycle.

Fig. 6.10 Hot Oil (Thermic Fluid) Cycle

For more information on design of process utility systems, refer to Chapter 11.
Air cooling Air cooled heat exchanger has certain advantages as well as
disadvantages over cooling water cooled heat exchanger using cooling water.
Advantages
1. Generally operating cost of Air Cooled Heat Exchanger (ACHE) is less than
the same of Water Cooled Heat Exchanger (WCHE). Operating costs of
WCHE includes cost of make-up water for the cooling tower, power required
for the cooling tower fans and for the circulating pump, cooling tower
maintenance, etc. While operating cost of ACHE includes cost of power
consumed by fans. Day by day cost of water increases and hence, the
difference in operating cost between WCHE and ACHE increases.
2. Inlet temperature of water to cooling tower (or outlet temperature of cooling
water from heat exchanger circuit) should not be greater than 60°C otherwise
it creates the scale formation and also affects the material of construction of
cooling tower. While for air practically there is no limit for temperature.
3. Fouling or scale formation does not take place with aircooling.
Disadvantages
1. Fixed cost of ACHE is higher than the same of WCHE. Heat transfer
coefficient of air is quiet less than heat transfer coefficient provided by
cooling water and hence, heat transfer area required by ACHE is more. Also
design inlet temperature of cooling water from cooling tower is always less
than design ambient temperature. It gives lesser Mean Temperature
Difference (MTD) for ACHE compared to the same for WCHE and hence it
increases heat transfer area required for ACHE.
2. Minimum driving force required for heat transfer, ΔTmin, is 10° to 15°C for
ACHE and 3 to 5°C for WCHE. Hence, minimum temperature that can be
achieved with ACHE is ambient temperature plus 10°C, while with WCHE it
is design temperature of cooling water from cooling tower plus 3°C. Hence, if
the fluid is to be cooled or condensed below 55°C in a tropical location,
ACHE cannot be used.
Based on atmospheric conditions, cost of heat exchangers and cost of water, one
can find out the temperature above which ACHE becomes cheaper than WCHE.
Fluid allocation
1. Fluid allocation in case of phase change: Condensing or vaporizing fluid is
normally taken on shell side as in the shell, liquid and vapour phases are
 easily separated. In case of a condenser, if condensation is carried out in tube
side, then the entire liquid condensate forms a film on heat transfer surface
which provides additional resistance to heat transfer. In case of reboilers
normally on both sides phase change take place and hence, fluid allocation
depends on other factors.
2. Fluid allocation in case of no phase change: If there is no phase change in
shell side and tube side fluid, then fluid allocation depends on the following
factors:
(a) Corrosion: Corrosive fluid is allocated tube side as the cheaper material like
mild steel can be used for costly shell, baffles, etc. Also, other costlier parts
like tubes and tube sheets require the use of non-corrosive or costlier
material of construction in case of corrosive fluid, whether it is on tube side
or in shell side. For example, in cooling of ethyliodide, it should be taken
on tube side as it requires the use of special material (Hastelloy-B).
(b) Fouling: In fixed tube sheet heat exchanger inside surface of tubes can be
easily cleaned and hence, the fluid which has the greatest tendency to foul
on heat transfer surface should be placed in the tubes. For example, cooling
water has a tendency to foul. Hence, normally it is passed through tubes.
(c) Fluid temperature: At very high temperature as well as at very low
temperature use of special alloy is required. Hence, very hot or very cold
fluid is placed on tube side to avoid the use of costlier material for the shell.
At moderate temperature, hotter fluid is better passed through tubes. If it is
placed on shell side more insulation is required not only to reduce the heat
loss but also for safety purpose.
(d) Operating pressure: High pressure stream should preferably be placed on
tube side as high pressure tubes are cheaper than high pressure shell.
Diameter of shell is very much greater than diameter of tubes hence, its
thickness is more sensitively changed with change in pressure as compared
to that of the tubes.
(e) Fluid flow rates and viscosity: Very low value of flow rate of fluid and high
value of viscosity of fluid give low value of Reynolds number. Fluid which
provides very low value of Reynolds number should be placed on shell side
as the dependency of shell side heat transfer coefficient on Reynolds
number is less as compared to the same of tube side heat transfer
coefficient. Hence, for the lower value of Reynolds number shell side heat
transfer coefficient is higher than tube side heat transfer coefficient. For
example, in cooling of viscous oil, it is allocated shell side as it provides
 poor Reynolds number.
6.6.3 Energy Balance and Heat Transfer Calculations
Establish energy balance or heat duty balance and based on that find the mass
flow rate of heating medium or cooling medium.
For example in case of condenser
ϕt = ṁλ = ṁwCLw(t′2 – t′1) kW  (6.10)
where, ṁ = Mass flow rate of vapour condensed, kg/s
λ = Latent heat of condensation kJ/kg
ṁw = Mass flow rate of cooling medium required kg/s
CLw = Specific heat of cooling medium kJ/(kg · °C)
t′2, t′1 = Outlet/inlet temperatures of cooling medium, °C

6.6.4 Mean Temperature Difference

Mean temperature difference (MTD) is calculated from equation
ΔTm = Ft · ΔTlm  (6.11)
where, ΔTlm = Logarithmic mean temperature difference, °C

ΔTlm =   (6.12)

where, ΔT2 and ΔT1 are terminal temperature differences.

For 1-1 heat exchanger (one shell side pass and one tube side pass), Ft = 1
When shell side passes and/or tube side passes are more than one, Ft must be
determined.
Ft = f (R, S ) where R and S are temperature ratios  (6.13)

R =   (6.14)

S =   (6.15)

where, t1 = Hot fluid inlet temperature, °C

t2 = Hot fluid outlet temperature, °C
t′1 = Cold fluid inlet temperature, °C
t′2 = Cold fluid outlet temperature, °C
Graphs of Ft vs S for the different values of R, for the different cases are given in
Ref. 1. Figure 6.11 gives values of Ft for even number of tube passes. If Ft <
0.75, economic design of shell and tube heat exchanger cannot be achieved. For
such a case, alternate type of heat exchanger design should be considered. In the
cases where temperature cross is likely to occur, multiple shell pass heat
exchanger or more numbers of heat exchangers in series, should be considered.

Fig. 6.11 Temperature Correction Factor: One Shell Pass; Even Number of Tube Passes1 (Fig. T-3.2A of
TEMA, 8th Edition)
(Reproduced with the Permission of Tubular Exchanger Manufacturers Association, Inc. USA.)

Empirical correlations6 are reported by Wales in literature for calculating Ft for

different passes on shell and tube sides.
6.6.5 Estimation of Overall Heat Transfer Coefficient
Assume the value of U, overall heat transfer coefficient. For this, use the
standard tables, in which the range of the values of U for the different cases are
given. Refer to Table 6.7 for the recommended values.
Table 6.7 Typical Overcell Coefficients for Shell and Tube Heat Exchangers
Shell side Tube side U, W/(m2 · °C)
 [A] Heat exchangers

1. Water Water 800 –1400

2. Organic solvent Water 300 –850
3. Organic solvent Brine 200 –500
4. Light oil Water 350 –900
5. Heavy oil Water 60 –300
6. Water Brine 600 –1200
7. Gas Brine 15 –250
8. Gas Gas 10 –50
9. Gas Water 20 –300
10. Naphtha Water 300 –400
11. Lube oil Water 150 –500
12. Gasoline Water 350 –550
13. Kerosene Water 150 –300
14. Water 10 to 30% 500 –1250
Caustic soda solution

[B] Condensers

1. Organic vapour Water 700 –1000

2. Organic vapour + Gas Water or Brine 100 –300
3. Low boiling hydrocarbons vapour mixture Water 450 –1100
4. Naphtha vapour Water 300 –425
5. Steam Water 2000 –5000
6. Alcohol vapour Water 500 –1000
7. Saturated Organic vapour Water or Brine 280 –680

[C] Vaporizers/Reboilers

1. Light organics Steam 800 –1200

2. Heavy organics Steam 600 –900
3. Aqueous solution Steam 1000 –1500
4. Chlorine Steam 700 –1500
5. Ammonia Steam 700 –1500
6. Water Steam 1250 –2000
7. Refrigerants Water 425 –850

Find out anticipated heat transfer area based on this selected or assumed value of
U.
A = (ϕt to be in W)  (6.16)

For the first trial calculation, this A is the heat transfer area provided or actual
heat transfer area of available or selected heat exchanger.
A = Apro = Ntπ doL m2  (6.17)
where, do = Outside diameter of tube, m
L = Tube length, m
Nt = Total number of tubes
Values of do and L are decided by designer and from Eq. (6.17), value of Nt is
determined. Then for the first trial, number of tube side passes and shell side
passes are decided. Tube arrangement and type of heat exchanger are now
selected.
6.6.6 Finding Shell Diameter
Based on all these informations find the inside diameter of shell. Select the type
of baffle. First choice is 25% cut segmental baffle. Fix the value of baffle
spacing. If there is cooling or heating on shell side then for the first trial baffle
spacing Bs could be 0.3 to 0.5 times shell ID and if there is condensation or
boiling on shell side, Bs could be equal to shell ID.

6.7 CLASICAL METHOD OF HEAT TRANSFER

AND PRESSURE DROP CALCULATIONS FOR
SHELL AND TUBE HEAT EXCHANGERS
6.7.1 Calculations of Tube Side Heat Transfer
Coefficient
1. For heating or cooling on tube side (no phase change), tube side heat transfer
coefficient is determined by Sieder–Tate equation. If Reynolds number; Re is
less than 2000.
Nu =   (6.18)

where, Nu = Nusselt number

In this equation
If Nu ≤ 3.5 then Nu is taken 3.5.
If Re > 4000, tube side heat transfer coefficient is determined by Dittus–Bolter
equation:
Nu =   (6.19)
Re = Reynolds number =

Pr = Prandtl number =

hi = Tube side heat transfer coefficient, W/(m2 · °C)

di = Tube ID, m
L = Length of tube, m
k = Thermal conductivity of fluid, W/(m · °C)
Cp = Specific heat of fluid, kJ/(kg · °C)
μ = Viscosity of fluid at the bulk fluid temperature, (N · s)/m2 or Pa · s
μw = Viscosity of fluid at tube wall temperature, (N · s)/m2 or Pa · s
C = 0.021 for gases
= 0.023 for non-viscous liquid
= 0.027 for viscous liquid
Gt = Tubeside mass velocity, kg/(m2 · s)

where, at = Tube side flow area =   (6.20)

ṁ = Tube side mass flow rate of fluid, kg/s

Nt = Number of tubes
Np = Number of tube side passes
ut = Tube side fluid velocity = Gt/ρ, m/s
ρ = Density of fluid, kg/m3
To calculate tube side heat transfer coefficient, first calculate tube side flow area
(at), next tube side mass velocity (Gt), then tube side Reynolds number (Re) and
finally tube side Prandtl number (Pr). Depending on the value of Re, use Eq.
(6.18) or Eq. (6.19) to calculate hi.
For the transition region or for 2000 < Re < 4000, no reliable equation is
available, Hence, for this region hi is found by using both Eqs.   (6.18) and
  (6.19) and the lesser of two values is considered.
Alternatively, tube side heat transfer coefficient hi can be calculated (by Kern’s
equation) from the value of “heat transfer factor; Jh”, for the entire range of
Reynolds number (from Re = 10 to 106). Equation relating hi and Jh is

Nu =   (6.21)

Graph of Jh versus Re is given in Fig. 6.12 for the different values of L / di.
 Fig. 6.12 Tube Side Heat-Transfer Factor 3 (Reproduced with the Permission of Elsevier, UK)

2. For condensation inside the tubes: Condensation coefficient depends on the

position of condenser. For condensation, shell side condensation with horizontal
position is the best which gives the maximum value of the coefficient. But if
condensation is carried out in tube side then for horizontal position, hi is
calculated by following two equations and higher of the two values is
considered:

(a) hci = 0.76 kL   (6.22)

This equation is called Nusselt equation, valid for stratified flow.

where, hci = Condensation coefficient, W/(m2 · °C)
kL = Thermal conductivity of liquid condensate, W/(m · °C)
ρL = Density of liquid condensate, kg/m3
ρV = Density of vapour, kg/m3
g = Acceleration of gravity = 9.81 m/s2
μL = Viscosity of liquid condensate, (N · s)/m2 or kg/(m · s) or Pa · s
τh = Horizontal tube loading or flow of condensate per unit length of tube, kg/(m
· s)
All liquid condensate properties kL, ρL and μL must be determined at mean
temperature of condensate film.
τh =   (6.23)

where, L = Tube length, m

ṁc = Total condensate flow, kg/s
Nt = Total number of tubes

(b) hci =   (6.24)

where, h′i = 0.021   (6.25)

This equation is known as Boyko–Kruzhilin equation, valid for annular flow.

Rec = Reynolds number for the condensate film =   (6.26)

Prc = Prandtl number of liquid condensate =   (6.27)

For vertical position, hci = 0.926 kL   (6.28)

where, τv =   (6.29)

ṁc = Mass flow rate of liquid condensate, kg/s

Rec =   (6.30)

This equation is also called Nusselt equation, valid for Rec ≤ 2000. For Rec >
2000, use Boyko–Kruzhilin equation, i.e., Eq. (6.24).
6.7.2 Calculation of Tube Side Pressure Drop
1. If their is no phase change in tube side fluid, then tube side pressure drop can
be calculated by following equation:

Δpt = Np   (6.31)

m = 0.25 for Re ≤ 2100

= 0.14 for Re > 2100
where, Δpt = Tube side friction pressure drop, N/m2 or Pa
Np = Number of tube side passes
Jf = Tube side friction factor can be obtained from Fig.   (6.13)
ρ = Density of tube side fluid, kg/m3
ut = Tube side fluid velocity, m/s
Δpt calculated by this equation is actually permanent pressure loss. Calculated
pressure drop (loss) should be less than maximum allowable pressure drop. In
some applications, maximum allowable pressure drop is decided by process
conditions. While in other applications, maximum allowable pressure drop is
actually optimum pressure drop (loss). Heat exchanger design means the balance
between two opposite factors, heat transfer coefficients related to fixed cost and
pressure drop related to operating cost. Increase in heat transfer coefficient by
modifying the heat exchanger design also increases pressure drop. Hence, ideally
actual pressure drop should be equal to optimum pressure drop which gives the
total minimum cost of heat exchanger (i.e., fixed cost + operating cost). As a
general guideline (Table 6.8) can be used to decide the value of maximum or
optimum pressure drop. As per Kern, optimum pressure drop for gases is 2 psi
(13.8 kPa) and for liquids it is 10 psi (69 kPa) (Ref. 5).
2. If condensation takes place inside the tubes then it is difficult to predict the
pressure drop as vapour mass velocity is changing throughout the condenser. A
common practice is to calculate Δp′t by Eq. (6.31) for inlet vapour flow rate and
conditions, and multiply it by factor 0.5.
Δpt = 0.5 Δp′t  (6.32)

6.7.3 Calculation of Shell Side Heat Transfer

Coefficient
1. For heating or cooling on shell side or for no phase change on shell side, heat
transfer coefficient ho can be calculated by following procedure:
 Fig. 6.13 Tube Side Friction Factors3 (Reproduced with the Permission of Elsevier, UK)

(a) Calculate shell side flow area by following equation:

As =   (6.33)
where, As = Shell side flow area, m2
Pt = Tube pitch, m
do = Outside diameter of tube, m
Bs = Baffle spacing, m
Ds = Shell inside diameter, m
(b) Calculate shell side mass velocity Gs and linear velocity (us).

Gs =   (6.34)

uS =   (6.35)

where, Gs = Shell side mass velocity, kg/(m2 · s)

ṁs = Shell side mass flow rate of fluid, kg/s
As = Shell side flow area
ρs = Density of shell side fluid, kg/m3
(c) Calculate the shell side equivalent diameter (de).
For square pitch arrangement.
de =

=   (6.36)

For triangular pitch arrangement,

de =

=   (6.37)

(d) Calculate shell side Reynolds number.

Reo =   (6.38)

where, μ = Viscosity of shell side fluid at average temperature, kg/(m · s)

Also calculate shell side Prandtl number
Pro =

Cp, μ and k are the properties of shell side fluid at average temperature.
(e) Calculate the shell side heat transfer coefficient by the following correlation
(Ref. 5):

Nuo = = 0.36 Re0.55Pr0.33   (6.39)

This correlation is valid for the range of Reynolds number from 2000 to 1000
000. Alternately, shell side heat transfer, coefficient can also be found out
from “shell side heat transfer factor; Jh” (by Kern’s method).

Nu = = Jh · Re · Pr0.33   (6.40)

Graph of Jh versus Res given in Fig. 6.14.

2. If condensation takes place on shell side then its coefficient depends on the
position of condenser.
Shell side condensation coefficient for horizontal position can be determined by
following modified Nusselt equation:

hco = 0.95 kL   (6.41)

where, hco = Shell side condensation coefficient, W/(m2 · °C)

kL = Thermal conductivity of liquid condensate, W/(m · °C)
ρL = Density of liquid condensate, kg/m3
ρV = Vapour density, kg/m3
g = Acceleration of gravity = 9.81 m/s2
μL = Viscosity of liquid condensate, N · s/m2 or Pa · s
τh = Horizontal tube loading or flow of condensate per unit length of tube
=   (6.23)

Nr = Average number of tubes in vertical tube row

Nr ≅   (6.42)

N′r = Number of tubes in central row

=   (6.43)

Db = Tube bundle diameter, m

Pt = Tube pitch, m
 Fig. 6.14 Shell Side Heat-transfer Factors for Segmental Baffles3 (Reproduced with the Permission of
Elsevier, UK)

For condensation outside the vertical tube bundle, shell side condensation
coefficient is given by Nusselt equation.
hco = 0.926 kL   (6.28)

where, τ v =   (6.29)

In Nusselt equation; Eq. (6.22) and Eq. (6.28), suffixes i and o to hc denote
inside and outside of tubes, respectively.
6.7.4 Calculation of Shell Side Pressure Drop
1. For no phase change on shell side or for heating or cooling on shell side, shell
side pressure drop Δps can be calculated by the following equation:

Δps =   (6.44)

where, Δps = Shell side pressure drop, N/m2 or Pa

Ds = Shell inside diameter, m
de = Shell side equivalent diameter, m
L = Tube length, m
Bs = Baffle spacing, m
ρs = Density of shell fluid, kg/m3
us = Shell side linear velocity, m/s
μ = Viscosity of fluid at the bulk fluid temperature, (N · s)/m2
μw = Viscosity of fluid at tube wall temperature, (N · s)/m2
Jf = Shell side friction factor for the different values of % baffle cut of segmental
baffle and for the different values of Re. It can be obtained from Fig.
6.15.
Shell side pressure drop Δps should be less than maximum allowable pressure
drop. Maximum allowable pressure drop for shell side is also decided either
based on process requirement or based on economic considerations.
Table 6.8 can be used to judge optimum pressure drop.
Table 6.8 Optimum Pressure Drop (Maximum Allowable Pressure Drop) Based on Economic
Considerations (Ref. 3)
Fluid Optimum, Δpt or Δps, kPa

Liquids of μL < 1 cP or mPa · s 35

 Liquids of μL = 1 to 10 cP or mPa · s 50 to 70

Gas or vapour at 1 to 2 atm 13.8

Gas or vapour at high vacuum (up to 60 Torr absolute pressure) 0.4 to 0.8 (3 to 6 Torr)
Gas or vapour at high pressure > 10 atm 0.1 times operating pressure

Fig. 6.15 Shell Side Friction Factors for Segmental Baffles3 (Reproduced with the Permission of Elsevier,
UK)
2. In the case of shell side condensation, no reliable correlation is available for
predicting Δps. Hence, Δps can be approximately calculated by following
equation:
Δps = 0.5 Δp′s  (6.45)
where, Δp′s = shell side pressure drop calculated by Eq. (6.44) based on inlet
vapour flow rate and inlet conditions.
This also should be less than maximum allowable pressure drop.
Calculated pressure drops on tube side and shell side do not account for nozzle
losses.
6.7.5 Calculation of the Overall Heat Transfer
Coefficients
Uo =   (6.46)

where, Uo = Overall heat transfer coefficient based on outside area of tubes,

W/(m2 · °C)
ho = Shell side heat transfer coefficient, W/(m2 · °C)
do = OD of tube, m
di = ID of tube, m
kw = Thermal conductivity of tube wall material, W/(m · °C)
hid = Fouling coefficient for tube side fluid, W/(m2 · °C)
hod = Fouling coefficient for shell side fluid, W/(m2 · °C)
Fouling refers to any unwanted deposit of material on a heat transfer surface, or
by extension, to any deposit that alters the desired fluid flow pattern. Fouling
creates an additional resistance to heat transfer and increases pressure drop by
narrowing flow passages.
For the various types of fluids, recommended fouling coefficients are given in
Table 6.9.
Table 6.9 Recommended Fouling Coefficients (Ref. 3, 6 & 7)
Fluid Fouling coefficient W/(m2 · °C) Fluid Fouling coefficient W/(m2 · °C)

Cooling water 3000–6000 Light Hydrocarbon 5000

Demineralized water 8000–10 000 Heavy Hydrocarbon 3000

Soft water 5000–8000 Aqueous Organic Solution 5000
Steam 4000–10 000 Vegetable oils 3000
Steam condensate 1500–5000 Caustic Solutions 5000
Brine 3000–5000 Gasoline 5000
Air 5000–10 000 Kerosene 3000–5000
Organic vapours 10 000 Light gas oil 3000–5000
Organic liquids 5000 Heavy gas oil 2000–3000
Heavy hydrocarbons 2000 Naphthas 5000
Aqueous salt solutions 3000–5000 Coal flue gas 1000

6.7.6 Calculating the Heat Transfer Area Required

Ar = (ϕt to be in W)  (6.47)

Also calculate actual heat transfer area Ao of selected heat exchanger.

Ao = Ntπ doL  (6.17)

≈ {1.1 to 1.2}

< {0.8 to 1}, < {0.8 to 1}

Thus an optimum design of shell and tube heat exchanger can be obtained.
A lube-oil cooler for 2 MW steam turbine is to be designed for
following operating conditions: Lube-oil, conforming to ISO VG-46 grade,
flows at the rate of 450 L/min and is to be cooled from 65°C to 45°C with
cooling water.
Kinematic viscosity of lube oil = 45.5 cSt at 40°C
Density of lube oil = 869 kg/m3
Specific heat of lube oil = 2.1413 kJ/(kg · °C)
Thermal conductivity of lube oil at 55°C = 0.13 W/ (m · K)
Viscosity of lube oil at 55°C = 15 cP or mPa · s
Cooling water inlet/outlet temperatures = 35°C/39°C
Solution
Heat duty, ϕt = ṁo CLo Δt
Volumetric flow rate of oil = 450 L/min = 450 × 60 × 10–3 m3/h = 27 m3/h
ṁ = 27 × 869 = 23 463 kg/h ≡ 6.5175 kg/s
ϕt = 23 463 × 2.1413(65 – 45)
ϕt = 1004 826 kJ/h ≡ 279.118 kW
≡ 279.118 × 103 W
 Fig. 6.16

Heat balance
ϕt = ṁwCLw Δt′
279.118 = ṁw × 4.1868(39 – 35)
ṁw = 16.667 kg/s (≡ 60 m3/h)
ΔT1 = 65 – 39 = 26°C
ΔT2 = 45 – 35 = 10°C

ΔTlm = = 16.745°C  (6.12)

Allocate lube oil on tube side and cooling water on shell side.
For the first trial calculations, let the number of tube side passes Np = 4

R = S =

t1 = 65° C, t2 = 45° C, t′1 = 35° C, t′2 = 39° C

R = = 5 S = = 0.1333

From Fig. 6.11, Ft = 0.95

Mean temperature difference, ΔTm = Ft ΔTlm
ΔTm = 0.95 × 16.745 = 15.908°C
Assume overall heat transfer coefficient U = 400 W/(m2·°C). (Ref.: Table 6.7)
Provisional heat transfer area for the first trial calculation.

A =

A = 43.8644 m2 = NtπdoL
do = 5/8 in = 15.875 mm, L = 6 ft = 1.8288 m

Nt = = 480.93 ≅ 480

Select BEM type heat exchanger with tube pitch Pt = 1.25 do = 19.8437 mm and
triangular pitch arrangement.
Tube bundle diameter Db =   (6.1)

From Table 6.2 for Np = 4, k1 = 0.175, n1 = 2.285

Db = 15.875 = 507.5 mm

Let diametric clearance between shell ID and Db, C = 12.5 mm

Shell ID, Di = Db + 12.5
= 507.5 + 12.5 = 520 mm
Evaluation of hi:
Tube side flow area:
di = 13.3858 mm (for 18 BWG tube)

at =   (6.20)

= (0.013 3858)2 = 0.016 887 m2

Tube side mass velocity Gt =

Gt = = 390.57 kg/(m2 · s)

Viscosity of oil:
μ = 15 cP = 15 × 10–3 kg/(m · s)

Ret = = 348.54 which is low.

To increase the Reynold number, let the number of tube side passes Np = 8
k1 = 0.0365, n1 = 2.675 from Table 6.2.

Db = 15.875 = 550 mm

Shell ID, Di = Db + 12 = 562 mm

Evaluation of hi:
at = (0.013 3858)2 = 8.4436 × 10–3 m2

Gt = 390.57 × = 781.14 kg/(m2 · s)

ut = = 0.9 m/s

Ret = 348.54 × 2 = 697 low but can be acceptable.

Pr =

= 247
Using Sieder–Tate equation (Ret < 2000)

Nu = = 1.86(Re · Pr · di/L)0.33   (6.18)

= 1.86 × 1

= 19.618 > 3.5

hi = 19.62 × = 190.53 W/(m2 · °C) (too low)

To improve oil side heat transfer coefficient, allocate lube oil on shell side and
water on tube side. Increase the tube length from 6 ft to 10 ft for the same value
of heat transfer area provided. It will decrease the shell side flow area and will
improve the shell side heat transfer coefficient.
L = 10 × 0.3048 = 3.048 m
Apro = 43.8644 m2 = NtπdoL = Nt × π × 0.015 875 × 3.048
Nt = 288.56 ≅ 288
Let number of tube side passes Np = 4
k1 = 0.175, n1 = 2.285 (Ref.: Table 6.2)

Db = 15.875 = 406 mm

Shell ID, Di = 406 + 12 = 418 mm

Evaluation of hi:

at = (0.013 3858)2 = 0.01 m2

Tube side mass velocity, Gt = = 1666.7 kg/(m2 · s)

Density of water at 37°C, ρ = 993.328 kg/m3 (Table 3.28 of Ref. 2)

ut = = 1.678 m/s (good velocity)

Re =

Viscosity of water at 37°C, μ = 0.73 cP = 0.73 × 10–3 kg/(m · s)

Re = = 30 561.8 > 4000

Thermal conductivity of water at 37°C, k = 0.628 W/(m · °C)

Prandtl number of water, Pr =

Pr = = 4.867

Using Dittus–Bolter equation

Nu = = 0.023 Re0.8 Pr0.33   (6.19)

hi = 0.023 × × (30 561.8)0.8 × 4.8670.33

= 7046.6 W/(m2 · °C)

Evaluation of oil side heat transfer coefficient (ho):
Shell side flow area
As = × Di × BS  (6.33)

Pt /do = 1.25, Di = 418 mm, Bs = = 83.6 mm

As = × 0.418 × 0.0836 = 6.989 × 10–3 m2

Shell side mass velocity, Gs =   (6.34)

Gs = = 932.537 kg/m2 · s

us = = 1.073 m/s acceptable velocity  (6.35)

Shell side equivalent diameter for triangular pitch

de =   (6.37)

de = [19.84372 – 0.907 × (15.875)2]

= 11.4465 mm
Shell side Reynolds number Res =

= = 711.62 (rather poor)

Pr =

= 247
For Re = 711.62, Jh = 2 × 10–2 (From Fig. 6.14 for 25% baffle cut)

Nu = = Jh Re Pr1/3   (6.40)

hs = 2 × 10–2 × × 711.62 × (247)1/3 × 1

= 1014.18 W/(m2 · °C)

Overall heat transfer coefficient by Eq. (6.46):
From Table 6.9,
hod = 3000 W/(m2 · °C), hid = 5000 W/(m2 · °C)
Tube material: steel or cupronickel
Thermal conductivity of tube material kw = 50 W/(m · °C)

Uo = 570.8 W/(m2 · °C)

Heat transfer area required, Ar =   (6.47)

= = 30.739 m2

% Excess heat transfer area =

= × 100 = 42.7% (sufficiently high)

Shell side pressure drop:

Δps = 8Jf   (6.44)

Ds = 0.418 m, de = 0.011 4465 m, L = 3.048 m

Bs = 0.0836 m, us = 1.073 m/s, ρs = 869 kg/m3
For Res = 711.62, Jf = 7.7 × 10–2 (From Fig. 6.15)

Δps = 8 × 7.7 × 10–2

= 410 281.7 Pa ≡ 410.282 kPa (too high)

To decrease the shell side pressure drop, baffles spacing can be increased from
83.6 mm to 160 mm.
Revised value of Bs = 160 mm = 0.16 m
Shell side flow area, As = 6.989 × 10–3 × = 0.013 376 m2
Shell side mass velocity, Gs = 932.537 × = 487.25 kg/(m2 · s)

us = = 0.56 m/s (rather poor)

Re = = 371.82

for Re = 371.82, Jh = 2.7 × 10–2 (from Fig. 6.14)

Nu = = 2.7 × 10–2 × 371.82 × (247)1/3 = 62.989

hs = 62.989 × = 715.37 W/(m2 · °C)

Uo = 462.15 W/(m2 · °C)

Ar = = 37.96 m2

% Excess heat transfer area = × 100 = 15.55% (adequate)

Revised shell side pressure drop:

For Ds = 0.418 m, Bs = 0.16 m, de = 0.011 4465 m, L = 3.048 m, us = 0.56 m/s,
ρs = 869 kg/m3

Δps = 8Jf   (6.44)

For Res = 371.82, Jf = 9.26 × 10–2 (From Fig. 6.15)

Δps = 8 × 9.26 × 10–2

Δps = 70 220.7 Pa ≡ 70.22 kPa (acceptable)

Tube side pressure drop:
Δpt = Np   (6.31)

Np = 4, L = 3.048 m, di = 0.013 3858 m, ut = 1.678 m/s

ρ = 993.328 kg/m3, μ/μw ≅ 1
For Ret = 30 561.8 Jf = 3.6 × 10–3 (From Fig. 6.13)

Δpt =

Δpt = 50 668 Pa ≡ 50.668 kPa (acceptable)

Comment: In this case, viscosity of lube oil is very high. Hence, it gives very
poor Reynolds number. Fluid, which provides poor Reynolds number must be
allocated shell side to get the higher heat transfer coefficient. Here also, by
changing the allocation of lube oil from tube side to shell side, lube oil side
coefficient increased considerably. In any (without phase change) shell and
tube heat exchanger for the fixed value of heat transfer area, increase in tube
length (correspondingly decrease in number of tubes and shell inside diameter)
increases the overall heat transfer coefficient but at the expense of pressure
drop or in other words increase in tube length for the fixed value of heat
transfer area decreases the fixed cost of heat exchanger but increases its
operating cost.

6.8 CRITERIA OF SELECTION BETWEEN

HORIZONTAL CONDENSER AND VERTICAL
CONDENSER
1. For only filmwise condensation, horizontal position gives higher condensation
coefficient than vertical position. In filmwise condensation, thickness of
condensate film over heat transfer surface decides the value of condensation
coefficient. With horizontal position and shellside condensation, the
condensate travels less distance over heat transfer surface before falling down
by gravity compared to vertical position. With vertical position, condensate
travels over the entire tube length before falling down by gravity. Hence,
average condensate film thickness obtained with horizontal position is less
than the same with vertical position and hence, it can provide higher
condensation coefficient.
2. In case of condensation with subcooling in a shell and tube heat exchanger,
condensation coefficient is higher with horizontal position while subcooling
coefficient is higher with vertical position. In case of subcooling with
horizontal position, small fraction of gravitational force is acting on the pool
of condensate and hence, condensate is flowing with almost no turbulence.
Hence, subcooling with horizontal position is natural convection heat transfer.
While in subcooling with vertical position, entire gravitational force is acting
on pool of condensate. Hence, it creates turbulence in pool of condensate and
provides higher subcooling coefficient. Subcooling with vertical position can
be considered as forced convection heat transfer.
Therefore, in the case of condensation with subcooling, shell and tube heat
exchanger is designed for both positions (horizontal and vertical) and the
position that gives higher value of overall coefficient is selected.
3. In case of condensation with non-condensables, selection of position depends
on the percentage of non-condensables present in inlet vapour for which
guidelines are given by Frank5.
(a) If non-condensables are < 0.5% (by mass), then the presence of non-
condensables is ignored in design calculations. Heat exchanger is designed
as a total condenser. For this case, horizontal position should be selected as
it provides the higher condensation coefficient.
(b) If non-condensables are > 70% (by mass), then for the entire flow rate,
without phase change correlation is applied to calculate the heat transfer
coefficient. However, in calculation of heat duty (ϕt), condensation is
considered. For this case, selection of the position does not depend on heat
transfer coefficient, as without phase change coefficient does not depend on
position of heat exchanger so in this case position is decided by other
factors like available area or available height.
(c) Between 0.5 to 70% non-condensables, heat transfer coefficient is
determined by considering both condensation as well as cooling of non-
condensables and vapours. For this case horizontal position should be
selected as it gives higher condensation coefficient, if condensation with
cooling is carried out on shell side. If the same is carried out on tube side,
then vertical position is better, because dry gas film over the heat transfer
surface is continuously swept away by incoming stream.
10 900 kg/h of nearly pure saturated methyl ethyl ketone
vapour at 13.73 kPa g is to be condensed and cooled to 60°C by cooling water
which is available in plant at 32°C. Consider pressure drops of 13.7 kPa for
vapour and 68.7 kPa for the water as permissible. Design the suitable shell and
tube heat exchanger for the heat duty.
Solution
Heat duty calculations
ϕt = ṁλ + ṁCLΔt = ϕc + ϕsub
where, ϕc = ṁλ = heat duty for condensation
ϕsub = ṁCLΔt = heat duty for subcooling
ṁ = 10 900 kg/h = 3.0278 kg/s
λ = Latent heat of vaporization of methyl ethyl ketone (MEK) at condensation
temperature
At condensation temperature of pure MEK vapour
pv = pt = 13.73 kPa g = 115.055 kPa a = 863 Torr a
Antoine equation for methyl ethyl ketone
ln pv = 16.5986 –

where, pv is in Torr and T in K

pv = pt = 863 Torr, T = 356.87 K t = 83.87°C
From Table 3-179 of Ref. 2,
λMEK at 78.2°C = 105.93 kcal/kg = 443.5 kJ/kg
Value of λMEK is required at 83.87°C.
Using Watson relation,

λ2 = λ1 (2.11)

Tr = reduced temperature = T/Tc

Tc of MEK = 535.6 K

λ at 83.87°C = λ78.2°C

λ83.87°C = 443.5 × 0.9882 = 438.27 kJ/kg

ϕc = ṁλ = 3.0278 × 438.27 = 1327 kW
Specific heat of MEK liquid at = 72°C:

CL = 2.298 kJ/(kg · °C) (From Table 3-184 of Ref. 2)

ϕsub = 3.0278 × 2.298 × (83.87 – 60) = 166 kW
ϕt = ϕc + ϕsub = 1327 + 166 = 1493 kW
Energy Balance
ϕ = ṁwCpw (t′o – 32)
Fixing the outlet temperature of cooling water (t′o) as 40°C,
mass flow rate of cooling water required
ṁw = = 44.5746 kg/s ≡ 160.5 m3/h

First trial calculations:

Using Table 6.7, let overall heat transfer coefficients be
for condensation Uc = 800 W/(m2 · °C)
for subcooling Usub = 200 W/(m2 · °C)
Assume that entire flow of cooling water is first utilized for subcooling and then
for condensation.
ϕsub = ṁwCpw(t′1 – 32)
where t′1 = intermediate temperature of cooling water
166 = 44.5746 × 4.1868 × (t ′1 – 32)
t ′1 = 32.89°C

Fig. 6.17(a) Condensation

LMTD for condensation:

ΔT1 = 83.87 – 32.89 = 50.98°C
ΔT2 = 83.87 – 40 = 43.87°C
ΔTlm = = 47.336°C

= ΔTmc
For condensation of pure component R = 0 and Ft = 1, hence LMTD = MTD

Fig. 6.17(b) Subcooling

LMTD for subcooling:

ΔT1 = 83.87 – 32.89 = 50.98°C
ΔT2 = 60 – 32 = 28°C

ΔTlm sub = = 38.349°C

= ΔTm sub
Assume that subcooling will take place in the perfect
counter current manner.
Ft = 1, hence LMTD = MTD
Area based on assumed values of U:
For condensation:
Ac = = 35.042 m2

For subcooling:
Asub = = 21.643 m2

Total area A = Ac + Asub = 56.685 m2

= = 0.3818 and Asub < Ac

Based on the selected values of overall coefficients, area required for subcooling
(Asub) is less than area required for condensation (Ac). Hence, to get the smaller
size of heat exchanger horizontal position is selected. Ideally in such a case, heat
exchanger should be designed for both positions and the position which requires
lesser heat transfer area should be selected.
Area provided for the first trial calculation
A = 56.685 m2 = NtπdoL
Choose 3/4 in OD (19.05 mm OD), 6 ft (1.83 m) long tubes.
Nt = = 518

Shell side fluid is a clean fluid, hence, select triangular pitch arrangement.
Pt = 1.25 do = 23.8125 mm
Number of tube side passes = 2 (for first trial calculations)
Tube bundle diameter:

Db = do   (6.1)

= 19.05 (Using Table 6.2)

= 607 mm
Select a fixed tube sheet heat exchanger.
Let diametric clearance between shell ID and Db, C = 13 mm
Shell inside diameter, Di = 607 + 13 = 620 mm
Calculation of tube side heat transfer coefficient:
at = (0.015 748)2 = 0.050 45 m2

(From Table 11.2, of Ref. 2 for 16 BWG tube di = 15.748 mm)

Tube side mass velocity,
Gt = = 883.54 kg/(m2 · s)

ut =
= = 0.89 m/s < 1 m/s

To avoid the excessive fouling it is recommended to keep the velocity of water

greater than 1 m/s.
Increase the number of tube side passes from 2 to 4.
for Np = 4, at = 0.025 22 m2

Db = 19.05 (Table 6.2)

= 630 mm
Let diametric clearance between shell ID and Db, C = 15 mm
Shell inside diameter = 630 + 15 = 645 mm (revised)
Gt = = 1767.43 kg/(m2 · s)

ut = 1.78 m/s acceptable

Viscosity of water at 36°C, μ = 0.72 cP = 0.72 ×10–3 kg/(m · s)
Ret = = 38 657.6

Prandtl number (Pr):

Thermal conductivity of water at 36°C, k = 0.6228 W/(m · °C)
Pr =

= 4.84
Using Dittus–Bolter equation
= 0.023 Re0.8 Pr0.33   (6.19)

Considering = 1

hi = × (38 657.6)0.8 × (4.84)0.33

= 7155.5 W(m2 · °C)

Shell side coefficient:
(a) For condensation zone, hco:
Calculation for mean temperature of condensate film:
Let tw = Tube wall temperature, °C
hco = 1500 W/(m2 · °C) (assumed for first trial calculation)
At steady state
Heat transfer rate through condensate film = overall rate of heat transfer
hco Ac(tc – tw) = Uc Ac(tc – tav)
where, tc = Condensation temperature of vapour on shell side, °C
tav = Average tube side fluid temperature for condensation zone, °C
1500(83.87 – tw) = 800

tw = 58.577°C
Mean temperature of condensate film
tav = = 71.22°C

Physical properties of liquid condensate at 71.22°C:

Viscosity of liquid MEK at 71.22°C, μL = 0.32 cP = 0.32 × 10–3 kg/(m · s)
Density of liquid MEK at 71.22°C, ρL = 805 kg/m3
Thermal conductivity of liquid MEK at 71.22°C, kL = 0.173 W/(m · °C)
Shell side condensation coefficient with horizontal position

hco = 0.95 kL   (6.41)

Acceleration of gravity g = 9.81 m/s2

Density of vapour ρv = (s for standard conditions)

ρv =

= 2.79 kg/m3
τh = = 3.194 × 10–3 kg/(m · s)   (6.23)

Average number of tubes in vertical rows  (6.42)

Nr = 2/3

= = 17.64 ≅ 18

hco = 0.95 × 0.173 × (18)–1/6

= 1864.86 W/(m2 · °C)

hco is close enough to assumed value of hco. Hence, correction in the value of tw
is not required.
(b) For subcooling zone, hosub:
With horizontal position, subcooling is taking place via natural convection only
for which reliable correlations are not available. Based on Kern’s
recommendation (Ref. 5), hosub for horizontal position.
hosub = 50 Btu/(h · ft2 · °F)
= 283.77 W/(m2 · °C)
Overall Heat Transfer Coefficient for Condensation Eq. (8.46):
From Table 6.7,
For organic vapours, hod = 10 000 W/(m2 · °C)
For cooling water, hid = 4000 W/(m2 · °C)
With methyl ethyl ketone and cooling water, stainless steel-304 is a suitable
material for the tube.
Thermal conductivity of SS-304 material, kw = 16.3 W/(m · °C)

Uoc = 820.38 W/m2 · °C)

Heat transfer area required for condensation
Acr =
Acr = 34.17 m2
Overall heat transfer coefficient for subcooling:

Uosub =

Uosub = 237.72 W/(m2 · °C)

Heat transfer area required for subcooling,
Asubr = = 18.21 m2

Total heat transfer area required

Atr = Acr + Asubr = 34.17 + 18.21 = 52.38 m2

= = 1.082

or Excess heat transfer area = (1.082 – 1) × 100 = 8.2%

% Excess heat transfer area should be at least 10%. Hence, to increase the value
of heat transfer area, increase the tube length. New value of tube length is
L ≅ 1.83 × = 1.95 m

Let new or revised tube length L = 2 m with same number of tubes

Heat transfer area provided Apro = × 56.685 = 61.95 m2

Revised value of shell side condensation coefficient

hoc ∝ L1/3

hoc = × 1864.86 = 1920.9 W/(m2 · °C)

New value of Uoc

=
Uoc = 831 W/(m2 · °C)

Aoc = = 33.73 m2

Asub = 18.21 m2
Areq = 33.73 + 18.21 = 51.94 m2

= = 1.1927

Excess heat transfer area = (1.1927 – 1) × 100 = 19.27%

If the equal percentage of excess heat transfer area (19.27%) in both zones are
provided then,
Heat transfer area provided for condensation:
Acpro = 33.73 × 1.1927 = 40.23 m2
Heat transfer area provided for subcooling:
Asubpro = 18.21 × 1.1927 = 21.72 m2

= = 0.3506

Hence, 35.06% of total heat transfer area should be provided for subcooling.
Assuming that tube will be uniformly distributed in the cross section of shell.

0.3506 =

x = × 0.3506 = 0.2754

From Table 1.19b of Ref. 2, for x = 0.2754, h/Di = 0.382

Hence, 38.2% of shell inside diameter should be submerged in the pool of
condensate to facilitate subcooling.
This can be achieved by providing inverted U-seal (Fig. 6.18) height of inverted
U-seal from the base of shell ID.
h = 0.382 × 645 = 246.4 mm
 Fig. 6.18 Detail of U-Seal

Shell side pressure drop Δps:

Shell side flow area
A′s = × x′

where x′ = 1 – = 1 – 0.382 = 0.618

Let baffle spacing Bs = Shell ID = 645 mm

As = × 0.645 × 0.645 × 0.618 = 0.051 42 m2

Shell side mass velocity, Gs =

Gs = 58.884 kg/(m2 · s)
us = (Gs/ρv) = = 21.105 m/s

Equivalent diameter for triangular pitch arrangement

de =   (6.37)

= (23.81252 – 0.907 × 19.052) = 13.736 mm

Res = = 73 530

(Viscosity of MEK vapour at 83.87°C = 1100 × 10–7 Poise)

From Fig. 6.15, for 25% cut segmental baffle
Shell side friction factor, Jf = 0.0358 for Re = 73 530

Δps = 0.5 × 8 × Jf   (6.44)

= 0.5 × 8 × 0.0358
= 129 56 Pa ≡ 12.956 kPa < Δps max (13.73 kPa)
To decrease the shell side pressure drop further increase the % baffle cut from
25% to 35%, then for 35% cut segmental baffle
Jf = 0.03
Δps = × 12.956 = 10.86 kPa < Δ psmax (13.73 kPa)

Tube side pressure drop, Δpt:

Δpt = Np   (6.31)

L = 2000 mm, Np = 4, di = 15.748 mm, ρ = 993.684 kg/m3,

ut = 1.78 m/s
Res = 38 657.6, Jf = 3.41 × 10–3 (from Fig. 6.13)

Δpt = 4 ×

Δpt = 37 558 Pa ≡ 37.558 kPa < Δptmax (68.6 kPa)

Resulting Heat Exchanger Specifications:
1-4, fixed tube sheet shell and tube heat exchanger (BEM type),
Shell ID = 645 mm, Baffles: 35% cut segmental,
Baffle spacing = 645 mm,
Tube OD = 19.05 mm, 16 BWG, Tube length = 2000 mm,
Δ Pitch arrangement, Pt = 1.25 do, Number of tubes, Nt = 518
Note: As shown in this example, in case of condensation with subcooling,
inverted U-seal or dam baffle is required to accumulate condensate within the
condenser.

6.9 CONDENSATION WITH NON-

CONDENSABLES IN SHELL AND TUBE HEAT
EXCHANGERS
For the design of this type of heat exchanger, methods available are: (i) Colburn
and Hougen method, and (ii) Porter and Jaffreys method. But these methods are
complex, requiring trial and error calculations. These methods consider both
resistances; resistance provided by heat transfer and diffusional resistance
provided by mass transfer.
Approximate methods are also available for the same. In approximate methods,
only the resistance, offered by heat transfer is considered while mass transfer
resistance is neglected. Hence, in these methods it is assumed that overall rate of
heat transfer in condensation with cooling of non-condensables is totally
controlled by heat transfer resistance. One of the approximate methods is
Gilmore’s equation5.

=   (6.48)

where, hcg = Heat transfer coefficient for condensation with cooling of non-
condensables, W/(m2 · °C)
hc = Average condensate film coefficient, determined by using single component
condensation correlations, at the average condensate composition and
total condensate loading, W/(m2 · °C)
hg = Average gas film coefficient determined for the average vapour gas mixture
flow rate, W/(m2 · °C)
ϕg = Total sensible heat transfer from vapour-gas mixture, kW
ϕt = Total heat transfer, kW
ϕt = Latent heat of vaporization + sensible heat for cooling the gas-vapour
mixture.
For non-condensables, present in vapour, guidelines given by Frank (Sec. 6.8),
may be followed.
Air saturated with n-butanol vapour at 49 kPa g pressure and at
100°C temperature is to be sent to a shell and tube heat exchanger for recovering
butanol. Air-butanol vapour mixture is to be cooled to 50°C by cooling water
which enters at 32°C and leaves the heat exchanger at 40°C.
Air flow rate (butanol free) = 500 Nm3/h
Check that whether the shell and tube heat exchanger with following
specifications is suitable for the given duty earlier:
(i) Type of heat exchanger: BEM as per TEMA
(ii) Material of Construction of heat exchanger: SS-304
(iii) Shell ID = 387 mm
Tube ID/OD = 15.748/19.05 mm
(v) Tube length = 1500 mm
(vi) Tube Pitch = 25.4 mm
(vii) Number of tubes = 134
(viii) Number of tube side passes = 4
(ix) Number of shell side passes = one
(x) Baffle type = 25% cut segmental
(xi) Baffle spacing = 120 mm
Properties of fluids and material
(a) Thermal conductivity of SS-304, kw = 16 W/(m · °C)
(b) Antoine equation for butanol
ln pv = 17.2160 – where pv is in Torr and T is in K.

(c) Latent heat of vaporization of butanol at 75°C, λ = 642 kJ/kg

(d) Physical properties of butanol condensate at average condensate film
temperature
Specific heat, CL = 2.8763 kJ/(kg · °C)
Viscosity, μL = 0.95 cP or mPa · s
Thermal conductivity, kL = 0.168 W/(m · °C)
Density, ρL = 810 kg/m3
(e) Physical properties of air-butanol vapour mixture of average composition at
75°C and at the operating pressure:
Specific heat, Cp = 1.3043 kJ/(kg · °C)
Viscosity, μ = 0.0174 cP or mPa · s
Thermal conductivity, k = 0.03 W/(m · K)
Density, ρ = 2.012 kg/m3
(f) Physical properties of cooling water
CLw = 4.1868 kJ/(kg · °C), μw = 0.72 cP
kw = 0.63 W/(m · K), ρw = 1000 kg/m3
(g) Fouling coefficients hod = hid = 5000 W/(m2 · °C)
Solution
Heat duty equation for the given case
ϕt =
where, ϕt = Heat duty, kW
ṁa = Mass flow rate of air, kg/s
ṁv = Average mass flow rate of butanol vapour, kg/s
ṁBC = Condensation rate of butanol, kg/s
Cpa = Specific heat of air, kJ/(kg · °C)
Cpv = Specific heat of butanol vapour, kJ/(kg · °C)
λ = Latent heat of vaporization of butanol, kJ/kg
Cpav = Specific heat of air-butanol vapour mixture, kJ/(kg · °C)
Density of air at normal condition
ρair = (s for standard conditions)

ρair = = 1.1853 kg/m3

ṁa = 500 × 1.1853 = 592.65 kg/h ≡ 0.1646 kg/s

ṁv = Average flow rate of butanol vapour
ṁv =

where, ṁvi = Mass flow rate of butanol vapour at inlet, kg/s

ṁvo = Mass flow rate of butanol vapour at outlet, kg/s
At inlet, air is saturated with butanol vapour at 100°C and at 49 kPa g.
Vapour pressure of butanol at 100°C = pv = p–v
where p–v = partial pressure of butanol vapour at inlet
Using Antoine equation, for T = 100 + 273.15 = 373.15 K, pv = 385.617 Torr
Mole fraction of butanol vapour in incoming air-butanol mixture:
yi =
pt = 49 kPa g = 49 + 101.325 = 150.325 kPa a ≡ 1127.53 Torr a
yi = = 0.342

If ṅa = molar flow rate of air, kmol/s

ṁvi = yi × × molar mass of butanol (C4H9OH)

= 0.342 × × 74 = 0.2183 kg/s

ṁvo = yo × molar mass of butanol

yo = Mole fraction of butanol vapour in air-butanol vapour mixture at outlet

At the outlet also air is saturated with butanol vapour.

p–v′ = pv′ = vapour pressure of butanol at 50°C temperature
Using Antoine equation, for T = 273.15 + 50 = 323.15 K, pv′ = 32.8336 Torr
pt′ = Operating pressure of air – Vapour mixture of butanol at outlet = pt – Δps
where, Δps = Shell side pressure drop.
Assume shell side pressure drop Δps = 13.8 kPa = 2 psi
pt′ = 150.325 – 13.8 = 136.525 kPa a ≡ 1024 Torr a
yo = = 0.032 06

ṁvo = 0.032 06 × × 74 = 0.0139 kg/s

ṁv = = 0.1161 kg/s (average)

ṁBc = Condensation rate of butanol = ṁvi – ṁvo

= 0.2183 – 0.0139 = 0.2044 kg/s
ϕt = (0.1646 + 0.1161) × 1.3043 (100 – 50) + 0.2044 × 642
= 149.53 kW
Heat balance:
ϕt = ṁwCpw Δt′
149.53 = ṁw × 4.1868 × (40 – 32)
ṁw = 4.4643 kg/s = 16.07 m3/h

Fig. 6.19

MTD: ΔTm = Ft × ΔTlm  (6.11)

ΔTln

ΔTlm = , ΔT1 = 100 – 40 = 60°C, ΔT2 = 50 – 32 = 18°C

ΔTlm = = 34.88°C

Ft = LMTD correction factor

R = and S =

R = = 6.25, S = = 0.1176

From Fig. 6.11, for 1–4 shell and side tube heat exchanger,
Ft = 0.96
ΔTm = 0.96 × 34.88°C = 33.485°C
Tube side heat transfer coefficient, hi:

Tube side flow area, at =   (6.20)

= (0.015 748)2 = 6.525 × 10–3 m2

Tube side mass velocity, Gt =

= 648.2 kg/(m2 · s)
Tube side linear velocity ut = = 0.6842 m/s (low)

Ret =

= 14 965
Pr =

Pr = 4.7849
= 0.023 Ret0.8Pr0.33   (6.19)

hi = 0.023 × × (14 965)0.8 × (4.7849)0.33 × 1

hi = 3374.9 W/(m2 · °C)

Shell side heat transfer coefficient, ho:

=   (6.48)

Mean condensate film coefficient, hc:

For horizontal position, modified Nusselt’s equation is applicable.

hc = 0.95 kL × Nr –1/6  (6.41)

τh = = 1.017 × 10–3 kg/(m · s)  (6.23)

Nr =   (6.42)

Nr = 10
Density of vapour, ρv =

ρv = 3.666 kg/m3

hc = 0.95 × 0.168 × (10)–1/6

hc = 2042.9 W/(m2 · °C)

Average gas film coefficient, hg:

Shell side flow area, As =   (6.33)

= × 0.12 × 0.387 = 0.011 61 m2

Equivalent diameter for triangular pitch arrangement

de = (0.02542 – 0.907 × 0.019 052)

= 0.018 247 m
Shell side mass velocity Gs:

GS = = 24.18 kg/(m2 · s)

Reynolds number Res =

= 25 357
Pr =

= 0.7565
Gas film coefficient can be determined by following correlation:
 = 0.36 Res0.55 Pr1/3   (6.39)

hg = 0.36 × × (25 357)0.55 × (0.7565)1/3 × 1

= 142.59 W/(m2 · °C)

=   (6.48)

where, ϕg = (ṁa + ṁv)Cp av Δt

= (0.1646 + 0.1161) × 1.3043 (100 – 50)
= 18.306 kW
ϕt = 149.53 kW ≡ 149.53 × 103 W

ho = 741.8 W/(m2 · °C)

Overall heat transfer coefficient Uo by Eq. (6.46).
Thermal conductivity of tube material (SS–304), km = 16 W/(m · °C)

Uo = 442.13 W/(m2 · °C)

Heat transfer area required, Aor:

Aor =

= 10.1 m2
Heat transfer area provided,
Aopro = NtπdoL = 134 × π × 0.019 05 × 1.5 = 12.0293 m2
Excess heat transfer area = × 100 = × 100 = 19.1%

Hence, heat transfer area provided by heat exchanger is sufficient for the given
duty.
Shell side pressure drop Δps:

ΔpS =   (6.44)

Res = 253 57, Jf = 4.2 × 10–2 (From Fig. 6.15)

L = 1500 mm, Bs = 120 mm, Ds = 387 mm, de = 18.247 mm
ρs = Density of shell side fluid (air–vapour mixture)
Average mole fraction of butanol vapour, yav =

yav = = 0.187 03

Mav = yav × molar mass of butanol + (1 – yav) × molar mass of air

Mav = 0.187 03 × 74 + (1 – 0.187 03) × 29 = 37.4163 kg/kmol

ρs =

ρs = 1.8537 kg/m3
us = shell side linear velocity

= = 13.04 m/s (acceptable)

Δps = 8 × 4.2 × 10–

= 14 039 Pa ≡ 14.04 kPa < 49 kPa (Available pressure drop)

Calculated shell side pressure drop (Δps = 14.04 kPa) is also close to optimum
pressure drop (Δpopt = 13.8 kPa). Hence, it is satisfactory.
Tube side pressure drop, Δpt:
Δpt =   (6.31)

Np = 4, L = 1500 mm, di = 15.748 mm, ρ = 1000 kg/m3, ut = 0.6842 m/s

for Re = 14 965 from Fig. 6.13 Jf = 4.3 × 10–3

Δpt =

Δpt = 5408.4 Pa ≡ 5.408 kPa << Δpt max (68 kPa)

Hence, tube side pressure drop is very much less than maximum or optimum
pressure drop.
Note: Considering % excess heat transfer area, Δps and Δpt, given shell and
tube heat exchanger is suitable for the required heat transfer duty. If the given
shell and tube heat exchanger is at design stage (not fabricated), then number
of tube side passes could be increased from 4 to 8 because of the following
reasons:
(i) Cooling water has a tendency to foul. To avoid the excessive fouling it is
better to keep the velocity of cooling water more than 1 m/s, preferably near
2 m/s.
(ii) Increase in tube side passes will increase the tube side heat transfer
coefficient and overall coefficient. Consequently, it will decrease the
required heat transfer area and hence the fixed cost. However, it will increase
the operating cost for pumping the tube side fluid. Also in the given case,
overall heat transfer coefficient is controlled by shell side heat transfer
coefficient and hence, increase in number of tube side passes will have little
effect on the value of overall heat transfer coefficient.

6.10 MULTICOMPONENT CONDENSATION

In condensation of pure component (Example: condensation of pure ethanol
vapour), condensation is taking place in isothermal manner or at constant
temperature (i.e., degree of freedom = 1). While in case of multicomponent
condensation, during the condensation, temperature decreases from dew point to
bubble point temperature, except in the following cases:
1. Condensation of the vapour mixture of two immiscible components
(immiscible in liquid phases). Example: Vapour mixture of toluene and water
vapour.
2. Condensation of the azeotropic vapour mixture.
Example, condensation of vapour mixture containing 89.4% (by mole) ethanol
vapour and 10.6% (by mole) steam at 1 atm pressure.
In condensation of vapour mixture of miscible components (miscible in liquid
phase), condensation starts from dew point temperature and terminates at bubble
point temperature. For the calculations of dew point, bubble point and
compositions of vapour phase and liquid phase in between dew point and bubble
point, vapour–liquid equilibrium data at operating pressure for the given system
is required.
For finding the vapour-liquid equilibrium data from dew point to bubble point at
operating pressure, there are three options.
(i) Find the VLE data by conducting the actual experiment. Particularly if the
operating pressure is atmospheric or vacuum, then determination of VLE data
by actual experiment is easier.
(ii) Find the VLE data from literature.
(iii) Find the VLE data by theoretical equation.
yi = Kixi   (6.49)
where, yi = Mole fraction of component i in vapour phase
xi = Mole fraction of component i in liquid phase
Ki = Equilibrium constant for component i
Ki is a function of operating pressure, temperature as well as of composition.
For an ideal solution (i.e., the solution following Raoult’s law),

Ki = = f (T, p, xi, yi)   (6.50)

where, pvi = Vapour pressure of pure component i at the operating temperature

pt = Operating pressure
For non-ideal vapour–liquid equilibrium, provide

Ki = = f (T, p, xi, yi)   (6.51)

where, γi = Activity coefficient of component i in liquid phase

φi = Fugacity coefficient of component i in vapour phase
If the solution can be considered as ideal solution then γi = 1. (Example: for the
liquid mixture of benzene and toluene γi = 1). If vapour mixture can be
considered as ideal gas mixture then φi = 1 (Example: vapour mixture of benzene
and hexane at 1 atm pressure).
In the design of multi-component condenser calculations, weighted temperature
difference is important instead of mean temperature difference.
To calculate weighted temperature difference, entire condensing range is divided
in different small intervals. The weighted temperature difference is calculated by
the following equation:

Weighted ΔT =   (6.52)

where, ϕi = Heat load for an interval, kW

ΔTavi = Average temperature difference for the interval, °C
In condensing multicomponent mixture from dew point to bubble point, vapour
and condensate are also cooled by sensible heat transfer. Vapour is cooled
simultaneously during condensation but cooling of condensate should be
facilitated by providing the inverted U-seal or by using a dam baffle (Fig. 6.2(f)).
Cooling of condensate formed at or near the dew point, heat the condensate
formed at or near the bubble point in condensate collecting vessel and regenerate
the vapour. Hence, it is necessary to cool the condensate simultaneously with
vapour mixture from the dew point to bubble point in the condenser to avoid the
regeneration of vapour from condensate and for the cooling of condensate
certain number of tubes must be submerged in the pool of condensate which is
assured by using inverted U-seal or dam baffle.
Design of multicomponent condenser is illustrated by the following example.
Overhead vapour flows at the rate of 500 kmol/h from a
distillation column at 4 atm a. It contains saturated hydrocarbons and has the
composition; 30% n–C3H8, 25% n–C4H10, 20% n–C5H12 and 25% n–C6H14
(mole basis). It is to be condensed in a condenser. Permissible pressure drop in
the condenser is 13.8 kPa.
(a) Determine the condensing range.
(b) Compute the condensing curve.
(c) Compute the weighted average temperature difference, and
(d) Design and specify a suitable condenser.
Table 6.10 Composition of Overhead Vapour
 mole % kmol/h kg/h
Component (z × 100) ṅi Molar mass ṁi mass %
i

n–C3H8 30 150 44 6 600 20.75

n–C4H10 25 125 58 7 250 22.80

n–C5H12 20 100 72 7 200 22.64

n–C6H14 25 125 86 10 750 33.81

Total 100 500 31 800 100.00

Solution
(a) Condensing range: Condensation starts from dew point and terminates at
bubble point.
Dew point:
At dew point, Σxi = = 1 and zi = yi   (6.49)

Value of equilibrium constants Ki for hydrocarbons can be obtained from Fig.

6.20.
This calculation requires trial and error.
Assume, t = 81°C and p = 4 atm a,
KC3H8 = 5.7, KC4H10 = 2.3, KC5H12 = 0.89, KC6H14 = 0.4

Dew point temperature of vapour mixture = 81°C at 4 atm a pressure

At bubble point, Σyi = ΣKixi = 1, where xi = zi  (6.49)
 Fig. 6.20 Ki Values for Hydrocarbons at High Temperatures9 (Reproduced with the Permission of the
American Institute of Chemical Engineers, USA.)

Assume t = 31°C
KC3H8 = 2.5, KC4H10 = 0.74, KC5H12 = 0.24, KC6H14 = 0.085
ΣKixi = (0.3 × 2.5 + 0.25 × 0.74 + 0.20 × 0.24 + 0.25 × 0.085) = 1.00
Bubble point of vapour mixture = 31°C at 4 atm a
Therefore, the condensation range is from 81°C to 31°C.
(b) Condensing curve:
Divide the condensation range in five intervals; say 81°C to 70°C, 70°C to 60°C,
60°C to 50°C, 50°C to 40°C and 40°C to 31°C.
Equilibrium constant Ki values (From Fig. 6.20):
Table 6.11 Equilibrium Constants Ki Values (from Fig. 6.20)
Ki
Component
81°C 70°C 60°C 50°C 40°C 31°C

C3H8 5.7 4.9 4.2 3.6 3.0 2.5

C4H10 2.3 1.75 1.5 1.2 0.95 0.74

C5H14 0.89 0.7 0.56 0.42 0.32 0.24

C6H14 0.4 0.3 0.23 0.17 0.12 0.085

Let F = Total molar flow rate of saturated vapour mixture entering to condenser
= 500 kmol/h
First interval:
Let, L1 = Molar flow rate of condensate at the exist of 1st interval (at 70°C)
V1 = Molar flow rate of vapour at the exist of 1st interval
F = L1 + V1
Let C1 = Molar flow rate of condensate formed in the interval
for first interval C1 = L1
Component balance F zi = L1xi + V1yi   (6.53)

L1xi =   (6.54)

Assume the value of V1/L1 and find the value of L1xi by Eq. (6.54). Then ΣL1xi =
L1, F – L1 = V1 , V1/L1 = ? Value of V1/L1 obtained at the end of calculations
should be same as the assumed one. So it requires trial and error calculations.
Table 6.12 Flash Calculations for Range 81°C to 70°C
Component Fzi Ki (t = 70°C) L1 xi C1 xi λi, kJ/kg Cmpi kJ/ (kmol · K) CLi kJ/ (kg · K)

C3H8 150 4.90 10.917 10.917

 230.27 84.253 3.12
C4H10 125 1.75 22.522 22.522 301.450 111.08 2.70

C5H12 100 0.70 35.460 35.460 303.540 137.08 2.51

C6H14 125 0.30 70.225 70.225 309.820 163.17 2.43

Total 500 139.124 139.124

Assume = 2.6

= 139.124

L1 = 139.124 kmol/h
V1 = F – L1 = 500 – 139.124 = 360.876 kmol/h

= 2.5939 ≈ 2.6

Values of λi at average temperature (75.5°C) are obtained from Fig. 3.9 of Ref. 2.
Specific heat of vapour are determined by using following equation:
C3H8 : Cmpi = –5.338 + 310.24 × 10–3 T – 164.64 × 10–6 T2 + 34.691 × 10–9 T3
C4H10 : Cmpi = –1.779 + 386.96 × 10–3 T – 193.25 × 10–6 T2 + 34.833 × 10–9 T3
C5H12 : Cmpi = –3.411 + 485.01 × 10–3 T – 251.94 × 10–6 T2 + 48.677 × 10–9 T3
C6H14 : Cmpi = –4.738 + 582.41 × 10–3 T – 310.64 × 10–6 T2 + 62.923 × 10–9 T3
where Cmpi are in kJ/(kmol · K) and T is in K (Table 5.1 of Ref. 10)
Specific heat of liquids are obtained from Fig. 4 of Ref. 5 at average temperature
(75.5°C).
Values of Cmp and CL are at 1 bar a in ideal state. However, these are not
significantly affected under pressure up to 10 bar.
For the interval 81°C to 70°C,
ϕ1 = ΣC1xiλi + Cmpi Δt + CLi Δt

ϕ1 = (10.917 × 230.27 × 44 + 22.522 × 301.45 × 58 + 35.46 × 303.54 × 72

+ 70.225 × 309.82 × 86) + × 84.253 × 11

+ × 111.08 × 11

+ × 137.08 × 11

+ × 163.17 × 11 + (3.12 × 44) × 11

+ × (2.7 × 58) × 11 + (2.51 × 72) × 11

+ × (2.43 × 86) × 11

= 3150 472.1 + 558 323.68 + 143 602.7

= 3852 398.5 kJ/h ≡ 1070.1 kW
Second interval:
Table 6.13 Flash Calculations for Range 70°C to 60°C
ComponentV1yi (= Fzi – L1xi)Ki(t = 60°C) L2xi V2yi (= Fzi – L2xi) C2xi'*

C3H8 139.083 4.2 24.834125.166 13.917

C4H10 102.478 1.5 44.64380.357 22.121

C5H12 64.540 0.56 59.81040.190 24.350

C6H14 54.775 0.23 97.96227.038 27.737

Total 360.876 227.25272.75 88.125

*x′ represents mole fraction of liquid component, condensed in the specified

i
interval.
For the second interval let L2 is total moles of condensate leaving at 60°C and C2
is moles of condensate formed in second interval.

L2 = ΣL2xi =

where, V2 = Total moles of vapour at 60°C, leaving the second interval

L2 = ΣL2xi =
Assume = 1.2

L2 = ΣL2xi = 24.834 + 44.643 + 59.81 + 97.962 = 227.25 kmol/h

F = V2 + L2
V2 = F – L2 = 500 – 227.25 = 272.75 kmol/h

= = 1.2 (Check)

V2yi = Fzi – L2xi, C2x′i = V1yi – V2yi

Table 6.14 Calculated Properties at 65°C
Componentλi, kJ/molCmpi, kJ/(kmol · K)CLi, kJ/(kmol · K)

C3H8 11 605.81 82.053 132.64

C4H10 18 212.58108.28 150.56

C5H12 22 608.72133.62 177.86

C6H14 28 445.12159.06 205.24

For the interval 70 to 60°C,

ϕ2 = ΣC2x′iλi + Cmpi Δt + CLi Δt

ϕ2 = (13.917 × 11 605.81 + 22.121 × 18 212.58 + 24.35 × 22 608.72

+ 27.737 × 28 445.12)
+ × 82.053 × 10 + × 108.28 × 10

+ × 133.62 × 10 + × 159.06 × 10

+ × 132.64 × 10 + × 150.56 × 10

+ × 177.86 × 10 + × 205.24 × 10

= 1903 903.2 + 342 435 + 331 589

= 2577 927.2 kJ/h ≡ 716.1 kW
 Table 6.15 Flash Calculations for Range 60°C to 50°C
Component V2yi Ki (t = 50°C) L2xi L3xi V3yi C3x′i

C3H8 125.1663.60 24.834 48.5751 101.42523.741

C4H10 80.357 1.20 44.643 73.7028 51.297 29.060

C5H12 40.190 0.42 59.810 80.4117 19.588 20.600

C6H14 27.038 0.17 97.962 113.7812 11.220 15.820

Total 272.750 227.250316.4708183.53089.220

Value of L3xi, V3yi and C3x′i of above table are obtained via following
calculation:
L3 is total mole of condensate at 50°C.

L3 = ΣL3xi =

L3 = ΣL3xi =

Trial: Assume = 0.58

L3 = ΣL3xi =

L3 = ΣL3xi = 48.5751 + 73.7028 + 80.4117 + 113.7812 = 316.4708 kmol/h

V3 = 500 – 316.4708 = 183.5292 kmol/h

= 0.5799 ≅ 0.58 (Check)

C3x′i = (L3xi)3 – (L2xi)2 = (V2yi)2 – (V3yi)3

Table 6.16 Calculated Properties at 55°C
Componentλi, kJ/kmolCmpi, kJ/(kmol · K)CLi, kJ/(kmol · K)

C3H8 12 895.3 79.932 128.95

C4H10 19 183.9 105.58 148.13

C5H12 23 513.1 150.29 173.33

C6H14 28 805.2 155.09 201.64

ϕ3 = ΣC3x′iλi + CmpiΔt + CLiΔt

ϕ3 = (23.741 × 12 895.3 + 29.06 × 19 183.9 + 20.6 × 23 513.1 + 15.82 × 28

805.2)
+ × 79.932 × 10 + × 105.58 × 10

+ × 150.29 × 10 + × 155.09 × 10

+ × 128.95 × 10 + × 148.13 × 10

+ × 173.33 × 10 + × 201.64 × 10

= 1803 699.6 + 234 646.85 + 469 986 = 2508 332.5 kJ/h ≡ 696.76 kW
Fourth interval:
Table 6.17 Flash Calculation for Range 50°C to 40°C
Component V3yi Ki(t = 40°C) L3xi L4xi V4yi C4x′i

C3H8 101.425 3.00 48.5751 85.714 64.28637.14

C4H10 51.297 0.95 73.7028 101.01 23.99027.31

C5H12 19.588 0.32 80.4117 92.593 7.407 12.18

C6H14 11.220 0.12 113.7812 121.36 3.640 7.58

Total 183.53 316.4708400.67799.32384.21

Value of L4xi, V4yi and C4x′i of above table are obtained by the following
calculations:

L4 = ΣL4xi =

∴ L4 = ΣL4xi =

Trial: Assume = 0.25 (Check)

L4 = 85.714 + 101.01 + 92.593 + 121.36 = 400.677 kmol/h

V4 = 99.323, = 0.248 ≅ 0.25

V4yi = Fzi – L4xi, C4x′i = (L4xi)4 – (L3xi)3

ϕ4 = ΣC4x′iλi + Cmpi Δt + CLi Δt

Table 6.18 Calculated Properties at 45°C

Componentλi, kJ/kmolCmpi, kJ/(kmol · K)CLi, kJ/(kmol · K)

C3H8 13 448.0 77.785 125.27

C4H10 19 912.4 102.850 143.27

C5H12 24 116.0 126.910 168.81

C6H14 29 525.3 151.440 198.04

ϕ4 = ΣC4x′iλi + CmpiΔt + CLiΔt

= (37.14 × 13 448 + 27.31 × 19 912.4 + 12.18 × 24 116 + 7.58 × 29 525.3)

+ × 77.785 × 10

+ × 102.85 × 10 + × 126.91 × 10

+ × 151.44 × 10 + × 125.27 × 10

+ × 143.27 × 10 + × 168.81 × 10

+ × 198.04 × 10

ϕ4 = (156 0801 + 131 547.2 + 588 128.9) = 2280 477.1 kJ/h ≡ 633.46 kW
Fifth (final) Interval:
Table 6.19 Flash Calculations for Range 40°C to 31°C
Component V4yi V5yi L4xi C5xi L5xi = Fzi

C3H8 64.286 0 85.714 64.286 150

C4H10 23.990 0 101.01023.990 125

 C5H12 7.407 0 92.593 7.407 100

C6H14 3.640 0 121.3603.640 125

Total 99.323 400.67799.323 500

Table 6.20 Calculated Properties at 35.5°C

Componentλi, kJ/kmolCmpi, kJ/(kmol · K)CLi, kJ/(kmol · K)

C3H8 14 553.3 75.72 123.43

C4H10 20 155.3 100.23 140.84

C5H12 24 718.9 123.67 165.8

C6H14 30 245.4 147.22 194.4

ϕ5 = ΣC5x′iλi + CmpiΔt + CLiΔt

ϕ5 = (64.286 × 14 553.3 + 23.99 × 20 155.3 + 7.407 × 24 718.9 + 3.64 × 30

245.4)
+

+ × 123.43 × 11 + × 140.84 × 11

+ × 165.8 × 11 + × 194.4 × 11

= 17 12285.2 + 47 806.1 + 77 4123.5 = 25 34214.8 kJ/h ≡ 703.95 kW

Total heat duty ϕt = ϕ1 + ϕ2 + ϕ3 + ϕ4 + ϕ5
= 1070.1 + 716.1 + 696.76 + 633.46 + 703.95 = 3820.37 kW
Assume chilled water as a cooling medium.
Let inlet temperature of chilled water, ti′ = 6°C
Outlet temperature of chilled water, to′ = 9°C
Energy balance:

ϕt = = 3820.37 kW = m·w × 4.1868 × (9 – 6)

Mass flow rate of chilled water required

m·w = = 304.16 kg/s

≡ 1094 976 kg/h ≡ 1095 m3/h

ΔTavi (LMTD) for 1st interval = = 66.79°C

Ft = 1
Table 6.21 Summary of Heat Balance Calculations
Interval (Based on vapour temperature)Temperatures of chilled waterHeat duty ϕ, kWΔTavi, (LMTD)

81°C to 70°C 8.16°C to 9°C 1070.1 66.79°C

70°C to 60°C 7.6°C to 8.16°C 716.1 56.99°C
60°C to 50°C 7.05°C to 7.6°C 696.76 47.52°C
50°C to 40°C 6.553°C to 7.05°C 633.46 38°C
40°C to 31°C 6°C to 6.553°C 703.95 29.02°C

Total 3820.37

Figure 6.21 gives the results in diagrammatic form.

Fig. 6.21 Condenser for Condensation of Hydrocarbon Vapours

Weighted ΔT =   (6.52)
Weighted ΔT =

= 45.385°C
Heat duty for subcooling:
ϕsub = 143 602.7 + 331 589 + 469 986 + 588 128.9 + 774 123.5
= 2307 430.1 kJ/h ≡ 640.95 kW ≡ 640.95 × 103 W
Heat duty for condensation:
ϕc = ϕt – ϕsub = 3820.37 – 640.95 = 3179.42 kW ≡ 3179.42 × 103 W
This value includes heat load required for cooling of vapour. Heat transfer
coefficient for the cooling of vapour in presence of condensation is quite high.
Also, cooling of vapour and condensation of vapour, both are taking place
simultaneously. Hence, as per the Kern’s suggestion, separate calculation for
determining heat transfer coefficient for cooling of vapour is not required.
Assume values of overall coefficients for the first trial calculation.
Overall coefficient for condensation Uoc = 800 W/(m2 · °C)
Overall coefficient for subcooling Uosub = 200 W/(m2 · °C)
Provisional heat transfer area:
For condensation, Acpro =

= = 87.568 m2

For subcooling, Asub, pro = = 70.61 m2

Total heat transfer area Apro = 87.568 + 70.61 = 158.18 m2

Let tube length L = 8 ft = 2.4384 m
Tube outside diameter do = 3/4 in = 0.019 05 m

Number of tubes, Nt =

Nt = 1084
Let tube pitch Pt = 1.25 do = 0.023 81 m, triangular pitch arrangement, number
of tube side passes = 4 (for first trial calculations)
Tube bundle diameter
Db = do   (6.1)

for Np = 4, k1 = 0.175, n1 = 2.285 (Table 6.2)

Db = 19.05 = 870 mm

Let the clearance between shell inside diameter and tube bundle diameter = 16
mm.
Shell ID Ds = 870 + 16 = 886 mm
Tube side heat transfer coefficient hi:
Allocate chilled water on tube side
Tube side flow area, at =   (6.20)

Tube inside diameter, di = 15.748 mm (For 16 BWG tube)

at = (0.015 748)2 = 0.052 785 m2

Tube side mass velocity Gt = = 5762.24 kg/(m2 · s)

Density of water at 7.5°C, ρ = 999.877 kg/m3 (From Table 3.28 of Ref. 2)

Tube side velocity, ut = = 5.76 m/s (very high)

Velocity of water in tubes should preferably be in between 1 to 3 m/s.

Let number of tube side passes = 1
For Np = 1, k1 = 0.319, n1 = 2.142 (From Table 6.2)

Db = 19.0 = 848 mm

Shell ID, Ds = 848 + 16 = 864 mm (Revised)

Tube side flow area, at = (0.015748)2 = 0.2111 m2

Gt = = 1440.83 kg/(m2 · s)

ut = = 1.44 m/s which is within acceptable range

Tube side Reynold’s number,

Viscosity of water at 7.5°C, μ = 1.3 × 10–3 kg/(m · s)

Ret = = 17 454

Prandtl number, Pr =

Thermal conductivity of water at 7.5°C, k = 0.596 85 W/(m · °C)

Pr = = 9.12

Dittus–Bolter equation

= 0.023 Re0.8 Pr0.3   (6.19)

hi = 0.023 × × (17 454)0.8 × (9.12)0.33 × 1

hi = 4473.85 W/(m2 · °C)

ho: Heat transfer coefficients for both condensation and subcooling must be
determined separately.
For condensation
For the first trial calculation let = 1500 W/(m2 · °C)
(ts – tw) = Uoc(ts – tt)
where, Uoc = 800 W/(m2 · °C)
tw = Tube wall temperature, °C
ts = shell side average temperature = = 56°C

tt = Tube side average temperature = = 7.5°C

1500(56 – tw) = 800(56 – 7.5)
tw = 30.13°C
Mean temperature of condensate film, tf = = 43°C

Physical properties of liquid condensate should be determined for the average

condensate composition. Here, properties are determined based on the
composition of condensate in the intermediate (third) interval.
Table 6.22 Calculated Properties of Liquid Condensate in Third Interval
Component C3x′i Mole fractionMass fraction ρL, kg/m3 μ*, cPk, W/(m · °C)

C3H8 23.741 0.2660 0.1874 582 0.10 0.1211

C4H10 29.060 0.3257 0.3024 579 0.15 0.1263
C5H12 20.600 0.2309 0.2661 626 0.20 0.1305
C6H14 15.820 0.1773 0.2441 659 0.27 0.1367

Total 89.221 1.0000 1.0000

* For viscosity, Fig. 14 of Ref. 5

ρL =   (6.55)

= = 609.84 kg/m3

Viscosity of nonpolar liquid mixture by Irving’s equation (Eq. (3.107) of Ref. 2)

ln μLmix = Σwi ln μi  (6.56)
= [0.1874 ln (0.1) + 0.3024 ln (0.15) + 0.2661 ln (0.2) + 0.2441 ln (0.27)]
μLmix = 0.1732 cP ≡ 0.1732 × 10–3 kg/(m · s)
kLmix = Σkixi = 0.266 × 0.1211 + 0.3257 × 0.1263
+ 0.2309 × 0.1305 + 0.1773 × 0.1367
= 0.1277 W/(m · K)
Average molar mass of vapour mixture,
Mav = ΣMiyi = 0.3 × 44 + 0.25 × 58 + 0.2 × 72 + 0.25 × 86 = 63.6 kg/kmol
Density of vapour mixture, ρv =

= 9.418 kg/m3

Position of condenser: Horizontal

Shell side condensation coefficient
= 0.95 k   (6.41)

τh = , Nr = ≅ 24

= (23.741 × 44 + 29.06 × 58 + 20.6 × 72 + 15.82 × 86)/3600

= 1.548 kg/s
τh = = 5.856 × 10–4 kg/(m · s)

= 0.95 × 0.127

= 0.95 × 0.1277 × 32 839.57 × 0.5888 = 2345.7 W/(m2 · °C)

For the second trial calculation if it is assumed = 2345.7 W/(m2 · °C)
(resulting value of first trial calculation) condensate film temperature, tf =
47.73°C (revised). Hence, minor change in the value of film temperature is
noted. Also, precise temperature wise physical properties of these components
(particularly density and thermal conductivity) are not available. So, second trial
calculation will not make any significant difference in the final value.
= 2345.7 W/(m2 · °C)
As stated earlier with horizontal position, subcooling of condensate is taking
place by natural convection only. No reliable correlation is available for natural
convection. Hence, based on Kern’s recommendation, assume
hosub = 50 Btu/(h · ft2 · °F) = 283.77 W/(m2 · °C)
Overall heat transfer coefficient for condensation, Eq. (6.46):
For light hydrocarbon mixture hod = 5000 W/(m2 · °C) and
For cooling water hid = 4000 W/(m2 · °C) (From Table 6.9)
Thermal conductivity of tube material
kw = 50 W/(m · °C) (Steel or cupronickel tube)

Uoc = 809.47 W/(m2 · °C)

Heat transfer area required for condensation
Acreq =

Acreq = 86.54 m2
Overall heat transfer coefficient for subcooling

Uosub = 230.78 W/(m2 · °C)

Asub =

Asubreq = 61.19 m2
Total heat transfer area required,
Areq = 86.54 + 61.19 = 147.73 m2
Total heat transfer area provided,
Apro = 158.16 m2

Excess heat transfer area = × 100 = 7.06 %

It is less than 10%.

If number of tubes is increased to 1200,
Db = 889, shell ID = 905 mm, Nt = 1200
at = × 0.2111 = 0.2337 m2

Gt = = 1301.5 kg/(m2 · s)

Re = × 17 454 = 15 766.18

ut = Gt/ρ = = 1.3016 m/s

hi = 4473.85 ×

hi = 4124.26 W/(m2 · °C)

∝

Nr = (2/3) = 24.9 ≅ 25

= 2345.7 × = 2410 W/(m2 · °C)

Uoc = 801.97 W/(m2 · °C)

Heat transfer area required for condensation

Acreq =

Acreq = 87.35 m2
Uosub = 229.57 W/(m2 · °C) (based on new value of hi)

Asubreq = = 61.52 m2

Total area required, Areq = 87.35 + 61.52 = 148.87 m2

Apro = Ntπ doL = 1200 × π × 0.019 05 × 2.4384 m = 175.12 m2
Excess heat transfer area = × 100 = 17.63% (adequate)

If equal % excess heat transfer area for condensation and subcooling are
considered then,
Area provided for condensation Ac pro = 1.1763 × Ac req
Acpro = 102.75 m2
Asub pro = 175.12 – 102.75 = 72.37 m2

= = 0.4133

Thus, 41.33% of total heat transfer area is available for subcooling and the same
area should be submerged in the pool of condensate. Assuming that tubes will be
uniformly distributed in the cross section of shell.

= 0.4133 =

= 0.4133 or x = 0.3246

From Table 1.19 b of Ref. 2:

for x = 0.3246, = 0.432

To facilitate the subcooling of condensate or to submerge 41.33% of total heat

transfer area in the pool of condensate, height of inverted U-seal (h) required
from the bottom most point of shell inside diameter.
h = 0.432 × 905 = 390.96 mm = 391 mm (as shown in Fig. 6.21)

6.11 TINKER’S FLOW MODEL

All the latest methods (Delaware method; Sec. 6.12), for the process design of
without phase change heat exchangers, are based on Tinker’s flow model. If
there is no phase change in shell side fluid then shell side heat transfer
coefficient and shell side pressure drop in all recent methods are calculated
based on this model. In old methods, like Kern’s method, it is assumed that
entire shell side fluid is flowing across the tube bundle and between the baffles.
But actually shell side fluid is flowing in various ways. In Tinker’s flow model
shell side flow is divided in five different steams.
 Fig. 6.22 Tinker’s Flow Model11

Stream A Stream A is the tube-to-baffle leakage stream or it is the fraction of

shell side fluid flowing through the clearance between tube hole in baffle and
tube outside diameter. This stream does not bypass the heat transfer area (outside
area of tubes) and hence, it does not create any adverse effect on the value of
heat transfer coefficient. However, it makes the significant difference in pressure
drop (loss). When stream A leaves this clearance it forms free flowing jet.
Hence, boundary layer separation occurs and considerable friction loss or
pressure drop takes place. Effect of stream A must be considered in the
calculation of pressure drop.
Stream B Stream B is the actual cross flow stream or it is the fraction of shell
side fluid which is flowing across the tube bundle and between the baffles. In
Kern’s method and other old methods it is assumed that entire shell side fluid is
flowing like stream B.
Stream C Stream C is bundle to shell bypass stream or it is the fraction of shell
side fluid flowing through the clearance area between shell inside diameter and
tube bundle. Stream C is the main bypass stream. (bypassing the heat transfer
area). This clearance area provides low pressure drop path for the shell side
fluid. Hence, % of shell side fluid bypassed through this clearance area is quite
significant. It is maximum with pull through floating head heat exchanger and
minimum with fixed tube sheet heat exchanger. Amount of stream C can be
reduced considerably by using sealing strips which are attached on inside surface
of shell (See Fig. 6.7). They provide the partial blockage or the additional
resistance in the path of stream C.
There is no stream designated as stream “D”.
Stream E Stream E is the baffle to shell leakage stream. It is a part of shell side
fluid flowing through the clearance between the edge of baffle and shell wall.
Like stream C, this stream is also bypassing the heat transfer area and hence,
reduces shell side heat transfer coefficient. But amount of stream E is lesser than
amount of stream C. Normally, the clearance between baffle outside diameter
and shell inside diameter is in the range of 1.6 to 4.8 mm while clearance
between tube bundle and shell inside diameter is in the range of 10 to 100 mm.
Stream F Stream F is the pass-partition bypass stream. In tube sheets where
pass partition plates are attached (for forming tube side passes), in that portion
tubes cannot be provided. In multipass heat exchangers, one can find more
number of gaps in tube bundle. Stream F is the fraction of shell side fluid
flowing through these gaps. If the gap is vertical it provides low pressure drop
path for fluid flow. Just like stream C and stream E, this stream is also bypassing
the heat transfer area and reduces shell side heat transfer coefficient. Amount of
stream F and its adverse effects are significant in multipass heat exchanger. To
reduce the amount of this stream sometimes dummy tubes are used.

6.12 DELAWARE METHOD FOR DESIGN OF

BAFFLED SHELL AND TUBE HEAT
EXCHANGERS
It was seen in Sec. 6.11 that flow path across the tubes and various bypasses on
shell side of a shell and tube heat exchanger are far more complicated than
assumed in Kern’s method. Provision of excess heat transfer area in Kern’s
method was just to account for the correction in the shell side heat transfer
coefficient (ho) due to the disturbed fluid flow path. In a competitive
environment, it became necessary to design a baffled shell and tube heat
exchanger with reasonable accuracy.
Prof. A. P. Colburn, Prof. O. P. Bergelin and their other colleagues undertook a
collaborative efforts at the University of Delaware, USA to define various
corrections for the fluid path B and other bypasses on shell side. Dr. K. J. Bell, a
Research Fellow of the program, compiled the finding and prepared a report15 on
the subject. These findings and correlations are updated from time to time16. The
method, developed on the basis of the report, is known as Delaware method
(also popularly known as Bell’s method). This method gives more satisfactory
predictions of shell side heat transfer coefficient (ho) and its pressure drop (Δps)
for cooling and heating of shell side fluid. An important change in definition of
Reynolds number (Reo) is based on mass velocity through the minimum cross-
sectional area in tube banks.
Reo =   (6.57)

where, do = Outside diameter of tube, m

Gs = Mass velocity on shell side, kg/(m2 · s)
=
μ = Viscosity of shell side fluid at average temperature, kg/(m · s)
us = Gs/ρ m/s  (6.58)
As is shell side flow area which is determined by following correlations:
For triangular layout of tube bundle,

As = Bs   (6.59a)

and for linear and rotated square tube layouts,

As = Bs   (6.59b)

where, Bs = Baffle spacing,

Di = Inside diameter of shell, m
Db = Tube bundle diameter, m
do = Outside diameter of tube, m
Pt = Tube pitch, m
Pn = Tube pitch normal to flow or distance between centres of adjacent tubes
normal to flows, m
Note that correlation, (Eq. 6.58) for As is different from As in Eq. (6.33) of
Kern’s method.
Based on Reo, defined in Eq. (6.57), flow regimes are defined as under which are
different from Reo defined in Eq. (6.38).
0 < Reo < 100 Laminar
100 < Reo < 4000 Transition
4000 < Reo Turbulent
In this book, correlations for Reo > 100 only are considered.

6.12.1 Shell Side Heat Transfer Coefficient

Heat transfer coefficient (ho) for no phase change (of fluid) is now defined as
ho = hoc Jc Jb Jl Js  (6.60)
where, hoc = Shell side heat transfer coefficient, calculated for cross flow (stream
B of (Fig. 6.22) over ideal tube bank with no leakage or bypassing,
W/(m2.°C).
Jc = Correction factor to consider the effect of baffle configuration
Jb = Bypass stream correction factor
Jl = Baffle leakage correction factor
Js = Correction factor for unequal baffle spacing at inlet/outlet
Heat transfer coefficient (hoc) can be calculated using following the correlation:

  (6.61)

where, Jh = Colburn J-factor for an ideal tube bank

Values of Jh can be read from Fig. 6.23.
To consider the effect of window zone on shell side heat transfer coefficient,
fraction (Fc) of total tubes that are in cross flow is found from Fig. 6.24.
Correction factor Jc can then be read from Fig. 6.25 as a function of Fc. For no
tubes in the window zone, Jc = 1.0.
To consider the effect of bypass streams C and F (see Fig. 6.22), fraction of
cross-flow area, available for bypass flow, is calculated by the following
correlation:

  (6.62)

where, npv = Number of vertical gaps in tube bundle created by vertical pass
 partition plates or number of pass partition lanes, through the tube field,
parallel to the direction of cross-flow stream
Wp = Width of above gaps or lanes, m
= Nominal pass partition plate thickness + 0.006 35 m
Bs = Central baffle spacing, m
Fig. 6.23 Correlation of Colburn Jh for Ideal Tube Banks with Plain Tubes16

Fig. 6.24 Estimation of Fraction of Tubes in Cross Flow16

Table 6.23 Nominal Pass Partition Plate Thickness
Nominal size of shell, mm Plate thickness, mm

Carbon steelAlloy material

 610 9.5 6.35
610 – 1524 12.7 9.5
1549 – 2540 15.9 12.7

Fig. 6.25 Correction Factor for Baffle Configuration Effects16

 Fig. 6.26 Correction Factor for Bypass Stream16

Adverse effect of bypass stream on heat transfer coefficient can be significantly

reduced by providing sealing strips (see Fig. 6.7).
Let Nss = Number of pairs of sealing strips, encountered by bypass stream (C in
Fig. 6.22) in cross-flow zone
Nc = Number of tube rows crossed during flow through one cross-flow section

= (for pitch arrangement)  (6.63a)

= (for pitch arrangement)  (6.63b)

= (for pitch arrangement)  (6.63c)

x = Fraction of baffle cut (see Fig. 6.2)

Bypass stream correction factor Jb can be read from Fig. 6.26, knowing Fsbp and
Nss/Nc ratio.
To consider the effect of leakage streams A and E on ho, tube to baffle leakage
area can be determined by the following correlation:
  (6.64)

where, Ctb = Diametrical clearance between tube hole in baffle and OD of tube
= 0.8 mm or 0.0008 m (normal value)
Fc = Fraction of total tubes in cross flow
Shell to baffle leakage area for one baffle, Asb, can be determined by the
following correlation:

Asb =   (6.65)

where, Csb = Diametrical clearance between shell and baffle, m

α = Baffle cut angle or angle subtended by intersection of the cut edge of the
baffle with inside surface of shell, radians
= 2 cos–1(1 – 2 x) radians  (6.66)
Calculations of Atb, Asb and As yield baffle leakage correction factor Jl from Fig.
6.27.
Correction factor, Js, for unequal baffle spacing at inlet and/or outlet can be
determined by the following correlation:

Js =   (6.67)

where, Nb = Number of baffles

Rsi = Bsi/Bs
Rso = Bso/Bs
Bs = Internal (central) baffle spacing, m
Bsi = Entrance baffle spacing, m
Bso = Exit baffle spacing, m
n = 0.6 for Reo > 100; i.e. for transition and turbulent flows

Nb =   (6.68)
Substitution of various parameters in Eq. (6.60) will yield ho.

Fig. 6.27 Correction Factor for Baffle Leakage Effects16

6.12.2 Shell Side Pressure Drop

Δps = [(Nb – 1) (Δpbi) Rb + Nb Δpwi]Rl

+ 2 Δpbi Rb   (6.69)
where, Ncw = Number of effective cross-flow rows in each window section

= for pitch arrangement  (6.70a)

= for pitch arrangement  (6.70b)

= for pitch arrangement  (6.70c)

Δpbi = Pressure drop during flow across one ideal cross-flow section, Pa
Δpwi = Pressure drop through one ideal window section, Pa
Rb = Correction factor for effect of bundle bypass
Rl = Correction factor for baffle leakage effect
Rs = Correction factor for unequal baffle spacing
Pressure drop during flow across an ideal cross-flow section, Δpbi, can be
determined using the following correlation:

Δpbi = 8 Jf Nc   (6.71)

where, Jf = Friction factor for ideal tube bank. It can be obtained from Fig. 6.28.
Nc = Number of tube rows crossed during one flow through one cross-flow
section
ρs = Density of shell side fluid, kg/m3
Equation (6.71) is similar to Eq. (6.44) with a few changes.
Pressure drop through one ideal window section, Δpwi, can be determined using
the following correlation:
Δpwi =   (6.72)

where,
Ncw = Number of effective cross-flow rows in each window section
Aw = Flow area of window, m2
= Awg – Awt  (6.73)
Awg = Gross window area, m2
=   (6.74)

Fig. 6.28 Correlation of Friction Factors for Ideal Tube Banks16

 Fig. 6.29 Correction Factor for Pressure Drop for Tube Bundle Bypass Flow16

Awt = Window area occupied by tubes, m2

=   (6.75)

Correction factor, Rb, for the effect of tube bundle bypass can be read from Fig.
6.29.
Correction factor, Rl, for the effect of baffle leakage can be read from Fig. 6.30.

Fig. 6.30 Correction Factor for Baffle Effect on Pressure Drop16

Correction factor, Rs, for unequal baffle spacing can be determined using the
following correlation:

Rs =   (6.76)

where, n' = 1.6 for turbulent flow (Reo > 100)

Substitution of various parameters in Eq. (6.69) yields predicted pressure drop
on shell side. In actual practice, it is observed15 that Δps, calculated by
correlation Eq. (6.69) ranged from 0.75 to 1.05 times the measured overall
pressure drop with an average value of 0.9. This pressure drop does not account
for nozzle (pressure) losses.
Delaware method is much more detailed than given in this Sec. 6.12 and
contains variety of correlations for different tube pitches, types of tubes (such as
finned) and other configurations. Reader is advised to refer Ref. 15 and 16 for
other correlations.
Synthesis make-up gas of ammonia plant is to be cooled in a
shell and tube heat exchanger from 320°C to 155°C by exchanging heat with
boiler feed water (BFW) entering at 125°C in countercurrent direction. Total
(wet) volumetric gas flow rate is 140 000 Nm3/h which flows through the tubes
at 27.35 bar a while BFW flows at the rate of 41 500 kg/h in the shell at 125 bar
a. Dry gas composition is 74% H2, 24.7% N2, 0.3% Ar and 1% CH4 by mole.
Stoichiometric calculations reveal that presence of water in the gas mixture is
0.0205 kmol/kmol dry gas. Following heat exchanger is available for the cooling
service. Check whether the heat exchanger will be adequately rated for the
service by Delaware method.
Type of heat exchanger: 1 – 1 shell and tube BEM type
Shell diameters = 964.4 mm (outside), 939 mm (inside)
No. of tubes = 1150
Details of tubes: MS, OD = 19.05 mm, ID 14.8336 mm (14 BWG)
Tube length = 9140 mm
Tube layout: Triangular pitch (Δ) = 25.4 mm
Tube bundle diameter (Db) = 925 mm
Baffles: 25% cut segmental
Baffle spacings: 300 mm (central), 370 mm (at inlet and outlet)
Diametrical clearance between tube and baffle = 0.8 mm
Thermodynamic and transport properties of gas mixture components and water
are listed in Table 6.24.
Solution
Basis: Total volumetric gas flow rate = 140 000 Nm3/h (wet)
Molar gas flow rate, ṅ = = 6246.096 kmol/h (wet)
Moisture content of gas, Hm = 0.0205 kmol/kmol dry gas
Thus 1.0205 kmol wet gas mixture contains 0.74 kmol H2, 0.247 kmol N2, 0.003
kmol Ar, 0.01 kmol CH4 and 0.0205 kmol H2O.
On this basis, mole fractions of the components of the wet gas mixture are
calculated to be 0.7251 H2, 0.242 N2, 0.0029 Ar, 0.0098 CH4 and 0.0202 H2O.
Nitrogen in the gas mixture, = 6246.096 × 0.242
= 1511.555 kmol/h
Similarly, = 6246.096 × 0.725
= 4529.044 kmol/h
= 6246.096 × 0.0029
= 18.114 kmol/h
= 6246.096 × 0.0098
= 61.212 kmol/h
= 6246.096 × 0.0202
= 126.17 kmol/h
Table 6.24 Thermodynamic and Transport Properties
Property value*
ComponentPhasePres., bar aTemp., °C Enthalpy** Viscosity Density Heat capacity
Thermal cond. (k), mW/(m · K)
(H), kJ/kg (μ), μPa · s (ρ), kg/m3 (Cp), kJ/(kg · °C)

Hydrogen Gas atm 155 5807.8 11.407 246.71 0.0574 14.492

M = 2.0159 320 8203.1 14.277 324.81 0.0414 14.546
Gas 19.8315 155 5920.4 11.438 247.85 1.1131 14.514
320 8218.1 14.250 325.70 0.8053 14.555
Nitrogen Gas atm 155 566.92 23.334 34.615
 0.7971 1.0473
M = 28.0134 320741.6429.345 44.4550.57631.0737
Gas 6.6187155560.8623.381 34.7475.19771.0513
320741.6229.370 44.5493.74951.0755
Methane Gas atm 1551224.314.982 54.7670.45682.6340
M = 16.0425 3201709.519.245 87.3560.32963.2479
Gas 0.26801551224.714.975 54.7330.12082.6323
3201709.719.239 87.3320.08723.2471
Argon Gas atm 155340.2030.289 23.7431.13700.5209
M = 39.948 320426.1138.672 30.3200.82070.5206
Gas 0.0793155340.2830.259 23.7210.09020.5204
320426.1430.665 30.3030.06510.5204
Water Gas atm 1552367.414.379 29.2620.51691.9827
M = 18.0153 3202695.821.122 45.6100.37092.0232
Gas 0.5525 1552370.514.401 28.9900.28081.9513
3202696.821.126 45.5440.20212.0181
Water Satd. liquid 2.322 125106.02222.080683.56939.024.2521
M = 18.0153 64.166 280817.8393.506 581.15750.285.2889
Compr. liquid125 125114.53225.270690.96945.164.2210
280814.8895.472 591.64759.905.1385

* Ref.: REFPROP Ver. 8.0, NIST Standard Reference Data Base 23, 2007
** Reference enthalpy of saturated liquid (h) = 0 kJ/kg at NBP
Average molar mass of the wet gas mixture,
Mmix = 0.7251 × 2.0159 + 0.242 × 28.0134 + 0.0029 × 39.948 + 0.0098 ×
16.0425 + 0.0202 × 18.0153
= 8.8778 kg/kmol
Properties of each component at respective partial pressure (pi) are required.
pi = yi p
where, p = total pressure = 27.35 bar
In Table 6.24, properties are listed at atmospheric pressure as well as at
respective partial pressure for each component. Study of the property data reveal
that except for the density (ρ), other properties are close enough at two pressures.
For heat transfer rate, evaluations are made at both pressures.
At atmospheric pressure,
ϕ = (8203.1 – 5807.8) 2.0159 × 4529.044 + (741.64 – 566.92) 28.0134 ×
1511.555
+ (1709.5 – 1224.3) 16.0425 × 61.212 + (426.11 – 340.20) 39.948 × 18.114
+ (2695.8 – 2367.4) 18.0153 × 126.171
= 30 552 721 kJ/h ≡ 8486.867 kW
At partial pressures,
ϕ′ = (8218.1 – 5920.4) 2.0159 × 4529.044 + (741.62 – 560.86) 28.0134 ×
1511.555
+ (1709.7 – 1224.7) 16.0425 × 61.212 + (426.14 – 340.28) 39.948 × 18.114
+ (2696.8 – 2370.5) 18.0153 × 126.171
= 29 912 374 kJ/h ≡ 8038.993 kW ≡ 8038.993 × 103 W
Gas mixture’s heat transfer rate can also be calculated with the use of ideal gas
heat capacity data and considering the mixture to be an ideal one (Ref. 10).
C°p mix =

= 28.8715 – 0.338 × 10–3 T + 4.0491 × 10–6 T2 – 1.1367 × 10–9 T3

ϕ″ =

= 30 514 829 kJ/h ≡ 8476.341 kW

ϕ′ differ by –2.14% from ϕ and –2.01% from ϕ″. Since actual enthalpy values
could be determined at respective partial pressures, heat transfer rate at partial
pressures (ϕ′) will be considered for the heat exchanger design.
For the first estimation of BFW outlet temperature (t′), consider Cp = 4.1868
kJ/(kg · °C) for water.

t′ =

= 297.16°C
However, it can be seen from Table 6.24 that overage Cp of compressed water (at
125 bar a) is 4.6798 kJ/(kg · °C).
With this value,

t′ =

= 279.02°C, say 279°C

If average Cp value for saturated liquid water would have been considered, t′
would be calculated as 276.09°C.
Above calculations indicate importance of correct thermodynamic and
transport properties in process design.
Following average values of properties will be considered for the heat exchanger
design:
Table 6.25 Average Thermodynamic and Transport Properties
Average property value
Component Phase
 Viscosity (μ), μPa · s Thermal cond. (k), mW/(m · K) Density (ρ) kg/m3 Heat capacity, (Cp) kJ/(kg · °C)

Hydrogen Gas 12.844 286.780 0.9592 14.5345

Nitrogen Gas 26.376 39.648 4.4736 1.0634
Methane Gas 17.107 71.033 0.1040 2.9397
Argon Gas 30.462 27.012 0.0777 0.5204
Water Gas 17.764 37.267 0.2415 1.9847
Water Compr. liquid 160.371 641.300 852.53 4.6798

Mass flow rate of gas mixture, ṁ1 = 6246.096 × 8.8778 (in tubes)
= 55 451.6 kg/h
Based on Eq. (6.20), at =

Tube side mass velocity, Gt =

= 77.5044 kg/(m2 · s)
Viscosity of gas mixture is calculated using Eq. (3.87) of Perry’s Handbook
(Ref. 2)

μmix =   (6.77)

Selecting simpler form of Θij,

Θij = (Mj/Mi)1/2   (6.78)
i represents any component while j represents component, other than i.
μmix =

= 39.2353 μPa · s
If μ′mix = Σ yi μi form would have been selected,
μ′mix = 16.3107 μPa · s [41.6% of μmix computed by Eq. (6.77)]
Reynolds number,
Ret =

= 29 302
Average heat capacity of gas mixture,
Cpmix =

= 3.269 kJ/(kg · °C)

If Cp mix would have been calculated by Cp mix = , it would have yielded
same answer.
Average thermal conductivity of gas mixture (Ref. 2),
kmix =   (6.79)
= (286.78 × 0.7251 + 39.684 × 0.242
+ 71.033 × 0.0098 + 27.012 × 0.0029 + 37.267 × 0.0202)
= 219.066 mW/(m · K)
≡ 0.219 W/(m · K)
Prandtl number, Pr =

= 0.585 66
Using Dittus–Bolter equation [Eq. (6.19)],

= 0.021 × Ret0.8 Pr0.33

hi = 0.021 ×

= 973.3 W/(m2 · °C)

Shell side coefficient by Delaware method:
BFW flows in the shell at 125 bar a. This will require much thicker shell.
ho = hoc Jc Jb Jl Js  (6.60)
For triangular pitch, referring Eq. (6.59a),
As = 300

= 72 146.25 mm2
≡ 0.072 146 m2
Gs = ṁ2/As

= 159.784 kg/(m2 · s)
Reo =

From Fig. 6.23, for Reo = 18 980.3,

Jh = 0.008
Pr =

= 1.1703
Referring Eq. (6.61)
hoc = 0.008

= 5386.7 W/(m2 · °C)

For 25% cut baffles, referring Fig. 6.24, Fc = 0.647
Referring Fig. 6.25, Jc = 1.018
Effect of bypass streams C and F:
For 1–1, heat exchanger, npv = 0
Referring Eq. (6.62),
Fsbp =

No. of sealing strips, Nss = 0 or Nss/Nc = 0

x = 0.25
Referring Fig. 6.26, Jb = 0.931
Effect of leakage streams A and E:
Referring Eq. (6.64),
Atb =

= 22 670.8 mm2
Eq. (6.66); α = 2 cos–1(1 – 2 × 0.25)
= 120° ≡ 2.0944 rad
Eq. (6.65); Asb =

= 4719.9 mm2
=

From Fig. 6.27, Jl = 0.685

No. of baffles, Nb = 28
Rsi = 370/300 = 1.2333
Rso = 370/300 = 1.2333
n = 0.6
Referring Eq. (6.67)

Js =

= 0.99
Substitute various parameters in Eq. (6.60).
ho = 5386.7 × 1.018 × 0.931 × 0.685 × 0.99
= 3462.15 W/(m2 · °C)
For calculating overall heat transfer coefficient, Uo, use Eq. (6.46).
hod = 10 000 W/(m2 · °C) for BFW (i.e. for clean water)
hid = 10 000 W/(m2 ·°C) for gas mixture (clean fluid)

Uo = 503.61 W/(m2 · °C)

ΔTlm =

= 35.205°C = ΔTm for 1-1 pass HE

Using Eq. (6.16),

Areq =

= 468.65 m2
Actual provided HTA,
Ao = ρ × 1150 × 0.019 05 × 9.14
= 629.055 m2
Excess HTA =

= 34.23% (highly adequate)

Tube side pressure drop:
For Ret = 29 302.4,
Jf = 0.0038 (Fig. 6.13)
Density of gas mixture,
ρ =

= 5.7173 kg/m3
ut = 77.5056/5.7173 = 13.556 m/s
Using Eq. (6.31),
Δpt =

= 11 154 Pa
≡ 0.111 bar (quite low)
Shell side pressure drop:
Pressure drop for an ideal cross-flow section can be calculated using Eq. (6.48).
For Reo = 18 980.27, Jf = 0.153 (Fig. 6.15)
Δpbi = 8 × 0.153 × 21 ×

= 348.8 Pa
For calculating shell side pressure drop by Delaware method, Eq. (6.69) is to be
used.
ṁ2 = (41 500/3600) = 11.5278 kg/s
Ncw =

= = 10.513

As = 0.072 146 m2
Using Eq. (6.74),
Awg =

= 135 385.46 mm2

Using Eq. (6.75),
Awt =

= 65 555.3 mm2
Referring Eq. (6.73),
Aw = 135 358.46 – 65 555.3
= 69 830.16 mm2
Referring Eq. (6.71),
Δpwi =
= 128.404 Pa
Correction factors Rb and Rt can be obtained from Figs. 6.29 and 6.30,
respectively.
For Fsbp = 0.0582 and Nss/Nc = 0,
Rb = 0.792(Fig. 6.29)

For = 0.1723 and = 0.3796,

Rl = 0.434 (Fig. 6.30)

Using Eq. (6.76),
Rs =

= 0.715
Using Eq. (6.69),
Δps = [(28 – 1) × 384.8 × 0.792 + 28 × 128.404]

× 0.434 + 2 × 384.8 × 0.792 × 0.715

= 5785.3 Pa
≡ 5.785 kPa (acceptable)
Notes: (i) Temperature ranges on shell and tube sides are high and hence,
thermal expansion analysis as per TEMA is required. A suitable expansion joint
may be necessary on the shell.
(ii) The example was also solved with Kern’s method and comparison of the
results are given below.
Delaware (Bell’s) Method Kern’s method
h0 3462.15 W/(m2 · °C) 3102.5 W/(m2 · °C)
Δps 5.785 kPa 8.867 kPa
Uo 503.61 W/(m2 · °C) 495.26 W/(m2 · C)
% Excess HTA 34.23 32

It may be noted that there is not much difference in the results. This is because
it is BEM type (i.e., fixed tube sheet type) 1-1 heat exchanger. Contributions of
bypass streams (C and F) are not significant to show any difference in the
results. However, if the type of heat exchanger is AET (i.e., pull through
floating head) with more number of tube side passes, contributions of bypass
streams; C and F would have been significant. In Exercise 6.5, the heat
exchanger is AET with 8 number of tube side passes is designed. For such a
case, h0, given by Kern’s method, could be 1.5 to 2 times the ho, calculated by
Delaware method. Also shell side pressure drop (Δps), given by Kern’s method
could have been 3 to 5 times the value given by Delaware method. Thus
Delaware method gives correct design for AET type heat exchanger with more
number of tube passes. Hence, for such heat exchangers, Kern’s method gives
lesser HTA than actually required.

6.13 PROCESS DESIGN OF REBOILERS AND

VAPORIZERS
Design correlations for reboilers and vaporizers are same. Reboilers are used
with distillation columns. In a reboiler, part of liquid feed is vaporized while in a
vaporizer the entire feed of liquid is vaporized.
Following types of reboilers are used in chemical industries:
1. Kettle-type Reboiler
2. Bundle-in-column Reboiler
3. Horizontal Thermosyphon Reboiler
4. Vertical Thermosyphon Reboiler
5. Forced Circulation Reboiler.
Process designs of Kettle-type Reboiler and Vertical Thermosyphon Reboiler are
given in subsequent sections.
6.13.1 Process Design of Kettle-type Reboiler
Kettle-type reboiler (see Fig. 6.1(e)) is normally used with a distillation column
and also as a vaporizer. Vapour leaving the kettle-type reboiler is always in
equilibrium with the liquid. In this type of reboiler, heating surface is immersed
in the pool of liquid. Hence, perfect pool boiling is taking place. In pool boiling,
maximum value of heat transfer coefficient is achieved in nucleate boiling
region. To achieve the nucleate boiling region, ΔT (Temperature difference
between temperature of heating surface and boiling temperature of liquid) should
be less than and close to critical temperature drop. Critical temperature drop for
water is 20 to 30°C and for light organics is 20 to 50°C at atmospheric pressure.
In kettle-type reboiler boiling fluid side pressure drop is negligible, hence, it is
preferred with high vacuum distillation column.
Process design steps
1. Calculation of heat duty In actual design calculation, heat duty of reboiler is
 calculated based on energy balance equation around distillation column.
ϕB = HDṅD + HW ṅW – HF ṅF + ϕC + ϕL   (6.79)
where, ϕB = Heat duty of reboiler, kW
HD = Enthalpy of distillate, kJ/kmol
ṅD = Molar flow rate of distillate, kmol/s
Hw = Enthalpy of residue, kJ/kmol
ṅW = Molar flow rate of residue, kmol/s
HF = Enthalpy of feed, kJ/kmol
ṅF = Molar flow rate of feed, kmol/s
ϕC = Heat duty of condenser, kW
ϕL = Heat loss of distillation system, kW
If the vaporization rate of reboiler is given then heat duty can be determined by
the following equation:
ϕB = ṁv λ × 1.05 kW (considering 5% heat loss to surrounding)  (6.79a)
ṁv = Vaporization rate, kg/s
2. Fix the value of mean temperature difference ΔTm. It should be less than and
close to critical temperature drop.
Average temperature of heating medium th = ΔTm + tB  (6.80)
where, tB = boiling temperature of liquid, °C
3. Select the suitable heating medium (Sec. 6.6.2).
4. Based on energy balance, find the mass flow rate of heating medium. If the
temperature of heating medium required is less than 180°C, then saturated
steam can be selected as heating medium. In this case,
ϕB = ṁsλs  (6.81)
where, λs = Latent heat of vaporization of steam, kJ/kg (to be obtained from
Steam Tables)
ṁs = Mass flow rate of steam required, kg/s
5. To get the first estimate of heat transfer area, assume the value of overall heat
transfer coefficient (Table 6.7). Find the heat transfer area based on the
assumed value of Uo.
A = = Aprovided = Nt π do L  (6.16 and 6.17)
6. Fix the value of tube outside diameter do and tube length. Find the number of
tubes Nt. Decide the tube arrangement. In reboiler, normally 1 in (25.4 mm
OD) tubes are used when tube side fluid is saturated steam. Square pitch
arrangement with U-tube-bundle is preferred. In kettle-type reboiler, higher
value of tube pitch (Pt = 1.5do to 2do) is preferred to avoid the “vapour
blanketting” of heat transfer surface inside the pool of liquid. With kettle-type
reboiler full baffles (% baffle cut = 0) are used as extra space is available for
the flow of liquid. Nucleate boiling coefficient or any phase change
coefficient does not depend on type of baffle or baffle spacing. Here, baffles
are provided only to reduce the vibration in tubes. Here, baffle spacing can be
kept equal to tube bundle diameter or even higher than that but within
permissible limit, specified in TEMA standard.
7. Calculate shell side boiling coefficient by Mostinski equation.

ho = hnB = 0.104   (6.82)

where, ho = hnB = Shell side nucleate boiling heat transfer coefficient in W/(m2 ·
°C)
= Heat flux of reboiler, W/m2

pc = Critical pressure of component or psuedo-critical pressure of the mixture of

components vaporized, bar
p = Operating pressure in reboiler, bar
This equation is reliable for single component boiling. For, a mixture, actual
coefficient is lesser than the same predicted by this equation.
8. Calculate tube side heat transfer coefficient, hi:
If saturated steam is used as heating medium, then calculation for hi is not
required. Value of hi = 6000 W/(m2 · °C) can be safely used. This value also
includes fouling resistance.
or   (6.83)

If hot oil is used as heating medium, then h′i is calculated by forced convection
correlation; Eq. (6.18) or   (6.19), depending on the value of Reynolds
number.
9. Calculate the overall heat transfer coefficient Uo, by Eq. (6.46).
10. Calculate the heat transfer area required using Eq. (6.47).
11. Calculate % excess heat transfer area. It should be in between 10 to 20%. If
not, then change the number of tubes or tube length. Repeat the calculations
until % excess area is in the desired range.
12. Based on the final or last value of heat transfer area provided, calculate the
actual heat flux.
Actual heat flux, ϕa = W/m2  (6.84)

13. Actual heat flux, ϕa ≤ 0.7 (ϕB × 103/A)c

where (ϕB × 103/A)c is the critical heat flux. Critical heat flux is the heat flux
corresponding to critical temperature drop. This condition must be satisfied.
Critical heat flux can be estimated by modified Zuber’s equation3 which can
be written as

= W/m2  (6.85)

Kb = 0.44 for square pitch arrangement

0.41 for triangular pitch arrangement
Pt = Tube pitch, mm
do = Tube outside diameter, mm
Nt = Total number of tubes in bundle
λ = Latent heat of vaporization, J/kg
ρv = Vapour density, kg/m3
σ = Liquid surface tension, N/m
ρL = Liquid density, kg/m3
g = Acceleration due to gravity = 9.81 m/s2
14. Find the shell inside diameter. Shell inside diameter of kettle-type reboiler
should be greater of following two values.
(a) Shell ID = Liquid level + 0.15 to 0.25 m
Liquid level in kettle-type reboiler = Tube bundle diameter (Db) + 50 to
100 mm
To find the tube bundle diameter for U-tube bundle, Eq. (6.2) and for tube
bundle with floating head, Eq. (6.1), can be used.
(b) Minimum Shell ID can be obtained from the following table:
Table 6.26 Minimum Shell Diameter of Kettle-type Reboiler
Heat flux, W/m2 Shell ID/Tube bundle diameter (Di /Db)

< 25 000 1.2 to 1.5

25 000 to 40 000 1.4 to 1.8
> 40 000 1.7 to 2.0

15. Check the liquid entrainment. To avoid the excessive entrainment, vapour
velocity at liquid surface should be less than vmax.

vmax = 0.2   (6.86)

where, vmax = Maximum permissible vapour velocity at liquid surface, m/s

Actual vapour velocity at liquid surface.
v =   (6.87)

Liquid surface area ≅ L × W   (6.88)

where, L = Tube length, m
W = width of liquid level, m
Boiling of Liquid Mixtures For the close boiling liquid mixture (boiling range is
less than 5°C), Eq. (6.82) can be used to find the nucleate boiling coefficient. If
for the given liquid mixture boiling range is wide then, Palen and Small equation
can be used.
(hnb) mixture = fm × (hnb) single component   (6.89)
where, fm = exp [–0.0083 (tbo – tbi)]   (6.90)
tbo = Temperature of vapour mixture leaving the reboiler, °C
tbi = Temperature of liquid entering the reboiler, °C
Kettle-type reboiler can be considered as one equilibrium stage. Hence, the
vapour leaving the reboiler is in equilibrium with the residue. Dew point
temperature of vapour leaving the reboiler (tbo) is equal to bubble point
temperature (tbp) of residue. To find tbi, composition of liquid entering to reboiler
must be determined. Material balance around reboiler gives composition of
liquid (x1i) entering to reboiler.
L– = ṅL = ṅV + ṅW   (6.91)
L–x1i = ṅL x1i = ṅV · yi + ṅW · xi   (6.92)
where, ṅW = Molar flow rate of residue, kmol/s
xi = Mole fraction of component i in residue
yi = Ki xi = Mole fraction of component i in vapour leaving the reboiler
x1i = Mole fraction of component i in liquid entering to reboiler.
V– = ṅV = Molar flow rate of vapour leaving the reboiler, kmol/s
L– = ṅL = Molar flow rate of liquid entering the reboiler, kmol/s
Bubble point of liquid entering to reboiler, gives the value of tbi.
A packed distillation tower with overhead condenser and
kettle-type reboiler is used for aromatics separation. Design conditions of the
column are as follows.
Table 6.27 Composition of Distillation Column Streams
Composition, mole %
Component
Feed Distillate Bottoms

Benzene 2.2 22.8 0

Toulene 7.4 72.2 0.5
Ethyl benzene 43.4 5.0 47.5
Styrene 47.0 0 52.0

Feed flow rate to distillation tower is 100 kmol/h.

A reflux ratio of 6.0 is selected for the design. Operating pressure at the base of
column is fixed at 21.3 kPa a (160 Torr). All the streams may be taken as
saturated liquids. Also assume that the mixtures are ideal mixtures. Saturated
steam at 0.2 MPa a pressure is used as heating medium in kettle-type reboiler.
Design the kettle-type reboiler for this distillation column.
Solution
Heat duty of kettle-type reboiler
ϕB = ϕC + HD ṅD + HW ṅW – HF ṅF + ϕL   (6.79)
To calculate the heat duty of reboiler ϕB by using above equation flow rates (ṅF,
ṅD and ṅW), latent heats at bubble point temperatures, specific heats of liquid
streams, bubble point temperatures of streams, dew point temperature of
overhead vapour, etc., must be determined.
ṅF = Feed flow rate = 100 kmol/h
ṅF = ṅD + ṅW = 100
Benzene balance:
0.228 ṅD = 2.2
or ṅD = 9.65 kmol/h
ṅW = 100 – 9.65 = 90.35 kmol/h
ϕC = Heat duty of condenser = (R + 1) ṅDλ
R = 6
Latent heat of vaporization must be determined at the average temperature of the
dew point and bubble point of distillate product. Assuming negligible pressure
drop in the tower, operating pressure is constant throughout the distillation
column and is equal to operating pressure at the base of column for heat duty
calculation, i.e., at 21.3 kPa a.
Assumption of constant pressure throughout the column is not true in real case
but for simplicity this assumption is made. In actual case, top pressure is to be
found by distillation calculations and dew point at the real pressure at the top is
to be calculated.
Bubble point of Distillate:
For ideal vapour liquid equilibrium,
at bubble point Σxi pv = pt and zi = xi.
Vapour Pressure Data:
Table 6.28 Antoine Constants10
Component A B C

Benzene 6.0306 1211.0 –52.35

Toluene 6.0795 1344.8 –53.65
Ethyl benzene 6.0821 1424.3 –59.95
Styrene 6.1911 1507.4 –58.15

Antoine equation log10 pv = (2.1)

where, pv is in kPa and T is in K.

At bubble point of distillate,
0.228 pvB + 0.722 pvT + 0.05 pvEB = pt
Trial for T = 328.3 K
pvB = 43.865 kPa, pvT = 15.244 kPa and pvEB = 5.95 kPa
0.228 × 43.865 + 0.722 × 15.244 + 0.05 × 5.95 = 21.3 kPa = pt
Bubble point of distillate, TBP = 328.3 K (55.3°C)
At dew point of distillate
= 1 and zi = yi

Try T = 334.7 K
pvB = 55.1565 kPa, pvT = 19.7054 kPa, pvEB = 7.909 kPa

× 21.3 = 1.003 (Check)

Dew point temperature of distillate, TDP = 334.7 K (61.7°C)

Latent heat of vaporization must be determined at average temperature.
Tav = = 331.5 K

Table 6.29 Properties of Components of Distillate

Component yi Critical temperature Tc, K T1, K λ at T1, kJ/kmol

Benzene 0.228 562.16 353.3 30 761

Toluene 0.722 591.79 383.7 33 179
Ethyl benzene 0.05 617.2 409.3 35 564

= (Watson equation)

For Benzene
=

λB = 31 943.68 kJ/kmol at 331.5 K

Similarly,
λT = 36 124.5 kJ/kmol at 331.5 K
λEB = 40 130.2 kJ/kmol at 331.5 K
λav = Σλvi yi
= 0.228 × 31 943.68 + 0.722 × 36 124.5 + 0.05 × 40 130.2
= 35 371.56 kJ/kmol
Heat duty of condensation
ϕC = (R + 1) ṅD λav
ϕC = (6 + 1) × 9.65 × 35 371.56 = 238 9348.9 kJ/h ≡ 663.7 kW
Let 331.5 K be the reference temperature for enthalpy calculations, i.e., T0 =
331.5 K
HD = 0, HD ṅD = ϕD = 0
Residue is bottom product of distillation and is a saturated liquid or at its bubble
point. To find the enthalpy of residue, first the bubble point of residue must be
determined.
At bubble point of liquid: Σxi pv = pt and zi = xi

Trial: At T = 363.55 K (90.55°C), = Σyi = 1

Vapour pressures: pvT = 54.959 kPa, pvEB = 24.588 kPa

pvS = 18 Pka, pt = 21.3 kPa
Table 6.30 Bubble Point Calculations for Residue

Component xi pvi, kPa yi =

Toluene 0.005 54.959 0.013

Ethyl benzene 0.475 24.588 0.548
Styrene 0.52 18 0.439

Total Σyi = 1 (check)

Bubble point of residue Tbp = 363.55 K (90.55°C)

Empirical heat capacity equation for liquids
Cml = a + bT + cT2 + dT3
where, Cml is in kJ / (kmol · K) and T is in K.
Table 6.31 Heat Capacity Equation Constants10
Component a b × 103 c × 106 d × 109

Toluene 1.8083 812.223 –1512.67 1630.01

Ethyl benzene 4.3143 900.174 –1450.05 1433.6
Styrene –38.019 1197.21 –2195.65 1933.12
Benzene –7.2733 770.541 –1648.18 1897.94

These heat capacity equation constants are used for calculation of heat-flow rates
of various streams.
Table 6.32 Heat Capacity Equation Constants for Residue
Component xi ai · xi bi · xi × 103 ci · xi × 106 di · xi × 109

Toluene 0.005 0.009 4.061 –7.563 8.15

Ethyl benzene 0.475 2.049 427.583 –688.774 680.96
 Styrene 0.52 –19.77 622.549 –1141.738 1005.222
Total 1.00 –17.712 1054.193 –1838.075 1694.332

Enthalpy of residue,
Hw = Σaixi (Tbp – T0) + Σ bixi × 10–3

Hw = –17.712(363.55 – 331.5) + 1054.193 × 10–3

– 1838.075 × 10–6 ×

+ 1694.332 × 10–9 ×

Hw = 6338.32 kJ/kmol
ϕW = HWW = 6338.32 × 90.35 = 572 667.2 kJ/h ≡ 159.074 kW
ϕF = HFF:
Table 6.33 Heat Capacity Equation Constants for Feed
Component xi ai · xi bi · xi × 103 ci · xi × 106 di · xi × 109

Benzene 0.022 –0.16 16.952 –36.26 41.755

Toluene 0.074 0.134 60.105 –111.93 120.621
Ethyl benzene 0.434 1.872 390.676 –629.322 622.182
Styrene 0.47 –17.869 562.689 –1031.956 908.566
Total 1.00 –16.023 1030.422 –1809.476 1693.124

At bubble point = Σyi = 1

Trial: At T = 357.75 K, pt = 21.3 kPa

Table 6.34 Bubble Point Calculations

Component xi pvi, kPa yi =

Benzene 0.022 116.227 0.12

Toluene 0.074 45.422 0.158
Ethyl benzene 0.434 19.923 0.406
Styrene 0.47 14.445 0.319

Total 1.00 1.003 (check)

Bubble point of feed, Tbp = 357.75 K (84.75 °C)
HF = –16.023(357.75 – 331.5) + 1030.422 × 10–3

– 1809.476 × 10–6

+ 1693.124 × 10–9

= 5079 kJ/kmol
ϕF =HF ṅF = 507 900 kJ/h ≡ 141.08 kW
Heat duty of kettle-type reboiler
ϕB = ϕC + ϕD + ϕW – ϕF + ϕL  (6.79)
ϕB = 663.7 + 0 + 159.074 – 141.08 + ϕL
ϕB = 681.7 + ϕL = 681.7 × 1.05 = 716 kW (considering 5% heat loss)
≡ 716 × 103 W
In continuous operation at steady state, boiling temperature of liquid in kettle-
type reboiler = Bubble point temperature of residue = Dew point temperature of
vapour leaving the reboiler = Steady state equilibrium temperature of kettle-type
reboiler = Tbp.
Tbp = 363.55 K (90.55°C)
Critical temperature drop of light organics is ranging from 20°C to 50°C. For the
precise calculations, value of critical temperature drop and critical heat flux at
the actual operating condition must be determined by experiment.
Temperature of heating medium th = tBP + ΔTm  (6.80)
ΔTm should be close to and less than ΔTc.
Let ΔTm ≅ 30°C
th = Temperature of saturated steam = 393.38 K = 120.23°C
ΔTm = 393.38 – 363.55 = 29.83°C (acceptable)
Saturated steam pressure = 0.2 MPa
Latent heat, λs = 2201.6 kJ/kg (Ref: Steam Tables)
Steam required:
ṁs = = 0.325 218 kg/s ≡ 1170 kg/h
For the first trial calculations, assume Uo = 1000 W/(m2 · °C)
Heat transfer area based on assumed value of Uo:

Ao = = 24 m2 = Ao provided

Ao provided = 24 m2 = Nt π do L
Nominal U-tube length = 3.6576 m (12 ft)
do = 25.4 mm, di = 21.1836 mm (14 BWG)
Ao provided = 24 m2 = Nt π × (0.0254) × 3.6576
Number of tubes = 82 (U-tubes)
Number of tube holes in tube sheet = 82 × 2 = 164
Use square pitch arrangement.
Tube pitch Pt = 1.5 do = 38.1 mm
Minimum U bend radius = 1.5 do = 38.1 mm
ho = hnbo = Nucleate boiling heat transfer coefficient for mixture = fm · hnb  
(6.89)
where, hnb can be determined by Mostinski equation2

hnb = 0.104 pc0.69   (6.82)

= 29 833.33 W/m2

p = 21.3 kPa = 0.213 bar

No reliable method is available for estimating critical pressure (pcm) of a
mixture.
pcm ≅ Σ yi pci (Kay’s rule)
fm = correction factor
Table 6.35 Composition in Reboiler
Residue composition Composition of vapour Composition of liquid entering reboiler
Component
xi n.Wxi, kmol/h yi = pvixi/pt n.Vyi, kmol/h n.Lxi, kmol/h
 x1i

Toluene 0.005 0.45175 0.013 0.856 1.30775 0.0084

Ethyl benzene 0.475 42.9163 0.548 36.07 78.9863 0.5058
Styrene 0.52 46.982 0.439 28.895 75.877 0.4859

Total 1.00 90.35 1.00 65.82 156.17 1.000

Total material balance around reboiler

L– = n.L = n.V + n.W
and component balance: n.Lx1i = n.W · xi + n.V · yi
where, correction factor
fm = exp[–0.0083(tbo – tbi)]   (6.90)
where, tbo = Temperature of vapour mixture leaving the reboiler = 90.4°C
tDP = Dew point temperature of vapour leaving the reboiler, °C
or tBP = Bubble point temperature of residue, °C
tbi = Temperature of liquid entering to the reboiler = Bubble point temperature of
liquid entering the reboiler.
At tbi
= 1

To find L–(= n.L) and x1i by Eq. (6.87) and   (6.88), value of V– is required.

Table 6.36 Properties of Residue Components

Component xi Critical temperature Tc, K T1, K λi, kJ/kmol at T1 K

Toluene 0.005 591.79 383.7 33 179

Ethyl benzene 0.475 617.2 409.3 35 564
Styrene 0.52 647 418.2 36 819
Total 1.00

Watson equation

To be precise, λ must be determined at average temperature of Tbo and Tbi and

also for average composition of liquid. Here, λav is determined at Tbo and for the
residue composition as there is a negligible change in temperature and
composition.
At T = Tbo = 363.55 K
Using Watson equation
λT = 34 365 kJ/kmol, λEB = 38 356 kJ/kmol, λS = 39 941 kJ/kmol
λav ≅ 0.005 × 34 365 + 0.475 × 38 356 + 0.52 × 39 941
= 39 160 kJ/kmol
= = 65.82 kmol/h

L– = n.V = n.L + n.W = 65.82 + 90.35 = 156.17 kmol/h

Values of x1i are calculated by using Eqs   (6.91) and   (6.92) and are
given in Table 6.35.
At Tbi = 363.1 K (Trial)
pvT = 54.1658 kPa, pvEB = 24.1972 kPa, pvS = 17.7 kPa
pt = 21.3 kPa

= 1 (check)
Tbi = 363.1 K
Correction factor fm = exp[–0.0083(363.55 – 363.1)]
fm = 0.996 27
ho = hnbmix = fm · hnb = 0.996 27 × hnb
pcm = Σpciyi (Kay’s rule)
Table 6.37 Critical Pressure Data10
Component pc, bar yi

Toluene 41.1 0.013

Ethyl benzene 36.1 0.548
Styrene 39.9 0.439

Total 1.000

pcm = 0.013 × 41.1 + 0.548 × 36.1 + 0.439 × 39.9

= 37.83 bar,
= 0.005 63 = prm
hnb = 0.104 × (37.83)0.69 × (29 833.33)0.7

= 1304.72 W/(m2 · °C)

ho = hnbmix = 0.996 27 × 1304.72 = 1299.85 W/(m2 · °C)
For the condensation of saturated steam consider
h′i = 6000 W/(m2 · °C)
This value also includes fouling resistance.
Shell side fouling coefficient, hod = 5000 W/(m2 · °C)
Thermal conductivity of steel tube material, kw = 50 W/(m · °C)
Using Eq. (6.46):

Uo = 822.86 W/(m2 · °C)

Heat transfer area required
Ao req = = 29.17 m2

Heat transfer area provided, Aopro = 24 m2

Ao req > Aopro (not acceptable)
Increase in heat transfer area will decrease the value of hnb and the value of Uo.
Value of overall coefficient removed [(Uo = 822.86 W/(m2. °C)] is within normal
range (See Table 6.7). Further decrease in the value of Uo is not required. To
decrease the value of Aoreq increase the pressure and temperature of saturated
steam.
Let saturation temperature = 404.35 K
Pressure = 0.28 MPa
λs = 2170.1 kJ/kg

Steam required

= = 0.33 kg/s ≡ 1188 kg/h

ΔTm = 404.35 – 363.55 = 40.8°C

Aoreq = = 21.327 m2

Excess heat transfer area =

= 12.533% (satisfactory)
Actual heat flux, = = 29 833.33 W/m2

To verify condition of ϕa ≤ 0.7 (ϕB)c, critical heat flux is determined by modified

Zuber’s equation.

=   (6.85)

Kb = 0.44, Pt = 38.1 mm, do = 25.4 mm

Nt = 82 × 2 = 164, g = 9.81 m/s2
λ = Latent heat of vaporization, J/kg
λs = 39 160 kJ/kmol
Average molar mass of residue
Mav = ΣMi xi = 0.005 × MT + 0.475 × MEB + 0.52 × MS
Mav = 0.005 × 92.141+ 0.475 × 106.168 + 0.52 × 104.152 = 105.05 kg/kmol
λ = × 103

= 372 774.9 J/kg

ρv = (s = standard conditions)

Average molar mass of vapour

Mav = ΣMiyi = 0.013 × MT + 0.548 × MEB + 0.439 × MS
Mav = 0.013 × 92.141 + 0.548 × 106.168 + 0.439 × 104.152 = 105.1 kg/kmol

ρv =

ρv = 0.74 kg/m3
ρL = Density of liquid mixture =

For Residue:
Table 6.38 Properties of Residue Components
wi (mass fraction)
mole fraction
Component xi Liquid density, ρLi kg/m3

Toluene 0.005 0.0044 867

Ethyl benzene 0.475 0.48 867
Styrene 0.52 0.5156 906

Total 1.00 1.00

ρL =

= 886.68 kg/m3
σ = surface tension of liquid mixture = σm
Surface tension of liquid mixture can be estimated by the following equation:
(Eq. (3.152) of Ref. 2)
= Σ[Pi](ρLmxi – ρVmyi)   (6.93)
where, σm = Surface tension of liquid mixture, dyn/cm
[Pi] = Parachor of component i to be obtained from Table 3.343 of Ref. 2
ρLm = Liquid mixture density, mol/cm3
ρVm = Vapour mixture density, mol/cm3
[Pi] for toluene = 189.6 (For C6 H5 – group) + 55.5 (For—CH3 group).
[Ptoluene] = 245.1
[PEB] = 189.6 + 55.5 + 40 (For —CH2— group)
= 285.1
[Pstyrene] = 189.6 + (15.5 + 9) + 40 = 254.1
ρLm = 886.68 kg/m3 ≡

≡ 8.44 × 10–3 mol/cm3

ρVm = 0.74 kg/m3 ≡ × 10–3 ≡ 7.041 × 10–6 mol/cm3

σ 1/4 = = 245.1(8.44 × 10–3 × 0.005 – 7.041 × 10–6 × 0.013)

+ 285.1(8.44 × 10–3 × 0.475 – 7.041 × 10–6 × 0.548)
+ 254.1(8.44 × 10–3 × 0.52 – 7.041 × 10–6 × 0.439)
= 2.2666
σ = 26.3936 dyn/cm
≡ 26.3936 × 10–3 N/m

= 0.44 × × (26.3936 × 10–3

× 9.81 × (886.68 – 0.74) × 0.742)0.25

= 64 317 W/m2

0.7 = 45 022 W/m2

Thus ϕa ≤ 0.7 (ϕB)c

To find the tube bundle diameter for square pitch arrangement with U-tube
bundle, following equation can be written:
= Nt P2t + Area of central portion

= Nt P2t + Db × Pt

(for Pt = 38.1 mm (1.5 in) = minimum U-tube radius)

– DbPt – NtP2t = 0

Db =

Minor correction in this equation:

Db =

= + 25.4 (82 U-tubes = 2 × 82 = 164 tubes)

= 600.74 ≅ 600 mm
Ideally tube bundle diameter must be determined based on actual tube sheet
layout as shown in Fig. 6.31.

Fig. 6.31 Tube Sheet Layout Drawing for Db = 600 mm, Nt = 162

Let the height of weir which maintains the liquid level in kettle-type reboiler be
700 mm.
Shell ID, Ds:
Ds = 700 + 250 = 950 mm
(where 250 mm is the vertical distance between shell ID and liquid level.)
or Ds = 1.6 Db (From Table 6.26)
Hence, Ds = 960 mm
Based on Tube sheet layout drawing for Db = 600 mm, Nt = 162
Let shell ID, Ds = 960 mm (greater of the two values )
Checking of liquid entrainment
To avoid the excessive liquid entrainment
v < vmax
where, v = Actual vapour velocity at liquid surface, m/s
vmax = Maximum permissible vapour velocity at liquid surface, m/s
v =

where, L′W = Liquid surface area

 Fig. 6.32 Checking of Liquid Entrainment

L′ ≈ Tube length = (U-tube)

L′ = 1.8288 m
From geometry, width of liquid surface,
Liquid level – = 220

= – 2202

W = 853.23 mm
Liquid surface area, L′W = 1.8288 × 0.853
= 1.5604 m2
v =

v =

v = 1.664 m/s

vmax = 0.2   (6.86)

vmax = 0.2 = 6.92 m/s

v < vmax Hence, there is no excessive liquid entrainment.

Shell side or boiling side pressure drop Δps in case of kettle-type reboiler is
negligible.
Tube Side Pressure Drop Δpt
On tube side, condensation of saturated steam is taking place.

Δpt = 0.5 Δ p′t = 0.5 Np   (6.32)

Here Np = 2, L = = 1.8288 m

di = 0.021 1836 m
ρ = Density of saturated steam at 0.28 MPa
= = 1.548 kg/m3

Tube side mass velocity, Gt =

Gt = = 11.56 kg/(m2 · s)

Tube side Reynolds number

Re = = 18 139.44

Viscosity of saturated steam at 404.35 K

μv = 1350 × 10–7 P = 1350 × 10–8 kg/(m · s))
Jf = 0.0041 (From Fig. 6.13)

ut =

Δpt = 0.5 × 2

= 230.15 Pa ≡ 0.23 kPa (acceptable)

6.13.2 Design of Vertical Thermosyphon Reboiler
As a vertical thermosyphon reboiler, BEM type fixed tube sheet heat exchanger
is normally used. It is installed in the perfect vertical position. Figure 6.33 shows
standard location of vertical thermosyphon reboiler in which normal level of
liquid in distillation column is same as level of top tube sheet of heat exchanger.
Inside the tubes, liquid is partially vaporized by using the heating medium like
saturated steam or hot oil on shell side. It decreases the density of fluid (vapour–
liquid mixture) in “hot” leg compared to the density of saturated liquid in
relatively “cold” leg and starts natural fluid circulation in the anticlockwise
direction as shown in Fig. 6.33 by arrows. In this type of reboiler, syphoning
effect is created by heating (thermally), hence it is called thermosyphon reboiler.
HTRI and HTFS have developed their own methods for the process design of
thermosyphon reboiler which are not available in open literature. The oldest and
the simplest method for the process design of vertical thermosyphon reboiler is
Kern’s method which is outlined as follows.

Fig. 6.33 Standard Arrangement of Vertical Thermosyphon Reboiler

1. Calculate the heat duty of reboiler based on energy balance equation for
distillation column using Eqs.   (6.8) and   (6.74)
2. Fix the value of mean temperature difference, ΔTm. It should be less than and
close to critical temperature drop. Then, average temperature of heating
medium th = ΔTm + tB.
3. Select the suitable heating medium. Based on the energy balance, find the
mass flow rate of heating medium required.
4. Assume the value of overall heat transfer coefficient, Uo = 1000 W/(m2 · °C).
 Find the heat transfer area based on assumed value of Uo.

Ao = = Apro = NtπdoL  (6.47 and 6.17)

Find or decide the required dimensions of the heat exchanger.

5. Assume the recirculation ratio equal to 4 for the first trial.
Recirculation ratio =

=   (6.94)

where, ṁLo = Mass flow rate of liquid at outlet of reboiler

ṁVo = Mass flow rate of vapour at outlet of reboiler.
6. Carry out the pressure balance. At steady state,
Available differential head = Pressure loss in the circuit
or Δpav = Pressure loss in the base of distillation column
+ Pressure loss in associated piping and nozzles
+ Pressure loss in reboiler (i.e., tube side pressure drop)
Velocity of liquid through the shell of distillation column is very low compared
to the velocity of liquid through pipes. Hence, pressure loss in the base of
distillation column may be neglected. If piping arrangement is smooth and
higher sizing for the pipes is provided, then pressure loss in piping may also
be neglected. Hence, Δpav ≅ Δpt, i.e., at steady state.
Δpt = Tube side pressure drop (loss) of reboiler, Pa.
Available differential head, Δpav is calculated by following equation:
Δpav = L (ρL – ρav) g   (6.95)
where, L = Length of tube, m
ρL = Density of liquid in cold leg, kg/m3
ρav = Average density of liquid vapour mixture in hot leg, kg/m3
g = Acceleration of gravity = 9.81 m/s2
ρav = (Ref. 5)  (6.96)

where, Vi = Specific volume of saturated liquid at the inlet of reboiler, m3/kg

Vi = 1/ρL
Vo = Specific volume of liquid-vapour mixture at the outlet of reboiler, m3/kg
Vo =   (6.97)

where, ṁVo, ṁLo = Mass flow rates of vapour phase and liquid phase at the outlet
of reboiler, kg/s
In Kern’s method, it is suggested to find Δpt by same equation that is used for no
phase change in the tube side fluid [Eq. (6.31)].
In this equation, ρ = ρav = average density of tube side fluid.
7. Compare the value of available differential head Δpav and pressure drop in the
system or tube side pressure drop, (Δpt), obtained based on assumed value of
recirculation ratio. There are total five possibilities:
(a) Δpav ≅ Δpt
(b) Δpav > Δpt
(c) Δpav > > Δpt
(d) Δpav < Δpt
(e) Δpav < < Δpt
(a) Δpav ≅ Δpt implies that the assumed value of recirculation ratio (4 for first
trial calculation) is equal to the actual value of recirculation ratio. Proceed
for the next step of calculation.
(b) Δpav > Δpt, implies that the assumed value of recirculation ratio is less than
the actual recirculation ratio. Hence, assume higher value of recirculation
ratio, repeat the calculations and again compare the new values of Δpav and
Δpt. By trial and error calculations, find the recirculation ratio for which
Δpav ≅ Δpt. Recirculation ratio can have a fractional value.
(c) Δpav > > Δpt or Δpav/Δpt > 50 implies that the assumed value of recirculation
ratio is very much less than actual recirculation ratio. In this case also, one
can find the value of recirculation ratio for which Δpav ≅ Δpt, by trial and
error calculations but very high recirculation ratio and hence, very high
tube side pressure drop may be obtained which cannot be permitted. Very
high tube side pressure drop will result in significant boiling point elevation
inside the tubes of the reboiler and decrease the mean temperature
difference. Very high tube side velocity gives erosion and vibrations in
tubes. Hence, in such a case one of the following options can be
 considered:
(i) Provide the control valve with flow meter in the inlet line to the reboiler.
Then, pressure balance is Δpav = Δpt + Pressure loss in flow meter +
Pressure drop offered by control valve.
Pressure drop of control valve can be adjusted to the desired value. This option
also provides flexibility in the actual operation. In actual plant operation,
recirculation ratio or circulation rate can be corrected depending on the
actual performance. This facility is not available with other options.
(ii) Elevate the level of top tube sheet of reboiler above the liquid level at the
base of distillation column as shown in Fig. 6.34.

Fig. 6.34 Increase of Elevation of Reboiler

In this option,
Δpav = L′ (ρL – ρav)g – (L – L′) ρavg  (6.95a)
Hence, this modification decreases the value of available differential head Δpav
and may equate it with Δpt at the lower value or reasonable value of
recirculation ratio.
(d) If Δpav < Δpt , then one of the following options can be considered:
(i) Increase the number of tubes and decrease the tube length. This modification
will decrease the values of both Δpav and Δpt. But, the percentage
decrease in the value of Δpt will be more compared to the same in the
value of Δpav. Hence, it may be possible to equate Δpav and Δpt for the
lower value of recirculation rate.
(ii) Δpav < Δpt, implies that the actual recirculation ratio is less than assumed or
minimum recommended value of recirculation ratio. In this case, actual
recirculation ratio can be increased by elevating liquid level at the base of
 the distillation column above the top tube sheet of reboiler as shown in
Fig. 6.35.
In this case, equation of Δpav is changed.
Δpav = ρLgh + L (ρL – ρav)g   (6.95b)
But this modification may increase the height of support of distillation column.
If increase in the height of support is not possible, then one can go for
underground construction. If none of these options is permitted, then use a
pump in the inlet line to the reboiler to increases the circulation rate. Then
this reboiler is called a forced circulation reboiler.

Fig. 6.35 Decrease of Elevation of Reboiler

(e) Δpav < < Δpt implies that the actual recirculation ratio is very low or very
much less than minimum recommended value, i.e., 4. This is possible when
the viscosity of bottom product of distillation column is very high. In this
case use a pump in the inlet line to the reboiler to get the desired value of
circulation rate and desirable value of tube side heat transfer coefficient.
But then this reboiler also becomes a forced circulation reboiler.
8. After finding the recirculation ratio for which available differential head
becomes nearly equal to pressure loss in the system, calculate the tube side
heat transfer coefficient. In the latest method, tube length is divided in
different sections. For the different sections, different regimes of two phase
flow are considered. Hence, for the different sections different values of heat
transfer coefficients are calculated by different correlations (Ref. 12, 13 and
14). But in the simplified Kern’s method, it is suggested to use the same
forced convection correlation which are actually derived for the sensible heat
 transfer (cooling or heating) of tube side fluid. Hence, in Kern’s method, hi is
determined by Eq. (6.18) or   (6.19). In addition to that Kern has specified
the maximum limit for the value of hi as under.
If hi ≥ 300 Btu/(h · ft2 · °F) [= 1700 W/(m2 · °C)],
take hi = 1700 W/(m2 · °C)
9. Calculate the shell side or heating medium side heat transfer coefficient. If
shell side heating medium is saturated steam then take
ho ′ = 6000 W/(m2 · °C)
This figure also includes fouling resistance. If hot oil is used as heating medium,
then use the Eq. (6.40) or Eq. (6.41) for the calculation of ho.
10. Determine the overall heat transfer coefficient by Eq. (6.46).
11. Calculate the required heat transfer area by using Eq. (6.47). If ratio of
provided HTA to required HTA is 10 to 20% design is acceptable. If it is not,
then change the value of Aopro., either by increasing the tube length or by
increasing number of tubes. An increase in tube length will keep the tube side
heat transfer coefficient unchanged but it will increase the value of Δpt.
Increase in number of tubes will decrease the value of hi and also of ho, If hot
oil is used as heating medium it will decrease the value of Δpt. In any case,
both pressure balance calculations and heat transfer coefficient calculations
must be repeated.
12. Find the shell side pressure drop (Δps). It should be less than the maximum
allowable value.
Design a vertical thermosyphon reboiler which must provide
10 000 kg/h of n-pentane vapour to a distillation tower at 10 bar a pressure.
Solution
Determination of boiling point of pentane:
Antoine equation9 for n-pentane
ln pv = A – , where, pv in Torr, T in K(2.1)

ln pv = 15.8333 –

pv = pt = 10 bar = 7500.6 Torr

T = 398.38 K ≡ 125.4°C
Heat duty, ϕ = ṁλ × 1.05 (5% heat loss)
where, λ = Latent heat of vaporization at the boiling point
Latent heat of vaporization of pentane at 125.4°C:
At 25°C, λv1 = 366.512 kJ/kg
Critical temperature of pentane, tc = 197°C, (Ch 3, of Ref. 2)
From Watson’s equation,

λ2 = = 366.512

= 366.512 × 0.7167
= 262.68 kJ/kg
Heat duty, ϕ = 1.05 × 10 000 × 262.68
ϕ = 2758 140 kJ/h ≡ 766.15 kW ≡ 766.15 × 103 W
Let ΔTm = 39.6°C. Assume that this value of ΔTm is less than and close to the
critical temperature drop.
Average temperature of heating medium
th = tB + ΔTm = 125.4 + 39.6 = 165°C
From Steam Tables, for the steam temperature ts = 165°C, saturation steam
pressure = 700 kPa and latent heat of condensation of steam, λs = 2064.9 kJ/kg.
ϕt = ṁsλs

ṁs = = 0.371 kg/s ≡ 1335.6 kg/h

For the first trial calculations, assuming overall heat transfer coefficient
U = 1000 W/(m2 · °C)
Apro =

Apro = 19.3472 m2 = Nt π do L
Let length of tube L = 6 ft = 1.8288 m
do = 25.4 mm
19.3472 = Nt × π × (0.0254) × 1.8288
Nt = 132.57 ≅ 133 tubes
For 25.4 mm OD tube, 31.75 mm tube pitch (Pt/do = 1.25) triangular pitch
arrangement and 1-1 shell and tube heat exchanger,
Shell ID, Di = 438 mm (From Table 6.1(f))
Assume recirculation ratio = 4
Available differential head
Δpav = L(ρL – ρav)g  (6.95)
ρL = Density of liquid pentane at 125.4°C = 626 kg/m3
Density of pentane vapour at 10 bar a pressure and at 125.4°C
ρv =

ρv = 21.724 kg/m3
Average density of liquid-vapour mixture
ρav =   (6.96)

Vi = Specific volume of liquid at inlet

Vi = = 1.597 44 × 10–3 m3/kg

For a recirculation ratio = 4

Flow rate of liquid at the outlet of reboiler
ṁLo = 4 × 10 000 kg/h = 40 000 kg/h
Flow rate of vapour at the outlet of reboiler ṁVo = 10 000 kg/h
Specific volume of vapour at outlet

Vo =

Vo = 0.010 4843 m3/kg

ρav = = 211.715 kg/m3

Δpav = 1.8288 × (626 – 211.715) × 9.81 = 7432.5 Pa

Assume that higher sizes of the piping will be used in between base of
distillation column and thermosyphon reboiler.
Pressure loss in the system ≅ Tube side pressure drop, Δpt.
Tube side pressure drop,

Np = 1, ≅ 1, ρav = 211.715 kg/m3

Tube inside diameter

di = 22.098 mm (16 BWG tube)
Tube inside flow area, Eq. (6.26)

at = = 0.051 m2

Tube side mass velocity

Gt =

= 272.33 kg/(m2 · s)
Tube side velocity, ut =

= 1.2863 m/s
Ret = = 50 149.6

(Kern suggests to use liquid viscosity, instead of viscosity of liquid–vapour

mixture)
Jf = 3.2 × 10–3 (From Fig. 6.13)

Equation (6.31): Δpt =

Δpt = 808.94 Pa
Δpav > Δpt
Assume the higher value of recirculation ratio (greater than 4) and by trial and
error calculations find the value of recirculation ratio for which Δpav ≅ Δpt.
Trial I: Assume circulation ratio = 15.5
ṁLo = 15.5 × 10 000 = 155 000 kg/h, ṁvo = 10 000 kg/h

Vo = = 4.2904 × 10–3 m3/kg

Vi = 1.597 44 × 10–3 m3/kg

ρav = = 366.874 kg/m3

Δpav = 1.8288 × (626 – 366.874) × 9.81 = 4648.9 Pa

Tube side mass velocity, Gt = = 898.693 kg/(m2 · s)

Tube side Reynolds number

Ret = = 165 494

Tube side fraction factor, Jf = 2.64 × 10–3 (From Fig. 6.13)

Tube side velocity ut = = 2.45 m/s

Tube side pressure drop

Δpt =

= 4677.2 Pa
Δpav ≅ Δpt
Hence, recirculation ratio of 15.5 is acceptable.
Heat transfer coefficients:
(a) Tube side heat transfer coefficient
Ret = 165 494
Kern suggests to use liquid properties.
Properties of liquid pentane at 125.4°C:
Specific heat CL = 2.2483 kJ/(kg · °C)
Viscosity μL = 0.12 cP
Thermal conductivity, kL = 0.124 56 W/(m · °C)

Pr = = 2.166
 = 0.023   (6.19)

hi = 0.023 × × (165 494)0.8 × (2.166)0.33 × 1

hi = 2503.5 W/(m2 · °C) ≥ 1702.6 W/(m2 · °C)

Take hi = 1702.6 W/(m2 · °C)
Shell side heat transfer coefficient ho ′ = 6000 W/(m2 · °C)
(This value includes fouling resistance)
Thermal conductivity of tube material kw = 45 W/(m · K) (mild steel tubes).
Tube side fouling coefficient hid = 5000 W/(m2 · °C)
Eq. (6.46):

Uo = 900.13 W/(m2 · °C)

Heat transfer area required
Areq = = 21.494 m2

= = 0.9 < 1.1

Assume new ratio of 1.2 for the tube length.

Revising tube length Lrev = 6 × = 8 ft ≡ 2.4384 m

Apro = Nt π do L
= 133 × π × 0.0254 × 2.4384 = 25.8786 m2
For the new tube length and with recirculation ratio = 15.5
Δpav = 4648.9 × = 6198.53 Pa

Tube side pressure drop,

Δpt = 5182.63 Pa
Δpav > Δpt, hence assume the higher value of recirculation ratio.
Trial: For the recirculation ratio = 17
ṁLo = 17 × 10 000 = 170 000 kg/h, ṁVo = 10 000 kg/h
Specific volume of liquid-vapour mixture at the outlet of reboiler

Vo = = 4.066 × 10–3 m3/kg

Average density, ρav =

= 378.46 kg/m3
Δpav = 2.4384 × (626 – 378.46) × 9.81
= 5921.33 Pa
Tube side pressure drop:
Tube side mass velocity
Gt = = 980.392 kg/(m2 · s)

Re =

= 180 539.2
Jf = 2.4 × 10–3 (From Fig. 6.13)

ut = = 2.59 m/s

Δpt = 1 ×

= 5862.75 Pa
Δpav < Δpt (acceptable)
For the recirculation ratio = 17
Rei = 180 539.2

hi = 2503.5 ×

= 2684 W/(m2 · °C) ≥ 1702.6 W/(m2 · °C)

Take hi = 1702.6 W/(m2 · °C)
Heat transfer area required Ared = 21.494 m2

Excess heat transfer area = × 100

= 20.4%
Shell side pressure drop Δps:
For the condensation of saturated steam, shell side pressure drop is calculated by
Eq. (6.45).
Shell side Friction Factor, jf:
To find this factor, first shell side flow area must be determined by Eq. (6.33).
Let Bs = Baffle spacing = Ds = Shell inside diameter
Bs = Ds = 0.438 m, do = 25.4 mm, Pt = 31.75 mm

As = × 0.4382 = 0.038 37 m2

Shell side mass velocity Gs = = 9.67 kg/(m2 · s)

Density of steam ρv =

From Steam Tables specific volume of saturated steam at 700 kPa pressure,
Vs = 0.272 68 m3/kg
ρv = 3.6673 kg/m3

us = = 2.64 m/s

Equivalent diameter, de =   (6.37)

de = (0.031 752 – 0.907(0.0254)2)

de = 0.0183 m
Shell side Reynolds number:
Res =
Viscosity of steam at 165°C and at 700 kPa pressure μv = 14.5 × 10–6 (N · s)/m2
(Table 3.302, of Ref. 2)
Res = = 12 204.2

Jf = 4.78 × 10–2 (For 25% baffle cut) (Ref. Fig. 6.15)

Δps = 0.5 × 8 × 4.78 × 10–2

Δps = 325.6 Pa ≡ 0.326 kPa < < 13.8 kPa

It is well within permissible limit.
6.13.3 Criteria of Selection between Kettle-type
Reboiler and Thermosyphon Reboiler
1. Boiling fluid side (process fluid side) pressure drop in kettle-type reboiler is
negligible. While the same is high and significant in thermosyphon reboiler.
Hence, with vacuum distillation, kettle-type reboiler is preferred.
Thermosyphon reboiler is not preferred for use with absolute pressure below
0.3 bar. For example, in separation of the mixture of monochloroacetic acid
and acetic acid by distillation, kettle-type reboiler is the only choice. Cracking
temperature of monochloroacetic acid is 120°C and is obtained as the bottom
product. Vapour pressure of monochloroacetic acid at 118°C is 60 Torr.
Hence, to restrict the boiling point of monochloroacetic acid below 120°C,
pressure drop in distillation column plus pressure drop in overhead condenser
plus pressure drop in reboiler plus pressure drop in associated piping must be
less than 60 Torr (0.08 bar). So, kettle-type reboiler is the only choice.
2. If the viscosity of bottom product of distillation column is very high, then
thermosyphon reboiler cannot be used. In such a case, forced circulation
reboiler is preferred.
In fatty acid, separation in a distillation tower under vacuum, forced circulation
reboiler is used due to viscous nature of fatty acids. However, for the vacuum
service and corrosive system with highly viscous bottom product kettle-type
reboiler is used. For example, separation of pure POCl3 (phosphorous
oxitrichloride) from treated crude POCl3, residue or bottom product of
distillation is highly viscous. Thermosyphon reboiler cannot be used for this
application.
3. With atmospheric and high pressure distillation column, thermosyphon
reboilers are preferred as size and fixed cost of thermosyphon reboiler is less
than the same of kettle-type reboiler. Because of the fluid circulation,
thermosyphon reboiler provides higher heat transfer coefficient than kettle-
type reboiler. Even for the same value of heat transfer area, fixed cost of
kettle-type reboiler is higher than the same of the thermosyphon reboiler.
4. If the ratio of vaporization rate required to mass flow rate of residue (bottom
product of distillation column) is 0.8 or above, kettle-type reboiler is
preferred.
Kettle-type reboilers are also use as “Vaporizers” in industry. Thermosyphon
reboiler cannot be used as vaporizer. Separation taking place in kettle-type
reboiler can be considered as equivalent to one theoretical stage. If Murphree
efficiency of the column is 50% (approx.), it can result in reduction of two
trays.

6.14 AIR COOLED HEAT EXCHANGERS AND AIR

HEATERS
6.14.1 Introduction
Atmospheric air (having temperature varying from 20 to 45°C, depending on
season) can be used as cooling medium if the process fluid is to be cooled up to
60°C as minimum driving force required for heat transfer in air-cooled heat
exchanger is 10°C.
Advantages and disadvantages of air as a cooling medium over water as a
cooling medium are given in section on air cooling.
As shown in Fig. 6.36, air cooled heat exchanger consists of banks of finned
tubes over which air is blown by forced draft fan mounted below or drawn by an
induced draft fan(s), mounted above the tube bundle. Tube bundle consists of
tubes with spiral wound transverse fins in the majority of applications. Forced
draft unit consumes less power. Induced draft provides comparatively even
distribution of air across the bundle since the air velocity approaching the bundle
is relatively low. Percentage recirculation of hot exhaust air is less in induced
draft unit compared to the forced draft unit since the exist air velocity is several
times that of forced draft unit. Induced draft arrangement gives more protection
to heat transfer surface against rain fall, snow fall, sleet, etc.
 Fig. 6.36 Forced Draft Air Cooled Heat Exchanger

If the difference between process fluid outlet temperature (required) and

ambient-air temperature is less than 30°C, induced draft unit is preferred because
of the less % recirculation of hot exhaust air. But if the same is more than 30°C,
forced draft unit is preferred as it consumes less power. 3/4 in (19.05 mm) OD
and 1 in (25.4 mm) OD tubes are more commonly used. Fin spacing from 3.6 to
2.3 mm (7 to 11 fins per linear inch) and triangular pitch from 43 to 63.5 mm are
common. Tube ends are left bare to permit the attachment or insertions of tubes
into appropriate holes of box-type headers. Fin material can be aluminum,
copper, mildsteel, stainless steel, etc.
The same heat exchanger is also used for heating the air by saturated steam or
cooling of the atmospheric air by chilled water. Hot air is used as drying medium
in the dryers while chilled air is used for air conditioning. In areas of water
scarcity, air cooled refrigerant condenser is considered. Refer to Exercise 6.10(c)
for design of such a condenser.
For all above-mentioned applications of the air heat exchangers, worst condition
or design condition is different.
1. If atmospheric air is used as cooling medium then for the design calculations
the maximum possible temperature of ambient air in the entire year must be
considered.
2. If air is heated by saturated steam to get the hot air for dryer, then for the
design calculations minimum possible temperature of ambient air in the entire
year must be considered.
3. When atmospheric air is cooled to lower temperature (say to 14°C) by chilled
water, then sizeable percentage of heat duty is utilized in the condensation of
water vapour which is always present in atmospheric air. Hence, for this case,
 for the design calculations atmospheric air with the maximum possible
absolute humidity (and with corresponding temperature) must be considered.
Fins are used to increase the heat transfer area and thereby to decrease the size of
heat exchanger required. Needless to say, fins should have good contact with the
tubes. Many techniques are used to fix fins on the tubes. For the heat transfer, fin
area is not as effective as bare tube area. Hence, in design calculations to
consider the effectiveness of fin area, fin efficiency is determined.
Design method for air side or finside is completely different from the same used
for shell and tube heat exchanger, but for finding tube side heat transfer
coefficient and tube side pressure drop same correlations are applicable that are
used for a shell and tube heat exchanger.
6.14.2 Method for Finding Air Side or Fin Side Heat
Transfer Coefficient and Pressure Drop
Method is detailed in Example 6.9 for calculating air side flow area (as).
Consider
Tube OD = 19.05 mm, Fin OD = 38 mm
Tube pitch = 43 mm, Fin height = 9.475 mm
Number of fins per liner inch = 8 or fin spacing = 3.15 mm
Fin thickness = 0.889 mm
Type of fin = Spiral wound transverse fin
Solution
(a) Fin area per unit length of tube,
Af = (Fin OD2 – Tube OD2) × 2 × No. of fins per one meter of tube
length  (6.98)
= (0.0382 – 0.019 052) × 2 × × 1000

= 0.534 86

Bare tube area/m of one tube = π do × 1 – πdotf · nf   (6.99)

where, do = OD of bare tube, m
tf = thickness of fin, m
nf = number of fins per 1 m of tube length
= 8 × = 315

Bare tube area,

Ab = π × 0.019 05 × 1 – π × 0.019 05 ×

= 0.043 09 m2/m of one tube

Fig. 6.37 Projection of Finned Tube

(b) Projected Perimeter:

= 2(Fin OD – tube OD) × nf + 2 × (1 – tf · nf)   (6.100)

= 2(0.038 – 0.01905) × 8 ×

= 13.377 m/m of one tube

Air or finside heat transfer coefficient:
Find the equivalent diameter by using following equation:
de =   (6.101)

de = = 0.0275 m

Shell side or air side flow area:

as = Air facing area – Area occupied by finned tubes
Air facing area = Length of area × Width of area
= Tube length × Width of area
In majority of applications, width of air facing area is kept equal to tube length.
 Fig. 6.38 Air Facing Area

For example, let tube length L = 1500 mm and

W = L ≅ 1500 mm
as = WL – np × do × L – np ×
(fin OD – tube OD) tf · nf L  (6.102)
where, np = Number of tubes in bank perpendicular to the direction of flow of air
nf = Number of fins per 1 m of tube length = 315
np = = 34.88 = 35

Let W = 1510 mm (revised) for np = 35 tubes in one bank

W = (np – 1) × pt + Fin OD + x
W = (35 – 1) × 43 + 38 + x or x = 10 mm
Hence, there will be 5 mm clearance on either side.
as = 1.51 × 1.5 – 35 × 0.019 05 × 1.5 – 35 × (0.038 – 0.019 05)
× (0.889 × 10–3) × 8 ×

= 0.986 31 m2
After calculating air side flow area (as), following method is recommended for
heat transfer calculations.
(c) After calculating de and as, calculate the air side mass velocity

Gs =   (6.103)

where, ṁa = Mass flow rate of air, kg/s

as = Air side or shell side flow area, m2
(d) Calculate air side Reynolds number.
Re =   (6.104)

where, de = Equivalent diameter, m

Gs = Mass velocity of air, kg/(m2 · s)
μ = Viscosity of air at average temperature, kg/(m · s)
(e) Calculate air side Prandtl number.
All properties must be determined at average temperature on shell side.
(f) Calculate fin side or air side heat transfer coefficient (clean and for 100% fin
efficiency).
hf = Jha Pr1/3   (6.105)

where, hf = Heat transfer coefficient, W/ (m2 · °C) (clean)

Jha = Factor for heat transfer coefficient
k = Thermal conductivity of air, W/(m · °C)
de = Equivalent diameter, m
Pr = Prandtl number of air
Jha , factor of heat transfer coefficient, can be calculated by following equation:
Jha = 0.085 2072 Re 0.7324   (6.106)
(g) Calculate dirty fin side heat transfer coefficient, hf ′ (for 100% fin efficiency)

=   (6.107)

where, hdo = Shell side fouling coefficient, W/(m2 · °C)

hf ′ = Dirty fin side heat transfer coefficient, W/(m2 · °C)
(h) Calculate dirty fin side heat transfer coefficient based on inside heat transfer
area and considering fin efficiency
= (Ω × Af + Ao ′ )   (6.108)

where, = Dirty fin side or air side heat transfer coefficient based on inside
heat transfer area, W/(m2 . °C)
Ω = Fin efficiency
Af = Total fin area, m2
Ao ′ = Total bare tube area, m2
Ai = Inside heat transfer area, m2
(i) Fin efficiency Ω can be determined from Fig. 6.39 (a) to (d) in which graph is
drawn for (re – rb) vs Ω for the different values of .

where, re = , m, rb = , m

Fig. 6.39 (a) Efficiency of Annular Fins of Constant Thickness

Fig. 6.39 (b) Efficiency of Annular Fins with Metal Area for Constant Heat Flux
 Fig. 6.39 (c) Fin Efficiency of Several Types of Straight Fin

yb =

km = Thermal conductivity of fin material, W/(m · °C)

hf ′ = Dirty fin side heat transfer coefficient determined by Eq. (6.107), W/(m2 .
°C)
In(u) = Modified Bessel function of the first kind and order n, dimensionless
Kn(u) = Modified Bessel function of the second kind and order n, dimensionless
u = A function
β = Constant
(j) Determine the tube side heat transfer coefficient hi as per correlations
 applicable for shell and tube heat exchangers. Decide the tube side fouling
coefficient hi.
Find the overall heat transfer coefficient based on inside heat transfer area, Ui

(d) Efficiency Curves for Four Types of Spine

Fig. 6.39 Fin Efficiencies for Various Types of Fins5

Ui =   (6.109)

(k) Then calculate the inside heat transfer area required

Ai =   (6.110)
where, ϕt = Total heat duty, W
ΔTm = Mean temperature difference, °C
= Ft ΔTlm  (6.11)
where, Ft = LMTD correction factor for cross flow can be determined, from Fig.
  (6.40)
 Fig. 6.40 (a) and (b) Cross Flow Temperature Difference Correction Factors5

Fig. 6.40 (c) and (d) Cross Flow Temperature Difference Correction Factors5

(l) Heat transfer area provided (inside surface)

Ai = Nt π di L   (6.111)
Provided area should be greater by 10 to 20% than the required area.
(m) Calculate, the tube side pressure drop by using the same correlation that is
used for shell and tube heat exchanger. It should be less than maximum
allowable pressure drop.
(n) Calculate the volumetric equivalent diameter by following equation:

Dev =   (6.112a)

Fig. 6.41 Finned Tube Arrangement

SL and ST are tube pitches (Pt).

Net free volume,
Vfv = W × L × x′ – np do 2L

– np × (Fin OD2 – do 2) × tf nf L  (6.112b)

where, W = Width of air facing area, m

L = Length of tube, m
x′ = Horizontal distance between two vertical bank (see Fig. 6.41)
= Pt sin 60 (For equilateral triangular pitch arrangement)
nt = Total number of tubes
np = Number of tubes in one bank, perpendicular to the direction of the flow of
air
do = Outside diameter of (bare) tube, m
tf = Thickness of fin, m
nf = Number of fins per 1 m of tube length
Af = Total fin area, m2
Ao ′ = Total bare tube area, m2
For Example 6.9
Net free volume, Vfv = 1.51 × 1.5 × 0.037 24 – 35 × (0.019 05)2 × 1.5

– 35 × (0.0382 – 0.019 052) × (0.889 × 10–3)

× = 0.0569 m3

Af = (0.534 86 m2/1 m of one tube) × L × nt

= 0.534 86 × 1.5 × nt = 0.802 29 × nt m2
Ao ′ = (0.043 09 m2/m of one tube) × L × nt = 0.064 635 nt m2

= (0.802 29 + 0.064 635) × 35 = 30.3424 m2

Dev = = 0.0075 m

Reynolds number for the calculation of pressure drop

Res =   (6.113)

where, Gs = Mass velocity of air based on flow area as, kg/(m2 · s)

μa = Viscosity of air at average temperature, kg/(m · s)
Calculate the air side (or shell side) pressure drop

Δps =   (6.114)

where, Δps = Pressure drop, N/m2 or Pa

Jfa = Air side friction factor can be calculated by following
equation.
= 1.085 58 × Re–0.128 025   (6.115)
SL = Center-to-center distance to the nearest tube in the next
bank, m (as shown in Fig. 6.41)
ST = Pitch in transverse beak, m (as shown in Fig. 6.41)
LP = Effective path length of pressure drop, m
ρa = Average density of air, kg/m3
LP can be determined by following equation:

Lp =   (6.116)

Calculated pressure drop should be less than maximum allowable pressure drop.
Mobil therm oil is used as heating medium (i.e., thermic fluid)
in chemical industry. Its operating range is from –1.1°C to 316°C.
It is required to cool 9000 kg/h of Mobil therm oil from 260°C to 200°C by
using atmospheric air as a cooling medium in air cooler. Design the suitable air
cooler.
Properties of Mobil therm oil at 230°C temperature
Density, ρ = 850 kg/m3
Specific heat, CL = 2.5 kJ/(kg · °C)
Viscosity, μ = 0.595 cP or mPa · s
Thermal conductivity, k = 0.1159 W/(m · °C)
Solution
Heat duty required
ϕt = ṁCL Δt = 9000 × 2.5 × (260 – 200)
= 1350 000 kJ/h
≡ 375 kW ≡ 375 × 103 W
Assume maximum possible temperature of atmospheric air in the area as 48°C.
Let outlet temperature of air = 75°C
Mass flow rate of air required
ṁa =

Specific heat of air at 61.5°C

Cpa = 0.25 kcal/(kg · °C) = 1.0467 kJ/(kg · °C) (Ref. 10)
ṁa = 13.2692 kg/s ≡ 47 769 kg/h
ΔTm Mean temperature difference
ΔTm = ΔTlm · Ft
Oil temperatures are very much greater than air temperatures, hence, forced draft
fan can be selected to reduce the power consumption. Atmospheric air will enter
from the bottom of tube bundle. Hot oil is also introduced from the bottom of the
header. Hence, overall there is a cocurrent contacts. But, actually there is cross
flow contact with multipass flow of oil.

ΔTlm = = 164.69 °C

R = = 2.222

S = = 0.127 36

Ft = 0.98 (From Fig. 6.40a)

ΔTm = 164.69 × 0.98 = 161.396°C
Evaluation of ho:
Equivalent diameter de =   (6.101)

Let tube OD = 19.05 mm,

Type of fin = Spiral wound transverse fin
Fin OD = 38 mm, Tube pitch = 43 mm (Equilateral triangular )
Number of fins per linear inch = 8
Fin thickness = 0.889 mm
Fin and tube material = mild steel
Fin area, Af1 = 2 × (fin OD2 – do 2) × No. of fins per 1 m length  
(6.100)

Af1 = 2 × (0.0382 – 0.019 052) × 8 ×

= 0.534 86 m2
Bare tube area Ao1 = π do × 1 – π dotf nf  (6.99)
(where nf = no. of fins/1 m of one tube)
Ao1 = π × 0.01905 × 1 – π × 0.01905 × (0.889 × 10–3) × 8 ×

Ao1 = 0.043 09 m2
Projected perimeter per 1 m of one tube
= 2 (fin OD – do) × nf + 2 (1 – tf nf)  (6.100)

= 2(0.038 – 0.019 05) × 8 × = 13.377 m

de = = 0.0275 m

Flow Area:
For the first trial calculation,
Let tube length = 900 mm = L
and width of air facing area, W = 900 mm
Let np = number of tubes in one bank, perpendicular to the flow of air

np ≅ ≅ ≅ 20.93

Let np = 21
W = (np – 1) pt + Fin OD + x
W = 20 × 43 + 38 + x
x = 2 mm
Let W = 910 mm
x = 12 mm (Revised)
Flow area
as = WL – np × do × L – np (fin OD – tube OD) tf nf L  (6.102)
= 0.91 × 0.9 – 21 × 0.019 05 × 0.9 – 21(0.038 – 0.019 05)
× 0.889 × 10–3 × 8 × × 0.9

= 0.358 67 m2
Mass velocity, Gs = = 36.995 kg/(m2 · s)
Reynolds number, Re =   (6.104)

Viscosity of air at 61.5°C

μa = 2000 × 10–7 P = 2000 × 10–8 kg/(m · s)
(Figure 3.42 of Ref. 2)
Re = = 50 868

Factor for heat transfer coefficient

Jha = 0.085 2072 Re0.7324  (6.106)
= 0.085 2072(50 868)0.7324
= 238.49
Clean heat transfer coefficient, for 100% fin efficiency
hf = Jha   (6.105)

Prandtl number of air at 61.5°C:

Thermal conductivity of air at 61.5°C, k = 0.028 784 W/(m · °C)
(Table 3.314, of Ref. 2)

Pr =

= 0.7273
hf = 238.49 × × (0.7273)1/3

= 224.49 W/(m2 · °C)

Dirty fin side heat transfer coefficient

=   (6.107)

hdo = 5000 W/(m2 · °C) (From Table 6.10)

hf ′ = 214.84 W/(m2 · °C)

Dirty fin side heat transfer coefficient based on inside heat transfer area and fin
efficiency.
hfi ′ = (Ω × Af + Ao ′)   (6.108)

Af = nt Af1 L, where, nt = Total number of tubes

Total fin area Af = nt × 0.53 486 × 0.9 = 0.481 374 nt m2
Ao ′ = Total bare tube area
Ao ′ = nt Ao1 L = 0.043 09 × 0.9 × nt = 0.038 781 nt
Ai = nt π di L
Tube inside diameter, di = 14.8336 mm (For 14 BWG tube)
Ai = π (0.014 8336) × 0.9 × nt = 0.041 94 nt m2
Fin efficiency (Ω):
Evaluation of function, (re – rb) :

where, re = = 19 mm ≡ 0.019 m

rb = = 9.525 mm ≡ 0.009 525 m

km = Thermal conductivity of fin material (mild steel)

= 45 W/(m · °C)
yb = (For fins of uniform thickness)

= = 4.445 × 10–4 m

Function = (0.019 – 0.009 525) = 0.98

= 1.994 75, Ω = 0.68 [from Fig. 6.39(a)]

=
= 1875.45 W/(m2 · °C)
Evaluation of hi:
Tube side flow area, at
Let number of tubes per pass = np = 21
at = 21 ×

= 21 × × (0.0148 336)2 = 3.629 × 10–3 m2

Tube side mass velocity,

Gt = = 688.9 kg/(m2 · s)

Tube side velocity, ut =

= 0.81 m/s (low)

Tube side Reynolds number
Ret = = 17 174.57
Prandtl number Pr =

= 12.8343
Using Dittus–Bolter equation

= 0.023 × × (17 174.57)0.8(12.8343)0.33

= 1019.15 W/(m2 · °C)

Ui =   (6.109)

Ui = = 583.29 W/(m2 · °C)

Heat transfer area (inside) required

Aireq = = 3.98 m2

Let heat transfer area to be provided

Aipro = 1.1 × Aireq = 4.378 m2
Heat transfer area per bank = np π di L = 21 × π × 0.014 8336 × 0.9 = 0.880 76
m2
Number of banks required = = 4.971

Let number of banks = 5

Excess heat transfer area = 10.65%
Total number finned tubes nt = 21 × 5 = 105
Total fin area Af = 105 × 0.534 86 × 0.9 = 50.544 m2
Total bare tube area Ao ′ = 105 × 0.043 09 × 0.9 = 4.072 m2
Total outside heat transfer area = 54.616 m2
Air side pressure drop
Net free volume, Vfv = W L x′ – np × L – np × (Fin OD2 – do 2) tf nf L
  (6.112b)
where, x′ = Horizontal distance between two vertical banks, mm
= Pt sin 60 (For equilateral triangular pitch arrangement), mm
= 43 mm × sin 60 = 37.239 mm
Net free volume, Vfv = 0.91 × 0.9 × 0.037 239 – 21 × (0.019 05)2 × 0.9

–21 × (0.0382 – 0.019 052) (0.889 × 10–3) × 0.9

= 0.020 62 m3
Volumetric equivalent diameter

Dev =   (6.112a)

= = 0.007 555 m

Reynolds number for calculating pressure drop

Res = = 13 974.86

Jfa = 1.085 58 × Re–0.128 025  (6.115)

= 1.085 58 × (13 974.86)–0.128 025 = 0.319 86

Δps =   (6.114)

where, Lp = Effective path length for pressure drop, m

Lp = = 0.037 239 × = 0.1862 m

Density of air ρ = = 1.056 kg/m3

Δps =
(For equilateral triangular pitch SL = ST = Pt)
Δps = 5096 N/m2 = 519.6 mm WC > Δpsmax = 350 mm WC
Discharge pressure of fans are generally less than 350 mm WC.
To decrease the pressure drop, flow area as must be increased
Let the revised tube length, L = 1200 mm and
width of flow area (air facing area), W = 1200 m
Number of tubes in one bank
np ≅ = 27.9 = 28

Let np = 28
W = (np – 1)pt + Fin OD + x
W = 27 × 43 + 38 + x = 1200 ⇒ x = 1 mm
Let W = 1210 mm and x = 11 mm (Revised)
as = WL – npdoL – np (fin OD – do)tf nf L  (6.102)
= 1.21 × 1.2 – 28 × 0.019 05 × 1.2
– 28(0.038 – 0.019 05) × 0.889 × 10–3 × 8 ×

= 0.633 64 m2
Mass velocity, Gs = = 20.941 kg/(m2 · s)

Reynolds number, Re =

Re = = 28 794

Factor for heat transfer coefficient

Jha = 0.085 2072 Re0.7324 = 157.2
Clean heat transfer coefficient, for 100% fin efficiency
hf = Jha   (6.105)

= 157.2 × × (0.7273)1/3
= 147.97 W/(m2 · °C)
Dirty finside coefficient
=

h′f = 143.72 W/(m2 · °C)

(re – rb) = (0.019 – 0.0095 25)

= 0.8
re/rb = 1.994 75, Ω = 0.75 (Fig. 6.31(a))
Dirty finside heat transfer coefficient based on inside heat transfer area and fin
efficiency
h′fi =   (6.108)

= (0.75 × 0.534 86 × 1.2 × nt + 0.043 09 × 1.2 × nt)

= (0.75 × 0.534 86 + 0.04309) ×

= 1370 W/(m2 · °C)

Evaluation of hi
np = 28
at = 28 × (0.014 8336)2 = 4.84 × 10–3 m2

Gt = = 516.53 kg/(m2 · s)

ut = = 0.608 m/s,

Re = = 12 877.3

hi = 0.023 × × (12 877.3)0.8(12.8343)0.33

= 809.445 W/(m2 · °C)
Overall coefficient based on inside heat transfer surface, Eq. (6.109):
=

Ui = 461.82 W/(m2 · °C)

Aireq = = 5.03 m2

Heat transfer area per one bank = πdiL × np = 1.5658 m2

Number of banks required = = 3.2124

Let number of banks = 4 =

Excess heat transfer area = × 100 = 24.5 %

Total number of finned tubes nt = 28 × 4 = 112

Tube OD = 19.05 mm
Fin OD = 38 mm
Number of fins per inch = 8
Thickness of fin = 0.889 mm
Type of fin = spiral wound transverse fin
Tube length = 1200 mm
Width of air facing area = 1210 mm
Tube pitch = 43 mm (equilateral triangular)
Total inside heat transfer area, Ai = 6.263 m2 (provided)
Total fin area Af = 71.885 m2
Total bare tube area, Ao ′ = 5.7913 m2
Air side pressure drop
Net free volume
Vfv = WLx ′ – np × do 2 L – np × (Fin OD2 – do 2) tf nf L  (6.112b)

= 1.21 × 1.2 × 0.037 239 – 28 × (0.019 05)2 × 1.2 – 28

× (0.0382 – 0.019 052) × 0.889 × 10–3 × 8 ×

= 0.0365 m3

Dev =   (6.112a)

= 0.007 518 m
Reynolds number for calculating pressure drop

Re =

= 7871.7
Friction factor for calculating pressure drop
Jfa = 1.085 58 × Re–0.128 025  (6.115)
= 1.085 58 × (7871.7)–0.128 025 = 0.344 25
Effective path length for pressure drop

Lp = = 0.037 239 × = 0.148 956 m  (6.116)

Δps =   (6.114)

= × 10.6

= 1410 N/m2 ≡ 143.8 mm WC < 350 mm WC (Δ psmax)

Tube side pressure drop
Tube side pressure drop can be calculated by following equation:

Δpt = Np   (6.31)
Np = No. of tube side passes = 4
For Re = 128 77.3, Jf = 0.0045 (From Fig. 6.13)
L = 1.2 m, di = 0.014 8336 m

ρ = 850 kg/m3, ut = 0.608 m/s, ≅ 1

Δpt =

= 3401.25 N/m2 ≡ 3.401 kPa

Calculated pressure drop is well within maximum allowable (68 kPa).
Hence, to optimize the design, tube side velocity and tube side heat transfer
coefficient can be increased. With a special arrangement in header, number of
tube side passes can be doubled.
Let number of tubes per pass = 14
at = 14 × (0.014 8336)2 = 2.4194 × 10–3 m2

Gt = 1033.3 kg/(m2 · s), Re = 25 760.75, ut = 1.2156 m/s

hi = 1409.6 W/(m2 · °C)
Ui = 610 W/(m2 · °C)

Aireq = = 3.809 m2

Number of banks required = = 2.4326

Let number of banks = 3 =

Excess area = × 100 = 23.32%

Total number of finned tubes nt = 28 × 3 = 84

Tube OD = 19.05 mm, Fin OD = 38 mm, Number of fins per inch = 8
Thickness of fin = 0.889 mm, Type of fin = Spiral wound transverse fin
Tube length = 1200 mm
Width of air facing area = 1210 mm
 Tube pitch = 43 mm (equilateral triangular)
Total fin area, Af = 53.9139 m2
Total bare tube area, Ao ′ = 4.343 47 m2
Number of tubes per pass = 14
Number of tube side passes = 6
Inside heat transfer area provided, Ai = 4.6974 m2
Air side pressure drop

Dev = = 0.007 518 m (no change)

Res = 7871.7, Jfa = 0.344 25 (no change)

Lp = = 0.037 239 × = 0.111 72 m (Revised)

Revised air side pressure drop

Δps = 1410 × = 1057.53 N/m2 ≡ 1.058 kPa

≡ 107.85 mm WC < Δpsmax (= 350 mm WC)

Tube side pressure drop (revised)
Np = 6, Ret = 25 760.75 Jf = 0.0038
ut = 1.2156 m/s

Δpt =

= 18 687 N/m2 ≡ 18.687 kPa < 68 kPa

Since both side pressure drops are within desired range, design is acceptable.

6.15 PLATE HEAT EXCHANGERS 3,4,11

6.15.1 Introduction
The Plate Heat Exchanger (PHE) consists of a frame in which closely spaced
metal plates are clamped between a head and a follower. The plates have circular
holes in a corner and these holes are sealed around by gasket which also attach
over the plate edges. Holes at the corner are known as ports. Typical plate and
gasketed plate heat exchanger is shown in Fig. 6.42.
Plates are pressed from stainless steel, titanium, Hastelloy C, Incoloy 825, nickel
200, Monel 400, aluminum brass, tantalum, etc. The thickness of plate ranges
from 0.5 to 3 mm. The gap between the plates ranges from 1.5 to 5 mm. This
gap is maintained by a gasket, which is sandwiched between two metal surfaces.
Plate is having corrugated surface. Corrugated surface design strengthens the
plates, increases the effective heat transfer area and produces turbulence in the
fluids flowing through gaps. Plates are easily cleaned and replaced. The heat
transfer area can be readily increased or decreased by adding or removing plates,
respectively. Hot and cold fluids are allowed to flow in alternate channels. For
any one plate, one side corrugated surface is in contact with cold fluid while the
other side of corrugated surface is in contact with hot fluid. Variety of flow
arrangements are possible.

Fig. 6.42 Plate-and-Frame Heat Exchanger. Hot Fluid Flows down between Alternate Plates and Cold
Fluid Flows up between Alternate Plates.

Heat transfer area per plate ranges from 0.03 to 1.5 m2. Height-to-width ratio of
plate ranges from 2 to 3. The upper limit of standard gasketed PHE is reported as
650 m2 of heat transfer surface. Total number of plates of this PHE is 400.
Variety of gasket materials (Table 6.39) are used in plate size of this PHE having
2.8 m height and 1.1 m width.
Table 6.39 Recommended Gasket Materials for PHE2
Gasket material Temperature limit, °C

Styrene butadine rubber 85

Medium nitrile rubber 135
Butyl rubber 150
Fluorocarbon rubber 175
Compressed asbestos fiber 260
 Fig. 6.43 Various Arrangements of Flows in Plate and Frame Heat Exchangers

6.15.2 Advantages and Disadvantages of Plate Heat

Exchangers Over Shell and Tube Heat Exchangers
Advantages
1. Plate heat exchanger provides higher heat transfer coefficient as compared to
shell and tube heat exchanger in available space.
2. Maintenance and cleaning is easier with PHE as compared to shell and tube
heat exchanger.
3. Fouling resistance is less with PHE.
4. Minimum driving force required for heat transfer is 1°C for PHE while the
same is 3 to 5°C for shell and tube heat exchanger.
5. PHE is more flexible. It is easy to increase or decrease the heat transfer area
by adding or removing the plates.
6. With highly viscous slurry, PHE is better as it can be easily cleaned.
Disadvantages
1. Fixed cost of plate and hence, of PHE is higher. PHEs are not as widely used
as shell and tube heat exchangers.
2. Gasketed PHE cannot be used if operating pressure is more than 30 bar g or
operating temperature is more than 250°C. While for the same case shell and
tube heat exchanger can be used.
3. Internal leakage or mixing of two fluids is more common with PHE compared
to shell and tube heat exchanger.
4. Liquid containing suspended particles tend to plug the flow area in PHE easily
and so frequent cleaning becomes necessary. In shell and tube heat exchanger,
chocking can be delayed/avoided by keeping higher velocity in tubes or by
selecting bigger size tubes.
6.15.3 Design of Plate Heat Exchanger
General process design steps3,4,17,18 for PHE are same as that for shell and tube
heat exchanger. Only differences are mentioned as follows:
1. LMTD correction factor (Ft) for PHE is determined from the value of number
of transfer units, (NTU).
Ft = f (NTU)
where, NTU = number of transfer units =   (6.117)

Δt = Temperature change required in process stream, °C.

ΔTm = Logarithmic mean temperature difference, °C
For series flow, Ft can be taken as 0.95. For 1-1 pass and higher passes (2-2, 3-3,
4-4) curves of Ft vs NTU are available (refer to Fig. 6.44).
2. If there is no phase change in the fluid, then forced convective heat transfer
coefficient in conduits or plate film coefficient

  (6.118)

Reported values of constant and exponents are in the range of

c = 0.15 to 0.4
a = 0.65 to 0.85
b = 0.3 to 0.45
x = 0.05 to 0.2
Typical values are c = 0.26, a = 0.65, b = 0.4, and x = 0.14.
where, hp = Plate film coefficient, W/(m2 · °C)
de = Equivalent diameter, m

Fig. 6.44 Log Mean Temperature Correction Factor for Plate Heat Exchanger3 (Reproduced with the
Permission of Elsevier, UK)

de = 2y
y = Gap between the plates, m
Re = Reynold number = (de Gp)/μ = (de up ρ)/μ  (6.119)
Gp = Mass velocity of fluid, kg/(m2 · s)
= ṁ/(nc · Af)
ṁ = Mass flow rate of fluid per channel, kg/s
Af = Cross-sectional area of channel or gap (i.e., flow area), m2
up = Channel velocity, m/s
Af = yW, m2
nc = No. of channels per pass
W = Effective width of plate, m
3. Fouling coefficients for PHEs are given in Table 6.40. Fouling coefficients of
PHEs are higher than the same for shell and tube heat exchangers.
Table 6.40 Recommended Fouling Coefficients for Plate Heat Exchangers
Fluid Fouling coefficient, W/(m2 · °C)

Cooling water 10 000

Steam 100 000
Demineralized water 100 000
Soft water 50 000
Lubricating oil 6 000
Light organic solvent 10 000
Process fluids 5 000 – 20 000

4. Pressure drop in plate heat exchanger can be calculated by the following

equation:
Δp = Δpp + Δppo   (6.120)
where, Δp = Total pressure drop, Pa
Δpp = Conduit or channel pressure drop, Pa
Δppo = Port pressure drop, Pa
Pressure drop occurs, when fluid is flowing through conduit or channel

Δpp = 8Jfp   (6.121)

where, de = Equivalent diameter for conduit, m

de = 2y = 2 × gap between plate, m
ρ = Density of fluid, kg/m3
up = Channel velocity, m/s
Lp = Path length, m
Lp = Plate length × Number of passes
Jfp = Friction factor = f (Re)
Value of Jfp depends on the design of plate used. For the turbulent flow
following approximate equation can be used:
Jfp = 0.6 Re–0.3 (Ref. 3)  (6.122)
Δppo = Pressure drop occurs due to the flow of fluid through ports

= per port  (6.123)

where, uh = Velocity of fluid through the holes, m/s

=   (6.124)

Ah = Area of hole = m2

dh = Diameter of hole, m
Np = Number of passes
Ethanol liquid at the rate of 25 kg/s is to be cooled from 78°C
to 40°C in gasketed plate heat exchanger. Operating pressure at inlet of heat
exchanger is 2 atm g. Cooling water, available in plant at 32°C, is used as a
cooling medium in heat exchanger. Design the suitable plate heat exchanger.
Solution
Heat duty, ϕt = ṁCLΔ t
CL = Specific heat of ethanol liquid at average temperature (59°C)
= 2.93 kJ/(kg · °C)
ϕt = 25 × 2.93 × (78 – 40) = 2783.5 kW ≡ 2783.5 × 103 W
Let outlet temperature of cooling water = 40°C

Fig. 6.45

ϕt = 2783.5 kW = ṁw × 4.1868 × (40 – 32)

ṁw = Mass flow rate of cooling water
= 83.1 kg/s ≡ 301 m3/h
Logarithmic mean temperature difference for the perfect counter current contact

ΔTlm = = 19.254°C

Number of transfer unit (NTU) based on the maximum temperature difference

NTU = = 1.9736  (6.117)

For 1 : 1 pass arrangement, from Fig. 6.44:

Ft = 0.96
ΔTm = Ft ΔTlm = 0.96 × 19.254 = 18.484°C
Assuming overall heat transfer coefficient as
U = 2000 W/(m2 · °C)
A = = 75.3 m2

Selecting the plate having effective width 0.5 m and effective length 1.5 m.
Effective area of one plate = 1.5 × 0.5 = 0.75 m2
For the first trial calculation
Area provided = 75.3 m2 = No. of plates × 0.75 m2
Number of plates = 100.4
Let number of plates = 101
Area provided, Apro = 101 × 0.75 = 75.75 m2
Number of channels per pass, nc = (101 – 1)/2 = 50
Let gap between successive plate, y = 3 mm
Equivalent diameter de = 2y = 6 mm = 6 × 10–3 m
Af = Flow area = Cross-sectional area of gap
Af = yW = 0.003 × 0.5 = 1.5 × 10–3 m2
Ethanol side heat transfer coefficient

=   (6.118)
Let c = 0.26, a = 0.65, b = 0.4 and ≅ 1

hp · = 0.26Re0.65 Pr0.4

For the flow of ethanol, Re =

Mass velocity of ethanol, GpE =

= = 333.33 kg/(m2 · s)

Channel velocity, upE = GpE/ρ

ρ = Density of ethanol at 59°C = 775 kg/m3
upE = = 0.43 m/s

Re = = 3333.3

Viscosity of ethanol at 59°C, μ = 0.6 cP

Pr =

Thermal conductivity of ethanol at 59°C, k = 0.147 W/(m · °C)

Pr = = 11.96

hpE = 0.26

hpE = 0.26 × × (3333.3)0.65(11.96)0.4

hpE = 3350.43 W/(m2 · °C)

Cooling water side heat transfer coefficient:
Mass velocity of water, GpW =
= = 1108 kg/(m2 · s)

Density of water at 36°C, ρ = 993.684 kg/m3

Viscosity of water at 36°C μ = 0.72 cP
hpW = = 1.115 m/s

Reynold number Re =

Re = = 9233

Thermal conductivity of water at 36°C, k = 0.6228 W/(m · °C)

Prandtl number for water at 36°C
Pr =

Pr = 4.84
hpW = 0.26 Re0.65Pr0.4

hpW = × (9233)0.65 × (4.84)0.4

hpW = 19 168.66 W/(m2 · °C)

Overall heat transfer coefficient:
=

hfE = Fouling coefficient for ethanol = 10 000 W/(m2 · °C)

tp = Plate thickness = 1 mm = 0.001 m
kp = Thermal conductivity of plate material
Let the plate material be titanium.
kp = 21 W/(m · °C)
hfW = Fouling coefficient for cooling water in PHE
= 10 000 W/(m2 · °C) (From Table 6.40)
 =

= 1671.52 W/(m2 · °C)

Area required = = 90 m2

Area required > Area provided

For the second trial calculations
Let number of plates be 121 with 2-2 pass arrangement.
Number of channels per pass = = 30

Area provided, Apro = 121 × 0.75 = 90.75 m2

For NTU = 1.9736 Ft = 0.95 (From Fig. 6.44)
ΔTm = 0.95 × 19.254°C = 18.2913°C
hpE
Mass velocity, Gpe = 333.33 × = 555.55 kg/(m2 · s)

Channel velocity, upe = 0.43 × = 0.7167 m/s

Re = = 5555.5

hpE = 3350.43 × = 4669.84 W/(m2·°C)

hpW
Mass velocity, GpW = 1108 × = 1846.67 kg/(m2 · s)

Channel velocity, upW = 1.115 × = 1.8583 m/s

Re = 9233.33 × = 15 389

hpW = 19 168.66 × = 26 717.5 W/(m2·°C)

Overall coefficient
 =

Uo = 2003.25 W/(m2·°C)

Area required, Areq = = 75.9646 m2

Area provided, Apro = 90.75 m2

Excess area = × 100 = 19.46% (acceptable)

Pressure drops
Ethanol side pressure drop
Δp = Δpp + Δppo  (6.120)
Channel pressure drop
Jfp = 0.6 Re–0.3  (6.122)
= 0.6(5555.5)–0.3 = 0.045 158
Path length, Lp = Plate length × Number of passes = 1.5 × 2 = 3 m
Δpp = 8 × 0.045 158 ×

= 35 953.5 N/m2 ≡ 35.954 kPa

Let port diameter (hole diameter) = 120 mm
Hole area = 0.122 = 0.011 31 m2

uh = velocity of ethanol through hole (port)

uh =

= 2.8522 m/s
Port pressure drop
Δppo = 1.3 × Np

= 1.3 × × 2 = 8196 N/m2 ≡ 8.196 kPa per port

Assuming 2 ports, ethanol side pressure drop

Δp = 35.954 + 2 × 8.196 = 51.746 kPa
It may be noted that port pressure drop is 31.7% of total pressure drop which is
significant.
Cooling water side pressure drop:
Re = 15 389
Jfp = 0.6 Re–0.3 = 0.6 × (15 389)–0.3 = 0.033 265
Lf = Plate length × No. of passes = 1.5 × 2 = 3 m
Channel pressure drop by Eq. (6.121):

Δpp = 8Jfp

= 8 × 0.033 265 ×

= 228 295.6 N/m2 ≡ 228.296 kPa per port (very high)

Trial-2:
To decrease the cooling water side pressure drop, let number of cooling water
side passes = 1, i.e., 2 : 1 arrangement. Ethanol side heat transfer coefficient,
velocity and pressure drop are same as that for the previous case (2 : 2
arrangement).
For 2 : 1 arrangement
up = 0.7167 m/s, hpE = 4669.84 W/(m2 · °C), Δp = 44.15 kPa (ethanol side)
For cooling water,
No. of plates = 121, No. of passes = 1
For the flow of water,
Number of channels per pass = = 60

Mass velocity, GpW = = 923.33 kg/(m2 · s)

Velocity of water through channel, up = = 0.929 m/s

Re = = 7694.5

hpW = 26 717.5 × = 17 026.54 W/(m2 · °C)

Overall heat transfer coefficient
=

Uo = 1921.26 W(m2 · °C)

Area required Areq = = 79.206 m2

Excess area = × 100 = 14.575%

Cooling water side pressure drop

Re = 7694.5
Jfp = 0.6 Re–0.3 = 0.6 × (7694.5)–0.3 = 0.040 95
Channel pressure drop

Δpp = 8Jfp   (6.121)

Path length Lp = Plate length × Number of passes = 1.5 m

Δpp = 8 × 0.040 95 × ×

= 35 118.3 N/m2 ≡ 35.118 kPa

Hole or port diameter = 120 mm
Hole area = 0.011 31 m2
uh = Velocity of water through hole (port)

uh = = 7.3942 m/s (high)

Increase the port size to 150 mm.

uh = = 4.732 m/s (acceptable)

Δppt = × Np

= 1.3 × × 1
= 14462.7 N/m2 ≡ 14.463 kPa per port
Assuming 2 ports, pressure drop for the flow of cooling water
Δp = 35.118 kPa + 2 × 14.463 kPa = 64.044 kPa
Here also, port pressure drop is 46.7% of total pressure drop.

6.16 SPIRAL FLOW HEAT EXCHANGERS

6.16.1 Introduction
Spiral flow heat exchangers provide a spiral flow path for one or both (cold and
hot) fluids being handled. They are of two types; (i) spiral tube exchanger and
(ii) spiral-plate exchanger. These heat exchangers offer the following advantages
over conventional shell and tube heat exchangers:
1. Compactness
2. Centrifugal force increases the heat transfer coefficient, particularly of highly
viscous liquid slurries or sludges.
3. Spiral plate heat exchangers foul at much lower rate than the shell and tube
heat exchanger because of the single flow passage and curved flow path
which does not allow easy settlement of solids.
4. Relative ease of cleaning.
5. Spiral configuration reduces stress associated with differential thermal
expansion.
6. Maintenance cost is less.
Disadvantages of spiral-tube and spiral-plate heat exchangers over shell and tube
heat exchangers are as follows.
1. Spiral flow heat exchangers are more expensive (requires higher fixed cost)
than shell and tube heat exchangers having the same heat transfer surface.
2. Design of spiral flow heat exchangers is not well established like the same of
shell and tube heat exchanger.
3. Pressure drop (loss) is higher than the same in shell and tube heat exchanger.
4. Maximum design pressure for spiral plate heat exchanger is 10 bar g as the
spiral construction limits the design pressure.
5. Gasket is special and assembly requires skill.
6.16.2 Spiral Plate Heat Exchanger
Spiral plate heat exchanger is fabricated from two relatively long strips of plate
which are spaced apart and wound around an open split center to form a pair of
concentric spiral passages. Spacing is maintained uniformly along the length of
the spiral by spacer studs welded to plate. Commonly used material for the
fabrication of spiral plate heat exchanger are carbon steel, stainless steel,
Hastelloys, nickel and nickel alloys, titanium, copper-alloys, etc.
Figure 6.46 shows spiral flow in both channels. Flow channels are closed by
welding alternate channels at both sides of spiral plate.

Fig. 6.46 Spiral Flow in Both Channels

Figure 6.47 shows flow is spiral in one channel, axial in other. One channel is
completely open on both ends and other closed at both sides of plate.
 Fig. 6.47 Spiral Flow in One Channel, Axial in Another

Figure 6.48 shows combination flow.

Fig. 6.48 Combination Flow

Spiral flow in both channels This type of flow arrangement is used for liquid-
to-liquid services. With this type of flow arrangement, spiral plate heat
exchanger is covered by flat heads on both sides. Usually two liquids flow
counter currently with the cold liquid entering at the periphery and flowing
toward the core and the hot liquid entering at the core and flowing toward the
periphery. Use of this type spiral heat exchanger is common in cooling hot
hydrogenated edible oil from deodorizer as it offers good heat transfer
coefficient for viscous oil.
Spiral flow in one channel and axial flow in other This type of flow
arrangement is used for condensing or boiling. With this type of flow
arrangement, spiral flow heat exchanger is covered by conical heads on one or
both sides. Condensation or boiling takes place in axial direction. This
arrangement is preferred where there is a large difference in the volumes of two
fluids. In case of condensation, difference between volumetric flow rates of
condensing vapour and cooling medium is always large. In case of boiling, if hot
oil is used as heating medium, then also the difference between volumetric flow
rate of boiling liquid and the same of hot oil is large.
Combination flow Combination flow is used to condense vapours. In this
arrangement, condensing vapour flows axially and then condensate flows
spirally. This arrangement is used for condensation with subcooling. Part of the
open spiral is kept closed at the top. Entering fluid (condensing vapour) flows
axially through the center part of assembly in downward direction. Condensate
at bottom flows spirally and coming out from the side bottom. With this
arrangement, spiral heat exchanger is equipped with conical head at top and flat
cover at bottom.
6.16.3 Process Design Steps for Spiral Plate Heat
Exchanger
General process design steps18,19 are as follows:
1. Calculate heat duty (ϕ in kW).
2. Select cooling medium or heating medium.
3. Based on energy balance calculate the mass flow rate of heating medium or
cooling medium.
4. Calculate the LMTD.
5. Assume the value of overall heat transfer coefficient for the first trial
calculations.
6. Determine the heat transfer area required based on the assumed value of
overall heat transfer coefficient for first trial calculations.
A =

7. Heat transfer area for spiral plate heat exchanger

A = 2 × L × H   (6.125)
where, A = Heat transfer area, m2
L = Length of plate, m
H = Width of plate = Height or length of heat exchanger, m
Fix the value of H and find the value of L
L =   (6.126)

Diameter of heat exchanger or diameter of outside spiral can be calculated by

following equation:
Ds = [1.28 × L(dc + dh + 2t) + c2)1/2   (6.127)
where, Ds = Outside spiral diameter, m
L = Length of plate, m
dc = Channel spacing of cold side, m
dh = Channel spacing of hot side, m
t = Plate thickness, m
c = Core diameter, m
Some dimensions of spiral plate heat exchanger are standardized and are as
follows.
(a) Plate width H : 101.6 mm (4 in), 152.4 mm (6 in),
304.8 mm (12 in), 457.2 mm (18 in), 609.6 mm (24 in), 762 mm (30 in)
914.4 mm (36 in), 1219.2 mm (48 in), 1524 mm (60 in), 1778 mm (72 in)
(b) Core Diameter, C : 203.2 mm (8 in), 304.8 mm (12 in)
(c) Channel spacing, dc and dh : 4.76 mm (3/16 in) (for 304.8 mm maximum
width)
6.35 mm (1/4 in) (for 1219.2 mm maximum width)
7.94 mm (5/16 in), 9.525 mm (3/8 in), 12.7 mm (1/2 in), 19.05 mm (3/4 in),
25.4 mm (1 in)
(d) Plate thickness, t : For carbon steel 3.175 mm (1/8 in),
4.76 m (3/16 in),
6.35 mm (1/4 in), 7.94 mm (5/16 in)
For stainless steel 14 to 3 US gauge
In spiral plate heat exchanger to arrive at the best compact design outside
diameter of spiral Ds and plate width (H) are kept approximately equal.
8. Calculate the equivalent diameter of flow channel by following equation:
De =

For hot fluid

Deh =   (6.128a)
For cold fluid
Dec =   (6.128b)

9. Calculate the Reynolds number by following equation:

Re =   (6.129)

where, ṁ = Mass flow rate of fluid, kg/s

H = Width of plate, m
μ = Viscosity of fluid, kg/(m · s) or Pa · s
Reynolds number is calculated for both; cold fluid and hot fluid.
10. Calculate the critical Reynolds number Rec. Critical Reynolds number is the
value of Reynolds number above which turbulent flow is achieved.

Rec =   (6.130)

where, De = Equivalent diameter of flow channel, m

Ds = Outside diameter of spiral, m
11. Calculate the cold fluid side and hot fluid side heat transfer coefficients by
using the suitable correlations. Various correlations for calculating heat
transfer coefficient for spiral plate heat exchangers are given as follows.
(a) For spiral flow with no phase change and Re > Rec ( i.e., turbulent flow):

=   (6.131)

For spiral flow with, no phase change and, Re < Rec (i.e., laminar flow)

=   (6.132)

where, h = Heat transfer coefficient, W/(m2 · °C)

Cp = Specific heat of fluid, J/(kg · °C)
G = Mass velocity of fluid, kg/(m2 · s)
Re = Reynolds number
Pr = Prandtl number
μ = Viscosity of fluid, kg/(m · s) or Pa · s
k = Thermal conductivity of fluid, W/(m · °C)
d = Channel spacing, dc or dh, m
Ds = Outside spiral diameter, m
Here mass velocity is calculated by following equation:
G =   (6.133)

where, ṁ = Mass flow rate, kg/s

d = dc or dh, m
H = Width of plate, m
(b) For axial flow with no phase change and Re > 10 000
= 0.023 Re– 0.2 Pr –2/3   (6.134)

12. Calculate the overall heat transfer coefficient, Uo, by following equation.

=   (6.135)

where, hc = Cold fluid side heat transfer coefficient, W/(m2 · °C)

hh = Hot fluid side heat transfer coefficient, W/(m2 · °C)
hcd = Cold fluid side fouling coefficient, W/(m2 · °C)
hhd = Hot fluid side fouling coefficient, W/(m2 · °C)
t = Thickness of plate, m
km = Thermal conductivity of plate material, W/(m · °C)
13. Calculate the heat transfer area required
Ar =

14. Calculate the % excess heat transfer area. Ideally it should be in between 10
to 20%.
15. Calculate the cold fluid side and hot fluid side pressure drops by using the
suitable correlations for spiral flow heat exchanger.
(a) Spiral flow with no phase change and Re > Rec

Δp =   (6.136a)
(b) Spiral flow with no phase change and 100 < Re < Rec

Δp =   (6.136b)

(c) Spiral flow with no phase change and Re < 100

Δp = 5.5256 × 10–5   (6.136c)

(d) Axial flow with no phase change and Re > 10 000

Δp =   (6.136d)

where, L = Total length of plate, m

ρ = Density of fluid, kg/m3
ṁ = Mass flow rate of fluid, kg/s
ds = Channel spacing (dc or dh), m
H = Width of plate, m
μf = Viscosity of fluid, kg / (m · s) or Pa · s
μb = Bulk fluid viscosity, kg/(m · s)
Calculated pressure drop must be less than or equal to maximum allowable
pressure drop or optimum pressure drop.
Design a spiral flow plate heat exchanger for cooling the
malamine suspension liquid from 47°C to 33°C by using cooling water as
cooling medium based on the following data. Cooling water is available at 32°C.
Data
(i)
Table 6.41 Composition of Malamine Suspension Liquid
Component ṁ, kg/h mass % mole %

Water 16 052 57.21 74.2

Carbon dioxide 3 151 11.23 5.49
Ammonia 3 055 10.89 14.96
Urea 1 329 4.74 1.84
Melamine 4 470 15.93 2.95

Total 28 057 100.00 100.00

(ii) Physical properties of melamine suspension liquid

(a) Specific heat CL = 3.349 kJ/(kg · °C)
(b) Thermal conductivity k = 0.4935 W/(m · °C)
(c) Viscosity: At 47°C, μf = 1.8 cP
At 33°C, μf = 2.3 cP
(d) Density: At 47°C, ρ = 1120 kg/m3
At 33°C, ρ = 1126 kg/m3
(iii) Suitable material of construction: SA 240 Gr 316 L
(iv) Maximum allowable pressure drops
For malamine suspension liquid = 1.5 bar
For cooling water = 1 bar
Solution
Heat duty ϕt = ṁCLΔt = × 3.349 × (47 – 33)

= 365.41 kW ≡ 365.41 × 103 W

Mass flow rate of cooling water
Energy balance
Energy balance ϕt = ṁw CLw Δt′ = ṁw × 4.1868 × (36 – 32)
For the first trial calculations let the outlet temperature of cooling water = 36°C
ṁw = = 21.819 kg/s ≡ 78.55 m3/h

Fig. 6.49

ΔTlm =

= 4.17°C = ΔTm
Assuming the value of overall heat transfer coefficient
U = 1500 W/(m2 · °C)
Heat transfer area

A = = 58.42 m2

Heat transfer area provided for the first trial calculations

Ap = 58.42 m2 = 2 × L × H  (6.125)
Let H = 24 in = 0.6096 m
L = = 47.9167 m

To get the best compact design of spiral plate heat exchanger, width of plate (or
length of heat exchanger) should be approximately equal to outside spiral
diameter (or outside diameter of heat exchanger).
Let other dimensions of heat exchanger:
dc = Channel spacing for cold side fluid = 12.7 mm
dh = Channel spacing for hot side fluid = 6.35 mm
c = Core diameter = 203.2 mm
t = Plate thickness = 3.175 mm
Outside spiral diameter
Ds = [1.28 × L(dc + dh + 2t) + c2]1/2  (6.121)
= [1.28 × 47.9167(0.0127 + 0.006 35 + 2 × 0.003 175) + 0.20322]1/2
= 1.2646 m
H ≠ Ds
For second trial calculations let H = 36 in = 914.4 mm
L = 31.944 m
Ds = [1.28 × 31.944(0.0127 + 0.006 35 + 2 × 0.003 175) + 0.20322]1/2
= 1.04 m
H ≅ Ds
H = 914.4 mm, Ds = 1040 mm, L = 31.944 mm
dc = 12.7 mm, dh = 6.35 mm, C = 203.2 mm
t = 3.175 mm
Note: These dimensions of heat exchanger are not the final dimensions. These
can be changed after heat transfer coefficient calculations and/or pressure drop
calculations.
Hot fluid side heat transfer coefficient calculations:
Equivalent diameter Deh =   (6.128a)

Deh = = 12.6 mm

Reynolds number Re =   (6.129)

μf =

Re = = 8315.3

Critical Reynolds number: Rec = 20 000 = 20 000

= 4871.9
Re > Rec, i.e., process fluid will flow in turbulent region.
For spiral flow with no phase change and Re > Rec

= (0.023 Re–0.2 × Pr–2/3)6.131)

Cp = 3.349 kJ/(kg · °C)

Prandtl number, Pr =

= 13.912
Mass velocity, G =

= 1342.234 kg/(m2 · s)
= × 0.023 × 8315.3–0.2 × 13.912–2/3

hh = 3063.9 W/(m2 · s)
Cold fluid side heat transfer coefficient calculations:
Dec =   (6.128b)

= = 25.052 mm

Viscosity of water at 34°C, μ = 0.738 cP (Ref. 2)

Reynolds number Re =   (6.129)

= 64 665.44
Critical Reynolds number Rec =

= 6070.28
Re > Rec
For spiral flow with no phase change and Re > Rec

=   (6.131)

Cp = 4186.8 J/(kg · °C)

Mass velocity G =

= 1878.9 kg/(m2 · s)
Thermal conductivity of water at 34°C, k = 0.62 W/(m · k)
Prandtl number Pr = = 4.9836

= × 0.023 × 64665.44–0.2 × 4.9836–2/3

hc = 7336.76 W/(m2 · °C)

Overall heat transfer coefficient

=   (6.135)

where, hc = 7336.76 W/(m2 · °C), hh = 3063.9 (m2 · °C),

hcd = 15 000 W/(m2 · °C), hhd = 15 000 W/(m2 · °C),
t = 0.003 175 m
km = 16.26 W/(m · °C) (of stainless steel)

Uo = 1263.78 W/(m2 · °C)

Heat transfer area required, Ar1 =

= = 69.34 m2

Required area, Ar1 > Area provided, Ap1 which is not acceptable.
For the second trial calculations let L = 42 m
Area provided Ap2 = 2 × L × H
= 2 × 42 × 0.9144 = 76.81 m2
Revised value of Ds (outside diameter of spiral)
Ds = [1.28 × L(dc + dh + 2t) + c2]1/2
= [1.28 × 42(0.0127 + 0.006 35 + 2 × 0.003 175) + 0.20322]1/2
= 1.186 m
Critical Reynolds number for hot fluid
Rec = 20 000 = 4671.34

Re > Rec
Revised value of hot fluid side heat transfer coefficient

hh =

= 3048.56 W/(m2 · °C)

Critical Reynolds number for cold fluid

Rec = 20 000 = 5820.4

Re > Rec
Revised value of cold fluid side heat transfer coefficient

hc = × 7336.76 = 7266.53 W/(m2 · °C)

Overall heat transfer coefficient

=

Uo = 1259.07 W/(m2 · °C)

Heat transfer area required
Ar2 = = 69.6 m2 < Ap2

Excess heat transfer area = × 100 = 10.36% (acceptable)

Malamine suspension liquid side pressure drop

Δp = 0.0789   (6.136a)

L = 42 m, ρ = 1123 kg/m3, W = 7.7936 kg/s

ds = dh = 0.006 35 m, H = 0.9144 m
μf = 2.05 × 10–3 kg/(m · s)

Δp = 0.0789 ×

= 55 004.5 Pa ≡ 55 kPa ≡ 0.55 bar < 1 bar (acceptable)

Cooling water side pressure drop
L = 42 m, ρ = 1000 kg/m3, W = 21.819 kg/s
ds = dc = 0.0127 m, H = 0.9144 m, μf = 0.738 × 10–3 kg/(m · s)
Δp = 0.0789 ×

= 52 023.5 Pa ≡ 0.52 bar < 1 bar (acceptable)

Comments: In the given case, malamine suspension liquid solution is a slurry
solution. Hence, spiral plate heat exchanger is a better choice as compared to
shell and tube heat exchanger because of the following reasons:
(i) Slurry solution must be allocated to shell side of shell and tube heat
exchanger to facilitate the ease of flow. Shell and tube heat exchanger must
be U-tube or floating head heat exchanger to facilitate removal of tube
bundle and for frequent cleaning. Also, malamine solution required the use
of special material, SS 316 L. Hence, shell, tubes, tube sheets, baffles,
floating head and/or one of the heads must be of SS 316 L.
(ii) Shell and tube heat exchanger provides less heat transfer coefficient for the
slurry solution as compared to spiral plate heat exchanger. Also, fouling
coefficients are lesser with shell and tube heat exchanger compared to spiral
plate heat exchanger. Hence, shell and tube heat exchanger requires more
heat transfer area.
(iii) Cleaning is easier with spiral plate heat exchanger but requires skill.
(iv) Because of the compactness spiral plate heat exchanger will occupy less
space as compared to shell and tube heat exchanger.
(v) Here cooling water is available at 32°C. It can cool the malamine solution up
to 33°C with spiral plate heat exchanger, wherein driving force is only 1°C.
Minimum driving force required for heat transfer in shell and tube heat
exchanger is 3°C.
Based on (i) and (ii) it can be concluded that the cost difference between spiral
plate heat exchanger and shell and tube heat exchanger is less.

6.17 BRAZED ALUMINUM PLATE-FIN HEAT

EXCHANGER
Typical brazed aluminum plate-fin heat exchanger handling multiple streams is
shown in Fig. 6.50. This type of heat exchanger is preferred for low temperature
(subzero temperature) cryogenic applications as it offers the following features:
1. It can handle temperature approaches down to less than 2°C. In cryogenic
 duties, operating cost is dominated by the cost of energy required to generate
the low temperatures; hence, such a close temperature approach is very
important.
2. It is the most compact and lightest heat exchanger and it provides the
maximum heat transfer area density.
3. It can operate with a thermal effectiveness up to 98%.
4. It has multistream capability. All cold streams produced in process can be
passed through single heat exchanger to cool the incoming warm streams.
5. It is an expensive heat exchanger.

Fig. 6.50 Brazed Aluminum Plate-fin Heat Exchanger20

6.17.1 Construction and Working

It consists of alternating layers of plates and corrugated fins. Flow passes are
formed between the consecutive plates (plates are known as parting sheets).
These sheets provide the primary surface and fins provide the secondary surface
for heat transfer.
Brazed or corrugated fins hold the heat exchanger together. At the end of heat
exchanger, pads of finning are laid at an angle and serve as distributors, while in
major part of heat exchanger finning are laid parallel to the axis. Streams are
flowing truly in countercurrent directions. Each stream flows in number of
layers, each of which is divided into numerous parallel nearly rectangular
subchannels by the fins. Fin heights and frequencies decide the size of these
subchannels. Fin heights are in between 5 and 9 mm and fin frequencies in the
main heat transfer region is 590 to 787 fins/m (15 to 20 fins/in). The equivalent
hydraulic diameters of these subchannels are only a few millimeters. These small
passages result in very high heat transfer area density in the range of 800 to 1500
m2/m3. For comparison, heat transfer area density of shell and tube heat
exchanger is about 50 to 250 m2/m3 and the same of plate and frame heat
exchangers are about 200 to 1000 m2/m3. Such high area density and with
aluminum construction make this heat exchanger the smallest as well as the
lightest among all other type of heat exchangers. Typical size of such a heat
exchanger can be 1.2 m wide, 1.2 m deep and 6.2 m long. This heat exchanger
can also be used as vaporizer or condenser.
Brazed aluminum plate heat exchangers can be used at pressures up to 100 bar
and within a temperature range of –269°C to 100°C. With appropriate alloys for
the headers and nozzles, they can be used up to 200°C. However, the maximum
permissible temperautre for aluminum alloys decreases rapidly with increasing
pressure.

6.18 OTHER TYPES OF HEAT EXCHANGE

EQUIPMENTS
Several other types of heat exchange equipments are in use in chemical process
industry. To name a few; serpentine coolers, scraped surface heat exchangers,
helical coil-type heat exchangers, etc. Each one has specific area in which they
are used extensively due to convenience. For example, serpentine coolers are
used for condensing ammonia vapour in cold storages. Similarly for cooling
Vanaspati (hydrogenated edible oils), crystallizing menthol from its solution,
wax cooling, manufacture of ice cream, and for freeze crystallization
applications, scrapped surface heat exchangers are extensively used. Heat
transfer in jackets and half coils (of CSTR) with agitators are covered in Chapter
10. Process engineer should evaluate all possible designs for selecting a best
suited equipment for a given application. Due to space limitations, it is not
possible to cover process design of all types of heat exchange equipments in this
book but reader is advised to refer literature for specific equipment design.
6.19 HEAT EXCHANGER NETWORKING FOR
ENERGY INTEGRATION
Energy requirement of a chemical plant can be considerably reduced by process
integration. For example, in acetaldehyde plant, exothermicity of the reaction
(e.g., partial oxidation of ethanol) is utilized as a part of heat duty required by
the reboiler of the distillation column of the same plant. This integration reduces
saturated steam required for the reboiler of distillation column and eliminates the
requirement of cooling medium for the reactor. Large scale chemical plant uses
large number of heat exchangers. These heat exchanges are either heaters and
vaporizers or coolers and condensers. Integration between few of them reduces
hot utility requirement (i.e., requirement of heating medium) as well as cold
utility requirement. (i.e., requirement of cooling medium). Systematic method is
developed by Linnhoff and others21 for the energy integration via systematic
heat exchanger networking which is known as pinch technology.
In heat exchanger networking via pinch technology, following facts are
considered:
1. The heat transferred from a hot stream must be equal to heat transferred from
cold stream (heat losses are neglected).
2. Heat can only be transferred from a hotter fluid to a colder fluid (second law
of thermodynamics). Therefore, the temperature of a cold fluid must be less
than that of a hot fluid at all points along the path of heat exchanger.
ΔTmin: In actual energy integration temperature of hot stream must be reasonably
greater than the same of cold stream to get an acceptable rate of heat transfer.
This acceptable minimum temperature difference is also called pinch
temperature difference. Choice of pinch temperature difference (ΔTmin) is
important. The lower value of ΔTmin decreases the energy requirement but
increases the fixed cost of heat exchangers. Minimum suggested value of ΔTmin
is 10°C. This implies that in any of the exchangers to be used in the network, the
temperature difference between the hot stream and cold stream will not be less
than 10°C. For any network, there will be a best value for the minimum
temperature difference (ΔTmin) that will give the lowest total annual operating
cost. It is usually in between 10°C to 20°C.
“Hot streams” and “Cold streams”: For heat exchanger networking total streams
considered are divided into two types. “Hot streams” are the streams that need
cooling (heat sources) while “Cold streams” are the streams that need heating
(heat sinks).
6.19.1 Representation of Heat Exchanger Network
Heat exchanger network is represented as a grid. The process streams are drawn
as horizontal lines, with the stream numbers shown in square boxes, as shown in
Fig. 6.51. Hot streams are drawn at the top of grid and flow from left to right.
The cold streams are drawn at bottom and flow from right to left.

Fig. 6.51 Grid Presentation

Coolers and heaters are represented as circles; (C ) and (H), respectively. Heat
exchangers which are used to exchange the heat between the two process
streams are marked by two circles and the two circles are connected by vertical
line. Two circles with vertical line connect the two streams between which heat
is being exchanged.
As shown in Fig. 6.51, in this case a is hot stream number, n is cold stream
number. A is a heat exchanger exchanging the heat between hot stream a and
cold stream n. Cooler C cools the stream further after it leaves the heat
exchanger A. Heater H heats the stream n after it leaves the heat exchanger A.
Calculations for finding minimum utility requirements for the given stream data:
First step for this is selection of the value of minimum temperature difference,
ΔTmin between hot streams and cold streams. Then for the chosen value of
ΔTmin, minimum utility required can be determined by the following method,
developed by Hohmann22 and Lockhart. This method is illustrated by following
example.
Synthesize the heat exchanger network for the following four
process streams such that resulting heat exchanger network will require the
minimum hot and cold utilities. Also, find the values of minimum utilities (hot
and cold) required.
 Table 6.42 Process Stream Data
Stream Type ṁCp, kW/°C tin, °C tout, °C

1 Hot 10 250 130

2 Hot 8 160 90
3 Cold 7 60 160
4 Cold 6 110 260

Solution
Select the value of ΔTmin = 10°C. (Selected value of ΔTmin, may not be equal to
optimum temperature difference.) For ΔTmin = 10°C, Table 6.43 is prepared by
calculations, presented below.
Method for Preparing the Table:
The hottest temperature in the data is cold stream 4 outlet temperature (260°C).
List it at the top of the cold stream temperature column (column 2). Note down
the corresponding hot stream temperature (270°C), which is 10°C (= ΔTmin)
hotter at the top of the hot stream temperature column. Next highest temperature
is of hot stream 1 temperature; 250°C. Corresponding cold stream temperature
based on ΔTmin = 10°C is 240°C and written second in column 2. Similarly,
temperatures of hot streams and cold streams are written in descending order
either in column 1 (if it is hot stream temperature) or in column 2 (if it is cold
stream temperature). While corresponding temperatures for ΔTmin = 10°C are
written in other columns.
There are seven temperature intervals in this problem. In column 3 available heat
by hot streams for each temperature interval is calculated. No hot stream is
available at 270°C, hence, available heat for seventh interval is zero. Since hot
stream temperature of 270°C is of academic interest only as it is shown in
brackets in column 1.
For sixth interval (in column 3) available heat is calculated as (ṁCp)1 (250°C–
170°C) = 10 (250 – 170) = 800 kW. So, in sixth interval hot stream 1 is cooled
from 250°C to 170°C. Stream 1 is also the only hot stream in interval 5,
contributing another 100 kW, when it cools from 170°C to 160°C. Interval 4 has
both hot streams 1 and 2. In interval 4 both streams are cooled from 160°C to
130°C. Hence available heat for interval 4 is calculated as follows.
[(ṁCp)1 + (ṁCp)2] (160°C – 130°C) = ((10 + 8) × 30) = 540 kW
In column 4 cascaded heat available indicates cumulative or total heat available
at temperature t, when hot streams cool from 250°C to temperature t°C.
In column 5 required heat by cold streams for each temperature interval is
calculated. Stream 4 is present only in interval 7 and requires 120 kW heat to
increase its temperature from 240°C to 260°C, (ṁCp)4 (260 – 240) = 6(260 –
240) = 120 kW.
In interval 5 both streams 3 and 4 are present. Hence, required heat for interval 5
is [(ṁCp)3 + ((ṁCp)4] (160 – 150) = ((6 + 7) × 10) = 130 kW. Cascaded required
heat by cold streams at different temperatures are shown in column 6.
Column 7 represents the net available heat for each interval.
Net available heat (column 7) of any interval = Heat available with hot streams
(column 3) in this interval – Heat required by cold streams (column 5) in the
same interval.
Column 8 represents the cascaded net available heat or cumulative net available
heat at the end of each interval, starting from the seventh interval to first interval.
Cascaded heat is the amount of heat available from the hot streams over that
required by cold streams at any temperature t, as one starts the calculation from
the highest temperature to the lowest temperature. Find the highest negative
number in column 8. In this example it is –120 kW. It implies that 120 kW is the
minimum amount of heat that must be supplied from hot utilities. The same
amount of heat (120 kW) is added in each number of eighth column. This gives
ninth column. The point in the
last column where the value of adjusted cascaded heat is zero is the pinch point.
Top number in last column (120 kW), is the minimum amount of heat that must
be supplied from hot utilities and the bottom number (280 kW) is the minimum
amount of heat that must be removed by cold utilities.
6.19.2 Grand Composite Curve (GCC)
It is an important curve or graph to understand heat exchanger network synthesis
problem. To draw this curve for the given problem, Table 6.43 must be prepared,
based on given data. Then GCC is the plot of adjusted cascaded heat obtained in
the last column (ninth column) vs hot stream/cold stream temperatures.
(Columns 1 and 2). GCC for this example is shown in Fig. 6.52.

Fig. 6.52 Grand Composite Curve for Example 6.13

In this curve each interval is represented by line segment. Each interval is either
consuming heat (heat sink) or producing heat (heat source). Interval which acts
as a heat sink, represented by a line segment that moves down and to the left.
Interval which acts as a heat source, represented by a line segment that moves
down and to the right. For example, seventh interval locally requires 120 kW of
heat input (heat sink), hence it is represented by a line segment which moves
downward and to the left. Interval 6 produces 320 kW of heat (heat source),
hence it is represented by line segment moving downward and to the right.
Whenever there is a heat source segment just above a heat sink segment right
facing noze appear in the Grand Composite Curve. Such right facing nozes can
be self-integrated. While the bold segment above the pinch point requires heat
from hot utilities and bold segment below the pinch up to the dashed line
represents the heat which must be removed by cold utilities.
Heat exchanger network synthesis for the maximum energy recovery:
Network design for the maximum energy recovery and minimum use of hot and
cold utilities can be divided in the following steps:
1. Select the value of ΔTmin.
2. Prepare the table (like Table 6.43) for drawing the Grant Composite Curve.
3. Calculate the minimum hot utility required and the minimum cold utility
required. Identify the pinch point.
4. For the maximum heat recovery do not use cold utilities above the pinch and
do not use hot utilities below the pinch. Also, do not transfer heat across the
pinch.
5. Estimate the number of exchangers above the pinch equal to N – 1, where N is
the number of streams above the pinch. Similarly, estimate the number of
exchangers below the pinch equal to M – 1, where M is the number of streams
below the pinch.
6. Develop a suitable network.
Applying the same steps to the given Example 6.13:
1. ΔTmin = 10°C (selected)
2. GCC is drawn based on the value of ΔTm (Fig. 6.52).
3. Minimum hot utility required is to supply 120 kW heat and the minimum cold
utility required is for removing 280 kW heat. Pinch point occurs at 250°C hot
stream temperature and equivalent cold stream temperature of 240°C.
4. Above the pinch only stream 4 exists. Hot oil (another stream) can be used to
increase the temperature of stream 4 from 240°C to 260°C. Thus N = 2 which
means one heat exchanger is needed for this heat duty. Heat duty of this
heater is 120 kW. Below the pinch all four process streams and one cold
 utility stream exists (M = 5). Thus there is a need of four exchangers for this
part or below the pinch.
5. One possible heat exchanger network for the maximum energy recovery is
shown in Fig. 6.53.
Hot stream 1 contains (ṁCp) Δt = 10 × (250 – 130) = 1200 kW heat. Part of this
heat can be utilised to increase the temperature of stream 4 from 110°C to 240°C
by using the heat exchanger A. Heat duty of heat exchanger A is 6 × (240 – 110)
= 780 kW. Stream 1 is divided in two parts a and b. Stream a is used to heat
stream 4.
(ṁCp)1a = = 6.5 kW/°C

(ṁCp)1b = 10 – 6.5 = 3.5 kW/°C

 Fig. 6.53 Heat Exchanger Network for Example 6.13

Hot stream 2 contains (ṁCp) Δt = 8 × (160 – 90) = 560 kW heat. The same can
be used to increase the temperature of cold stream 3 from 60°C to 140°C, i.e., [7
× (140 – 60)] = 560 kW. For this heat exchange, heat exchanger B is used.
To heat the stream 3 from 140°C to 160°C, heat duty required is [7 × (160 –
140)] = 140 kW. Same can be furnished by hot stream 1b. This is done in heat
exchanger C. Outlet temperature of hot stream 1 from heat exchange C is to.
Then (ṁCp)b (250 – to) = 140
3.5 (250 – to) = 140
to = 210°C
Remaining heat of hot stream 1b is [3.5 × (210 – 130)] = 280 kW. If there exists
an opportunity to heat demineralised water for boiler feed make-up purpose, 280
kW heat exchange can be effected in an economiser (EC). If such an opportunity
is not available, stream 1b can be cooled to 130°C with the help of cooling water
in a cooler.
Stream 4, available at 240°C from heat exchanger, can be heated to the desired
temperature of 260°C in a heater (H) with the help of circulating thermic fluid.
In all there will be 5 heat exchangers in the network; A, B, C, H and EC, the last
two use hot and cold utilities, respectively.
Comments:
Here stream 1 is splitted into two parts. Stream splitting has one advantage and
two disadvantages.
Advantage: Stream splitting removes the use of one extra heat exchanger in
network. Here if the stream 1 is not splitted, then total number of heat
exchangers required in network will be 6 instead of 5.
Disadvantages:
(i) In stream splitting, control of the flows in two branches is required.
(ii) Splitting a stream means each branch has a lower flow rate than that for the
entire stream. In case of no phase change heat transfer, lower flow rate
means lower heat transfer coefficient and larger exchanger area.
Suggested heat exchanger network design is based on the value of ΔTmin = 10°C.
It is not the optimum design. To arrive at the optimum network design detailed
cost analysis is required for the different values of ΔTmin. The value of ΔTmin
which gives the minimum total cost is the optimum ΔTmin, and design of heat
exchanger network based on that is optimum network design. A simple computer
program can be developed for carrying out iterative calculations, varying ΔTmin
values.
Design the suitable heat exchanger network for atmospheric
crude fractionation unit based on the data given in Table 6.4422, 23.
Solution
1. Selection of pinch temperature difference: Fractions of crude petroleum oil
provides relatively lower heat transfer coefficients. Hence, the optimum value
of ΔTm is relatively higher. Select the value of ΔTm = 20°C.
Table 6.44 Table of Stream Data (Ref. 23, 24)
Stream No. Name of stream Temperature in, °C Temperature out, °C

1 Fuel oil 349 90

2 Gas oil 341 65
3 Kerosene 268 38
4 Reflux 251 77
5 Heavy Naphtha 235 38
6 Light Naphtha (Main) 168 71
7 Light Naphtha (Pre flash) 136 71
8 Desalted feed 15.6 121
9 Pre-flash feed 120 194
10 Crude tower feed 189 368

Table 6.45 ṁCp Data of Streams (Ref. 23, 24)

Stream No. Temperature range, °C ṁCp, MW/°C

1 349 to 243 0.215

243 to 213 0.197
213 to 167 0.178
167 to 90 0.168
2 341 to 210 0.105
210 to 172 0.1
172 to 111 0.087
111 to 65 0.076
3 268 to 135 0.065
135 to 38 0.054
4 251 to 169 0.105
169 to 77 0.091
5 235 to 127 0.008
127 to 38 0.007
6 168 to 136 0.6
136 to 118 0.478
118 to 108 0.41
108 to 71 0.303
7 136 to 118 0.256
118 to 108 0.21
108 to 71 0.159
8 15.6 to 121 0.379
9 120 to 122 0.4
122 to 163 0.422
 163 to 186 0.6
186 to 194 0.725
10 189 to 237 0.477
237 to 265 0.496
265 to 368 0.66

2. Table is to be prepared for drawing Grand Composite Curve.

Sample calculations for available heat in column 3 for the interval of
temperature change of hot streams from 214°C to 168°C: In this interval first
five streams exist (all hot streams from 1 to 5). Cold stream 10 also exists in this
interval. But, it cannot be considered in calculations of available heat (with hot
streams).
Available heat within this interval
= 0.197 (214 – 213) + 0.178 (213 – 168) (of Stream 1) + (0.105 (214 – 210) +
0.1 (210 – 172) + 0.087 (172 – 168)) (of Stream 2) + (0.065 (214 – 168)) (of
Stream 3) + [0.105 (214 – 169) + 0.091 (169 – 168)] (of Stream 4) + 0.008 (214
– 168) (of Stream 5)
= 20.949 MW.
In column 4, available heat indicates the cumulative or total heat available from
the hot streams at temperature t, when all hot streams (Stream No. 1 to 7) cool
from the hottest temperature 388°C to the temperature t°C.
In column 5, required heat by cold streams for each temperature interval is
calculated.
Sample calculations for the required heat in column 5 for the interval of
temperature from 148°C to 194°C of cold streams are as follows. In this interval
two cold streams; No. 9 and No. 10 exist. Required heat by cold streams in this
interval
= [(194 – 186) × 0.725 + (186 – 163) × 0.6 + (163 – 148) × 0.422] (of Stream 9)
+ [(194 – 189) × 0.477] (of Stream 10)
= 28.315 MW
In column 6, cascaded required heat by cold streams at temperature t′ indicates
the total heat required by the cold streams to increase the temperature of cold
streams from t′ to the maximum required temperatures:
Column 7 represents the net heat available for each interval.
Net heat available (column 7) = Heat available with hot streams (column 3)
– Heat required by cold streams (column 5)
Column 8 represents the cascaded or cumulative net heat available at any
temperature t. The most negative number in column 8 is –60.344 MW. It implies
minimum amount of heat that must be supplied from the hot utilities. The same
amount of heat 60.344 MW, is added in each number of 8th column. Resulting
numbers are written in the last column (column 9) as adjusted heat in MW. The
point in the last column where the adjusted cascaded heat is zero, represent the
pinch point. Topmost number in the last column, 60.344 MW is the minimum
amount of heat that must be supplied from the hot utilities and the bottommost
number of last column; 44.0974 MW is the minimum amount of heat that must
be removed by cold utilities.
3. Grand Composite Curve (GCC):
For this example, GCC is shown in Fig. 6.54. Pinch point occurs at 168°C of hot
streams temperature and 148°C of cold streams temperature.
4. Hot streams and cold streams exist above the pinch and below the pinch are
shown in Fig. 6.54.
For the maximum heat or energy recovery cold utilities cannot be used above the
pinch and hot utilities cannot be used below the pinch. Also, heat cannot be
transferred across the pinch.
Figure 6.55 also shows heat load (heat available or required) above and below
the pinch.
Sample calculation of heat load above the pinch for Stream No. 1:
Heat load above the pinch for Stream No. 1 = Σ(ṁCp)i Δti
= 0.215 (349 – 243) + 0.197 (243 – 213) + 0.178 (213 – 168)
= 36.71 MW
 Fig. 6.54 Grand Composite Curve for Example 6.14

Above the pinch design

For the maximum energy recovery cold utility cannot be used above the pinch.
But hot utility can be used above the pinch. One of the possible network designs
above the pinch is shown in Fig. 6.56. Here five hot streams and two cold steams
exist above the pinch. Heat available with Stream No. 5 above the pinch is 0.536
MW and is very low compared to heat loads (heat available or required) of other
streams existing above the pinch. Hence the heat exchange above the pinch heat
load of Stream No. 5 is not considered but the same load is transferred or
included in the design of network below the pinch. Here, Stream No. 9 (cold
stream) is splitted into four streams to exchange the heat with four hot streams.
Stream No. 10 (cold stream) is splitted into two streams to exchange the heat
with two hot streams. Streams are splitted near the pinch while heater H is
designed far away from the pinch.
Table 6.46 Cascaded Heat Flow Calculations for Example 6.14
1 2 3 4 5 6 7 8 9
Hot Cold Avai- Cas- Requi- Cas- Net Cas- Adju-
stream stream lable caded red caded heat caded sted
tempe- tempe- heat heat heat heat heat cascaded heat
rature rature
°C °C MW MW MW MW MW MW MW
(388) 368 0 0 0 60.344
0 25.74 –25.74
349 (329) 0 25.74 –25.74 34.604
1.72 5.28 –3.56
341 (321) 1.72 31.02 –29.3 31.044
23.36 45.392 –22.032
268 (248) 25.08 76.412 –51.332 9.012
6.545 8.318 –1.773
251 (231) 31.625 84.73 –53.105 7.239
7.696 7.632 0.064
235 (215) 39.321 92.362 –53.041 7.303
10.08 10.017 0.063
214 194 49.401 102.379 –52.978 7.366
20.949 28.315 –7.366
168 (148) 70.35 130.694 –60.344 0
27.523 11.372 16.151
(141) 121 97.873 142.066 –44.193 16.151
1.019 0.779 0.24
(140) 120 98.892 142.845 –43.953 16.391
4.076 1.516 2.56
136 (116) 102.968 144.361 –41.393 18.951
48.165 17.434 30.731
90 (70) 151.133 161.795 –10.662 49.682
8.97 4.927 4.043
77 (57) 160.103 166.722 –6.619 53.725
3.594 2.274 1.32
71 (51) 163.697 168.996 –5.299 55.045
0.822 2.274 –1.452
65 (45) 164.519 171.27 –6.751 53.593
1.647 10.233 –8.586
38 (18) 166.166 181.503 –15.337 45.007
 — 0.9096 –.9096
(35.6) 15.6 166.166 182.4126 –16.2466 44.0974

Fig. 6.55 Division of Hot Streams and Cold Stream around Pinch Point for Example 6.14

Below the pinch design

For the maximum energy recovery hot utility cannot be used below the pinch
(Refer to Fig. 6.57). If the stream splitting is required, then split it near the pinch.
Design the heat exchangers which utilize the utilities far away from the pinch.
Here, below the pinch, Stream No. 9 (cold stream) is again splitted at the pinch
into four streams to exchange the heat with the same four streams, from Stream
No. 1 to 4. Another cold stream, Stream No. 8, is also splitted into two streams
to exchange the heat with Stream No. 1 and Stream No. 6. At the end, Stream
No. 9 exchanges the heat with Stream No. 5. Total six coolers are designed, far
away from the pinch, at the end of network to meet the process requirements.
Complete design
Complete design of heat exchanger network for the maximum energy recovery is
obtained by merging the two network designs, above and below the pinch. To
reduce the number of heat exchangers used for Stream No. 9 above the pinch, C,
D, E and F are combined with C2, D2, E2 and F2, respectively. C2, D2, E2 and
F2 are the heat exchangers used below the pinch. For these combinations minor
adjustments are done in the temperatures of branch streams. Combined design or
completed design is shown in Fig. 6.58.
Fig. 6.56 Network Design above the Pinch for Example 6.14
Fig. 6.57 Network Design below the Pinch for Example 6.14
 Fig. 6.58 Combined Network for Example 6.14

6.19.3 Limitations of Heat Exchanger Networking

Heat exchanger networking is normally carried out on the basis of available data
on heat sources/sinks. However, in actual practice, certain overriding
considerations need attention which may exclude certain services from
networking. Following limitations need be carefully evaluated for each service
before being considered for networking:
1. Networking makes the system more complex and therefore control of the
system is more difficult.
2. Networking requires additional connecting piping and fittings. These pipings
may introduce additional pressure drops in the systems; thereby networking
of long distance heat exchangers may not be economically feasible.
3. Failure of any one heat exchanger can deteriorate the performance of entire
network.
4. Networking may call for lower ΔTm of heat exchangers which may call for
high fixed cost due to bigger size of heat exchangers.
5. ‘Odd’ heat exchanger should not be included in networking. Any heat
exchanger requiring special or costlier material of construction or that
provides very low or very high heat transfer coefficients in comparison to
other heat exchangers of network should not be included in the network.
6. Many networking of heat exchangers involve stream splitting (refer to
Example 6.14). Stream splitting of liquid-vapour mixed stream (i.e., two-
phase flow) is difficult. Other disadvantages of stream splitting are mentioned
in Example 6.13.
7. A specific heat exchanger of network may be needed during start-up or shut-
down. During this period, heating/cooling service of the heat exchanger may
not be available and therefore additional service will be required to be
provided in it. This will increase the cost of networking system.

6.20 HEAT TRANSFER IN SOLIDS

Unlike heat transfer in liquids and gases, heat transfer in solids takes place
chiefly by conduction. Cooling of solids is often required after a dryer, a
calciner, a roaster or a combustor. Metal powders, burnt lime, coal, etc., are
cooled from high temperature (in excess of 500°C). Urea, furnace grade carbon
black, etc., require cooling after drying. Pharmaceutical products (fine powder)
require cooling after drying. In food processing, coffee beans are cooled after
roasting. Beans, corn and peas are frozen by cooling with refrigerated air for
preservation.
Cooling of hygroscopic solids is required for more than one reason. Granulated
or prilled fertilizer and carbon black with some moisture may agglomerate on
natural cooling due to moisture migration in silo. This results in lump formation.
These lumps are quite hard and when broken by mechanical means produce
dusty material making it unsuitable for end use. However, caking tendency is not
observed in the same material if it is cooled (say below 60°C) in absence of
moisture.
Several technologies are available for cooling solids. Although compact plate-
type heat exchangers are developed for cooling of solids, their application is
limited due to cost and possible adherence of sticky solids to the heat exchanger
surface.
Direct contact coolers with or without indirect cooling (with cooling water) are
popular for cooling of solids. Adham25 has described various solids cooling
technologies at length. Heat recovery from hot gases, coming out of the cooler is
possible depending on temperature of gases and its possible use elsewhere in the
plant. Chiefly, four types of coolers are in use for most applications of solids
cooling.
1. Stationary cooler
2. Rotary cooler
3. Fluid bed cooler
4. Conveyor cooler
6.20.1 Stationary Coolers
Simple version of a stationary cooler (Fig. 6.59) has a bin with internal cooling
surface. It is recommended for free flowing material and is unsuitable for wet,
very fine and sticky powders. Overall heat transfer rate in the cooler is quite low
and hence, larger cooling surface is required.
 Fig. 6.59 Stationary Cooler with Water Cooled Plates

Solids travel in a near plug flow manner and a temperature gradient from top to
bottom is observed. Stationary coolers are recommended for moderate inlet
temperature (say 300 to 400°C) and for heat duties not exceeding 150 kW. Free
flowing powder up to 100 μm size can be cooled in such a cooler. Barring these
limitations, stationary coolers are compact, require minimum or no auxiliary
equipment, require no off-gas treatment and are simple to operate. Cooling up to
60 to 70°C should be possible with this cooler with circulating water maintained
at a temperature of 30 to 35°C.
6.20.2 Rotary Coolers
This versatile cooler can handle different type of solids and can be used as direct
or an indirect cooler with or without heat recovery. Although it can be used for a
wide range of solids, it is a preferred equipment for coarse particles. In this types
of coolers, hot air (or gas) flows in counter current direction of solids, increasing
the thermal efficiency. Cooling coils can be embedded in the bed to provide
indirect cooling. Heat from the hot air, coming out from the cooler, can be
recovered in a number of ways.
Figure 6.60 is a simple sketch of a rotary cooler. Diameter of the cooler is
restricted below 3 m for the considerations of fabrication and transportation.
Drive mechanism for the cooler also imposes restrictions on the dimensions.
Length-to-diameter ratio is kept in the range of 4 to 10. To avoid passing of
solids without coming in contact with air, depth of material is kept in the range
of 30–40% of radius. Such equipment can be used for drying as well as cooling.
In carbon black industry, ceramic industry, metal powder industry, etc., this
equipment is popular.

Fig. 6.60 Rotary Cooler with Direct Air and Spray Water Cooling

6.20.3 Fluid Bed Cooler

Coarse small particles (below 2–2.5 mm) can be fluidized with air or gas.
However, very fine powders (less than 10 μm) resists smooth fluidization. So
fluidized bed cooler can be considered for the intermediate size solids. In Fig.
6.61, four-stage fluid bed cooler with embedded cooling coils is shown.
Spray cooling is an option both in rotary as well as fluid bed coolers but it has to
be executed with caution. Hot air wet bulb temperature should not be allowed to
exceed 80°C and fluid or moving bed temperature is not allowed to fall below
150°C to prevent moisture absorption by the solids.
Fluid bed cooler can handle free flowing and non-sticky solids. Heat transfer rate
is excellent in a fluid bed cooler among all solid-cooling equipments. However,
it needs an auxiliary equipment such as a cyclone or a bag filter for off-gas
treatment to trap fines. This equipment is relatively simple in operation and easy
to maintain.
 Fig. 6.61 Four Stage Fluidized Bed Cooler with Indirect Water Cooling

Removal of fines from the solids may be considered as an advantage. The

product after cooling is free from dust and its particle size distribution is in the
narrow range making it a preferred marketable product.
6.20.4 Conveyor Coolers
Variety of conveyor coolers are available. Vertical spiral cooler, moving grate
cooler, vibrating pan cooler, screw conveyor coolers, etc., are a few of them. The
screw conveyor cooler is normally used for indirect cooling of solids with
cooling water which flows in the conveyor shell jacket and the hollow shaft.
Screw conveyor coolers can be used for free flowing solids, normally coarse in
nature (0.1–12 mm). Screw conveyor coolers can be advantageously used for
cooling as well as conveying. Single or multiple screw designs are available but
double screw design is popular up to a maximum length of 8.5 m.
Process design of cooling equipments of solids can be well explained by the
following example.
Urea prills from a prill tower in a fertilizer plant have a
temperature of 100°C and moisture content of 1.5 to 2% (by mass). They need to
be cooled to 60°C before sending to silo for storage to prevent caking and lump
formation. Average sieve analysis of prills is given in Table 6.47.
Table 6.47 Typical Distribution of Urea Prills by Size
Size of prills, mm mass %

+2.4 0.8
+2.0 9.0
+1.7 56.3
+1.4 26.6
+1.0 7.3

Physical properties of urea:

Particle density, ρs = 1.335 kg/L
Bulk density, ρsb = 0.758 kg/L (without compaction)
Thermal conductivity, ks = 80 W/(m · °C)
Heat capacity, Cs = 1.775 kJ/(kg · °C)
I. Cooling Mediums
(a) Circulating condensate is proposed to be used in the stationary cooler to
avoid possibility of fouling in the plates. Supply and return temperatures may
be taken as 35 and 45°C, respectively. Hot condensate is passed through an
external shell and tube heat exchanger in which it is cooled to 35°C with the
help of cooling water of 30°C which is recirculated.
(b) Ambient air at 44°C is to be used in rotary and fluid bed coolers. Outlet air
temperature from the coolers may be fixed at 75°C.
Work out process design of stationary, rotary and fluid bed coolers to process 60
t/h urea prills. While there is hardly any change in moisture content in the
stationary cooler, moisture content of prills gets reduced to 0.5 mass % in
rotary and fluid bed coolers.
Overall heat balance:
Average temperature of air =

= 59.5°C
≡ 332.65 K
Molar volume of air =

= 27.296 m3/kmol
Density of air, ρg = = 1.061 kg/m3

Moisture removal of urea = 60 000 (0.015 – 0.005) = 600 kg/h

Total heat load of prill cooling in rotary and fluid bed coolers,
ϕ1 = 60 000 × 1.775 (100 – 60) + 600 × 2644
= 4260 000 + 1586 400
= 5846 400 kJ/h
≡ 1624.0 kW ≡ 1624.0 × 103 W
Average heat capacity of air at 59.5°C,
Cpa = 1.005 kJ/(kg · °C)
Air is heated from 44°C to 75°C.
Mass flow rate of air, qma =

= 187 655 kg/h

Volumetric flow rate of air, qva = 187 655/1.061
= 176 866 m3/h
Viscosity of air at 59.5°C, μa = 0.072 kg/(m · h)
In case of stationary cooler, heat lead,
ϕ2 = 4260 000 kJ/h (no evapouration load)
≡ 1183.33 kW
Flow rate of circulating condensate,
qmc = = 101 748 kg/h

≡ 101.7 m3/h
Process design of stationary (bulk flow) cooler:
Refer to Fig. 6.59.
Assume 2b = 5 cm
b = 2.5 cm ≡ 0.025 m
In a stationary cooler, solids travel in plug flow and, therefore, a temperature
gradient in bulk solids is observed. Following equation25 relating to the
unsteady-state heat transfer from a solid, can be used to calculate cooling time
of the solids.

  (6.137)
where, t1 = Inlet solid temperature, °C
t2 = Outlet solid temperature °C
t′1 = Inlet circulating condensate temperature, °C
t′2 = Outlet circulating condensate temperature, °C
α = Thermal diffusivity in solid, m2/h
= ks/(ρs · Cs)  (6.138)
ks = Thermal conductivity of solid, kJ/(h · m · °C)
Cs = 1.775 kJ/(kg · °C) for urea
θ = cooling time, h
Above equation assumes that convective transfer coefficient of cooling water
and thermal conductivity of the plate material (SS in this case) is much larger
than ks. This assumption may be considered valid for urea prills.
ks = 80 W/(m · °C)
= 288 kJ/(h · m · °C)
α =

= 0.121 54 m2/h
LHS of Eq. (6.137) =

= 0.272 73
Spreadsheet or Mathcad can be useful in solving Eq. (6.134). Substituting values
of b and α in RHS of the equation and solving by trial and error,
θ = 0.0023 h = 8.28 s
Hold-up required for solids cooling,
= = 182.06 m3

Single silo or bin of such large operating volume with cooling plates would not
be desirable for operational and maintenance reasons. Cooling height of at
least 2 m is recommended.
Overall heat transfer coefficient of 175 W/(m2 · °C) will be safe to design the
cooling surface.
ΔT1 = 100 – 45 = 55°C
ΔT2 = 60 – 45 = 15°C
ΔTlm = = 30.79°C

Heating surface requirement, A1 =

= 219.61 m2 (min.)
Four or more silos may be provided. Distance between plates for cooling water
flow should be kept at least 2 cm. Silos and plate coolers with HTA,
exceeding 220 m2, can be easily sized for the required heat duty.
For four silos,
Hold-up volume of solids = 182.06/4 = 45.515 m3 per silo
Assume cooling (plate) height = 3 m
Cross-sectional area of silo = 45.515/3 = 15.172 m2 (each)
Consider square construction of the silo.
Each side = = 3.895 say 3.9 m
Plate length = 3.5 m
Width between two plates = 5 cm
No. of pairs of plates =

= 35
HTA per pair of plates = 3.5 × 3 × 2
= 21 m2
HTA per silo = 21 × 35 = 735 m2
Total HTA = 735 × 4 = 1960 m2
= = 13.38 (much in excess)

Note: In order to avoid adherence of urea dust, present in the feed of silo, a
small quantity of dry air is injected from bottom. The exhaust air from the
cooler can be directed to prill tower bottom.
II. Process Design of Rotary Cooler
Assume diameter of rotary cooler as 2.75 m. Air is passed in countercurrent
(horizontal) fashion to solids flow.
Cross-sectional area of cooler, A2 = (π/4) (2.75)2
= 5.939 57 m2
Velocity of air = = 8.272 m/s

This is a very high velocity and will blow dust and fine area prills. Normally in a
rotary cooler, velocity up to 1.5 m/s is preferred. This will call for six rotary
coolers to be used in parallel.
Flow of urea prills/cooler =

= 10 000 kg/h ≡ 10 t/h per silo cooler

Flow of air through each cooler,
qv air =

= 29 477.7 m3/h at inlet

Velocity of air at inlet =

= 1.379 m/s (< 1.5 m/s)

Rotary cooler length can be calculated by the following equation:
L = Nt Ht   (6.139)
where, L = Rotary cooler length, m
Nt = Number of transfer units
Ht = Height (length) of transfer unit, m
Nt =   (6.140)

where, ΔtG = tg, out – tg, in, °C

ΔTlm =

= = 20.17°C

Nt =
Preferred range of Nt is 1.5 to 2.5.

Ht =   (6.141)

where, G = Mass flow rate per unit area of air, kg/(m2 · s)

Ca = Heat capacity of air = 1.005 kJ/(kg · s)
Uv = Overall volumetric heat transfer coefficient, kJ/(m3 · s · °C)
= 237 G0.67/D   (6.142)
D = Rotary cooler diameter, m
G =

= 1.462 69 kg/(m2 · s)
Uv = 237 (1.462 69)0.67/2.75
= 111.2 kJ/(m3 · s · °C)
Note that units of Uv are volumetric and different from normal units.

HT =

= 13.2
L = 1.537 × 13.2 = 20.29 m or 20.3 m
L/D = 20.3/2.75 = 7.38
Rotary coolers, commonly in use, have L/D ratio in the range of 4 to 10. Bed
depth of about a meter is recommended in the cooler. Above design is for
cooling with air only. If cooling surface is provided across the length for indirect
cooling with cooling water, heat transfer rate can be estimated by considering
overall heat transfer coefficient of 55–110 W/(m2 · °C).
III. Process Design of Fluid Bed Cooler
For determining the fluidizing velocity, mean particle diameter (dp) is required.
Mean particle diameter,

dp =   (6.143)

where, dp = Mean particle diameter and

wi = Mass fraction of particle diameter, di
 =

= 0.6425
dp = 1/0.6425 = 1.556 mm
For estimating fluidization velocity, use of following equation25 is
recommended:

Rem =   (6.144)

where, Rem = Reynolds number at minimum fluidization velocity

dp = 1.556 mm = 0.001 556 m
ρg = 1.061 kg/m3
ρs = 1.335 kg/L = 1335 kg/m3
g = gravitational acceleration
= 9.81 m/s2
μg = 0.072 kg/(m · h)

Rem =

= (1135.65 + 5044.89)1/2 – 33.7

= 44.92
Minimum fluidization velocity is related with Reynolds number as 44.92:

um =   (6.145)

= 0.54 m/s at ambient temperature

At the minimum velocity, fluidization starts. Terminal velocity (ut) is the
velocity above which the solids particles will be pneumatically conveyed.
Reynolds number (Ret) at the terminal velocity is derived by

CdRet 2 = 1.6 × 107 (Ref. 26)  (6.146)

 where, Cd = Drug coefficient which depends on sphericity (ϕs) of the particles
ϕs = 0.6 for broken solids and the same can be taken for urea prills
Cd = 6 for ϕs = 0.6

Ret 2 = = 26 906

Ret = 164

ut =   (6.147)

= = 1.987 m/s

Actual fluidization velocity will be between um and ut. Select u = 1 m/s for
normal fluidization.
Volumetric flow rate of air, qva = 176 866 m3/h
Area of fluidization, A3 =

= 49.129 m2
Selected width, W = 3 m
Length, L = 49.129/3 = 16.376 m
To improve thermal efficiency, total length will be divided into a number of
stages. Select 4 stages for fluidization.
Length of each stage = = 4.094 m, say 4.1 m

Staging will be beneficial for crossflow of solids. Bed depth of 0.4 to 0.5 m is
adequate in each stage. Assuming bed depth of 0.5 m,
Hold-up of urea in fluid bed cooler = 49.129 × 0.5 × 758
= 18 620 kg
This means residence time (θ) of urea prills in the fluids bed cooler will be 18.62
min; or approximately 20 min.
Peak pressure drop in the bed26,
=   (6.148)

where, Δ pp = Peak pressure drop, kN/m2 or kPa

H = Height of bed = 0.5 m
ρb = Bulk density = 0.758 kg/L
g = 9.81 m/s2
V = Angle of internal friction.
For granular solids, consider it to be 55°.
Di = Fluid inlet orifice diameter
= 2 cm = 0.02 m (assumed)
Dc = diameter of equivalent cross section

= = 7.909 m

= 0.812 – 0.107 = 0.705

Δpp = 2.62 kPa ≡ 267 mm WC
Note: In normal course, pressure drop will vary between 30 and 50% of peak
pressure drop (i.e., 80–135 mm WC) across the bed. To this add friction loss in
the ducts which will give the fan pressure requirement.
During fluidization, broken and prills with poor mechanical strength will break
and form dust due to attrition. This is expected to give higher dust load than the
other two types of coolers. Dust will be separated from exhaust air and will be
allowed to fall in a dust dissolving tank for reprocessing. Urea prills, coming
out from the fluid bed coolers will be of near uniform size and with definite
mechanical strength which will not break during storage in silo or during
transportation.
It is recommended to test the desired product in a pilot batch fluidizer. The
fluidization velocity, orifice size, bed depth, pressure drop across the bed and
other design parameters can be verified in the pilot unit. Process design, based
on the verified parameters, will give an optimum design.

EXERCISES
6.1 Select the correct answer.
(a) For an exchanger with no phase change service design conditions are: Hot
fluid is cooled in counter current manner from 213°C to 57°C with a cold
fluid which is heated from 29.5°C to 40.5°C. Other parameters being
 maintained the same, if hot fluid inlet temperature is reduced to 201°C and
cold fluid inlet temperature is increased to 35°C, then
(i) Hot fluid and cold fluid outlet temperatures will change to 60°C and 45°C,
respectively.
(ii) Hot fluid outlet temperature will change to 60°C while cold fluid outlet
temperature will remain as 40.5°C.
(iii) Hot fluid outlet temperature will remain 57°C while cold fluid outlet
temperature will change to 45°C or
(iv) Data not sufficient to predict.
(b) In the heat transfer equation for heat exchanger, the heat transfer rate appears
to be directly proportional to the heat transfer area. Hence (other
parameters remaining the same) for double the heat transfer area, the heat
transfer rate will be
(i) doubled
(ii) more than doubled or
(iii) less than doubled.
(c) A shell and tube, horizontally installed heat exchanger with 2-tube passes,
has pressure gages (P1 and P2) installed, respectively on the inlet pipe to the
lower nozzle and on the outer pipe from the upper nozzle of the channel.
The pressure gages are approximately 1.4 m apart. The difference in the
readings (p1 – p2) is 0.42 bar. Each nozzle accounts for 0.07 bar pressure
drop. Estimate the difference in pressure reading (as p1 – p2) for the
following cases:
(i) Tube side passes are changed from 2 to 4 and
(ii) Tube side flow rate is doubled.
[Answer: (a) 1.42 bar, (b) 0.85 bar or (c) 0.42 bar]
6.2 A fired tube waste-heat boiler generate steam at a pressure of 10.34 bar g. It
is designed for a gas flow of 80 000 kg/h from an incinerator. Gas is designed
to be cooled from 600°C to 260°C. On a particular day, the gas flow is 55 000
kg/h and inlet temperature to the boiler is 500°C and steam is generated at
10.34 bar g. What will be the expected gas outlet temperature if the overall
heat transfer coefficient is proportional to mass flow rate of gas to the power
of 0.8? Assume other properties of flue gases to be nearly the same.
Assume overall thermal efficiency of the boiler to be 95% in both the cases.
Average heat capacity of flue gases may be taken as 1 kJ/(kg · K).
Demineralized water enters steam drum at 35oC. Calculate steam generation
from the boiler in both the cases.
Data: Saturation temperature of steam at 10.34 bar g = 185.5°C
Enthalpy of saturated steam at 10.34 bar g = 2780.8 kJ/kg
Enthalpy of water at 35°C = 146.6 kJ/kg
Reference state: Enthalpy of saturated liquid water at 0.01°C = 0 kJ/kg
6.3 120 000 kg/h of ethanol is to be cooled from 78°C to 40°C by cooling water
as coolant. Cooling water enters at 32°C and leaves at 40°C. Design shell and
tube heat exchangers for the different tube lengths 1.5 m, 3 m, 4.5 m and 6 m.
6.4 3000 kg/h of nearly pure saturated phosphorous oxychloride (POCl3) vapour
at 15 kPa g is to be condensed and cooled to 60°C by cooling water which is
available in plant at 32°C. Design the suitable shell and tube heat exchanger
for the given duty using following data:
Tube material: Graphite
Tube OD : 25 mm
Tube ID : 20 mm
Thermal conductivity of graphite: 150 W/(m · K)
Molar mass of POCl3: 153.5 kg/kmol
Specific heat of liquid POCl3: 0.92 kJ/(kg · °C)
Latent heat of vaporization of POCl3: 229 kJ/kg
Thermal conductivity of liquid POCl3: 0.173 W/(m · °C)
Viscosity of liquid POCl3 = 1.1 cP or mPa · s
Liquid density of POCl3 = 1675 kg/m3
Condensation temperature of POCl3 at 15 kPa g = 110°C
Viscosity of POCl3 vapour, μv = 1000 × 10–7 Poise
6.5 Determine shell side heat transfer coefficient and pressure drop by Kern’s
method and Delaware method for following heat exchange duty and compare
results:
(i) Type of HE: AET as per TEMA (pull through floating head heat exchanger)
(ii) Shell side fluid: Oil Flow rate = 10 000 kg/h
(iii) Inlet/Outlet temperatures: 110/50 °C
(iv) Properties of oil at average temperature
CL = 2.16 kJ/(kg · °C)
μL = 8 cP or mPa · s
kL = 0.14 W/(m ·°C)
ρL = 850 kg/m3
(v) Tube data:
Number of tubes, Nt = 918
Tube OD/ID = 19.05/15.748 mm
Tube length = 3 m
MOC of tube: Carbon steel
Type of tube layout: Triangular
Tube pitch, Pt = 25.4 mm
Tube bundle diameter, Db = 846 mm
Shell ID = 926 mm
No. of tube passes = 8
Pass partition plate thickness = 12.7 mm
No. of vertical pass partition plate = 1 (Hence, 2 tube passes)
(vi) Type of baffle: 25% cut segmental
Baffle spacing = 200 mm (central)
Entrance and exit baffle spacing = 300 mm
Clearance between shell ID and baffle diameter = 4.8 mm (diametric)
Clearance between tube hole in baffle and tube OD = 0.8 mm (diametric)
6.6 Determine the shell side heat transfer coefficients and shell side pressure
drops by Kern’s method and Delaware method for following heat exchange
duty and compare results.
(i) Type of HE: AET as per TEMA (pull through floating head heat exchanger)
(ii) Shell side fluid: Liquid methanol Flow rate: 100 000 kg/h
(iii) Inlet/Outlet temperatures: 95°C/40°C
(iv) Properties of methanol at average temperature:
CL = 2.84 kJ/(kg · °C)
μL = 0.34 cP or mPa · s
kL = 0.19 W/(m · K)
ρL = 750 kg/m3
(v) Tube data:
Number of tubes, Nt = 918
Tube OD/ID = 20/16 mm
Tube length = 4880 mm
MOC of tube: Cupro-nickel
Type of tube layout: Triangular
Tube pitch, Pt = 25 mm
Tube bundle diameter, Db = 826 mm
Shell ID = 894 mm
No. of tube passes = 2
Pass partition plate thickness = ?
(v) Type of baffle: 25: cut segmental (horizontal)
Baffle spacing, Bs = 356 mm
Clearance between shall ID and baffle diameter = 4.8 mm (diametric)
Clearance between tube hole in baffle and tube OD = 0.8 mm (diametric)
6.7 Design vertical thermosyphon reboiler for acetic acid dehydration column
based on following data:
(a) Feed is liquid at 30°C.
Feed flow rate = 12 000 kg/h
Feed composition = 80% acetic acid and 20% water (by mass)
(b) Distillate is saturated liquid at 100.6°C
Flow rate of distillate = 3000 kg/h
Distillate composition = 20% acetic acid and 80% water (by mass)
(c) Residue is nearly pure saturated acetic acid at 138°C and at 179.5 kPa a.
(d) Reflux ratio, R = 4.2
(e) Latent heat of vaporization
λAA = 25 732 kJ/kmol at 293.8 K
λW = 40 677 kJ/kmol at 373.15 K
(f) Crtical temperature of acetic acid = 592.71 K
Crtical temperature of water = 647.11 K
(g) Heat capacity equation for liquids
Cml = a + bT + cT2 + dT3
where, Cml = Liquid heat capacity, kJ/(kmol · K)
T = Absolute temperature, K
Table 6.48 Heat Capacity Data or Liquids10
ai bi × 103 ci × 106 di × 109

Acetic acid –36.0814 604.681 –393.957 –561.602

Water 18.2964 472.118 –1338.78 1314.24

6.8 50 000 kg/h cold air is required at 18°C temperature. Chilled water with 6°C
supply temperature and 8°C return temperature is to be used as cooling
medium. Design a suitable finned tube heat exchanger with box type shell.
Hint: Atmospheric air contains water vapour. Hence, condensation load of
water vapour must be considered in the calculation of heat duty. Air saturated
with water vapour at 40°C can be considered as design condition.
6.9 Design a kettle-type reboiler which must provide 2000 kg/h of ethylene
glycol vapour to a distillation column at 30 kPa a pressure at bottom.
6.10 A chilling plant is designed for achieving 703 kW (≈ 200 TR) heat
exchange capacity to supply aqueous ethylene glycol brine at –25°C by
evaporating liquid ammonia in the evaporator at –33.3°C (and 1 atm).
Ammonia vapour at –28°C leaves the evaporator and sent to a compressor.
(a) In one alternative, a shell and tube-type condenser is considered. Superheated
vapour at 120°C and 16.435 bar a from the compressor enters the shell of
the condenser and condenses at 42°C. Cooling water enters the tubes at
32°C and leaves at 38°C. For this duty design a BEM type horizontal
condenser. Consider tube OD and length as 19.05 mm (16 BWG) and
2.4383 m (8 ft), respectively. Triangular (Δ) pitch equals 1.25 do.
(b) In another alternative, an evaporative condenser is considered for condensing
ammonia. As shown in Fig. 6.62 evaporative condenser is basically a
rectangular tube bundle installed in a cooling tower.
Cooling water falls over the tubes at a required rate at essentially a constant
temperature; 4.5oC above the design wet bulb temperature of air. Ambient air
at 40°C dry bulb (DB) temperature and 28°C wet bulb (WB) temperature is
supplied from sides of the tower. An induced draft fan sucks the air from the
louvers and discharges it from top with 95% relative humidity (RH).
Ammonia gas enters at 120°C the tubes and condenses at 1.5°C higher than
the cooling water temperature. Design the tube bundle to be placed in the
evaporative condenser. For the evaporation of water film over tube bundle,
heat transfer coefficient can be determined by Kallam’s equation.
 Fig. 6.62 Evaporative condenser

ho = 7.3 × 10–9 × N0.05 [(G.0.3do YZ)/(μ PV)]4.4 (Ref. 5)  (6.149)

where ho = water side film coefficient, W/(m2 · °C)
G = ṁc/(L · B)  (6.150)
ṁc = circulation rate of cooling water, kg/s
L = Length of tube = 2.4384 m
B = Width of tube bundle, m
μ = Viscosity of water, kg/(m · s)
PV = Vertical tube pitch = 0.030 48 m
do = OD of tube = 19.05 mm
N = Number of horizontal tube rows
Y = (PH – do)/do  (6.151)
Z = (PH – do)/PH  (6.152)
PH = Horizontal tube pitch = 0.0381 m
Here, evaporation temperature of water film can be taken as 34°C.
Evaporation rate of water can be taken as 2% of circulation rate.
(c) In a water scarce area, an air cooler is considered for condensing ammonia.
Ambient air at 45°C is supplied to the condenser while ammonia is
 condensed at 60°C in the tubes. Compressed ammonia gas enters the
condenser at 130°C. A pressure drop of 15 kPa is permitted in the
condenser. Condenser will have two-pass arrangement in finned tubes and
leaves from top at 50°C. Each horizontal tube row can be considered as a
pass.
Effective tube details: SS 304 with OD = 19.05 mm and triangular (Δ) pitch of
43 mm
Fin details: Type of fin: Spiral wound transverse fins, SS 304 MOC, Fin OD =
38 mm, 8 fins per linear inch and fin thickness = 0.889 mm
Tube length = 3 m
Width of air cooler = 3 m
Design an air cooler for ammonia condensation duty.
Properties of ammonia are listed in Table 6.49.
Table 6.49 Properties of Ammonia**
Gas/Vapour
Temp. Pressure Enthalpy*, kJ/kg
o State Cp k μ
C bar a
Liquid Vapour kJ/(kg · K) mW/(m · K) μPa · s

–33.3 1.013 25 Satd, liq. 48.8 1418.3 2.2969 20.968 8.054

–28 1.013 25 Sup. vap. – 1428.1 2.2707 21.171 8.197
34 13.124 Satd, liq. 361.2 1488.0 3.3488 27.433 10.126
42 16.435 Satd, liq. 400.5 1490.4 3.5683 28.713 10.393
60 26.156 Satd, liq. 492.0 1489.3 4.2084 32.262 11.049
120 13.25 Sup. vap. – 1723.7 2.5166 37.461 13.641
130 26.30 Sup. vap. – 1714.1 2.7897 40.862 14.030

* Reference state: Enthalpy of saturated liquid at 0oC = 200 kJ/kg

** Ref.: REFPROP Ver. 8.0, NIST Standard Reference Data Base 23, 2007.
6.11 Synthesize the heat exchanger network for the following five process
streams such that resulting heat exchanger network will require the minimum
hot and cold utilities. Also find the values of minimum utilities (hot and cold)
required.
Table 6.50 Process Stream Data27
Stream Type ṁCp, kW/°C tin, °C tout, °C

1 Hot 228.5 159 77

2 Hot 20.4 267 88
3 Hot 53.8 343 90
4 Cold 93.3 26 127
5 Cold 196.1 118 265
 REFERENCES
1. TEMA, 8th Ed., Standard of the Tubular Exchanger Manufacturers
Association, New York, USA, 1988.
2. Perry, R. H., and D. Green, Perry’s Chemical Engineers’ Handbook, 6th Ed.,
McGraw-Hill, USA, 1984.
3. Sinnott, R. K., Coulson and Richardson’s Chemical Engineering, Vol. 6,
Revised 2nd Ed., Asian Publishers Books Pvt. Ltd., New Delhi, 1998.
4. Pfaudler Heat Exchanger Databook, Pfaudler International GmbH, Basel,
Switzerland.
5. Kern, D. Q., Process Heat Transfer, McGraw-Hill, USA, 1950.
6. Wales, R.E., Chem. Engg., 86(19), 1979, p. 122.
7. Ludwig, E. E, Applied Process Design for Chemical and Petrochemical
Plants, Vol. 3, 3rd Ed. Gulf Publishing, USA, 2001.
8. Smith, R.A., Vaporisers: Selection, Design and Operation, Longmans, UK,
1986.
9. Dadyburjor, D. B., Chem. Engg. Progr., 74(4), 1978, p. 86.
10. Bhatt, B.I. and S. M. Vora, Stoichiometry, 4th Ed., McGraw-Hill Education
(India), New Delhi, 2004.
11. Tinker, T., Trans. ASME, 80, 1958, p. 36–52.
12. Hughmark, G. A., Chem. Eng. Progr. 60(7), 1964, p. 59.
13. Fair, J. R., Chem. Engg., 70(14), July, 1963, p. 119.
14. Frank, O. and Prickett, R. D., Chem. Engg., 80(20), September 3, 1973, p.
107.
15. Bell, K. J., Final Report of the Cooperative Research Program on Shell and
Tube Heat Exchangers, Bulletin No. 5, University of Delaware, USA, 1963.
16. Private Communication with Regents Professor Kenneth J. Bell and Updated
Report on Delaware Method by Prof. Bell (2012).
17. Buonopane, R. A., Troupe, R. A. and J. C. Morgan, Chem. Engg. Progr.,
59(7), 1963, p. 57 to 61.
18. Bhatia, M. V. and P. N. Cheremisinoff, Process Equipment Series, Heat
Transfer Equipment, Vol. 2, Technomic Publishing Company, Inc., USA,
1980.
19. Minton, P. E., Chem. Engg., 77(10) May 4, 1970, p. 103.
20. Wadekar, V. V., Chem. Eng. Progr, 96(12), 2000, p. 39 to 49.
21. Linnhoff, B., Townsend, D. W. and G. W. Hewitt, User Guide on Process
Integration for the Efficient Use of Energy, Rev. Ed., Institution of Chemical
 Engineers, London, UK, 1994.
22. Biegler, L. T., I. E. Grussmann and A. W. Westerberg, Systematic Methods
of Chemical Process Design, 1st Edition, Prentice Hall PTR, USA, 1999.
23. Bagajewicz, M. and J. Shuncheng, Ind. Eng. Chem. Res., 40(2), 2001, p.
617–626.
24. Bagajewicz, M. and J. Soto, Ind. Eng. Chem. Res., 40(2), 2001, p. 627–634.
25. Adham, K., Chem. Engg. Progr., 96(8), 2000, p. 37.
26. McKetta J. J. and W. A. Cunningham, (Ed.), Encyclopedia of Chemical
Processing and Design, Vol. 23, Marcel Dekker, Inc., USA, 1985, p. 51 and
76 to 79.
27. Linnhoff, B. and T. N. Tjoe, Chem. Engg., 93(8), April 28, 1986. p. 47 to 60.
28. Chopey, N. P., Handbook of Chemical Engineering Calculations, 2nd Ed.,
McGraw-Hill, Inc., USA, 1994, Ch. 7.
29. McAdams, W. H., Heat Transmission, 2nd Ed., McGraw-Hill, Inc., USA
1942.
 7
Process Design of Liquid-Liquid
Extractors
7.1 INTRODUCTION
Liquid-liquid extraction is the separation of the components of a liquid solution
by contacting it with another insoluble liquid solvent. One of the components or
certain group of the components of the solution (feed) is selectively solubalised
in the solvent. It is called solute. After the contact, components of liquid feed are
distributed between two immiscible liquids; called extract phase and raffinate
phase. Solvent rich phase is called extract phase and solvent lean phase is called
raffinate phase. A component which is left behind in the raffinate phase after
extraction or non-solute part of feed is called diluent.
Solvent extraction finds many applications in industry. Take an example of
separation of solution containing 10% by mass acetic acid and 90% by mass
water. Separation of these two is difficult by ordinary distillation as this mixture
provides very low relative volatility. In liquid-liquid extraction, acetic acid-water
solution is contacted with a solvent such as ethyl acetate. The solvent makes the
immiscible or partially miscible liquid mixture with water. Acetic acid is
selectively extracted in ethyl acetate. At the end of single stage extraction two
immiscible liquid layers are obtained. Ethyl acetate rich layer is called extract
phase and water rich layer is called raffinate phase.
In another application, n-hexane is used as a solvent to extract sterol (a
nutracentical) from deodorised distillate, derived from soybean oil. Extract phase
is rich in sterol, leaving behind a raffinate, containing mixed tocopherols
(antioxidants), squalline, etc.
Liquid-liquid extraction is an important unit operation and finds applications in
refinery, pharmaceutical industries, metallurgical industries and other industries.
In refinery, after distillation, liquid-liquid extraction is the most widely used unit
operation.
Invariably extraction will have to be followed by another unit operation, such as
distillation or crystallisation for separation of solute from the extract phase.

7.2 IMPORTANT INDUSTRIAL APPLICATIONS

OF LIQUID-LIQUID EXTRACTION1, 2
1. Separation of 10% to 30% acetic acid in aqueous solution by using ethyl
acetate or ethyl ether as solvent.
2. Recovery of uranium from ore rich liquor. Solvent for this case is alkyl
phosphate in kerosene.
3. Separation of long chain fatty acids from vegetable oil by using propane as
solvent.
4. Propane deasphalting. Separation of lube oil from residuum (residue of
vacuum distillation column of refinery) by using propane as solvent.
5. Separation of tantalum and niobium. It involves dissolution of this mixutre in
hydrofluoric acid, followed by extraction with Methyl Isobutyl Ketone
(MIBK). Similarly, liquid-liquid extraction is used for the metal separations
such as uranium–vanadium, hafnium–zirconium and tungsten–molybdenum
mixtures.
6. Phosphoric acid and boric acid are finally purified by liquid-liquid extraction.
7. Liquid-liquid extraction is used for the separation of the mixture of aromatic
and paraffinic hydrocarbons of nearly the same molar mass and having close
boiling points (e.g., a mixture of benzene and haxane). For this case, solvents
used are liquid sulphur dioxide, diethylene glycol or sulpholane. Extractive
distillation is also used for the same application.
8. Many pharmaceutical products are produced in the form of complex mixture.
From this mixture, pure product is separated by liquid-liquid extraction; e.g.,
penicillin, resorcinol, etc.
9. In the petroleum refinery the separation of pure benzene and toluene from
aromatic rich feed stock is carried out by extraction. In this case, sulpholane
is used as solvent.
10. For dearomatisation of raw hexane fraction (or for the production of food
grade hexane) using N–methyl pyrrolidone (NMP) as solvent.
11. For dearomatisation of the straight run kerosene fraction (or for the
production of superior kerosene) using sulpholane as solvent.
12. For the purification of lube oil using furfural or NMP as solvent.
13. In an aqueous effluent containing very high COD, organic solvent (such as
toluene or xylene) is effectively used to extract out COD causing components.
14. In the manufacturing process of MTBE, methanol is separated from the
mixture of methanol and C4 stream by using water as solvent.
15. For extraction of unsaturated fatty acids from a mixture of fatty acids using
 fulphural as solvent.

7.3 THEORY OF LIQUID-LIQUID EXTRACTION

OPERATIONS
7.3.1 Continuous and Dispersed Phases
In extraction operation, two immiscible liquid phases are brought in contact with
each other. In Section 7.1, differentiation of extract and raffinate phases was
clarified. But before these two phases are separated from an extractor, two
immiscible liquids are brought in contact which are known as continuous and
dispersed phases. Dispersed phase is a liquid which forms the droplets and are
dispersed in a continuum of the other; known as a continuous phase. Ordinarily,
the liquid with smaller flow rate will be dispersed in another. For a batch
operation, the liquid in which impeller is immersed when at rest is usually a
continuous phase. Normally, dispersed phase has higher viscosity and lower
interfacial tension. In continuous extraction, interchanging of dispersed phase
and continuous phase is possible. Also inflammable liquid should preferably be a
dispersed phase. A few industrial examples illustrate both phases.
1. Refer to Table 8.78. A heterogeneous azeotrope of n-butanol and water when
condenses, two phases are separated. Bottom aqueous layer contains 6 to
7.5% (by mass) n-butanol, depending on temperature, which should be
recovered for economical reasons. Toluene or xylene is used as solvent for the
extraction purpose. In this case, aqueous layer; i.e., feed (dense phase) is a
continuous phase while organic solvent (light phase) is a dispersed phase.
2. Acetic acid is extracted from acetic acid-water (feed) mixture (dense phase)
with the help of ethyl acetate as solvent (light phase). In this case, heavy
phase is dispersed in light phase.
3. Feed mixture of dimethyl formamide (DMF)-water (light phase) is contacted
with methylene chloride (dense phase) for extraction of DMF in which light
phase is a dispersed phase.
4. Feed mixture of benzene-n-hexane (light phase) is contacted with liquid
sulphur dioxide (heavy phase) for extraction of n-hexane. In this operation,
light phase is a dispersed phase.
7.3.2 Phase Equilibrium
Understanding of a phase diagram is necessary before attempting design of
liquid-liquid extraction system. Brief discussions in this section will be limited
to a binary feed and a solvent. Thus the system comprises three components.
Ternary phase diagram acts as a map for liquid extraction applications. Ternary
phase diagrams are unique in that they show all three components of the system
on one plot. The following general principles govern ternary phase diagrams.
Each apex point of the triangle represents one of the pure components. Any point
of a side of the triangle represents a binary mixture. Lines may be drawn parallel
to the sides of the equilateral triangle for the plotting of the compositions. Any
point in the triangle represents a ternary mixture and its composition can be read
easily from the apex lines. The ternary phase diagram can be constructed from
experimental data. Figure 7.1 represents the ternary diagram for water-ethanol-
benzene system1 at 25°C and atmospheric pressure.
The saturation curve or bimodal curve (miscibility boundary) on Fig. 7.1 is
represented by DEFGHIJKK′J′I′H′G′F′E′D′. On this diagram DD′, EE′, ..., JJ′
represent tie lines which connect two equilibrium points on the miscibility
boundary. For example, a mixture representing point L can be prepared which
has composition 45% benzene, 35% ethanol and rest water (by mass) by mixing
a feed containing water (43.75%) and ethanol (56.25%) and nearly pure benzene.
After shaking it, if it allowed to equilibrate (similar to that in a mixer-settler), it
will separate in two layers. Top light layer (point I′) is known as the raffinate
phase which will contain 84.4% benzene and 13.85% ethanol
 Fig. 7.1 Phase Diagram for Benzene–Ethanol–Water at 25°C
(Reproduced with the Permission of the American Chemical Society, USA)

Note: All percentages are mass %

while bottom heavy layer (point I) is called the extract phase which will contain
13.7% benzene and 52.2% ethanol. Point M is called plait point, representing
critical composition; 37.4% ethanol, 52.0% benzene and 10.6% water. The plait
point represents ternary homogeneous liquid mixture, saturated with both;
solvent and diluent. Curve on the right of the plait point is called saturated
extract phase while curve on the left is called saturated raffinate phase. Extract
phase is a ternary homogeneous liquid mixture, saturated with diluent, while
raffinate phase is the ternary homogeneous liquid mixture, saturated with
solvent. The area under the binodal curve is the region where two liquid phases
will exist. Above the curve, there will be only one liquid phase.
Material balance equations can be established to find quantities of the separated
layers, knowing their compositions. Alternately using line ratio principle of
geometry, one can find the quantities from the diagram (Fig. 7.1).
IL/LI′ = Quantity of raffinate phase/Quantity of extract phase
= 0.7125 linear units/0.885 linear units = 0.805 : 1
Using above relation, one can find a mass ratio of extract to raffinate phases as
44.6:55.4. In addition to the above considerations, equilibrium constraints must
be satisfied.
Distribution coefficient or partition coefficient K or m at equilibrium is defined
as
K or m =   (7.1)

In liquid-liquid extraction, all the components are present in both extract and
raffinate phases at equilibirum. At equilibirum, acitivity of any component is
same in both phases.
År = Åe ⇒ γr x = γe y  (7.2)
where, År = activity of component in raffinate phase
Åe = activity of component in extract phase
γr, γe = activity coefficients of components in raffinate/extract phase
y = mole fraction of component in extract phase
x = mole fraction of component in raffinate phase
K′ =   (7.3)

where, K′ = distribution coefficient in mole fraction units

Large collections of experimental equilibrium data are now available for liquid-
liquid systems. One source is “Liquid-liquid Equilibrium Data Collection” by
Sorenson and Arit, DECHEMA, Frankfurt, Germany. In Perry’s Chemical
Engineers’ Handbook (Ref. 4), distribution coefficients for a few ternary systems
are given (Table 15.5 of 6th Ed.)

7.4 DESIRABLE SOLVENT PROPERTIES OR

CHOICE OF SOLVENT
7.4.1 Selectivity
The relative separation or selectivity is defined as the ratio of ratios of solute to
nonsolute components in extract phase to the same in raffinate phase at
equilibrium. It is represented as α (or β).
For all useful extractions, α > 1. If α = 1, it means no separation or no extraction.
It represents the separation power of the solvent. Higher the selectivity, lesser is
the number of theoretical stages required for the desired separation. Dilute solute
concentration generally gives the highest selectivity.
7.4.2 Distribution Coefficient
Distribution coefficient K or m is the ratio of mass fraction of solute in extract
phase to mass fraction of solute in raffinate phase at equilibrium.
It is not necessary that K should be greater than 1. K can be less than 1. But
higher value of K is desirable because higher the distribution coefficient, lesser is
the amount of solvent required for desired separation. Unfortunately solvents
which provide the higher value of K normally provide poor value of α.
7.4.3 Recoverability
It is always necessary to recover the solvent from extract phase for reuse. Hence,
selected solvent must be easily recoverable from the extract phase. In most of the
cases, solvent is separated from extract phase by distillation. If selected solvent
is to be separated or recovered by distillation than it should preferably not form
an azeotrope with the solute and also it should provide higher value of relative
volatility. Lesser value of latent heat of vaporisation of solvent is also desirable
for low cost of recovery. Water is not an attractive solvent for extraction as it
forms azeotrope or non-ideal solution with many organics and it has very high
value of latent heat of vaporisation.
7.4.4 Capacity
Capacity of a solvent to extract the solute is measured in terms of maximum
solubility of the solute in solvent or in terms of solubility limit. Higher the
solubility limit, lesser is the amount of solvent required. Solute should be
significantly soluble in the selected solvent.
7.4.5 Insolubility of Solvent in Raffinate
Low solubility of solvent in the raffinate phase is desirable. Solvent which is
more insoluble is preferred. More insoluble solvent can separate wide range of
composition of the feed mixture. Also, it gives less solvent loss and adds less
impurity in the final raffinate. Generally, a solvent which is more insoluble
provides higher selectivity.
7.4.6 Density
Higher density difference between solvent and diluent or extract and raffinate
phase is desirable. It makes the decantation easier. In most of the systems,
density difference decreases to zero at plait point, but in some system it can
become zero at an intermediate solute concentration and can invert the phases at
higher concentration. For such a system, in the desired range of separation in the
extractor, if density difference passes through zero then use of continuous
contact equipment like packed tower is not preferred but mixer-settler-type
stagewise extractor can be used. Other option is to use two continuous contact
extractors in series.
7.4.7 Interfacial Tension
A high interfacial tension promotes rapid coalescence and generally requires
high mechanical energy for agitation to produce small droplets. Conversely, low
interfacial tension requires lesser mechanical energy for agitation to produce
small droplets but at the same time it provides slow coalescence rates. Usually
interfacial tension decreases with increase in solute concentration and increase in
mutual solubility and falls to zero at plait point. Very high and very low
interfacial tension of selected solvent is not desirable, otherwise higher collision
rate is more important, hence moderately high interfacial tension is preferred.
7.4.8 Chemical Reactivity
The solvent should be inert towards the other components of extraction system
and to the common materials of construction.
7.4.9 Viscosity, Vapour Pressure and Freezing Point
All these properties of solvent should be low for ease in handling, transportation
and storage.
7.4.10 Toxicity and Flammability
The solvent should preferably be non-toxic, non-flammable and of low cost.

7.5 DESIGN OF COUNTER CURRENT

MULTISTAGE EXTRACTOR
Process design of extractor can be divided in the following steps:
1. Selection of a suitable solvent and determination of liquid-liquid equilibrium
data. Actually both these steps are interrelated.
2. Determine the minimum amount of solvent required for the desired
separation.
3. Determine or decide the actual amount of solvent. Actual amount of solvent
must be the optimum amount of solvent. Then complete the material balance.
4. Determine the number of theoretical stages (for stagewise extractor) or
number of transfer units (for differential continuous contact extractor)
required for the desired separation.
5. Find the tower diameter of tower-type extractor or carry out the process
design of agitator for agitated vessel-type extractor and also decide the height
and diameter of agitated vessel.
6. Find the stage efficiency for stagewise extractor or height of transfer units for
differential or continuous contact-type extractor.
7. Find the actual number of stages for stagewise extractor or height of tower for
tower-type extractor.
(a) First step is explained in previous Section 7.4.
(b) Sm: Minimum amount of solvent is the maximum amount of solvent for
which infinite number of stages are required for desired separation.

Fig. 7.2 Multistage Countercurrent Extraction

F + Sm = E1m + RNP   (7.4)

RNP – Sm = F – E1m = ΔRm   (7.5)
ΔRm is a difference point. It represents net flow outward at the last stage NP and
is equal to net flow inward at first stage. On an equilateral triangular
coordinates, when operating line from ΔR matches with tie line, infinite number
of stages are required for the desired separation. To find Sm, extend the tie lines
and allow them to intersect . A line joining the equilibrium compositions of
extract and raffinate phases is called a tie line.
Among all such points, a point that is farthest from S, if all points are on the left-
hand side of S or a point which is nearest to S, if all points are on the right-hand
side of S, represents ΔRm, the difference point for minimum solvent.
 (a) ΔRm Nearest from S for m < 1

(b) ΔRm Farthest from S for m > 1

Fig. 7.3 Minimum Solvent for Countercurrent Extraction4

In most of the cases but not in all cases, tie line which is on extension passes
through F gives the location of ΔRm. After finding the location of ΔRm, locate
E1m. Draw the line which intersects line F—S at XMm.
F + Sm = Mm  (7.6)
FXF + SmYs = XMm Mm = XMm (F + Sm)  (7.7)
By substituting the value of XMm from the graph, one can find Sm using Eq.
(7.7).
If A and B are insoluble liquids or can be considered as insoluble liquids at
operating temperature and in desired concentration change, then solute balance
is given by following equation:
A(X′F – X′NP) = B(Y′1 – Y′S)  (7.8)
where, X′ = and Y ′ =

=   (7.9)

=   (7.10)

where, Bm = Minimum amount of solvent

Y′1k is an equilibrium with X′F. From equilibrium data, one can find Y′1k that is in
equilibrium with X′F. Then from Eq. (7.10), one can find Bm. Operating line
starts from point (X′NP, Y′s).

Fig. 7.4 Determination of Bm

8. To decide the actual amount of solvent, find the number of theoretical stages
required for the desired separation for the different values of actual amount of
solvent, like
S = 1.1 Sm, S = 1.2 Sm, S = 1.3 Sm … and so on.
On increasing the values of S from Sm, value of N will decrease. In initial range,
N will significantly decrease with increase in S, but after a certain increase in S,
N may not decrease appreciably.
For example, in this case Fig. 7.4 one can say optimum value of S is very close
to 1.4 Sm.

Fig. 7.5 Determination of Sopt

Ideally actual amount of solvent = Sopt

But to find optimum values of S detailed costing calculations are required.
9. Procedure to find number of theoretical stages for desired separation N from
the value of S:
(a) If A and B are not completely insoluble liquids:
Refer to Fig. 7.2.
Overall Material Balance.
F + S = E1 + RNP  (7.11)
Solute balance:
FXF + SYs = E1Y1 + XNP RNP = MXM  (7.12)
F – E1 = RNP – S = ΔR = Net flow outward = Net flow inward  (7.13)
 Fig. 7.6 Determination of Operating Curve

Procedure:
Find Xm = . Locate point M on graph which lies on F—S line. Extend

the line that will intersect the binodal curve at E1. Draw the lines

and . These will intersect at the point ΔR. Draw the random lines from
ΔR that give the points of operating curve (Xs, Ys + 1).
Operating curve starts from point (XF, Y1) and terminates at point (XNP, Ys). Start
the stagewise construction from (XF, Y1) to (XNP, Ys) in between equilibrium
curve and operating line and find number of theoretical stages, required for
the desired separation.
(b) If A and B are immiscible liquids then overall material balance will be
A(X′F – X′NP) = B(Y′1 – Y′s)  (7.14)
By using Eq. (7.14), for the given actual amount of solvent B and its
composition Y′s, one can find Y′1. Operating line starts at (X′F, Y′1) and
terminates at (X′NP, Y′s).
Stagewise construction starts from point (X′F, Y′1) and terminates at the point (X
′NP, Y′s) of the operating line.
In a special case, where equilibrium curve is also the straight line then the
number of theoretical stages required for the desired separation can be
obtained from Eq. (7.16a) or Eq. (7.16b) (Ref. 3).
 Fig. 7.7 Determination of Equilibrium Stages

Fig. 7.8 Countercurrent Extraction with Insoluble Solvent

Let m′ = = slope of equilibrium line

Ratio of the slope of equilibrium line to the slope of operating line is given by
 following equation:
ε =   (7.15)

If, ε ≠ 1

N =   (7.16a)

When ε = 1
N = – 1  (7.16b)

Multistage counter current extraction requires less number stages for the given
amount of solvent than the cross-current extraction or it requires less amount of
solvent for the fixed number of stages than cross-current extraction.
Number of Transfer Units (Ref. 4)
In case of differential or continuous contact-type extractor, instead of number of
theoretical stages, number of transfer units are determined. Then the working
height of continuous contact-type extractor is given by the equation
Z = Nt Ht   (7.17)
where, Nt = Number of transfer units
Ht = Height of a transfer unit
Number of overall mass transfer units based on raffinate phase is given by
equation

NtoR =   (7.18)

where, XF = Mass fraction of solute in feed

XNP = Mass fraction of solute in final raffinate
r = Ratio of molar masses of non-solute to solute
X* = For any point (Xa, Ya) of operating curve, X*a is the value of X that is in
equilibrium with Ya.
 Fig. 7.9 Determination of Number of Transfer Units

If solute component or fraction of feed (A) and solvent (B) can be considered as
completely immiscible liquid mixture, then it is better to use mass ratio
concentrations, rather than mass fractions. For this special case NtoR is given by
the equation

NtoR =   (7.19)

where, X ′F = Mass ratio of solute to non-solute in feed

X ′NP = Mass ratio of solute to non-solute in final raffinate
X*a = For any point (X′a, Y′a) of operating line, X*a is the value of X′ that is in
equilibrium with Y′a
r = Ratio of molar masses of non-solute to solute
In another special case where operating line and equilibrium curve both are
straight lines, number of overall mass transfer units based on raffinate phase
(NtoR) is given by the equation,
NtoR = (if ε ≠ 1)  (7.20)

or
NtoR = [(X ′F – Y′s/m′)/(X ′NP – Y′s/m′)] – 1 (if ε = 1)  (7.21)

where, ε = and m′ =

A feed stream having flow rate of 200 kg/h and containing 20

mass % acetic acid in water is to be extracted at 25°C with 400 kg/h of recycled
MIBK (Methyl IsoButyl Ketone) that contains 0.1 per cent acetic acid and 0.01
per cent water. The aqueous raffinate is to be extracted down to 1 per cent acetic
acid. How many theoretical stages will be required and what will the extract
composition be? Extraction is to be carried out in counter current manner.
Table 7.1 Equilibrium Data4 of Water – Acetic Acid – Methyl Isobutyl Ketone at 25oC
mass % in Raffinate mass % in Extract
Water Acetic acid MIBK X′ Water Acetic acid MIBK Y′

98.45 0 1.55 0 2.12 0 97.88 0

95.46 2.85 1.7 0.0299 2.8 1.87 95.33 0.0196
85.8 11.7 2.5 0.1364 5.4 8.9 85.7 0.1039
75.7 20.5 3.8 0.2708 9.2 17.3 73.5 0.2354
67.8 26.2 6.0 0.3864 14.5 24.6 60.9 0.4039
55 32.8 12.2 0.5964 22 30.8 47.2 0.6525
42.9 34.6 22.5 0.8065 31 33.6 35.4 0.9492

(Ref. Table 15.1 of Ref. 4)

X′ = mass ratio of solute to diluent (C/A) in raffinate phase
Y′ = mass ratio of solute to extraction solvent (C/B) in extract phase
 Fig. 7.10 Ternary Diagram for AcOH-H2O-MIBK System4

Solution
F + S = E1 + RNP = M  (7.11)
200 + 400 = M = 600 kg/h
Xm =

= = 0.067 33,

Y1 = 0.088 (From Fig. 7.10)

MXM = E1Y1 + XNP RNP
40.4 = 0.088 (600 – RNP) + 0.01RNP
RNP = 158.97 kg/h, E1 = 441.03 kg/h
Random lines from ΔR give the points of operating line (Xs, Ys + 1).
Xs 0.2 0.14 0.08 0.065 0.01
Ys + 1 0.085 0.06 0.035 0.025 0.001
Equilibrium data:
X 0 0.0285 0.117 0.205 0.262 0.328 0.346
Y* 0 0.0187 0.089 0.173 0.246 0.308 0.336
Stepwise construction from XsF (Xs corresponds to XF) to XNP gives number of
theoretical stages required for desired separation.
 Fig. 7.11 Determination of Number of Stages for Example 7.1

From graph (Fig. 7.11), N = 4.62.

Solve Example 7.1 considering MIBK (B, solvent) and water
(A, non-solute component) of feed as completely immiscible liquids and
equilibrium curve as a straight line in terms of mass ratio concentrations.
Solution
Overall solute balance
A(X′F – X′NP) = B(Y′1 – Y′s)  (7.14)
A = 200 (1 – 0.2) = 160 kg/h
B = 400 (1 – 0.001) = 399.6 kg/h
= 0.0101

= 0.001

160 (0.25 – 0.0101) = 399.6 (Y′1 – 0.001)

Y′1 = 0.097
X′F = 0.25,
For operating range X′ < 0.25 slope of equilibrium curve

m′ = = 0.7571

ε = = 0.7571 × = 1.89

N =   (7.16a)

N =

N = 4.13
For the extraction, described in Example 7.1, calculate
(a) NtoR considering MIBK and water as partially miscible liquids.
(b) NtoR considering MIBK and water as completely immiscible liquids and
equilibrium curve as straight line (in terms of mass ratio concentration).
Solution
(a) Number of overall mass transfer units based on raffinate phase is given by
the equation

NtoR =   (7.18)

XF = 0.2, XNP = 0.01

Data for equilibrium curve
X 0 0.0285 0.117 0.205 0.262 0.328 0.346
Y 0 0.0187 0.089 0.173 0.246 0.308 0.336
Data for operating curve
Xs 0.2 0.14 0.08 0.01
Ys + 1 0.085 0.06 0.035 0.001
Table 7.2 Data for Finding NtoR [from Fig.   (7.12)]
 X 0.2 0.15 0.1 0.05 0.025 0.01
X* 0.113 0.089 0.063 0.032 0.014 0.001
X – X* 0.087 0.061 0.037 0.018 0.011 0.009
1/(X – X*) 11.494 16.393 27.027 55.555 90.909 111.111

Fig. 7.12 Equilibrium Data for AcOH – Water – MIBK System

 Fig. 7.13 Determination of NtoR

From Fig.   (7.13), = 1.7236 square units = 7.0351 (Integral value)

r = = 0.3

NtoR = 7.0351 +

= 7.0351 + 0.106 55 + 0.0719 = 7.213 55

(b) If water (A) and MIBK (B) are immiscible liquids and equilibrium curve is
straight line (in mass ratio concentration) for the given operating range

NtoR =   (7.20)

X′ 0 0.0299 0.1364 0.2708

Y′ 0 0.0196 0.1039 0.2354

X′F = 0.25, m′ =

= 0.7571
ε = = 0.7571 × = 1.89

NtoR =

= 5.584
For the separation of dimethylformamide (DMF) from its
dilute solution in water, liquid-liquid extraction may require lower operating cost
than atmospheric distillation. Methylene chloride is considered to be a preferred
solvent for extracting DMF from its aqueous solution.
DMF-water solution having a flow rate of 1000 kg/h and containing 20% DMF
by mass is to be counter currently extracted with methylene chloride to reduce
the DMF concentration to 1% in the final raffinate. Determine (a) the minimum
amount of solvent which can be used and (b) the number of theoretical stages if
actual amount of solvent is double than the minimum required.
Data: Phase equilibrium equation6 is given by equation Y = 0.5555 X at 25° C
where, Y and X are mass fractions of solute.
Solution
Mutual solubilities of water (A) and methylene chloride (B) in each other at
room temperature are very low. Hence, to simplify the calculations, water (A)
and methylene chloride (B) may be assumed as completely insoluble liquids.
(a) For completely immiscible liquids, solute balance is given by the equation
A(X′F – X′NP) = B(Y′1 – Y′s)  (7.14)
Table 7.3 Determination of Data for Equilibrium Line
Y 0 0.013 89 0.0278 0.042 0.055 55 0.0694 0.0833 0.1111 0.1389 0.166 65
X 0 0.025 0.05 0.075 0.1 0.125 0.15 0.2 0.25 0.3
 Y′ 0 0.014 0.0286 0.04384 0.0588 0.0746 0.09 0.125 0.1613 0.2
X′ 0 0.0256 0.0526 0.081 0.1111 0.1429 0.1765 0.25 0.3333 0.4286

Fig. 7.14 Determination of Equilibrium Stages for Example 7.4

From graph Y′1m = 0.125 [From Fig.   (7.14)]

Solute balance:
A(X′F – X′NP) = Bm(Y′1m – Y′s)
A = F(1 – XF) = 1000 (1 – 0.2) = 800 kg/h, X′NP = 0.01/0.99 = 0.0101
800 (0.25 – 0.0101) = Bm (0.125 – 0)
Bm = 1535.36 kg/h
(b) Actual amount of solvent B = 2 × 1535.36 = 3070.72 kg/h
Solute balance:
A(X′F – X′NP) = B(Y′1 – Y′s)
800 (0.25 – 0.0101) = 3070.72 (Y′1 – 0)
Y′1 = 0.0625
From Fig. 7.14, N = 3.85
If equilibrium curve is considered as a straight line in terms of mass ratio
concentration, then number of theoretical stages required for desired separation
is given by the following equation:
N =   (7.16a)

ε =

m′ =

= 0.523
ε = = 2

N =

N = 3.687
For the liquid-liquid extraction system, given in Example 7.4,
calculate NtoR for the actual amount of solvent B = 2 Bm.
Solution
Water (A) and methylene chloride (B) are considered as completely immiscible
liquids. Also, as shown in Fig. 7.14, equilibrium curve is a straight line in terms
of mass ratio concentration. With these assumptions, formula for finding NtoR is

NtoR =   (7.20)

B = 2Bm = 3070.72 kg/h, A = 800 kg/h,

X′f = 0.25, Y′s = 0, m′ = 0.523
ε = = 2, X′NP = 0.0101
NtoR =

NtoR = 5.11

7.6 INDUSTRIALLY IMPORTANT EXTRACTORS

Many types of extractors are available for achieving mass transfer (Ref. 4, 5, 7,
8, 9, 10). Each has its own particular advantage. As per general classification,
there are two fundamental types of extractors; stagewise and differential. The
stagewise extractors consist of discrete units, in each of which the phases are
brought into contact, mixed, allowed to separate and sent to next unit.
Differential extractors provide continuous contact and mass transfer along the
full length of the device and phases are separated only at the end.
7.6.1 Categorizing Industrially Important Extractors
The decisive criteria for the selection of the contactor device is the physical
properties of the system feed/solvent as it is characterised by the density
difference and interfacial tension, and the difficulty of the separation in terms of
required number of theoretical stages (NTS) to reach the desired raffinate and
extract concentrations. On this basis, industrial extractors can be divided into
four categories:
1. Mixer-settler for low NTS requirement.
2. Unagitated static columns for low NTS requirement.
3. Agitated columns, including pulsating equipments, for high NTS requirement.
4. Centrifugal extractors for difficult separations and for low NTS requirement.
Salient features of various extractors are discussed below.
Mixer-settlers
Each mixer-settler unit provides a single stage of extraction. The two phases
enter the mixing section where they are mixed using an impeller. Later the two-
phases are allowed to separate by gravity due to their density differences.
Typical mixer settler has mixing time of the order of a few minutes and settling
times of several minutes.
Advantages
1. Provides better contact of extract and raffinate phases and thus provides
 higher rate of mass transfer.
2. Higher efficiency or attain separation closer to equilibrium.
3. Requires limited instrumentation.
4. Reliable scale-up.
5. Requires low head room.
6. Low capital (fixed) cost.
7. Provides more stagewise flexibility.
8. Offers wide range of capacities and simple operation.
Disadvantages
1. Requires large hold-up.
2. Higher power requirement.
3. High solvent inventory.
4. Requires large floor space.
If number of theoretical stages, required for desired separation, is less than or
equal to four, mixer settler is probably the most economical amongst all types of
extractors. In this extractor, turbines or propellers are used as agitators. Amongst
all types of agitators, pitched blade turbine agitator [Fig. 10.2(iv)] is found to be
most effective for mixing of two immiscible liquids. For smaller vessels,
propellers are selected for mixing.
Types of mixer settlers There are basically two types of mixer settlers. Batch
vessels are used for mixing solvent and feed solution. After mixing for a few
minutes, two phases are allowed to settle. Top organic layer (generally light) is
decanted out.
Figure 7.15 depicts a continuous mixer-settler system. Solvent and feed solution
are pumped in the vessel with help of an axial pump. Mixer is typically equipped
with slow speed, large diameter, low shear pump-mixer-type agitator(s). Pump
mixers commonly have variable speed drives and are designed to avoid fine
droplet formation and air entrainment which hinder coalescence in the settler. If
required, auxiliary mix boxes with agitators can be used for large plants to avoid
mixing inefficiencies which can occur with large mix boxes.
 Fig. 7.15 Continuous Mixer Settler

In the adjoining settler, the phases coalesce and separate by gravity and are
removed via separate weir boxes as shown in the figure. Mix boxes can be
cylindrical or square. Settlers are generally rectangular. Devices such as flow
distributors, picket fences and various coalescence aids are installed in the settler
to improve performance. Phase separation rates in the settlers are relatively slow,
and the specific area is typically 3 to 5 m3 per h (of combined aqueous and
organic phases) per m2 of settler area, depending on operating temperature,
aqueous feed composition and concentration of the extract phase in the organic
phase.
Mixer settlers have become popular in nuclear industry. For this purpose, Krebs
mixer-settler is a popular design. This design is similar to Fig. 7.15 but the
agitator is fitted with a low head, high volume, low shear conical pump,
consisting of a static tulip shaped stator and a trapezoidal, typically six bladed
rotor. Settler is of compact design and thus occupies less floor space. It is usually
designed for flow rate of 8 to 12 m3/(m2 · h).
Unagitated static columns There are two types of columns that are used for
extraction operation; packed column and sieve plate column.
In industry, the use of packed columns (random or structured) for liquid-liquid
extraction applications will be in most cases found limited to operations where
the density differences between the carriers is higher than 30–50 kg/m3, the
volumetric phase ratio (φ) between dispersed drop phase and continuous phase is
in the magnitude of 0.5 < φ < 5 and the number of theoretical stages (NTS) ≤ 10.
Figure 7.16 schematically represents variety of packed columns for this service.
It may be noted that arrangements for heavy or light dispersed phases are
different for packed beds.

Fig. 7.16 Typical Column Arrangement for Packed Columns for Liquid Extractions

There is a different function of the column internals in vapour/liquid services

like distillation or absorption, where the specific surface of the packing provides
the required area to perform mass transfer. In contrast, liquid-liquid extraction
columns the mass transfer area is almost independent of the packing surface, but
is directly linked to the drop phase holdup in the tower. The main function of the
packing is to provide an increased flow path length and more residence time for
the drop phase. Coalesced drops are redispersed at the sharp edges of the
packing and the residence time distribution of both phases should be kept as
narrow as possible (plug flow) by minimizing any axial back mixing effect. It is
preferred to select a packing that is wetted by continuous phase. Historically
random packings were the choice in packed towers but with technical
developments, structured packings also find applications in extraction due to
better turn down ratio and higher throughputs.
Typical column internals for extraction service are also shown in Fig. 7.17. It
may be seen that dispersion tray and heavy phase inlet distributor are somewhat
different from those used for distillation/absorption service. Two designs of
coalescers are included. This piece of internals plays an important role as both
phases should separate well before they are allowed to come out from the
extractor. These coalescers are produced by two dissimilar filaments, knitted
together to form the mesh. One filament is a metallic wire made of stainless steel
or an alloy and the other is made of PP, PTFE, glass fibres, etc. In cartridge
form, the liquid flows from the centre radially outwards. For extraction columns,
mesh is made up of say 75 μm size filaments.
 Fig. 7.17 Packed Tower Internals for Liquid-Liquid Extraction

Advantages
1. High specific throughput capacity.
2. Easy operation and maintenance because no moving parts are involved.
3. For corrosive media there are almost unlimited options for the material
selection available, from stainless steel to high-alloyed metal, from
polypropylene to PTFE, technical ceramics or pure carbon.
4. A simple operation also at high pressure or temperature.
Disadvantages
1. Although empirical equations are available for design of packed columns, it is
still advisable to conduct pilot studies for fixing mass transfer coefficient
(KLa) and deciding on flooding velocity.
2. Packed columns are costly as compared to mixer settlers.
3. Selection of internals plays important role in its functioning. Hence, design of
packed columns is required to be carried out with care.
Sieve tray column of special design can also be used for liquid-liquid extraction
(Fig. 7.18). Unlike sieve tray column for distillation/absorption, it has
coalescer/disperser below each sieve tray and weirs are not provided on the
trays. Sieve tray column is less preferred to a packed column for extraction
service.
One of the more common liquid-liquid extraction applications is the removal of
sulphur compounds from liquefied petroleum gas (LPG) using an amine
solution. A packed tower sizing may be based on an amine superficial velocity of
5 m3/(m2 · h). Normally LPG is a dispersed phase in this application.
Another application of a packed column extractor is of organic contaminants
removal from waste water. When organic compound’s boiling point is more than
that of water, liquid-liquid extraction offers a better option. Acetic acid, pyridine,
aniline, nitrobenzene, etc., are some organic contaminants in waste water that
can be extracted with a solvent and recovered with low energy requirements.
Agitated columns Several designs of agitated columns are available in industry
for liquid-liquid extraction service. In this category, two types of extractors are
considered. One is pulsed column and another is a rotating disc contactor.
 Fig. 7.18 Sieve Plate Extractor

1. Pulsed Columns: The pulsed column [Fig. 7.19(a)] consists of a vertical

cylindrical contacting section with settlers at the top and bottom. The contacting
section is fitted with either sieve plates or “disc and doughnut” plates. Generally,
organic phase is continuous and aqueous phase is dispersed. Both are
continuously fed in countercurrent directions. Energy is applied to pulse the
liquids in the column up and down. One way is done by injecting pressurized air
into a pulse leg that pushes liquid into the column, then venting the pulse leg to
fill the pulse leg with solution from the column. Alternately energy can be
supplied by mechanical devices, such as pistons, that move back and forth to
displace the solution in the column. The pulse action lifts and lowers the solution
in the column, usually only a fraction of a metre.
Advantages
1. Potentially lower organic inventory than conventional mixer settlers.
2. Small floor space area.
3. Ability to operate with the solution with the smaller flow as the continuous
phase.
4. No phase recycle needed.
5. Can incorporate a large number of stages in one column.
6. Totally enclosed system, excludes dust, minimises evaporative losses,
improves fire protection, and (excludes air in mechanically agitated
extractor).
7. Has greater tolerance for suspended solids.
8. No internal moving parts; low shear; low entrainment; low maintenance.

Fig. 7.19(a) Agitated Extractors

9. Ability to vary droplet size, by varying pulse frequency and amplitude.

10. Opportunity to use inexpensive FRP for construction.
Disadvantages
1. HTS requirement is high.
2. High head room requirement.
3. Less flexible for large feed flow variations.
4. Pilot scale testwork needed to develop scale-up data.
5. Less accessible for clean-out.
6. Visual monitoring of process difficult.
7. Power is required for pulse generation.
2. Rotating Disc Columns (RDC) Two designs are figured in this category.
Scheibel extractor is shown in Fig. 7.19(b). The impellers are turbine type and
disc and doughnut baffles surround them. In another design [Fig. 7.19(c)], a
simple disc is provided in the centre and horizontal baffles are provided near the
wall. The disc rotates at much higher speed than turbine-type impellers. In Kuhni
design (third design), instead of baffles, sieve plates are provided between two
discs.

(b) Scheibel Extractor

 (c) Rotating Disc Extractor
Fig. 7.19(b) and 7.19(c)

The disc contactor consists of a cylindrical column that is divided into number of
compartments formed by a series of stator rings. Each compartment contains a
centrally located, horizontal rotor disk that creates a high degree of turbulence
inside the column. Typically, the disk diameter is 1/3rd to 2/3rd of the column
diameter. The disk disperses the liquid and forces it outward toward the wall
where the stator rings create quite zones where the two phases can separate.
Most extraction takes place in the mixing section but some does take place in the
calming section so that efficiency of each mixer settler unit is sometimes greater
than 100%. Typically, each mixer-settler compartment is 300 to 600 mm high.
Advantages
1. Good throughput.
2. High efficiency means more stages per metre height.
3. Less head room and less floor space.
Disadvantages
1. Corrosive liquids can damage the internal moving parts.
2. Pilot study is recommended for scale-up.
Centrifugal extractors
Centrifugal contactors are high-speed rotary machines with the advantage of
very low residence time. Centrifugal contactors utilise a mechanical device to
agitate the mixture to increase the interfacial area and decrease the mass transfer
resistance. In this machine, heavier phase is transported to the periphery and the
lighter phase to the centre which means centrifugal force is responsible for the
countercurrent flow of the phases. These are ideal for systems in which the
density difference is less than 4%.
Centrifugal extractors are expensive but provide many theoretical stages of
separation with very short hold-up time (about 4 s). While it is possible to obtain
many theoretical stages in a single machine, 3 to 5 theoretical stages are more
common. They are therefore used to extract valuables and heat sensitive
products such as vitamins and antibiotics.
Figure 7.20(a) and (b) show single stage and a four stage centrifuges for
extraction service, respectively.
Advantages
1. High throughput
2. Less amount of solvent is required
3. Short residence time by fast reaching the equilibrium
Disadvantage
1. High initial cost
Centrifugal extractors have become popular in pharmaceutical industry.
(a) Single Stage Centrifugal Extractor
 (b) Four Stage Centrifugal Extractor
Fig. 7.20 Centrifugal Extractors

7.6.2 Comparison of Various Industrial Extractors

With availability of different extractors, it is necessary to compare them for their
efficiencies and their processing capability. Process engineer can use Fig. 7.21
for selection of the best suited extractor for his application.

Fig. 7.21 Comparison Plot of Various Commercial Extractors4

 (Reproduced with Permission from Perry’s Chemical Engineers’ Handbook, 7th Edition)

Selection criteria and performance of various extractors are qualitatively

compared in Tables 7.4 and 7.5.
Table 7. 4 Selection Criteria for Continuous Extractors
Type of extractor Number of stages Flow rate Residence time Interfacial tension Viscosity Density difference Floor space

Mixer-settler Low High Very high Moderate to high Low to high Low to high High
Static (unagitated) columns Moderate Moderate Moderate Low to moderate Low to moderate Low to moderate Low
Agitated columns High Moderate Moderate Low to high Low to high Low to high Low
Centrifugal extractor High Low Very low Low to moderate Low to moderate Low to moderate Moderate

Table 7.5 Performance Comparison of Extractors

HETS, Superficial velocity of continuous and dispersed Liquid capacity of combined streams, Stage
Type of Extractor
m phases, m/h m3/(m2 · h) efficiency, %

Mixer settler (single 0.75–1 3–5 75–100

stage)
Sieve tray tower 0.5– 27–75 3–60 6–24
3.6
Packed Tower 0.6– 12–40 (Pall rings) 6–45 –
1.5 65–90 (Structured)
Pulsed column 0.8– 25–35 – –
1.3
Rotating disc 0.3– 7–30 15–30 80–100
contactor 0.5
Scheibel extractor 0.2– 6–14 15–30 80–100
0.4
Centrifugal extractor – 75–110 – 300–1000

7.6.3 Process Design of Mixer

It includes two steps:
(i) Determining the volume of the mixer.
(ii) Process design of agitator; e.g. deciding shape factors and power
requirement.
(iii) Mixer settlers are operated in batchwise mode and also in continuous
manner. For batch mixer, volume of mixer is decided by the total volume of
liquid mixture, charged to the mixer. For continuous mixer, volume of the
mixer depends on desired hold up time.
Ideally, hold up time in a continuous mixer is the time required to reach
equilibrium. It also depends on the rate of mass transfer. It can be decided on
the basis of pilot plant or lab scale experiments of the given system.
(iv) Process Design of Agitator (Ref. 2 and 11)
The work of Miller and Mann and of Olney and Carlson have established that to
calculate power required for agitating two liquid phases one can use the
 power data available for agitating single phase liquids, if density and
viscosity of two phase liquid mixture are determined by following equations.
ρav = ρC ZC + ρD ZD  (7.22)

μav =   (7.23)

where, μav = Average viscosity of liquid mixture cP or mPa · s

ρav = Average density of two-phase liquid mixture, kg/m3
ρC = Density of continuous phase, kg/m3
ρD = Density of dispersed phase, kg/m3
Z = Volume fraction of appropriate phase
μC = Viscosity of continuous phase, cP or mPa · s
μD = Viscosity of dispersed phase, cP or mPa · s

Fig. 7.22 Agitator Power Correlations4

Curve 1: Six Flat Blade Turbine, Da/Wi = 5, Four Baffles Each DT/12, (Refer to Fig. 10.2 (v))

Curve 2: Six Flat Blade Open Turbine, Da/W = 8, Four Baffles Each Equal to DT/12,
(Refer to Fig. 10.2 (ii)
Curve 3: 45° Pitched Blade Turbine, Da/Wi = 8, Four Baffles, Each DT/12, (Refer to Fig. 10.2 (iv). but with
 six blades)
Curve 4: Propeller Pitch Equal to 2Da, Four Baffles, Each DT/12, (Refer to Fig. 10.2 (i))

Curve 5: Propeller Pitch Equal to Da, Four Baffles, Each DT/10, (Refer to Fig. 10.2 (i))

(Reproduced with the Permission of McGraw-Hill Education, U.S.A.)

Power data for agitating single phase liquids in baffled vessel are available in the
form of graph Np vs Re [Fig.   (7.22)]. The same can be used to calculate
the power required for mixing two phase liquid mixture.
Np =   (7.24)

Re =   (7.25)

where, Np = Power number

Re = Reynolds number
P = Power required for agitation, W
gc = Newton’s law of gravitational constant, (gc = 1 for SI system)
n = Impeller rotational speed, s–1
Da = Impeller diameter, m
ρav, μav = Density and viscosity of two-phase liquid mixture, can be
determined by Eqs.   (7.22) and   (7.23), respectively.
7.6.4 Process Design of Batch Settlers (Ref. 3, 4, 11)
Settlers are also known as decanters. They are used to separate two immiscible
liquids or two partially miscible liquids. The mixture of liquids leaving a mixer,
is a cloudy dispersion which must be settled, coalesced and separated into its
liquid phases. It requires sedimentation and coalescence of the drops of the
dispersed phase. Decanters used are of the various shapes as shown in Fig.   
(7.23). It can be horizontal or vertical.
Volume of settler is estimated by taking hold-up time equal to 5 to 10 min.
Laboratory experiments for deciding hold-up time could prove valuable.
 Fig. 7.23 Batch Settlers

Diameter or size of the decanter is decided on the basis of fact that velocity of
continuous phase should be less than settling velocity of the droplets of
dispersed phase. Velocity of continuous phase is calculated by the equation
uc = < ud  (7.26)

where, uc = Velocity of continuous phase, m/s

qc = Continuous phase volumetric flow rate, m3/s
Ai = Area of interface, m2
ud = Settling velocity of the dispersed phase droplets, m/s
Stokes law is used to determine the settling velocity of the droplets.
ud =   (7.27)

where, dd = Droplet diameter, m

ud = Settling velocity of the dispersed phase droplet, m/s
g = Acceleration due to gravity = 9.81 m/s2
ρc = Density of continuous phase, kg/m3
ρd = Density of dispersed phase, kg/m3
μC = Viscosity of continuous phase, N · s/m2 or kg/(m · s)
Droplet size (dd) can be assumed equal to 150 μm. In decanter, feed droplet sizes
are normally well above 150 μm. If the calculated settling velocity is greater
than 4 × 10–3 m/s, then a figure of 4 × 10–3 m/s is to be used.
For vertical decanter, take L = 2 Di, thickness of dispersion band = 0.1 L and
residence time of droplets in the dispersion band = 0.1 L / ud. The residence time
of droplets in dispersion band should be 2 to 5 min.
For a horizontal cylindrical decanter, the interfacial area Ai depends on the
position of the interface.
Ai = WL  (7.28)
W = 2   (7.29)
where, W = Width of interface, m
Z = Height of the interface from the base of the vessel, m
L = Length of cylinder, m
r = Radius of cylinder, m

Fig. 7.24 Geometry of Horizontal Separator

In the approximate method, for horizontal decanter,

take L/Di = 4 and Di = 8.4   (7.30)
where, Di = Inside diameter, m
qd, qc = Volumetric flow rate, m3/s (for horizontal decanter)
For the extraction system described in Example 7.1, design a
mixer-setler-type extractor. Following data are obtained from pilot study:
(i) Residence time (hold-up time) of liquid mixer in pilot plant mixer, θ = 6 min
(ii) Stage efficiency = 0.92
(iii) Type of agitator = 45° pitch blade turbine
(iv) Shape factors of pilot plant mixer: Da/Wi = 8, four baffles, DT/J = 12, six
blades, H/DT = 1, Da / DT = 0.33
where, Da = Diameter of agitator, m
DT = Inside diameter of mixing vessel, m
J = Width of baffle, m
H = Depth of liquid tank, m
Wi = Width of agitator blade, m
(v) Power required per unit volume in pilot plant = 220 W/m3
Solution
Treybal2 demonstrated that for geometrically similar agitated vessels with equal
holding time and equal power per unit volume on the two scales, the stage
efficiency is likely to increase on scale-up. Based on this proposition, residence
time in commercial scale mixer can safely be taken as 6 min and stage efficiency
is 0.92.
Total number of mixer settlers = 4.62/0.92 ≅ 5
Working volume of mixer = Hold up time × Volumetric flow rate of liquid
mixture
Vw = θ × qv
qv = (Mass flow rate of feed + Mass flow rate of final extract)/ρav
ρav = ρC ZC + ρD ZD
ρMIBK = 801 kg/m3, ρwater = 1000 kg/m3, ρAcetic acid = 1049 kg/m3
This mixer–settler will be operated in continuous manner. If first mixer is
considered, then extract phase (MIBK rich phase) is the lighter phase and feed
(water rich phase) is the heavier phase. Feed density is estimated as
ρF = (6.55)

where, w1, w2 = Mass fractions of components

ρ1, ρ2 = Densities of components

ρF = = 1009.4 kg/m3

ρE = = 821.36 kg/m3

Volumetric flow rate of feed = 200/1009.4 = 0.198 m3/h

Volumetric flow rate of final extract = 441.03/826.36 = 0.5337 m3/h
Final extract is having higher volumetric flow rate than feed. Hence, feed or
water-rich layer is the dispersed phase and final extract (solvent rich layer) is the
continuous phase.

ZD = = 0.27

ZC = 1 – ZD = 0.73
ρav = ρC ZC + ρD ZD
= 821.36 × 0.73 + 1009.4 × 0.27 = 872.13 kg/m3
qv = (200 + 441.03)/872.13 = 0.735 m3/h

= 1, Vw = 0.735 = 0.0735 m3

= 0.0735

= 0.0735 m3

DT = 0.454 m
For 45° pitch blade turbine, Da/DT = 0.33, where Da = diameter of agitator, m
Da = 0.1498 m
Power required per unit volume = 220 W/m3
P = 220 × 0.0735 = 16.17 W
For first trial calculations, let tip velocity of agitator, v = 200 m/min = πDan
π × 0.1498 × n = 200
n = 424.98 rpm = 425 rpm
For pure components:
μMIBK = 0.5 mPa · s, μwater
= 0.8 mPa · s, μAcetic acid = 1.1 mPa · s
For organic liquid mixture12,
=   (7.31)

= = 1.889

μC = 0.5294 mPa · s

= = = 1.1818

μD = 0.846 mPa · s

μav = = 0.9059 mPa · s

Re = = = 153 025

Np = = 1.5 [From Fig.   (7.22)]

P =

= 35 W > > 16.17 W

Hence, decrease the value of tip velocity to match this value.
Let for second trial calculation, tip velocity = 155 m/min
n = = 329.36 rpm = 329 rpm

Re = 153 025 × = 118 459

Np = 1.5 (From Fig. 7.22)

P = = 16.27 W, close to 16.17 W

Use 45° pitch blade turbine, having diameter Da = 0.1498 m and rotational
speed, n = 330 rpm.
(b) Process design of settler
(i) Let position of settler be horizontal.
Approximate diameter of horizontal settler
Di = 8.4   (7.30)

qc = = 1.4825 × 10–4 m3/s

qd = = 5.503 82 × 10–5 m3/s

D = 8.4 = 0.1197 = 0.12 m

This equation is valid for L/Di = 4.
Length of horizontal decanter L = 0.48 m
D and L by other method:
Settling velocity of the droplets of dispersed phase; Eq. (7.27)
dd = 150 × 10–6 m, g = 9.8 m/s2, ρd = 1009.4 kg/m3
ρc = 821.36 kg/m3, μc = 0.5294 mPa · s

ud =

= 0.004 355 m/s > 4 × 10–3 m/s

Hence, uD = 0.004 m/s

Amin = = = = 0.037 m2
Amin = (Wi L)min
If the position of interface is fixed at the centreline of settler then Wi = Di, let L =
4Di.
Amin = 4Di2 = 0.037 m2
Di min = 0.096 m < Di = 0.12 m, L = 0.48 m
To fix the position of interface at centreline of decanter Zi = Di/2, locations of
outlet nozzles can be determined based on the pressure balance.

Fig. 7.25 Location of Nozzles for Horizontal Decanter

Pressure balance:
(zh – zi) ρH = (z1 – zi) ρL
ρH = density of heavier phase = 1009.4 kg/m3
ρL = density of lighter phase = 821.36 kg/m3
zi = Di/2 = 0.12/2 = 0.06 m
Let z1 = 0.8 Di = 0.8 × 0.12 m = 0.096 m
(zh – 0.06) × 1009.4 = (0.096 – 0.06) × 821.36
zh = 0.0893 m
Vertical decanter (other option):
Aimin = (π /4) D2i min = 0.037 m2
Di min = 0.217 m
Let Di = 0.25 m L = 2 × 0.25 = 0.5 m
Width of dispersion bend = 0.1 L = 0.05 m
Residence time of the droplets of dispersed phase in dispersion band,
θ = = = 11.481 s

As per the guidelines, this should preferably be 120 s. For 120 second residence
time, L = 5.226 m (very high). This guide line is given for higher flow rates
(1500 kg/h to 600 kg/h).
Hence, let Di = 0.25 m and L = 1 m which means θ = 23 seconds.
To fix the interface at the centre of vertical decanter, locations of nozzles must
be determined.
 Fig. 7.26 Location of Nozzles for Vertical Decanter

Pressure balance:
(zl – zi) ρLg + Zi ρH g = zh ρH g

zh = (zl – zi) + zi

zi = 0.5 m, zl = 0.9 L = 0.9 m

zh = (0.9 – 0.5) × + 0.5 = 0.8255 m

Two criteria may be used for scale-up of agitated vessel; (a) constant tip velocity
and (b) constant power per unit volume.
Criteria of constant tip velocity is used for miscible liquids where heat is
transferred from coil or jacket or where suspended solids are involved.
Criteria of constant power per unit volume can be used for immiscible liquid
systems gas-liquid systems, emulsions and pastes.
Incorporation of coalescer pads, such as that shown in Fig. 7.17, can improve the
separation of phases in the decanters.
7.6.5 Process Design of Packed Tower-Type Extractor
(Ref. 4, 7, 8, 9)
In packed tower-type extractor, standard commercial packings, used in vapour-
liquid system, are used. This includes structured packings, Raschig and Pall
rings, Berl and Intalox saddles and other random packings. Packing reduces the
vertical back mixing of the continuous phase and thereby reduces height
required for mass transfer. It is recommended to redistribute the dispersed phase
at about every 1.5 to 3 m distance to generate new droplets. Size of packing
should preferably be less than one-eighth of the tower diameter to minimise the
wall effect. Flooding velocity in packed tower is the velocity of dispersed phase
or continuous phase at which dispersed phase is coming out from the outlet pipe
of continuous phase or vice versa.
1. Tower diameter: Diameter of packed tower is determined based on value of
flooding velocity. To find the flooding velocity in packed tower many
correlations are available. One of the best correlations; Crawford–Wilke
correlation, is represented in Fig. 7.27.
 Fig. 7.27 Flooding in Packed Towers, Crawford-Wilke Correlation3 (Reproduced with the Permission of
McGraw-Hill Education, U.S.A.)

where, σc = Interfacial tension of continuous phase, lbf/ft

gc = Newton’s law gravitational constant
= 4.18 × 108 (lb · ft)/(lbf · h2)
ρC = Density of continuous phase, lb/ft3
μC = Viscosity of continuous phase, lb/(ft · h)
Δρ = Difference in density, lb/ft3
ap = Specific packing surface, ft2/ft3
ε = Fraction void volume in packed section
vCF, vDF = Superficial flooding velocity, ft/h
Actual velocity of the continuous phase and the same of the dispersed phase
should be less than or equal to 50% of the flooding values.
2. Height of packed tower: Height of packed tower H is given by equation.
H = NtoR · HtoR = NtoE · HtoE  (7.32)
Value of HtoR or HtoE should be decided by actual experiments at pilot plant
 scale. Value of HtoR in packed tower depends on type of system, type of
packing, size of packing, tower diameter, etc. Hence, it is difficult to find the
suitable correlation for finding HtoR for the given system.
For example, equation for finding HtoR for acetic acid-water-MIBK system, in
which water rich phase is the raffinate phase and MIBK rich phase is the
extract phase, is given by equation1.

HtoR =   (7.33)

where, HtoR = Height of equivalent overall transfer unit based on raffinate phase,
m
vD = Actual superficial velocity of dispersed phased, m/s
vC = Actual superficial velocity of continuous phase, m/s
This equation was derived by Sherwood et al.1 for the packed tower having 0.09
m diameter, 12.7 mm carbon rings at 25oC for the certain range of flow rates.
Hence, the use of this equation for the same system with larger diameter
packed tower, different type and size of packing and for different values of
flow rates, may not be reliable. So the equation of HtoR or HtoE should be
determined on the same system at pilot plant scale with similar type of
equipment and with similar conditions like same values of superficial
velocities, same operating temperature, etc. However, Eq. (7.33) may be used
for the preliminary design of the pilot plant equipment.
3. Criteria of selection: If number of theoretical stages required for desired
separation is less and limited floor area is available in the plant than packed
tower can be selected as extractor. Packed tower consumes much less power
than mixer settler.
Another method for finding height of packing for the packed-type extractor or
height of other continuous extractor, based on HETS (height equivalent to
theoretical stage), is as follows. The height of extraction zone cane be
determined by the following equation:
H = HETS × N  (7.34)
where, H = Height of extraction zone, m
N = Number of theoretical stages required for desired separation
HETS value can be selected from Table 7.6.
Table 7.6 Approximate HETS Values of Extractors
 Type of Extractor HETS, m Superficial velocity of continuous and dispersed phase, m/h

Sieve tray tower 0.5–3.6 27–75

Packed tower – Pall rings 0.6–1.5 12–40
Packed tower – Structured packing 0.6–1.3 65–90
Rotating disc contactor 0.3–0.5 7–30
Scheibel extractor 0.2–0.4 6–14
Pulsed sieve column 0.8–1.3 25–35

For the design of pilot plant extractor, Table 7.6 and Eq. (7.34) can be used.
However, the same cannot be used directly for the design of commercial scale
extractor.
For the extraction system described in Example 7.1, design the
packed tower-type extractor.
(a) Tower Diameter:
Let type of packing: SS Pall rings
Size: 16 mm (for first trial calculations)
From Table 9.1 of Chapter 9,
ap = specific packing surface = 340 m2/m3
ε = fraction of void volume = 0.92
To find the flooding velocities, first the value of following function must be
determined:
f =   (7.35)

where, σC = Surface tension of continuous phase, lbf/ft

From Example 7.6, volumetric flow rate of feed = 0.198 m3/h
Volumetric flow rate of final extract = 0.5337 m3/h.
Volumetric flow rate of extract is very much greater than the same of feed, hence
extract is the continuous phase and feed (heavy phase) is the dispersed phase.
Figure 7.16(b) represents the system to be adopted.
Final extract contains 89.2% MIBK, 8.8% acetic acid and 2% water by mass.
Surface Tension of MIBK:
Structure of MIBK:
(Molar mass = 12 × 6 + 16 + 12 = 100)
For hydrogen bonded compound, Sugden equation4 can be used to predict
surface tension of a pure compound.
σ 1/4 = [P] (ρL – ρG)  (7.36)
where, σ = Surface tension, dyn/cm
ρL = Liquid density, mol/cm3
[P] = Temperature independent parachor value
For MIBK,
ρL = 801 kg/m3
ρG =

ρG = 4 kg/m3

ρL = = 0.008 mol/cm3

ρG = = 4 × 10–5 mol/cm3

From Table 3.343, Ref. 4,

[P] = 171.9 + 55.5 + 29 = 256.4
σM1/4 = 256.4 (0.008 – 4 × 10–5) = 2.0409
σM = 17.35 dyn/cm
For acetic acid,
ρL = 1049 kg/m3
ρG = = = 2.41 kg/m3

ρL = = 0.017 48 mol/cm3

ρG = = 4.016 × 10–5 mol/cm3

[P ] = 55.5 + 73.8 = 129.3
σA1/4 = 129.3 (0.01748 – 4.016 × 10–5)
σA = 25.86 dyn/cm
For water, σW = 71.4 dyn/cm (Ref. 4)
Surface tension of liquid mixture4 is given by the following equation:
= Σσi1/4xi  (7.37)
where, xi = mole fraction of component i in liquid phase

xMIBK = = 0.7758

xAcOH = = 0.1276

xH2O = 1 – 0.7758 – 0.1276 = 0.0966

σC1/4 = 17.351/4 × 0.7758 + 25.861/4 × 0.1276 + 71.41/4 × 0.0966
σC = 21.443 dyn/cm
≡ 0.021 443 N/m
From Example 7.6,
ρC = 821.36 kg/m3, μC = 0.5294 × 10–3 kg/(m · s)
gc = 1, Δρ = ρD – ρC = 1009.4 – 821.36 = 188.04 kg/m3
To use Fig. 7.20 (Crawford – Wilke correlation), all variables must be in US
customary units.
σC = 0.021 443 × × = 1.47 × 10–3 lbf/ft

gc = 4.18 × 108 lb · ft/(lbf · h2)

ρC = 821.36 × = 51.276 lb/ft3

μC = 0.5294 × 10–3 × = 1.28 lb/(ft · h)

Δρ = 188.04 × = 11.739 lb/ft3

ap = 340 m2/m3 = 340 × = 103.632 ft2/ft3

ε = 0.92

f = = 852.81

From Fig. 7.27 and from Example 7.6,

= 175

= = = 452.72 ft/h

qC = 1.4825 × 10–4 m3/s

qD = 5.503 82 × 10–5 m3/s

= = = 2.6936

= 452.72

= = 0.038 33

vDF = 0.0055 m/s and vCF = 0.0148 m/s

Let actual velocities be 50%.
vC = 0.5 vCF and vD = 0.5 vDF
vD = 0.002 75 m/s, vC = 0.0074 m/s

vC =

= = 0.02 m2

Di = 0.16 m
Let Di = 0.2 m
Total flow in the packed column = qC + qD
= 2.0329 × 10–4 m3/s
= 0.731 84 m3/h
Cross-sectional area of column = (0.2)2 π/4
= 0.031 42 m2
Superficial velocity (capacity) = 0.731 84/0.031 42
= 23.3 m3/(m2 · h)
Based on curve FK in Fig. 7.21, this is acceptable.
Revised values of actual velocities based on final value of inside diameter.
vC = 0.0047 m/s, vD = 0.001 75 m/s
Height of packing:
Total height of packing H = NtoR · HtoR
NtoR = 7.213 55 (From Example 7.3)
HtoR for the given system
HtoR = 0.2347 (vC /vD)0.648
where, HtoR is in m.

HtoR = 0.2347 = 0.4452 m

H = 7.213 55 × 0.4452 = 3.211 m

Results: Type of packing: 16 mm, SS Pall rings, inside diameter of tower: 0.2 m,
total height of packing section = 3.25 m
How to find HtoR or HtoE from experiment (Ref. 12, 13)
To find the HtoR from the experiment at pilot plant or at lab scale, one can use
the following equation for the packed tower-type extractor:
HtoR = =   (7.37)

where, KRa = Overall mass transfer coefficient based on raffinate phase

=

ΔClm = Logarithmic mean driving force

Similarly, HtoE =   (7.39)
KEa =

In the extraction of acetic acid from an aqueous solution by

benzene in a pilot plant scale packed tower type extractor, following data were
measured:
(i) Cross-sectional area of tower = 0.0045 m2
(ii) Height of packing = 1.4 m
(iii) Concentration of acetic acid in feed = 0.69 kmol/m3
(iv) Concentration of acetic acid in final raffinate = 0.684 kmol/m3
(v) Flow rate of extract phase = 5.6 × 10–6 m3/s
(vi) Concentration of acetic acid in solvent = 0.004 kmol/m3
(vii) Concentration of acetic acid in final extract = 0.0115 kmol/m3
Determine HtoE from above data.
Equilibrium relationship for this system13 is C*B = 0.0247 Cw,
where C*B, Cw = concentration of acetic acid, kmol/m3
Solution
Total acetic acid transferred to extract (benzene) phase
= qVE (CB2 – CB1)
= 5.6 × 10–6 (0.0115 – 0.004) = 4.2 × 10–8 kmol/s
Equilibrium concentration of acetic acid in benzene phase at inlet and outlet,
C*B1 = 0.0247 × 0.684 = 0.0169 kmol/m3
C*B2 = 0.0247 × 0.69 = 0.0171 kmol/m3
Driving force at bottom, ΔC1 = C*B1 – CB1 = 0.0169 – 0.004 = 0.0129 kmol/m3
Driving force at top, ΔC2 = C*B2 – CB2 = 0.0171 – 0.0115 = 0.0056 kmol/m3
Logarithmic mean driving force

ΔClm = =

= 0.008 748 kmol/m3

HtoE =

VE = =

= 1.244 × 10–3 m/s

KEa =

= = 7.62 × 10–4 s–1

HtoE = = 1.6325 m

For the extraction system described in Examples 7.4 and 7.5,

design the packed tower-type extractor.
Solution
Feed flow rate F = 1000 kg/h. It contains 80% DMF and 20% water.
At steady state it is contacted with final extract E1.
E1 = B (1 + Y′1)
E′1 = 3070.72 (1 + 0.0625) = 3262.64 kg/h

Y1 = = = 0.0588

ρF =   (7.30)

ρDMF = 944.5 kg/m3, ρwater = 1000 kg/m3

ρF = = 955.1 kg/m3

ρCH2Cl2 = 1336 kg/m3 ρDMF = 944 kg/m3

ρE = = 1304.2 kg/m3

qvF = = = 1.047 m3/h

qvE = = = 2.502 m3/h

Tower Diameter
Let type of packing: SS Pall rings, size = 25 mm
ap = surface area of packing = 205 m2/m3
ε = fraction of void volume = 0.94
To find the flooding velocities by Crawford – Wilke correlation, first the
following function must be determined:

f =   (7.34)

σC1/4 = Σxi σi1/4   (7.37)

σDMF = 35.2 dyn/cm
Molar mass of DMF, M = 73.09
For pure DMF, σ1/4 = [P] (ρL – ρG) (Sudgen’s equation)
ρL, ρG in mol/cm3
[P] = 40 + 55.2 × 2 = 150.4 (Table 3–343, of Ref. 3)
ρL = 1336 kg/m3
ρG = = = 3.474 kg/m3

ρL = = 0.0157 mol/cm3

ρG = = 4.087 × 10–5 mol/cm3

σ1/4CH2Cl2 = 150.4 (0.0157 – 4.087 × 10–5) = 2.355

σCH2Cl2 = 30.765 dyn/cm
Mole fraction of DMF in extract phase,
x1 =

σ 1/4 = x1 σ1/4DMF + (1 – x1)σ1/4CH2Cl2   (7.37)

σ C1/4 = 0.0677 × 35.21/4 + (1 – 0.0677) × 30.7651/4
σC = 31.05 dyn/cm ≡ 0.031 N/m
For organic mixtures12,
=

μDMF = 0.802 mPa · s, μCH2Cl2 = 0.6 mPa · s

μC = 0.61 mPa · s ≡ 0.61 × 10–3 kg / (m · s)

Δρ = ρc – ρd = 1304.2 – 955.1 = 349.1 kg/m3
To use Fig.   (7.27), all variables must be in US customary units.
σC = 0.031 05 × = 2.128 × 10–3 lbf/ft

gc = 4.18 × 108 lb . ft / (lbf · h2)

ρC = 1304.2 × = 81.42 lb/ft3

μC = = 1.4756 lb/(ft · h)

Δρ = 349.1 × = 21.7936 lb/ft3

ap = 205 m2/m3
= 205 × = 62.484 ft2/ft3

ε = 0.94

f =   (7.35)
=

= 235.66
From Fig.   (7.27),
= 600

= = 679.45

= =

= = 2.39

= 679.45
6.482 vDF = 679.45
vDF = 104.82 ft/h and vCF = 250.52 ft/h

or vDF = 104.82 × × = 0.009 m/s

and vCF = 0.0215 m/s

Let vD = 0.5 vDF = 0.0045 m/s,
vC = 0.5 vCF = 0.010 75 m/s

= = 0.065 m2

Di = 0.288 m
Total volumetric flow (capacity) of tower = 2.502 + 1.047 = 3.549 m3/h
Superficial flow rate = 3.549/0.065 = 54.6 m3/(m2 · h)
In Fig. 7.21, area FK represents random packed columns. Maximum superficial
capacity can be read as 30 m3/(m2 · h) which is less than calculated value 54.6
m3/(m2 · h). Hence, packed column diameter will have to be increased to achieve
30 m3/(m2 · h).
Cross-sectional area required of the column = = 0.1183 m2

Column diameter, Di = 0.3881 m

Let Di = 0.4 m = 400 mm
Since column diameter is increased, there is no need to check superficial
velocities, etc.
Packing height:
H = NtoR × HtoR
NtoR = 5.345
Suitable correlation for HtoR is not available in open literature for the given
system. It must be determined by pilot plant study. Here HtoR is determined by
the same equation that is derived for acetic acid – water – MIBK system.

HtoR = 0.2347

= 0.2347

HtoR = 0.41 m
H = 5.345 × 0.41 = 2.19 m
Let H = 2.5 m
Resulting data:
Type of packing: 25 mm SS Pall rings
Inside diameter of tower = 300 mm
Height of packing section = 2.5 m
Feed density (955.1 kg/m3) is lower than solvent density (1336 kg/m3). Also
feed flow rate (1000 kg/h) is lower than solvent flow rate (1535.36 kg/h). Hence,
lighter phase is dispersed in heavier phase. System shown in Fig. 7.16(a) can be
adopted.

7.7 SUPERCRITICAL EXTRACTION (SCE)

In liquid-liquid extraction solvents like water, ethylene glycol, chloroform,
hexane, ethyl acetate, etc., are contacted with feed mixture to separate the
desired product. Desired product is transferred to liquid solvent. Then desired
product is to be separated from solvent. In short by liquid-liquid extraction from
a given feed solution one can make a new solution (solvent + solute). If
separation of new solution is easier than the same of the given solution, then
extraction is called useful extraction.
Separation of solute from solvent is normally done by distillation. It consumes
large amount of energy. Also it is difficult to get 100% pure solute; absolutely
free from solvent by distillation.
In supercritical extraction (Ref. 14 to 21), supercritical fluid is used as a solvent.
In Sections 7.1 to 7.6, it was discussed that the fluid can exist as liquid or gas.
But in addition to that a fluid can also exist as supercritical fluid (known as
SCF).
Supercritical fluid is any fluid at a temperature higher than critical temperature
and at a pressure higher than critical pressure.
In Fig. 7.28 and Fig. 7.29, phase diagrams of carbon dioxide and water are
given. Supercritical regions for both the compounds are shown in the figures.
7.7.1 Properties of Supercritical Fluid Solvents
(SCFS) (Ref. 13, 14, 22)
Physical properties of supercritical fluids are between those gaseous and liquid
states. Typical values of important physical properties are listed in Table 7.7.
Table 7.7 Comparison of Physical Properties of Gas, Liquid and Supercriitical Fluids22
Physical property Gas SCF Liquid
3 0.6 – 2 200 – 500 600 – 1000
Density, ρ, kg/m
Dynamic viscosity, μ, mPa · s 0.01 – 0.3 0.01 – 0.03 0.2 – 3
Kinematic viscosity, γ × 106, m2/s 5 – 500 0.2 – 0.1 0.1 – 5

Thermal conductivity, k, W/(m · K) 0.01 – 0.025 High* 0.1 – 0.2

Diffusion coefficient, D × 105, m2/s 10 – 40 0.07 – 0.08 0.0002 – 0.002

* Thermal conductivity values are high in the near critical region.

1. In supercritical state, density of fluid is just like liquid density and hence
solvent power for extraction is also equivalent to liquid solvents. In
supercritical region density of the fluid is sensitively changing with pressure.
Slight change in pressure creates a dramatic change in density.
2. Viscocities of SCFs are equivalent to viscocities of gases. Hence, they
consume less power for transportation. Also they provide lesser resistance to
heat transfer and mass transfer.
3. Molecular diffusivities of SCFs are intermediate to liquids and gases.
DL is in the order of C × 10–9 m2/s (DL = diffusivity of liquid).
DG is in the order of C × 10–5 m2/s (DG = diffusivity of gas).
Where C is a numerical value.

Fig. 7.28 Phase Diagram (p-t Diagram) of Carbon dioxide

Based on Data Contained in Thermodynamic Properties of Carbon dioxide, by: R. Span and W. Wagner, J.
Phys. Chem. Ref. Data, 25(6), p. 1509, 1996.

Advantages and disadvantages of SCF solvents over liquid solvents

Advantages
1. SCF solvents are easily and completely separated from extracted products.
This means SCF solvents give the product, nearly 100% free from solvent.
2. Reduction in pressure or increase in temperature separates the SCF from
extracted products. Hence, they require lesser energy for separation. Thus
 with SCE, near 100% recovery of solvent from extracted product is possible
with low cost, compared to liquid-liquid extraction with liquid solvents.
3. SCF solvent has very low viscosity and higher diffusivity compared to liquid
solvent. Hence, it provides lesser resistance to mass transfer and also it
provides lesser friction during transportation.
4. SCFs provide very high value of selectivity, compared to conventional liquid
solvents.

Fig. 7.29 Phase Diagram (p-t Diagram) of Water

 Based on Data Contained in
1. Vacuum Technology, Its Foundations, Formulae & Tables, by Leybold AG, Germany, 1987.
2. 1980 SI JSME STEAM TABLES, Japan.
3. Publication No. NRISTIR 5078 of National Institute of Standards & Technology, USA, 1998.

Selectivity:
α =

α =

Extraction is possible only if α > 1. If α >> 1, it implies that lesser number of

theoretical stages are required for desired separation. Hence, with SCF, lesser
number of theoretical stages are required for desired separation. Consequently
they require lesser height of extractor as compared to conventional liquid
solvents.
5. With common SCFs like CO2, solvent recovery is possible at room
temperature or at temperature close to the room temperature.
6. Common SCFs like CO2 is easily available with relatively low price as
compared to conventional liquid solvent.
7. SCFs like CO2 (dry) is non-corrosive and non-toxic in nature.
Disadvantages
1. It is found by experiments that supercritical fluids, provide poor value of
distribution coefficient for extraction as compared to conventional liquid
solvents.
Distribution coefficient,
m =   (7.1)

Hence, solvent/feed ratio, required with SCFs, is very much higher than
conventional liquid solvents.
2. Except the few solvents like CO2, for other solvent like hexane, ethylene
glycol, etc., it is very difficult to achieve the supercritical state of fluid due to
the practical limitations. Choice of solvent is narrow with SCFs.
3. Certain SCFs are flammable (like propane, butanes, etc.) and hence, adequate
safety measures are required in design.
Properties of CO2, H2O and n-hexane in supercritical region are given in Ref.
23. Properties of ethane, propane and butanes in supercritical region can be
found in Ref. 24.
Table 7.8 List of Selected Supercritical Solvents and their Properties25
Fluid tc (°C) pc (bar) Remarks

Carbon dioxide, (CO2) 31.00 73.77 Most Favourable

Dinitrogen monoxide, (N2O) 36.45 72.40 Not stable

Xenon, (Xe) 16.60 58.68 High price, poor availability

Ethane, (C2H6) 32.27 48.80 Inflammable

Ethylene, (C2H4) 9.19 50.39 Inflammable

Water, (H2O) 374.15 221.20 High temperature and pressure operation, corrosive, high latent heat of vaporisation

Propane (C3H8) 96.67 42.47 Inflammable

n–Butane (C4H10) 152.03 37.97 Inflammable

i–Butane (C4H10) 135.00 36.48 Inflammable

Among all these solvents, CO2 and propane are used in commercial plants.
Water is also a useful SCF but requires a very high pressure operation.
CO2 is most preferred as SCE solvent. Reasons are:
(a) It has low critical temperature (tc = 31.0oC).
(b) It has relatively low critical pressure (pc = 73.77 bar).
(c) It is non-toxic and non-corrosive in nature (when dry).
(d) It is easily available at low cost. It is available as a byproduct from many
industries; such as fermentation of molasses, ammonia plant, etc. It is present
in substantial amount in atmosphere as green house gas.
Advantages and disadvantages of SCE over conventional unit operations
Advantages
1. It consumes lesser energy compared to liquid-liquid extraction and distillation,
combined together. Hence, it requires lesser operating cost as compared to
extraction and distillation. It requires only electricity as a utility to drive a
compressor and a positive displacement liquid pump. With CO2 or N2O,
chilled water or brine is required for condensation.
2. Compared to the liquid-liquid extraction it gives purer products, nearly 100%
free from solvent or foreign impurities.
3. SCF provides very high value of selectivity as compared to conventional
liquid solvents. Hence, supercritical extractor requires lesser number of
theoretical stages and lesser height, compared to liquid-liquid extractor.
4. Compared to (atmospheric) distillation, supercritical extraction can be
 operated at lesser temperature, closer to room temperature. Liquid-liquid
extraction requires higher temperature in solvent recovery step. This
advantage is beneficial particularly for heat sensitive products.
5. Products obtained by biotechnology route is usually a dilute solution (< 10%)
in water. For the separation of desired product from water by distillation
requires very high amount of energy and operating cost. While for the same
separation, SCFE consumes very less energy and has low operating cost.
Several examples of herbal and heat sensitive products are given in Ref. 21 with
the use of CO2 as SCF.
One research paper mentions that operating cost required to separate ethanol
from its dilute solution by distillation is 5 times higher than the same by
SCFE while fixed cost of distillation unit could be half that of the SCFE unit.
Disadvantages
1. It requires higher capital investment as this operation is carried out at high
pressure. Fixed cost of compressor is relatively higher than cost of other
equipments. Also the cost of the extractor is high. There are only a few
suppliers of the technology and equipments of the plant of SCFE at present.
Hence, the specialisation of technology further increases fixed cost of plant at
present.
2. It provides very low value of distribution coefficient hence large amount of
solvent is required for separation.
3. Limited experience with its use in pilot and commercial applications.
However, now pilot studies with CO2 as SCF are available at certain research
institutes with sufficient data for scale-up.
4. Limited knowledge of design data. Also thermodynamics of supercritical
system is not fully understood till date.
7.7.2 Flow Sheet of Batch Supercritical Extraction
(Ref. 14, 15)
Figure 7.30 is the flow sheet of SC Extraction plant which is operated in
batchwise manner. Operational steps are as follows.
1. In bath extraction, feed is initially charged before starting carbon dioxide
circulation.
2. In CO2 vessel, CO2 is in compressed gas form but p < pc and temperature is
equal to room temperature.
3. From CO2 vessel, compressed gas passes through chilled water condenser or
 brine condenser where CO2 is liquefied.
4. Liquefied CO2 is pumped by metering pump (a positive displacement pump)
to high pressure. At the discharge of pump, CO2 is liquid with p > pc and t <
tc.
5. Compressed liquefied CO2 which is in liquid state at p > pc and at temperature
less than tc is converted to SCF by increasing the temperature above critical
temperature.
6. Supercritical CO2 is contacted with feed which was initially charged in the
batch extractor. Sufficient contact time is allowed to reach equilibrium
7. From the top of the extractor, extract phase (SC CO2 + solute) is obtained. It is
heated to the temperature at which soluble solute is separated into insoluble
solid or liquid.

Fig. 7.30 Flow Sheet of Batch SCE Plant

8. Then pressure is decreased and SC CO2 is converted to gaseous CO2 with p <
pc and at temperature which is close to tc.
9. CO2 is circulated till the desired degree of extraction is obtained. A level
indicator can be provided on the separator.
10. Once desired degree of extraction is obtained, system is depressurised.
Raffinate phase is separated from the extractor at atmospheric pressure. New
feed then can be charged to the extractor.
7.7.3 Application of SCE (Ref. 13 to 20)
SCE finds its application in food, pharmaceutical, petroleum refineries, solid
fuel industries and polymer industries. Extensive research work is going on other
areas of applications.
1. Established Commercial Applications
(a) Decaffenation of coffee
(b) Deasphalting of oil
(c) Extraction of essence from fragrant materials
(d) Separation of aromas and oils from spices
(e) Spray painting and coating
(f) Chromatographic analysis of polymer
(g) Supercritical fluid polymerisation, e.g., manufacturing of
polytetrafluoroethylene
(h) Extraction of herbs
(i) Purification of active pharmaceutical intermediates (API)
(j) Isolation of eico sapentanoic acid (EPA) from fish oils
(k) Purification of a vitamin intermediate
(l) Purification of a reactive multifunctional acrylic monomer
(m) Extraction of cholesterol from butter
(n) Extraction of oligomers from high vacuum sealants.
(o) Extraction of sterols from deodorised distillate of edible/non-edible oils.
2. Applications under Research and Development
(a) Extraction of various chemicals from coal and wood
(b) Extraction of organic chemicals from its dilute solution in water. (This
application is useful in biotechnology route and also in the recovery of
chemicals from wastewater stream)
(c) In separation of liquid hydrocarbon mixtures which are difficult to separate
by conventional techniques
(d) Separation of aromatics from waste solid material such as tobacco dust
(e) Regeneration of adsorbent materials
(f) Extraction of oil from oil containing conventional and unconventional seeds,
 e.g., soybeans, ground nuts, jojoba beans, coriander, etc.
(g) Deodorisation of vegetable and animal oils.
(h) Extraction of oil from potato chips
(i) Recovery of citrous oil from peelings
(j) Extraction of anticancer drugs and other substances from plants (herbs)
(k) Refining of pyrethrum to obtain pyrethrin insecticides
(l) Destruction in hazardous waste and production of biocrude from organic
waste (water is used as SCF)
(m) Extraction of chemicals from used automotive tires
(n) Separation of monoglycerides from the mixture of glycerides
(o) Paper deacidification and preservation
(p) Plastic recycling
(q) Textile dyeing
(r) Generation of nano particles
(s) Supercritical phase oxidation
Many other applications are under R&D and can be found from literature.
7.7.4 Use of Cosolvents and Surfactants (Ref. 16)
One of the major limitations of SCF is availability of very few suitable solvent.
Most suitable solvent is supercritical CO2 which is non-polar in nature.
Many non-volatile polar substances cannot be dissolved easily in non-polar SC
CO2. Most of the organic compounds are polar in nature, e.g., ethanol, ethylene
glycol, etc.
Polar solvents have high critical temperature. Hence, they involve very high cost
in extraction. Also other problems like thermal cracking may take place of such
solvent as SC solvent results in its rejection. It is found by experiment that an
addition of small amount of polar liquid solvent (1 to 10 mole %) in SC CO2,
dramatic improvement in the solubility can be achieved.
Take for example case of hydroquinone which dissolves very little in pure SC
CO2. Addition of 2.8 mole % methanol increases hydroquinone’s solubility by a
factor of 6. Addition of 2 mole % tributyl phosphate (TBP) increases
hydroquinone’s solubility by a factor of 250.
Hence, TBP is much more effective than methanol as cosolvent. Selection of the
correct cosolvent requires detailed knowledge of chemistry and molecular
structure of polar and non-polar compounds. Hence, selection of right cosolvent
is very important. Cosolvent is also known as modifier.
7.7.5 Few Commercial Applications
It will be interesting to study two case studies of supercritical
extraction/processing.
Decaffeination of coffee (Ref. 18) Coffee industry is an important food grade
industry. With industrialisation and increasing demand of refreshers, there has
been an increase in consumption of coffee in most of the parts of world. Coffee
seeds contain 1–3% caffeine. Consumption of large dosage of caffeine produces
adverse effect on human body, but at moderate level no major harmful effects are
observed.
Pure caffeine is a white powder, soluble in water and various organic solvents.
Hot water is an excellent solvent for caffeine but the same cannot be selected as
a solvent for extraction as other ingradients of the coffee beans are also soluble
in hot water while caffeine is selectively soluble in SCF CO2. There are many
plants in the world which extract caffeine from raw coffee. One of them is coffee
decaffeination plant in Houston, USA, built by Kraft General Foods Ltd. The
process, used by Kraft, is shown in Fig.   (7.31).

Fig. 7.31 Decaffeination Process

Process description The moist green coffee beans are charged to an extraction
vessel having 7 ft (2.13 m) diameter and 70 ft (21.33 m) height. Here moisture of
water act as modifier. Carbon dioxide is dried and charged from the bottom of
extractor in supercrifical state. From supercritical extract, caffeine is then
recovered by countercurrent water wash, carried out nearly at the same pressure.
The depleted carbon dioxide is dried and recycled to the extractor. Caffeine rich
aqueous stream is concentrated by reverse osmosis.
The Lipton tea decaffeination process is similar to Kraft process (in Germany)
except that the caffeine is removed from the carbon dioxide by passing the
solution through a bed of activated carbon.
The caffeine recovery from the extract is the key part of the process. As the
decaffeinated green coffee is finally contacted with solvent, equilibrium
distribution of caffeine between beans and solvent decide the percentage of
unrecovered caffeine in the beans.
If caffeine concentration in the solvent (SC CO2) is reduced to 0.002 mass % in
recovery step, then it gives equilibrium concentration of caffeine remaining in
the processed beans to about 0.16 mass %.
Hence, it is very much necessary to remove caffeine from extract phase as much
as possible. One way is to reduce the pressure of SC extract to atmospheric
pressure, but it is not an economical option.
It is suggested not to try to reduce the concentration of caffeine from coffee
below 0.16% as it is not economical. Hence, the selection of recovery process
and maximum extraction of caffeine is very important.
ROSE process (Deasphalting of oil) (Ref. 25) Solvent extraction has been used
to separate valuable components from heavy oil and residuals from less desirable
material. One of them is propane deasphalting. In the refinery, bottom product of
vacuum distillation column is known as residuum. Propane deasphalting is used
to separate the residuum into three fractions; (i) lube oil (ii) resin, and (iii)
asphalt.
 Fig. 7.32 ROSE Process

It uses liquid propane as solvent to extract lube oil and resin from residuum and
to reject high molar mass asphalt, as reffinate. Solvent recovery in conventional
propane deasphalting consists of series of flash vessels and steam stripping.
Entire propane is recovered as vapour at lower pressure. It is converted into high
pressure liquid propane by compression and liquefaction and recycled back to
extractor. Thus, energy consumption in conventional propane deasphalting is
very high and it demands the use of new technique. The ROSE (Residuum Oil
Supercritical Extraction) process is a new process developed by Kerr-McGee’s
R&D group. ROSE process produces identical products from residuum (lube oil,
resin, and asphalt) but it consumes less energy.
Details of three products obtained from residuum are as follows:
1. Asphalt: It is the heaviest, most viscous and the most hydrogen deficient
fraction of crude petroleum oil. At room temperature it is a dark brown to
black friable solid. Its ring and ball softening point ranges from 70 to 200°C.
Carbon to hydrogen ratio is 9 to 11.6 by mass. It is extremely difficult to
remove sulphur, nitrogen or metals from asphalt. Hydrotreating and cracking
catalysts are rapidly coked, deactivated and poisoned by asphalt.
2. Resin: It is an intermediate fraction, lighter than asphalt and heavier than
deasphalted oil or lube oil. Resin can be hydrotreated, hydrocracked,
thermally-cracked or even coked to produce more volatile products. Resins
are less viscous, low in metals, sulphur and carbon. Resins have less C/H
ratio than asphaltene.
3. DAO or Lube Oil: Deasphalted oil (DAO) or lube oil is a clean product
containing substantially lower metals and carbon residue than either asphalt
or resin. DAO can be sent for hydrotreating, hydrocracking or catalytic
cracking, without any further purification.
Process description The residuum is charged to extractor where it is contacted
with liquid propane at elevated pressure. In ROSE process also, extraction is
liquid-liquid extraction, similar to conventional propane deasphalting. It is not
the supercritical extraction. As an extractor, either Mixer-settler or Rotating Disc
Contractor (RDC) is used. From the extractor, asphalt is obtained as raffinate
phase, but it is contaminated by very less amount of propane. Propane vapour is
separated from raffinate phase by flashing and stripping in tower T1 which is a
flash vessel cum stripper.
Extract phase, containing large amount of liquid propane, resin and lube oil, is
passed through heater E1 where it is heated to the temperature at which resin
becomes insoluble and is separated from the extract phase. Propane rich layer
and resin layer are separated in decanter D1. Traces of propane from resin layer
is separated by flashing and stripping in the tower T2.
The remaining propane-oil solution is heated to higher temperature by two
heaters E2 and H2 in series. At this temperature liquid propane is converted into
supercritical propane. DAO or lube oil is not soluble in supercritical propane,
hence it is separated from propane in decanter D2. Traces of propane from DAO
or lube oil is separated by flashing and stripping in tower T3. Supercritical
propane from D2 is passed through heat exchangers E2 and E1, where its sensible
heat is utilised for heating. Then in heat exchanger C2 it is liquefied. Liquefied
propane is recycled back to extractor by pump P1, which develops only
differential pressure necessary to overcome system pressure drop.
Small amount of propane vapour, stripped off from the towers T1, T2 and T3, is
condensed in condenser C1. Condensate is sent to tank S1. This intermediate
hold-up tank is maintained at lower temperature and at lower pressure, where
make-up liquefied propane is also added. Small amount of propane from this
tank (7 to 15% of the total required in extractors) is pumped by pump P2 and
sent to suction line of pump P1. Pressure difference around pump P2 is quite
higher than the same around pump P1.
Thus in ROSE process, extraction is carried out in subcritical region or as liquid-
liquid extraction. Only properties of supercritical fluid are utilised in solvent
recovery step.
Advantages of Rose process over conventional propane deasphalting are as
follows.
1. 85 to 93% of the extraction solvent is recovered as supercritical fluid in the oil
separator without evaporation and pressure reduction. Also, the concept of
energy integration is applied in two heat exchangers (E1 and E2) to save the
utilities. Hence, energy required by ROSE process is substantially lower than
the same required for conventional propane deasphalting.
2. ROSE process requires less maintenance cost.
3. Increase in propane to feed ratio improves the quality of products. With ROSE
process, higher solvent to feed ratio is affordable, hence it results in better
quality of products.
ROSE process is competitive, well established and commercially proven
technology. It is licensed by more than 15 companies worldwide.
Decaffeination of coffee and ROSE process were discussed in detail to bring
forth the process engineering problems and heat integration aspects. For
development of other solvent extraction systems, process engineer should
consider similar aspects so that an efficient system can be designed. Useful
guidelines for process design of SCF extractor obtained from Ref. 21.
Conclusion
In the early stages of development of the Supercritical Fluid Extraction, the
requirement of expensive equipments, lack of proper and appropriate techniques
were the main hurdles of adoption and commercialisation of the technology.
But now in the changed scenario of improvement in the SCE techniques,
increasing cost of energy and increasing cost of conventional solvents makes this
unit operation more feasible and attractive. Still there is a large scope for the
development of supercritical extraction technology. More research work is
required to improve the distribution coefficient of SCE system. In future, it is
certain that this unit operation will be applied in many processes.

EXERCISES
7.1 A feed stream having flow rate of 1000 kg/h and containing 40 mass %
styrene and 60 mass % ethyl benzene is to be extracted with ethylene glycol
as solvent. Extraction is to be carried out in multistage countercurrent
extractor. Final raffinate is to be extracted down to 1% styrene. Determine (a)
minimum amount of solvent required for the desired separation and (b)
number of theoretical stages required for desired separation if the actual
amount of solvent is 5000 kg/h.
Table 7.9 Liquid-liquid Equilibrium Data4 of Ethyl Benzene-Styrene- Ethylene Glycol at 25°C
mass % in Raffinate mass % in Extract
Ethyl benzene Styrene Ethylene glycol Ethyl benzene Styrene Ethylene glycol

90.56 8.63 0.81 9.85 1.64 88.51

80.4 18.67 0.93 9.31 3.49 87.2
70.49 28.51 1.00 8.72 5.48 85.8
60.93 37.98 1.09 8.07 7.45 84.48
53.55 45.25 1.2 7.35 9.25 83.4
52.96 45.84 1.2 7.31 9.49 83.2
43.29 55.32 1.39 6.3 12.00 81.7
41.51 57.09 1.4 6.06 12.54 81.4

(From Table 15.2 of Ref. 4)

7.2 For the extraction described in Exercise 7.1, if a packed tower is used as
extractor then (i) find the value of NtoR for S = 5000 kg/h, and (ii) determine
the tower diameter.
7.3 A feed mixture containing 60 mass % methyl cyclohexane and 40 mass % n-
heptane is contacted with pure aniline as solvent. Extraction is carried out in
countercurrent manner in four mixer-settler vessels. Stage efficiency of
mixer-settler may be assumed equal to 1. Solvent to feed ratio (mass basis) is
kept 4. Feed flow rate is 200 kg/h. Find the resultant product flow rates and
product compositions.
Table 7.10 Equilibrium Data at 25°C
Hydrocarbon phase, mass % Aniline rich phase, mass %
Methyl cyclohexane n-Heptane Methyl cyclohexane n-Heptane

0.0 92.6 0.0 6.2

9.2 83.1 0.8 6.0
18.6 73.4 2.7 5.3
22.0 69.8 3.0 5.1
33.8 57.6 4.6 4.5
40.9 50.4 6.0 4.0
46.0 45.0 7.4 3.6
59.0 30.7 9.2 2.8
67.2 22.8 11.3 2.1
71.6 18.2 12.7 1.6
73.6 16.0 13.1 1.4
83.3 5.4 15.6 0.6
88.1 0.0 16.9 0.0

(Ref. 8, p. 325)
(Hint: Assume n-heptane and aniline as completely immiscible liquids to simplify the calculations.)

7.4 Design a mixer settler for Exercise 7.3.

Residence time required in mixer to achieve equilibrium = 8 min
Type of agitator of mixer = 45° pitched blade turbine
Tip velocity of agitator = 200 m/min
7.5 A feed mixture containing 20 mass % acetic acid and 80 mass % water is
contacted with nearly pure butyl acetate as solvent to recover acetic acid.
Extraction is carried out in a packed tower. Feed flow rate is 1000 kg/h. Final
raffinate leaving the packed tower should contain less than or equal to 1% (by
mass) acetic acid. Determine the following:
(i) Minimum amount of solvent required for desired separation.
(ii) Value of NtoR for actual amount of solvent S = 2 Sm.
(iii) Tower diameter required for packed tower.
(iv) Determine the height of packing. Assuming that Eq. (7.32) is valid for acetic
acid-water-butyl acetate system for finding HtoR.
Data:
Table 7.11 Liquid-Liquid Equilibrium Data (mass %) at 25°C
Solvent rich layer Water rich layer
Water Butyl acetate Acetic acid Water Butyl acetate Acetic acid

1.8 94.8 3.4 90.5 0.8 8.7

3.7 84.6 11.5 78.9 1.1 20.0
4.7 82.0 13.3 75.6 1.2 23.2
7.8 73.0 19.2 68.0 1.6 30.4
8.9 68.0 23.1 65.0 2.3 32.7
14.7 56.0 29.3 56.3 5.6 38.1

All percentages in Table 7.11 are mass %.

7.6 A feed stream having flow rate of 400 kg/h and containing 30% ethanol in
water is to be extracted at 25°C with 600 kg/h of recycled benzene that
contains 0.5% ethanol. The aqueous raffinate is to be extracted down to 1%
ethanol. How many theoretical stages will be required? Extraction is to be
carried out in counter manner in a packed tower. Design the extractor.
Table 7.12 Equilibrium Data of Benzene-Ethanol-Water System1 at 25°C.
mass % in Extract mass % in Raffinate
Benzene Ethanol Water Benzene Ethanol Water

99.35 0.60 0.05 0.13 11.20 88.67

98.00 1.85 0.15 0.28 22.00 77.72
96.30 3.50 0.20 0.53 30.90 68.57
95.25 4.45 0.30 1.02 36.00 62.98
92.00 7.40 0.60 3.35 43.50 53.15
90.75 8.45 0.80 4.80 45.90 49.30
84.40 13.85 1.75 13.70 52.20 34.10
80.00 17.45 2.55 20.45 52.15 27.40
70.60 24.75 4.65 33.20 48.00 18.80
64.10 29.30 6.60 40.95 44.00 15.05

Figure 7.1 represents phase diagram for the system.

7.7 Sulphonation of various aromatic compounds is carried out with either
sulphuric acid (98%) or oleum, having different strengths of SO3 in normal
course. Generation of spent dilute sulphuric acid with high chemical oxygen
demand (COD) is a serious effluent problem.
Consider following sulphonation reactions.
(a) Sulphonation of p-nitrotoluene
(b) Sulphonation of aniline

(c) Sulphonation of Tobias Acid

(d) Sulphonation of 2-Naphthol

As a part of Green Engineering, experiments are planned to carry out reactions

with liquid sulphur trioxide (SO3) at supercritical conditions. Assume that all
reactions will be carried out with stoichiometric proportions of the reactants
and with 100% conversions. For this process research suggest a suitable
supercritical fluid for each of the above reactions with reasonings from Table
7.8. Also consider whether a cosolvent is recommended for any of the above
reactions. Why?

REFERENCES
1. Varteressian, K. A., and M.R. Fenske, Ind. Engg. Chem., 28(8), 1936, p. 928.
2. Treybal, R. E., Liquid Extraction, McGraw-Hill Book Company, Inc., USA,
1951.
3. Rawat, B. S., Aromatic Extraction, Indian Institute of Petroleum, Dehradun,
2000.
4. Perry, R. H., and Green D. Perry, Chemical Engineers’ Handbook, 6th Ed.,
McGraw-Hill, USA, 1984.
5. Treybal, R. E., Mass Transfer Operations, 3rd Ed., McGraw-Hill, USA, 1980.
6. DMF Recovery and Purification, Technical Catalogue of Dupont Co., USA.
7. Nanoti, S. M., Industrial Extraction Equipment Selection and Process Design,
Indian Institute of Petroleum, Dehradun, 2000.
8. Brown, G. G., Unit Operations, John Wiley and Sons, Inc., New York, 1950.
9. Schweitzer, Handbook of Separation Techniques for Chemical Engineers,
McGraw–Hill, USA, 1979.
10. McKetta, J. J., Encyclopedia of Chemical Processing and Design, Vol. 21,
Marcel Dekker Inc., USA, 1984.
11. Nagata, S., Mixing, Kodansha Ltd., Japan, 1975.
12. Coulson, J. M., and J. F. Richardson, Chemical Engineering, Volume 2,
Pergamon Press, New York, USA, 1987.
13. Backhurst, J. R., and J. H. Harker, Chemical Engineering, Volume 5,
Pergamon Press, New York, USA, 1989.
14. Awasthi, A., R. A. Trivedi, Chem. Engg. World, Vol. XXXII, No. 10 Oct. 97,
p. 65.
15. Bork, M., Chem. Engg. World, Vol. XXXII.
16. Chrastil, J., J. Phys. Chem., 86, 1982, p. 3012
17. De Hann, A. B, De Graauw, J., Ind. Eng. Chem. Res., 30 1991, p. 2463.
18. Kohn P. M., and P. R Savage, Chem. Engg., 86, (6), March 12, 1979, p. 41.
19. Laheire, R. J., J. R. Fair, Ind. Engg. Chem. Res., 26, 1987, p. 2086.
20. Moore S., and S. Samdini, Chem. Engg., 32, March 1994.
21. Mukhopadhyay, M., Natural Extracts using Supercritical Carbon Dioxide,
CRC Press LLC, USA, 2000.
22. Savage, P. E., S. Gopalan, T. L. Mizan, C. J. Martino and E. E. Brock,
AICHE J. 41(7), 1995, p. 1723.
23. Bhatt, B. I., Design Databook, Properties of Steam, Selected Refrigerants,
Hexane and Brines, CBS Publishers and Distributors, New Delhi, 2007.
24. Younglove, B. A., and J. F. Ely, J. Phy. Chem. Ref. Data, 16(4), 1987, p. 577.
25. Bhatt, B. I., and S. B. Thakore, Stoichiometry, 5th Ed., McGraw Hill
Education, 2010.
26. Nelson, R. R., Chem. Engg. World, 24(7), 1989, p. 47.
 8
Process Design of Distillation
Columns
8.1 INTRODUCTION
In simple terms, vaporisation followed by condensation is known as distillation.
It is a method of separating components of a solution which are relatively
volatile. Distillation is the most widely used separation method in chemical
industries. Simple distillation or batch distillation was known in the first century.
Multistage countercurrent distillation was invented in 1830 by Aeneas Coffey of
Dublin1. He used tray tower for separating 95.6% ethanol (by mass) from its
dilute solution in water. But as of today 100% reliable method for designing of
distillation column, which can be applied to all systems without any limitation, is
not available. All methods have limitations. Hence, process design engineers use
more than one method for the design of distillation column. For an entirely a
new mixture, it is better to verify the design of its distillation column first on a
pilot plant scale column before applying it at commercial or large scale.

8.2 CRITERIA OF SELECTION

For a given mixture, if all the components are relatively volatile and are required
to be separated in pure form, then distillation is the first choice.
If the relative volatility of the two adjacent components is very high, then
evaporation is preferred. For example, concentration of orange juice.
If relative volatility of two adjacent components is very low: close to 1 (e.g., 10
to 30%, by mass acetic acid in water) or equal to 1 (e.g., azeotropic mixture,
95.6% by ethanol in water), special type of distillation like Azeotropic
Distillation or Extractive Distillation is used. For the same case, liquid-liquid
extraction, membrane separation or crystallization may be more economical.
Examples, (a) For the separation of pure acetic acid from 10% (by mass)
aqueous acetic acid solution, liquid-liquid extraction using ethyl acetate as
solvent is more economical than azeotropic distillation of the same solution
using butyl acetate or ethyl acetate as an entrainer, (b) For the separation of pure
ethanol from its azeotropic mixture with water [95.6% ethanol (by mass) in
ethanol-water solution], pervaporation unit consumes less than 20% of the
energy than that used in azeotropic distillation. Also for the same separation,
pervaporation unit requires low pressure steam (to maintain 80°C temperature
which can be obtained by flashing the condensate of high pressure steam, and (c)
For the separation of p-xylene from the mixed xylenes, crystallization is used.
If the separation of only one component is required rather than a separation of all
the components of the mixture, then absorption, stripping or liquid-liquid
extraction may be more economical than distillation.
Examples:
1. For the separation of penicillin from the complex mixture, liquid-liquid
extraction is used.
2. For the removal of hydrogen and methane from aromatics, stripping is used.
3. For the removal of propane from the mixture called asphalt, stripping is used.
Volatility of propane is very high compared to volatilities of other
components of the asphalt.
4. For the removal of hydrogen sulphide from the natural gas mixture, absorption
using ethanolamines as solvent is used.
5. For removal of ammonia from lean aqueous solution, steam stripping is
preferred.
6. If all components of air are to be separated in pure form (nitrogen, oxygen,
argon, etc.) cryogenic distillation is preferred. But if only nitrogen in pure
form is to be separated from air then pressure swing adsorption (PSA) is
preferred.

8.3 SELECTION OF EQUIPMENT FOR

DISTILLATION1,2
In most of the cases, distillation is carried out either in a tray tower or in a
packed tower. Criteria of selection between tray tower and packed tower are as
follows.
1. For high vacuum distillation, a packed tower is selected. Packed tower
generally provides lower pressure drop as compared to the tray tower.
Pressure drop in packed tower is usually in the range of 1/5th to 1/8th of the
same in tray tower. For example, in the distillation of the mixture of
monochloroacetic acid and acetic acid, absolute operating pressure in the
reboiler should be less than 50 Torr a to avoid thermal cracking of bottom
product (monochloroacetic acid). For this case, packed tower should be
selected. Packed tower provides lower residence time for liquid as compared
to tray tower. The rate of thermal cracking is directly proportional to
 residence time. For highly heat sensitive product, decrease in boiling point
means improvement in the quality. The rate of thermal cracking increases
exponentially with the temperature. For very high vacuum distillation, short
path distillation is used. For example, separation of vitamin E from its crude
mixture (such as soybean oil deodoriser distillate) is carried out in short path
distillation equipment. Operating absolute pressure in short path distillation
can be as low as 0.05 mbar or 50 μbar.
For low pressure distillation applications, liquid flow rates tend to be very low.
For such applications, packed column can be opted provided liquid rates are
above minimum wetting rates for the selected packing.
2. For very small tower diameter packed tower is preferred. With large tower
diameter, tray tower is preferred. With very small diameter (less than 150
mm), fabrication of trays is difficult while with large diameter, liquid
distribution is difficult in the packed tower.
3. For handling higher ratio of liquid flow rate to vapour flow rate, packed
towers are preferred. For very low liquid flow rate and for very low L/V ratio,
tray tower is selected. Packed tower requires minimum wetting rate to work
efficiently.
4. For corrosive system, packed tower with plastic packings like polypropylene
packings, glass fibre reinforced plastic packings, etc., is more economical
than tray tower made from special metal alloys. For example, for the
separation of methyl iodide and acetic acid, packed tower with glass fibre
reinforced plastic packings is more economical than tray tower which should
be made entirely from Hastalloy–B.
5. If the side products are to be drawn from the distillation column, like the
distillation columns of refinery, tray tower is preferred.

8.4 DISTILLATION COLUMN DESIGN

The design of a distillation column can be divided into the following steps:
1. If a feed mixture contains more than two components, then select two key
components from all the components of the feed mixture. For a binary feed
mixture determine more volatile and less volatile components.
2. Decide the operating pressure of distillation column. At the operating
pressure, find the vapour–liquid equilibrium data.
3. Specify the degree of separation required. In case of a binary distillation,
decide the top (distillate) and the bottom (residue) product compositions. In
case of multicomponent distillation, decide the distribution of key
 components in distillate and residue. Also, find the distribution of non key
components in the same. Complete the material balance.
4. Determine the minimum reflux ratio Rm.
5. Determine the optimum reflux ratio R and find the number of theoretical or
equilibrium stages required for the desired separation.
6. Select the type of tower: tray tower or packed tower (as discussed in Section
8.3). Select the type of tray in tray tower or type of packing in packed tower.
7. Find tower diameter and pressure drop across the tower. Decide the top
disengagement and bottom disengagement spaces.
8. Determine the tray efficiency and find the actual number of trays in case of
tray tower or find HETP (height equivalent to theoretical plate) and decide the
height of packing in packed tower. In case of packed tower, design the liquid
distributor, redistributor, packing support, etc.
9. Select the material of construction for all parts. Design the suitable insulation
system.
10. Design the condenser and reboiler.
11. Decide the control methodology for the distillation column, line sizes and
locations of all nozzles.
12. Mechanical design of distillation column is to be carried out which also
includes design of suitable supports for the distillation column.
Calculation and equations for the first five steps are different for binary and
multicomponent distillation. In this chapter first eight steps [(1) to (8)] are
discussed in detail.
8.4.1 Step 1: Selection of Key Components1,2
If the feed mixture to distillation column contains only two components, then it
is called binary distillation. In many cases, feed mixture actually contains more
than two components, but it is considered as binary mixture because the amounts
of other components (except two key components) are negligible.
When a feed mixture to distillation column contains more than two components,
it is called multicomponent distillation. In such a distillation, it is not possible to
obtain more than one pure component in single column. For the separation of N ′
component mixture by distillation, if all the components are to be separated in
pure form, then at least N ′–1 distillation columns are required.
In multicomponent distillation, one cannot specify the complete composition of
the top and bottom products independently. For example, a feed mixture to
distillation column is 100 kmol/h containing 20 kmol/h ethane, 30 kmol/h
propane and 50 kmol/h of butane. One cannot fix distillate composition: 20
kmol/h ethane, 29 kmol/h propane, 1 kmol/h butane and residue composition: 1
kmol/h propane, 49 kmol/h of butane. One can fix that distillate will contain 29
kmol/h of propane and 1 kmol/h of butane and residue will contain 1 kmol/h of
propane and 49 kmol/h of butane.
Based on the degree of freedom analysis for the multicomponent distillation, one
can fix only distribution of any two components of feed mixture in distillate and
residue but not the distribution of all components3. Before starting the design of
multicomponent distillation, two key components are selected. Key components
are the two components of the feed mixture between which one likes to make the
sharp separation. Two key components are further divided as light key
component and heavy key component. More volatile key component is called
light key and less volatile component is called heavy key.
Light key component It is a component of the feed mixture which is desired to
be kept out of the bottom product. It is the component of feed which is present in
residue in finite amount, while components lighter than light key are either not
present or present in very small amount. If all components of feed are present in
residue with sufficient concentrations, then most volatile component of residue is
the light key component.
Heavy key component It is a component of feed mixture which is desired to be
kept out of the top product. It is the component of feed which is present in
distillate in finite amount, while all components heavier than the heavy key are
not present or present in negligible amount. If all components of feed are found
in distillate with sufficient concentrations, then the least volatile is the heavy key
component.
If the light key and heavy key components are selected as per the order of
volatility then they are known as adjacent keys. If any other component lies in
between them, then they are known as split keys. If a component has very low
concentration in feed, then it is not selected as key component.
Components of the feed mixture other than key components are known as non-
key components. The non-key components which are distributed in top and
bottom products are known as distributed components. Any component which
does not appear in either top or bottom product up to significant extent is known
as non-distributed component.
8.4.2 Step 2: Selection of Operating Pressure and
Determination of Vapour-Liquid Equilibrium Data
Selection of operating pressure In many cases of high pressure distillation
column, operating pressure in distillation column is fixed by the temperature of
the cheapest cooling medium (e.g., cooling water), which must be able to
condense the distillate vapour. For this, bubble point of distillate (in case of total
condenser) should be sufficiently greater than cooling water temperature. In case
of partial condenser, bubble point of reflux (is equal to dew point of distillate
vapour) should be sufficiently greater than temperature of cooling water. But if,
this operating pressure (based on cooling water temperature) is calculated to be
very high, then chilled water or brine can be considered as cooling medium for
the overhead condenser. Operating pressure in distillation column is always less
than the critical pressure of top product or temperature of cooling medium of
overhead condenser cannot be greater than critical temperature of top product. In
many cases, distillation is carried out under vacuum to avoid the thermal
degradation of heat sensitive product or to facilitate the use of cheaper heating
medium like saturated steam in reboiler or to alter the vapour-liquid equilibrium
data.
Examples for the selection of operating pressure
1. Distillation of acetaldehyde-ethanol-water mixture is carried out at about 3
atm g pressure. Pure acetaldehyde is the top product. Normal boiling point (tB
at 1 atm) of acetaldehyde is 20.2°C. To facilitate the use of cooling water as a
cooling medium in the overhead condenser, boiling temperature or
condensation temperature of acetaldehyde should be sufficiently greater than
the design temperature of cooling water from cooling tower (say 32°C). Also
the product acetaldehyde is stored in bullet-type storage tanks which are
exposed to atmosphere. Hence, to reduce the cost of storage or to avoid the
use of insulation around storage tank and to avoid the use of any refrigeration
system for the storage tank, boiling point of acetaldehyde should be
sufficiently greater than atmospheric temperature (50°C maximum).
Considering all these aspects, condensation temperature (is also equal to
boiling temperature of pure component) of acetaldehyde is fixed at around
60°C. For getting 60°C as a condensation temperature of acetaldehyde,
 required operating pressure at the top of distillation column is around 3.6 atm
g (vapour pressure of acetaldehyde at 60°C). It also eliminates the use of
pump for the transfer of acetaldehyde to storage tank.
2. For the separation of methyl chloride from the mixture of chloromethanes,
normal boiling point or condensation temperature of pure methyl chloride is –
24°C. At 9 atm a pressure, its condensation temperature is 40°C. Hence, to
facilitate the use of cooling water as a cooling medium in overhead
condenser, this distillation column can be operated at 9 atm a pressure.
Similarly, in case of distillation of ammonia (Example 8.7), saturation
pressure of ammonia at 40°C is 1555.5 kPa a. Hence, the column is operated
at about 15 bar g to facilitate condensation of ammonia at 40°C in the
condenser with cooling water at 32°C.
3. For the separation of ethane from the mixture of ethane, propane and butanes
by distillation, cooling water cannot be used as a cooling medium in the
overhead condenser because critical temperature of ethane (tc = 32.3°C) is not
sufficiently greater than the temperature of cooling water (say 32°C).
Minimum driving force required for heat transfer in shell and tube-type
overhead condenser is 3 to 5°C. At 10°C, vapour pressure of ethane is 30 atm
a, hence chilled water (at 5 to 7°C) can be used as a cooling medium in the
overhead condenser, but the required operating pressure is 30 atm a, which is
high particularly for larger diameter column. At 20 atm a pressure,
condensation temperature of pure ethane is –6.4°C, hence, suitable brine
system can be used as a cooling medium in the overhead condenser, but it
will require higher cost of refrigeration and insulation. In the plant, if this
column is following a demethaniser column, then top product stream of
demethaniser can be used as a cooling medium (heat pump arrangement) in
the overhead condenser of deethaniser column and operating pressure can be
reduced to very low value, close to atmospheric. Optimum solution should
therefore be worked out.
4. In some high pressure distillation columns, operating pressure is limited by
undesirable side reactions which otherwise take place at higher pressure and
higher temperature. For example, to avoid the polymerisation of butadiene,
operating pressure of debutaniser, depropaiser, de-ethaniser, etc., columns of
naphtha or natural gas cracker units are kept at lower pressure.
Vacuum distillation For heat sensitive material or to facilitate the use of cheaper
heating medium in reboiler or to alter the vapour-liquid equilibrium data,
distillation column is operated under vacuum. Advantages and disadvantages of
vacuum distillation are given below.
Advantages
1. It prevents thermal decomposition of heat sensitive material. For example, in
separation of monochloroacetic acid and acetic acid by distillation, mono-
chloroacetic acid is obtained as bottom product. Boiling point of monochloro-
acetic acid must be restricted below 120°C, otherwise thermal degradation
can take place. If the boiling point of monochloroacetic acid is kept at 109°C,
then operating pressure at bottom of column must be equal to 40 Torr a.
2. It allows the use of more economical heating medium such as saturated steam
against hot oil in high capacity plant. For example, for the separation of
ethylene glycol-water mixture by distillation, ethylene glycol is obtained as
bottom product. Its normal boiling point is 197°C. Use of saturated steam as a
heating medium could be economical up to 180°C. Hence, this column is
better to be operated under vacuum to facilitate the use of saturated steam
instead of use of hot oil.
3. It requires fewer stages or lower reflux ratio or both in many cases as relative
volatility increases with decrease in the operating pressure.
4. In some case vacuum breaks the azeotrope and make the separation easier. For
example, if the distillation of ethanol–water mixture is carried out below 70
Torr absolute pressure, azeotrope does not form.
5. It reduces undesirable side reactions, if they are taking place. For example,
distillation of ethylbenzene-styrene mixture under vacuum reduces
polymerisation.
6. Vacuum distillation provides safety in the distillation of toxic or hazardous
material. For example, for the separation of nearly 100% pure nitric acid from
the azeotropic mixture of nitric acid and water, extractive distillation is used.
This extractive distillation uses concentrated Sulphuric acid as solvent
(Example 8.21). Material of construction of this column is borosilicate or
high quality glass. This column is operated under slight vacuum for safety
purpose. Here, vacuum also facilitates the use of saturated steam in reboiler.
7. It increases product recovery when dealing with maximum temperature
constraints. For example, refinery vacuum tower.
Disadvantages
1. It requires larger column diameter and larger diameter vapour lines.
2. Because of the reduction in condensation temperature, costlier cooling
 medium may be required for condensation.
3. Presence of air as a non-condensable reduces the efficiency of main
condenser. Also air leaving the condenser is saturated with vapour of top
product.
4. Increases the possibility of contamination of products by air or other
components of atmosphere.
Determination of vapour-liquid equilibrium (VLE) data
VLE data must be determined for the given mixture for distillation and at the
operating pressure of distillation column. To find the VLE data, there are three
options. Priority-wise these three options are given below.
1. VLE data can be determined by an experiment. If the operating pressure is
atmospheric or vacuum, then VLE data-experiment is easier. Only difficult
part is analysis. But if VLE data are required at high pressure, then this
experiment is difficult. This is the most reliable option or method for finding
the VLE data for a given mixture because the given feed mixture obtained
from lab scale or pilot plant scale process for distillation may contain some
unknown impurities which may not be considered or ignored in other two
options or methods given in the coming section.
2. Find the VLE data for the given system from the literature, if they are
available at the same operating pressure. In the literature, VLE data are
available for the several thousand binary and many multicomponent systems.
All these data were determined by actual experiments. Major sources for
these data are as follows.
(i) Vapour-Liquid and Liquid-Liquid Equilibrium Data Collection,
DECHEMA Chemistry Data Series4, Volume-1, Parts 1 to 10, by Gmehling
and Onken (1977).
(ii) Vapour-Liquid Equilibrium Data5 by Chu, Ju-chin et al. (1956).
(iii) Vapour-Liquid Equilibrium Data Bibliography6 by Wichterle, Linek and
Hala, Elsevier, The Netherlands, 1976.
(iv) Vapour-Liquid Equilibrium Data at Normal Pressures7 by Boublik,
Polak, Wichterle ad Hala, Pergamon, Oxford, UK, 1968
(v) Perry’s Chemical Engineers Handbook8, 7th Edition by Perry and Green,
McGraw-Hill, USA, 1998.
(vi) Azeotropic Data9, 3 Vol., 2nd Ed. by Gmehling, J., Wiley-VCH, Germany,
2004.
3. VLE data can be determined theoretically by using phase equilibrium
 equation. In the given feed mixture for distillation, if all components are from
the same group and operating pressure for distillation is atmospheric then
usually vapour-liquid equilibrium for such mixture can be assumed as ideal
vapour-liquid equilibrium. For example, feed mixture of benzene and toluene
at atmospheric pressure.
For the ideal VLE, Roult’s law is applicable.
yi pt = xi pv (Roult’s law)   (8.1)
where, yi = Mole fraction of component i in vapour phase
xi = Mole fraction of component i in liquid phase
pt = Operating pressure, kPa
pv = Vapour pressure of pure liquid i at the given temperature, kPa
General form of vapour-liquid equilibrium equation is yi = Kixi   (8.2)
where, Ki = Equilibrium constant for component i
Ki – values for some hydrocarbons for the range of operating pressure from
101.3 kPa to 6000 kPa and for the range of operating temperature from –70°C to
200°C, can be obtained-from Dadyburjor chart (refer to Fig. 6.20).
For non-ideal system VLE data is determined by following equation:
yiφi pt = xi γi pv   (8.3)
where, φi = Fugacity coefficient of component i
γi = Activity coefficient of component i
Fugacity coefficient φi represents the deviation of vapour phase from ideal gas
mixture. If the operating pressure is close to atmospheric pressure then vapour
phase can be assumed as ideal gas and φi is unity. φi can be calculated from an
appropriate equation of state like Redlich–Kwong equation, Peng–Robinson
equation, etc.
Activity coefficient γi represents the deviation of the behaviour of component in
given solution from the behaviour of same component in ideal solution. γi is a
function of temperature and liquid composition.
γi = f (xi, T)   (8.4)
γi is not dependent on pressure.
Activity coefficient γi can be determined by van Laar equation, UNIFAC method
or UNIQUAC method or by NRTL equation. Group contribution method like
UNIFAC method or UNIQUAC method are more reliable, but group
contribution parameters are available for limited number of groups in literature.
For example, for the mixture of ethane and propane, if VLE data are required at
10 atm a pressure, then φi ≠ 1, but γi can be assumed equal to 1 for both
components.
For the mixture of ethanol-ethyl iodine, if VLE data required are at 1 atm
pressure, then for both components
φi ≅ 1, but γi ≠ 1.

8.4.3 Step 3: Degree of Separation

In case of binary distillation, specify xD and xW considering equilibrium
limitations in mind, (where xD = mole fraction of more volatile component in
liquid distillate and xW = mole fraction of more volatile component in residue).
Find the molar flow rate of distillate D and molar flow rate of residue W, by
using following two material balance equations:
F = D + W   (8.5)
FzF = DxD + W xW   (8.6)
where zF represents mole fraction of more volatile component in feed. Since feed
can be saturated liquid, saturated vapour or a mixture of liquid and vapour,
symbol zF is used. If it is only liquid, symbol xF can be used.
To establish the material balance in multicomponent distillation is not easy. In
case of multicomponent distillation, specify the distribution of two key
components in distillate (top product) and in residue (bottom product),
considering equilibrium limitations in mind. Exact distribution of non-key
components in top product and bottom predict is not possible at the starting of
design. But, the same can be determined approximately by Hegstebeck and
Geddes equation10.

log = A + C log αi  (8.7)

where, di = Moles of ith component in distillate

bi = Moles of ith component in residue
αi = Average relative volatility of ith component with respect to heavy
key component
A, C = Constants
Material balance calculations (approximate) for multicomponent distillation is
illustrated by following example.
Feed mixture to a distillation column contains 30% n-hexane,
32% n-propane, 20% n-butane and 18% n-pentane (by mole). Total flow rate of
the feed is 100 kmol/h. If butane and pentane are selected as light key and heavy
key components, respectively then (a) fix the operating pressure of distillation
column, and (b) find the product compositions.
Solution
Generally, the mixture of propane and butane is stored and sold as LPG
(Liquefied Petroleum Gas). Hence in that case, top product (mixture of propane
and butane) must be in liquid phase or a total condenser must be used. To
facilitate the use of cooling water as cooling medium in overhead total
condenser, bubble point temperature of distillate must be sufficiently greater
than design temperature of cooling water from cooling tower (say 32°C). Also to
facilitate the storage of liquid distillate without using insulation and refrigeration
system, bubble point temperature must be sufficiently greater than atmospheric
temperature. Here distillate is highly flammable liquid. Hence, to avoid the use
of a pump between distillation column and a storage tank, let bubble point of the
distillate be 60°C.
For the first trial calculations, assuming that non-key components, propane will
appear in distillate only and hexane will appear in residue only. For the first trial
calculations material balance is assumed as follows.
Table 8.1 Material Balance of Distillation Column (1st Trial Calculations)
Component Distillate, kmol/h Residue, kmol/h

n-Propane 32.0 —
n-Butane 19.8 0.2
n-Pentane 0.2 17.8
n-Hexane — 30.0

Total 52.0 48.0

Bubble point temperature of distillate = 60°C. At bubble point, ΣKi xi = 1

KPr xPr + KPe xPe + KBu xBu = 1

xPr = = 0.6154, xPe = = 0.0038, xBu = = 0.381

0.6154 KPr + 0.0038 KPe + 0.381 KBu = 1(a)

Here values of equilibrium constants are to be determined at the temperature
60oC and operating pressure pt which is unknown. Hence by trial and error
calculations find the value of pt which will satisfy Eq. (a). Values of equilibrium
constants are determined by using Fig. 6.20.
At pt =1500 kPa and t = 60°C
KPr = 1.4, KPe = 0.21, KBu = 0.52
ΣKi xi = 0.6154 × 1.4 + 0.0038 × 0.21 + 0.381 × 0.52 = 1.06
At pt = 1600 kPa and 60°C,
KPr = 1.35, KPe = 0.2, KBu = 0.5
ΣKixi = 0.6154 × 1.35 + 0.0038 × 0.2 + 0.38 × 0.5 = 1.022 = 1
Operating pressure at top of column = 1600 kPa a
Assume pressure drop in column = 30 kPa
Operating pressure at the base of column = 1630 kPa a
To decide the distribution of non-key components Hegstebeck–Gedde’s equation
can be used.
log = A + C log αi  (8.7)

where, αi = αavi = (αitop · αibottom)1/2   (8.8)

To find out the average value of relative volatility, top and base (bottom)
temperatures of distillation column are to be determined.
For distillation column with total condenser, composition of overhead vapour
and distillate are same. Topmost temperature of distillation column is equal to
dew point temperature of overhead vapour.
Topmost temperature
At dew point, Σ = 1

= 1

At operating pressure pt = 1600 kPa and at temperature t, values of equilibrium

constants must be determined, where t is unknown.
For the first trial calculations, Let t = 75°C
At 75°C, KPr = 1.65, KBu = 0.65 KPe = 0.26
Σ = = 0.974

At 72°C, KPr = 1.55, KBu = 0.6, KPe = 0.25

Σ = ≈ 1

Bottommost temperature
Bottommost temperature of distillation column is the bubble point of residue.
Operating pressure at the base of column, p′t = 1630 kPa.
At bubble point Σ Ki xi = 1
For residue, xBu = = 0.004 17, xPe = = 0.37, xHe = = 0.625

For the first trial calculations, let t = 150°C (and at 1630 kPa)
KBu = 1.9, KPe = 0.95, KHe = 0.52
ΣKixi = 0.004 17 × 1.9 + 0.37 × 0.95 + 0.625 × 0.52 = 0.684
Let t = 180°C. KBu = 2.5, KPe = 1.3, KHe = 0.79
ΣKi xi = 0.004 17 × 2.5 + 0.37 × 1.3 + 0.625 × 0.79 = 0.99 ≈ 1
Table 8.2 Relative Volatilities
Feed Component α72° C α180° C αavi = (α72 · α180)0.5

n-Propane = 6.20 = 3.270 4.500

n-Butane (LK) = 2.40 = 1.923 2.150

n-Pentane (HK) 1.00 1.000 1.000

n-Hexane = 0.44 = 0.6077 0.517

At 72°C, KHe = 0.11 and at 180°C, KPr = 4.25.

For butane (light key component):
log = A + C log αi

log = A + C log (2.15)

1.9956 = A + 0.3324 C
For pentane (heavy key component)
log = A + C log (1)

–1.949 39 = A + 0
Solving two equations
A = –1.949 39 and C = 11.8682
For propane:

log = –1.949 39 + 11.8682 log (4.5)

log = 5.803

= 635 331 (very high)

di = 32 kmol/h
bi = 5 × 10–5 kmol/h ≈ 0 kmol/h
For hexane:
log = –1.949 39 + 11.8682 log (0.517)

= 4.47 × 10–6

bi = 30 kmol/h
di = 1.34 × 10–4 kmol/h ≈ 0 kmol/h
Propane does not appear in the residue and hexane does not appear in the
distillate. Hence, assumption made before deciding the distribution of
components is valid. No correction is required. So Table 8.1 shows the final
product composition.
8.4.4 Step 4: Minimum Reflux Ratio
Determine the minimum reflux ratio Rm. It is the maximum reflux ratio for
which infinite number of trays are required for the desired separation. It
corresponds to the minimum condenser cooling load and the minimum reboiler
heating load required for the desired separation. In other words, on decreasing
reflux ratio, a value of reflux ratio obtained for which desired separation
becomes impossible, is called minimum reflux ratio; Rm.
Determination of Rm for binary distillation For determination of Rm, it is
important to know quality of feed which is characterised by a parameter q. It is
defined as moles of liquid flow to stripping section of the column per mole of
feed and it can be determined by the following equation:
  (8.9)

where, HG = Enthalpy of saturated vapour, kJ/kmol

HF = Enthalpy of feed, kJ/kmol
HL = Enthalpy of saturated liquid, kJ/kmol
Equation (8.9) can be modified for calculations as given in Eq. (8.16).
If 1. feed is below bubble point, q > 1
2. feed is at bubble point, q = 1 (i.e., saturated liquid)
3. feed is partially vaporised, q < 1
4. feed is at dew point, q = 0 and
5. feed is superheated vapour, q < 0.
(a) In case of binary distillation with ideal vapour-liquid equilibrium curve or
with the equilibrium curve which is always concave downward, Rm can be
calculated by any of the following equations1:
(i) For feed liquid at the bubble point or with, the saturated liquid feed (i.e., for q
= 1)
Rm =   (8.10)

where, α = Average relative volatility

xD = Mole fraction of more volatile component in distillate
xF = Mole fraction of more volatile component in feed liquid
(ii) For the saturated vapour feed or for feed vapour at dew point (i.e., for q = 0)
Rm =   (8.11)

where, yF = Mole fraction of more volatile component in feed vapour

(iii) For any other condition of feed (for any value of q), Rm can be determined
by solving the following equation:
=   (8.12)

(b) In case of binary distillation with ideal or non-ideal vapour-liquid

equilibrium curve, minimum reflux ratio Rm can be determined by McCabe–
Thiele method (Ref. 1, 11). Following assumptions are made in McCabe–
Thiele method:
(i) For any equilibrium stage, molar latent heat of vaporization of liquid on tray
is equal to molar latent heat of condensation of vapour leaving the tray.
(ii) Heat losses and heat of solution (or say mixing) are negligible.
(iii) Total condenser is used but there is no subcooling. Condenser removes all
the latent heat from the overhead vapour but does not cool the resulting
liquid further.
Because of the first two assumptions, molar flow rate of liquid and molar flow
rate of vapour in each section (enriching section and stripping section) is
assumed to remain constant throughout the section. Because of the third
assumption, composition of overhead vapour is equal to composition of distillate
and also equal to composition of reflux.
L, G, L–, G– = constant
xD = yi = xo
where, L, G = Molar flow rates of liquid and vapour, respectively in rectification
or enriching section, kmol/h.
L–, G– = Molar flow rates of liquid and vapour, respectively in stripping or
exhausting section, kmol/h.
xD = Mole fraction of more volatile component in overhand vapour
xo = Mole fraction of more volatile component in reflux.
Method for determining Rm by McCabe Thiele method:
(i) Draw the vapour–liquid equilibrium curve, y vs x. Also, draw the 45°
diagonal (x = y) line.
(ii) Mark the points x = y = xD, x = y = zF and x = y = xW on the 45° diagonal line.
(iii) Draw the q-line. It starts from the point x = y = zF. The equation of q-line is

Thus, y =   (8.13)
Slope of q-line is .

(iv) Draw the line from x = y = xD point to the point of intersection of q-line and
equilibrium curve. Extend the same line to y-axis. Intercept of this line gives
the value of from which the value of Rm can be calculated. But this
procedure is valid for ideal vapour-liquid equilibrium curves and also for non-
ideal vapour-liquid equilibrium curves except the curves as shown in Figs.
8.1(b) and (c). For the non-ideal VLE curve like Fig. 8.1(b) tangent line of
equilibrium curve decides the value of Rm. For the non-ideal VLE curve as
shown in Fig. 8.1(c), a tangent operating line of stripping section may set the
value of Rm.
In case of multicomponent distillation, Rm can be determined by Underwood’s
method1,8,12. In this method, Rm is determined by solving the following two
equations:
Σ = Rm + 1   (8.14)

Σ = 1 – q   (8.15)

where, αi = Average relative volatility of each component i with respect to heavy

key
xiD = Mole fraction of component i in distillate
ziF = Mole fraction of component i in feed
q = Heat required to convert 1 mole of feed from its condition to a saturated
vapour divided by latent heat of vaporisation [Eq. (8.9)].
ϑ = Constant
Here αi = (αitop · αibottom)1/2  (8.8)
 Fig. 8.1 Determination of Minimum Reflux Ratio for Binary Distillation

Topmost temperature of distillation column is dew point temperature of

overhead vapour. Bottommost temperature of distillation column is bubble point
temperature of residue.
For accuracy purpose, it is suggested that ϑ, must be determined correct up to
four decimal places13. Equation (8.15) may give many values of ϑ, but correct
value of ϑ is that belongs to the interval starting from 1 to the value of αLK.
ϑ ∈ {αHK, αLK} or αLK > ϑ > αHK
where, αLK = Relative volatility of light key component with respect to heavy
key component.
αHK = Relative volatility of heavy key component with respect to heavy key
component = 1
Determine the minimum reflux ratio for the binary distillation at
standard atmospheric pressure based on the following data:
Feed = 100 kmol/h
Feed mixture: benzene-toluene
Mole fraction of benzene in feed, xF = 0.4
Condition of feed = Saturated liquid
Mole fraction of benzene in distillate required, xD = 0.99
Mole fraction of benzene in residue required, xW = 0.02
Average relative volatility, α = 2.25
Solution
Benzene-toluene mixture behaves as an ideal mixture.

Rm =   (8.10)

Rm =

Minimum reflux ratio Rm = 1.95

In the previous example, if the feed is at room temperature
(30°C), what is the value of Rm?
Solution
To find the value of Rm, first value of q must be determined using Eq. (8.9).

q =   (8.16)

tF = Temperature of feed = 30°C

tb = Bubble point temperature of feed
At bubble point:
Σxi Ki = 1

Σxi = 1 (for ideal VLE)

Antoine equations14
For benzene, ln pvB = 15.9008 –
For toluene, ln pvT = 16.0137 –

For the first trial calculations let tB = ΣtBi xi

where, tBi = normal boiling point of pure component i
Normal boiling point of benzene = 80.1°C
Normal boiling point of toluene = 110.6°C
For first trial calculations, tB = 0.4 × 80.1 + 0.6 × 110.6 = 98.4°C (371.4 K)

ln pvB = 15.9008 – = 7.16

pvB = 1287.54 Torr, similarly

pvT = 527.4 Torr
Σxi pvi = 0.4 × 1287.54 + 0.6 × 527.4 = 831.456 Torr ≠ 760 Torr
After trial and error calculations,
at tB = 95.3°C, pvB = 1181.72 Torr, pvT = 479.1 Torr
Σxi pvi = 0.4 × 1181.72 + 0.6 × 479.1 = 760.15 Torr
Bubble point of feed mixture, tBP = 95.3°C
Specific heat must be determined at the average temperature
tav = = 62.65°C

Table 8.3 Property Data of Benzene and Toluene

Component CL at 62.65°C, kJ/(kmol · °C) λ at 95.3°C, kJ/kmol

Benzene 146.96 29 391.3

Toluene 173.33 34 666.7

(Figure 3.11 and Table 3–179, of Ref. 14)

Average molar heat capacity, CL = ΣCLi xi
CL = 146.96 × 0.4 + 173.33 × 0.6
= 162.78 kJ/(kmol · °C)
Average latent heat
λ = Σxi λi = 0.4 × 29 391.3 + 0.6 × 34 666.7
= 32 556.54 kJ/kmol
q =
= 1.3265
Rm can be determined by Eq. (8.12).

After the trial-and-error calculations, for Rm = 1.67, LHS (1.9514) = RHS (1.95)
Use of a mathematical software such as Mathcad® or Maple® can be made to
solve such an equation.
Minimum reflux ratio, Rm = 1.67
A distillation column is to separate 4750 mol/h of feed
composed of 37% n-butane, 32%. iso-pentane, 21% n-pentane and 10% n-
hexane. The column operates at an average pressure of 2 atm a and will produce
a distillate product containing 95% n-butane and 5% iso-pentane. The bottom
product is allowed to contain no more than 570 mol/h of n-butane. Use
Underwood’s method to determine the minimum reflux for the required
separation. Feed is 25% (by mole) vapour. Assume ideal vapour–liquid
equilibrium. All compositions are mole %.
Solution
In this example, distillate composition is given. Hence, residue composition can
be determined by material balance. Distillate composition indicates that n-butane
is light key component and iso-pentane is heavy key component.
Overall material balance:
F = D + W
4750 = D + W
n-butane balance:
0.37 × 4750 = 0.95D + 570
D = 1250 mol/h and W = 3500 mol/h
Table 8.4 Feed and Product Compositions
Feed Distillate Residue
Component
mole % mol/h mole % mol/h mole % mol/h

n-butane 37 1757.5 95 1187.5 16.3 570.0

iso-pentane 32 1520.0 5 62.5 41.6 1457.5
 n-pentane 21 997.5 — — 28.5 997.5
n-hexane 10 475.0 — — 13.6 475.0

Total 100 4750.0 100 1250.0 100.0 3500.0

In Underwood’s method, to find average relative volatilities, topmost and

bottommost temperatures of distillation column are required.
Topmost temperature = Dew point temperature of overhead vapour
At dew point
Σ = 1 (for ideal VLE) – Raoult’s law

yi = xiD (for total condenser)

× 2 = 1

where, pvnB = Vapour pressure of n-butane

pviP = Vapour pressure of iso-pentane.
Antoine equations10:
n-butane: ln pvnB = 15.6782 –

iso-pentane: ln pviP = 15.6338 –

where, T is temperature in K and pvnB is in Torr.

After a trial-and-error calculations dew point temperature can be determined.
For t = 22°C, T = 295 K
pvnB = 1650 Torr = 2.17 atm
pviP = 615 Torr = 0.8 atm

LHS = × 2 = 1 = RHS

Dew point temperature of overhead vapour = 22°C.

Bottommost temperature = Bubble point of residue
At bubble point:
Σxi pisat = pt (for ideal VLE)
0.163 pvnB + 0.416 pviP + 0.285 pvnP + 0.136 pvnH = 2
Antoine equations10:
n-Pentane: ln pvnP = 15.8333 –

n-hexane ln pvnH = 15.8366 –

After a trial and error calculations, final T = 320 K or t = 47°C

pvnB = 3403.3 Torr = 4.478 atm
pviP = 1399.95 Torr = 1.842 atm
pvnP = 1084.1 Torr = 1.426 atm
pvnH = 361.6 Torr = 0.476 atm
LHS = 0.163 × 4.478 + 0.416 × 1.842 + 0.285 × 1.426 + 0.136 × 0.476
= 1.97 ≈ 2 = RHS
Table 8.5 Relative Volatility Calculations

Component α22 = α47 = αav = (α22 × α47)1/2

n-butane (LK) = 2.7125 = 2.43 2.567

iso-pentane (HK) = 1 1 1

n-pentane = 0.75 = 0.774 0.762

n-hexane = 0.216 = 0.258 0.236

Feed contains 25% vapour.

HF = 0.25 × HG + 0.75 × HL

q =   (8.9)

q = 0.75
Underwood’s equations; Eq. (8.14) and Eq. (8.15):
Equation (8.15) = 0.25

For ϑ = 1.6045, f(ϑ) = 0.250 25 = LHS ≈ 0.25 = RHS

Also ϑ ∈ {1, …, 2.567} (check)
Using Eq. (8.14)
= Rm + 1

Rm + 1 = 2.450 95
Rm = 1.450 95

Fig. 8.2 Determination of Optimum Reflux Ratio

8.4.5 Step 5: Determination of Optimum Reflux Ratio

Ideally the reflux ratio which should be used for a new design should be
optimum or the most economical. Optimum reflux ratio is the value of reflux
ratio for which the total cost of distillation is minimum. Value of optimum reflux
ratio depends on the type of system or feed mixture, specified separation and
operating conditions. Change in desired purity of products changes the value of
optimum reflux ratio and also the minimum reflux ratio. Based on cost analysis,
optimum reflux ratio is defined as that value of reflux ratio which causes the
total variable annual cost (annual fixed charges plus annual operating costs) to
be a minimum.
At the minimum reflux ratio, distillation column requires infinite number of
trays or infinite packing height and consequently the fixed cost is infinite but
minimum operating cost (cost for the heating medium of reboiler, cooling
medium for the condenser, power for reflux pump, etc., minimum). As the reflux
ratio increases above the minimum, number of trays required or packing height
required rapidly decreases but column diameter, sizes of condenser and reboiler,
and size of reflux pump increases. Despite the increase in column diameter and
increase in sizes of condenser and reboiler, the annual fixed cost for the
distillation goes down as the increase in reflux ratio above minimum results in
saving in tower height which more than offsets other fixed costs. This is not
continued as the reflux ratio increases further. At certain value of reflux ratio the
tray requirement or height of packing requirement approaches a near constant
value, characteristic of the minimum trays or height requirement, while the
diameter and sizes of heat exchangers continue to increase and hence, at some
value of reflux ratio, fixed cost passes through minimum point. Increase in the
value of reflux ratio further increases the fixed cost again. Operating cost always
increases with increase in reflux ratio. The total annual cost, which is sum of
operating and fixed costs, must therefore pass through a minimum at the
optimum reflux ratio.
For many hydrocarbon distillations, optimum reflux ratio lies between 1.2 to 1.5
times the minimum reflux ratio. But for the difficult distillation, like distillation
of acetic acid-water system, optimum reflux ratios are higher than 1.5. For
azeotropic mixtures, reflux ratio is usually high. For the optimum design of
distillation column, optimum values of reflux ratio, operating pressure and purity
of products must be determined.
Determination of theoretical stages To determine number of theoretical stages
or equilibrium stages, required for desired separation, many methods are
available. But they can be broadly divided in two groups:
1. Methods valid for binary distillation
2. Methods valid for multicomponent distillation
1. Methods Valid for Binary Distillation: Both shortcut methods and detailed
methods (rigorous methods) are available. Shortcut methods available for binary
distillation are Fensky’s equation1, Smoker’s equation10, etc. Detailed methods
available for binary distillation are McCabe–Thiele method11, Ponchon–Savarit
method1, etc. Also to find the number of theoretical stages for high purity
products, Kremser equation1, Robinson and Gilliland equation10, etc., are
available.
(a) McCabe–Thiele Method1, 11 Assumptions made in McCabe–Theile method
are discussed in the Section 8.4.4. McCabe–Thiele method cannot be used for
the following cases:
(i) Cases where heat of solution or heat losses are significantly large. For
example, separation of ammonia from dilute aqueous solution by distillation.
(ii) For the cases in which high purity of products are required. In such cases,
 other method is used to support the McCabe–Thiele method. However, for the
same cases one can use only McCabe–Thiele method, but more than one
graph are to be drawn at different scales or a software like AutoCad® can be
used.
Stepwise procedure for applying McCabe–Thiele method is as follows.
• Draw the vapour liquid equilibrium curve from the VLE data available at the
operating pressure of distillation column.
• Mark the point xD, zF, and xW on 45° diagonal line.
• Draw the q-line. It starts from x = y = zF point and its slope is q/(q –1).
• Determine the minimum reflux ratio Rm.
• Select the value of actual reflux ratio R. Draw the enriching (or rectification)
section (nth tray) operating line. The equation of this line is
yn+1 =   (8.17)

This line starts from x = y = xD point. Its intercept is xD /(R + 1).

• Draw the stripping or exhausting section operating line. It starts from the point
x = y = xW passing through the point of intersection of enriching section
operating line and the q line.
Operating line of the exhausting (or stripping section) is represented by the
following equation:

ym+1 =   (8.18)

where ym+1 is the vapour composition of light component of the tray above m-th
tray and xm is the liquid composition of light component on m-th tray.
= L + F · q  (8.19)
and V– = L– – W  (8.20)
• Starting from xD to xW or from xW to xD, carry out the stepwise construction
between operating lines and equilibrium curve. Total number of steps is equal
to number of theoretical or equilibrium stages required for the desired
separation.
Acetic acid is to be separated from a process stream containing
80% acetic acid and 20% water (by mass) by continuous distillation at
atmospheric pressure. Concentration of water in bottom product (pure acetic
acid) should not be greater than 50 ppm. Top product (distillate) contains 80%
water and 20% acetic acid (by mass). The feed is liquid at 30°C. Estimate the
number of theoretical stages required.
Table 8.6 Equilibrium Data for Acetic Acid-Water Data at 101.3 kPa
Temperature, °C Mole fraction of water Relative volatility
x X = x/(1 – x) y Y = y/(1 – y) αi = Y/X

118.3 0 0
110.6 0.1881 0.2317 0.3063 0.4415 α1 = 1.905 851

107.8 0.3084 0.4459 0.4467 0.8073 α2 = 1.810 489

105.2 0.4498 0.8175 0.5973 1.4832 α3 = 1.814 312

104.3 0.5195 1.0812 0.6580 1.9240 α4 = 1.779 539

103.5 0.5824 1.3946 0.7112 2.4626 α5 = 1.765 768

102.8 0.6750 2.0769 0.7797 3.5393 α6 = 1.704 09

102.1 0.7261 2.6510 0.8239 4.6786 α7 = 1.764 862

101.5 0.7951 3.8804 0.8671 6.5245 α8 = 1.681 374

100.8 0.8556 5.9252 0.9042 9.4384 α9 = 1.592 925

100.8 0.8787 7.2440 0.9186 11.2850 α10 = 1.557 838

100.5 0.9134 10.5473 0.9409 15.9205 α11 = 1.509 43

100.2 0.9578 22.6967 0.9708 33.2466 α12 = 1.464 821

100.0 1.0000 1.0000

(Date source: Table 13.1 of Ref. 8)

Solution
Feed contains 20% water and 80% acetic acid by mass. Molar mass of water is
18 and molar mass of acetic acid is 60. Water is the lighter component.
Mole fraction of water in liquid feed:

zF = xF = = 0.4545

Mole fraction of water in distillate:

xD = = 0.9302
Mole fraction of water in bottom product:

xW = = 1.666 × 10–4

q-line
Feed is liquid at 30°C temperature. From t-x-y data, bubble point of feed can be
easily determined. t = bubble point temperature of feed when x = xF.
For x = 0.4545, corresponding value of t = 105.14°C.
Bubble point of feed = 105.14°C
Latent heat of water at 105.14°C, λW = 40 392 kJ/kmol
Latent heat of acetic acid at 105.14°C, λA = 20 096.6 kJ/kmol
Specific heat at = 67.57°C,

CpW = 75.36 kJ/(kmol · °C)

CpA = 131.88 kJ/(kmol · °C)
Latent heat of feed, λF = 0.4545 × 40 392 + (1 – 0.4545) × 20 096.6 = 29 320.86
kJ/kmol
Mean specific heat of feed,
CpF = 0.4545 × 75.36 + (1 – 0.4545) × 131.88
= 106.19 kJ/(kmol · °C)
q =

= = 1.272

Equation of q-line:
y =   (8.13)

y =

= 4.676 47x – 1.67

q-line starts from x = y = zF or x = y = 0.4545 point.
To find another point, when y = 0.6, x = 0.4854 using q-line equation.
q-line passes through (0.4854, 0.6) point.
Draw the q-line in the graph. From Fig.   (8.3), intercept for the operating
line at minimum reflux,
0.304 =

= 0.304

Rm = 2.1
Let R = 2 Rm = 2 × 2.1 = 4.2

= = 0.1789

Fig. 8.3 McCabe–Thiele Diagram for Acetic Acid-Water System

On stagewise construction starting from x = y = xD, as shown in Fig. 8.3, up to x
= xr = 0.02, number of theoretical or equilibrium stages required is 16 in which
8th theoretical stage from the top is corresponding to feed tray.
Similarly, for the different values of R, number of equilibrium stages, N, required
for the same separation are determined and listed in Table 8.7.
Table 8.7 Theoretical Stages (N) for Different Values of R/Rm
R/Rm 2 2.5 3 3.5 4

N 16 15.0 14 14.0 13.5

On increasing the value of R/Rm above 2, decrease in the value of N is very

marginal. Hence, R = 2 Rm is considered to be nearer to the optimum value.
To find the number of equilibrium stages required to increase the purity of
bottom product (acetic acid) from x = 0.02 to xw = 1.666 × 10–4, Robinson and
Gilliland equation10 can be used.
For the stripping section, Robinson and Gilliland correlation is

Ns = + 1  (8.21)

where, Ns = Number of ideal stages required from the reference point x′r to xW
xW = Mole fraction of the more volatile component in the residue (bottom
product)
x′r = Mole fraction of more volatile component at the reference point
S′ = Slope of the stripping section operating line
K′ = Equilibrium constant for more volatile component
For the enriching section, Robinson and Gilliland correlation is

Nr =   (8.22)

where, Nr = Number of ideal stages required from reference point xr to x′d

x′d = Mole fraction of least volatile component in distillate
xr = Mole fraction of least volatile component at reference point
K = Equilibrium constant for the least volatile component
S = Slope of enriching section operating line
Here, in this example, Eq. (8.21) can be used to find the number of theoretical
stages required to purify the bottom product (acetic acid) from xr′ = 0.02 to xW =
1.666 × 10–4.
S′ = Slope of stripping section operating line = 1.2
K′ = Slope of equilibrium line in the portion from 0.02 to 0. [From Fig. 8.3]
= 2.5

Ns =

Ns =

As discussed earlier, second graph with different scale can be drawn for finding
theoretical stages in the stripping section below xr = 0.02. Actual plotting also
gives additional stages of 7.
Total number of theoretical stages required for desired separation = 16 + 7.33 =
23.33, say 24.
Note: It can be seen that nearly 30% theoretical stages are required to reduce
the water content in the bottom product from 0.02 mole fraction to 50 mg/L.
As stated earlier for the separation of pure acetic acid from its dilute aqueous
solution containing less than 30% by mass acetic acid, liquid-liquid extraction
and azeotropic distillation are more economical than ordinary distillation. Figure
8.4 shows that for separating nearly pure water and pure acetic acid from
25.18% by mass acetic acid solution, 22 + 7 = 29 theoretical stages required for
the value of reflux ratio R = 3.5823. Comparison between the separation of 80%
acetic acid solution and 25.18% acetetic acid solution is given in Table 8.8.
 Fig. 8.4 McCabe–Thiele Diagram for Acetic Acid-Water System
Table 8.8 Comparison between Two Cases
Feed flow rate, kg/h % by mass acetic acid in feed R N Top vapour kg /h λav kJ/kg Condenser duty, kW Reboiler duty, kW

12 000 80 4.2 23 3000 1886.7 8 175.7 9 000

12 000 25.18 3.5813 29 8978 2256.9 25 785.7 29 000

(Condenser duty, ϕC = (R + 1) D λAC)

Table 8.8 shows that for the separation of dilute solution, about 20% more stages
and approximately three times higher condenser duty and reboiler duty are
required. Hence, for the separation of dilute aqueous acetic acid solution liquid-
liquid extraction or azeotropic distillation is more economical.
(b) Use of Open Steam in Distillation While distilling an aqueous solution, if
non-aqueous component is more volatile then water is removed as residue
product. In such a case, instead of using separate reboiler at bottom, heat
required can be provided by admission of steam directly to the bottom of tower.
For the same value of reflux ratio and the given degree of separation, use of open
steam requires more number of trays or equilibrium stages. However, it is
usually cheaper than the separate reboiler system as it avoids fixed cost and
cleaning cost of reboiler. It also facilitates the use of lower pressure steam as
heating medium but with the use of open steam, condensate cannot be recycled
back to boiler feed tank.
Overall material balance,
F + S = D + W   (8.23)
where, S = Molar flow rate of saturated steam entering at bottom, kmol/h
More volatile component balance
FzF = D xD + W xW  (8.6)
In this case stripping section operating line starts from (x = xW, y = 0) point and
has a slope equal to L–/S or W/S.
where, L– = Molar flow rate of liquid in stripping section, kmol/h
= W, molar flow rate of residue, kmol/h
If superheated steam is used as open steam, then S and L– are calculated by the
following equations:
S = Ss   (8.24)

L– = S – Ss + W   (8.25)
where, Ss = Molar flow rate of superheated steam, kmol/h
Hs = Enthalpy of superheated steam, kJ/kmol
Hsat = Enthalpy of saturated steam, kJ/kmol
λ = Latent heat of vaporisation at operating temperature, kJ/kg
M = Molar mass of water = 18.0153 kg/kmol
Design of distillation column with open steam can be better explained with
following example.
Aqueous solution of ethanol contains 6% by mass ethanol. It is
to be concentrated in the distillation column to 30% ethanol (by mass). Open
steam, saturated at 80 kPa g pressure will be used for the distillation column.
Feed flow rate is 5000 kg/h and it is saturated liquid at its bubble point. Residue
should not contain more than 0.02 mass per cent ethanol. Sparging steam flow is
0.2 kg per kg of feed. Calculate the number of theoretical stages required for the
desired separation.
Solution
Basis: 5000 kg/h feed
F =

= 6.52 + 261.11 = 267.63 kmol/h

S = (0.2 × 5000)/18 = 55.556 kmol/h
F + S = D + W  (8.23)
D + W = 323.186 kmol/h
zF = = 0.024 37

xD = = 0.1436

xW = = 7.827 × 10–5

FzF = D xD + W xW  (8.6)
267.63 × 0.024 37 = 0.1436 D + 7.827 10–5 W
6.522 = 0.1436 D + 7.827 × 10–5 (323.186 – D)
D = 45.266 kmol/h
W = 277.92 kmol/h
Feed is saturated liquid at its bubble point. Hence, q = 1
L– = L + Fq = L + F = W
G– = G + F(q – 1) = G
S = G– = G = 55.556 kmol/h
G = (R + 1) D
55.556 = (R + 1) 45.266
R = 0.2273
Table 8.9 Equilibrium Data for Ethanol-Water System at 101.3 kPa
x 0.019 0.0721 0.0966 0.1238 0.1661

y 0.17 0.3891 0.4375 0.4704 0.5089

(Table 13.1 of Ref. 8)

Intercept of enriching section

operating line = xD/(R + 1) = 0.1436/(0.2273 + 1) = 0.117
Slope of stripping section operating line = L–/S = W/S = = 5
Stripping section operating line starts from (xW, 0) point.
From Fig. 8.5, number of equilibrium stages required is N = 7.

Fig. 8.5 McCabe–Thiele Diagram for Steam Stripping of Ethanol-Water Solution

(c) Ponchon–Savarit Method This method is more rigorous than McCabe–Thile

method. This method requires detailed enthalpy data and it is particularly applied
to a binary system in which heat of solution is significant. This method is
explained by following illustration.
Determine the minimum reflux ratio and number of theoretical
stages required for the distillation column of ammonia-water absorption
refrigeration plant based on following data. Also find the heat duties of overhead
condenser (ϕC) and reboiler (ϕB) for the given value of reflux ratio R.
Data: (Example 8.4 of Ref. 14)
(a) Feed:
Feed is strong aqueous (SA) ammonia solution which enters the distillation
column at 102.5°C.
Enthalpy of feed = 325.2 kJ/kg
Ammonia in SA = 36.3% (by mass)
Feed flow rate = 7523.28 kg/h
(b) Distillate: 100% pure ammonia in the form of saturated liquid at 40°C
(c) Residue: Weak aqueous (WA) solution of ammonia, having 29.8% by mass
(d) Operating pressure of column = 1555.5 kPa a (assumed constant throughout
the column)
(e) Cooling water at 32°C is used as cooling medium in overhead condenser and
saturated steam at 3 bar a is used as heating medium in generator (reboiler).
(f) H–x–y data (Fig. 5.19, of Ref. 8)
Table 8.10 Enthalpies of Saturated Liquid and Vapour Mixtures
x, y HL, kJ/kmol HG, kJ/kmol

0 15 480.7 50 310.7
0.15 10 887.3 48 194.8
0.30 7 519.0 45 665.1
0.40 5 808.0 43 985.9
0.45 5 614.9 42 997.0
0.50 4 900.0 42 078.8
0.60 4 523.2 40 024.9
0.70 5 189.7 38 234.3
0.80 6 449.3 36 122.0
0.90 7 181.6 33 334.7
1.00 9 159.6 27 832.4

[Reference states: Enthalpies of pure saturated liquids at triple points (0.01°C for
water and –77.8°C for ammonia) = 0 kJ/kmol]
where, HL = Enthalpy of saturated liquid at 1555.5 kPa a pressure, kJ/kmol
HG = Enthalpy of saturated vapour at 1555.5 kPa a pressure, kJ/kmol
x, y = Mole fraction of ammonia in ammonia-water liquid solution and in
ammonia–steam mixture, respectively.
(g) VLE data at pt = 1555.5 kPa a
(Table 3.23 and Table 3.24 of Ref. 8)
Table 8.11 Equilibrium Data for Ammonia-Water System
x y

0 0
0.30 0.9120
0.35 0.9456
0.40 0.9690
0.45 0.9800
0.50 0.9880
0.60 0.9950
0.75 0.9980
 1.00 1.0000

(h) Actual or operating reflux ratio, R = 0.3137 which is nearly equal to 0.08 kg
per kg ammonia (as NH3) in the feed.
Solution
Material balance
Feed flow rate, ṁF = 7523.28 kg/h

xF = = 0.3763

Average molar mass of feed

Mav = Σ Mi xi
Mav = 17 (0.3763) + 18 (1 – 0.3763) = 17.6237 kg/kmol
F = = 426.88 kmol/h

xW = = 0.31

xD ≅ 1
F = D + W or 426.88 = D + W(a)
Ammonia balance
FxF = D xD + W xW
426.88 × 0.3763 = D + 0.31 W(b)
Solving Eq. (a) and (b),
D = 41.0176 kmol/h, W = 385.8624 kmol/h
Minimum reflux ratio, Rm
Refer to Fig. 8.6(a) and (b). To find Rm, in H – x, y diagram, draw the tie–lines.
A line connecting equilibrium values of x and y is called a tie line. Extend the tie
lines and allow to intersect x = xD and x = xW lines. Intersection point with x = xD
is represented as ΔD and with x = xW as ΔW. Extended tie lines which give the
farthest locations of ΔDm and ΔWm and are corresponding to minimum reflux
ratio. But if x, y equilibrium curve is everywhere concave downward, then a tie
line which when extended passes through F corresponds to minimum amount of
reflux.
Here x, y equilibrium curve is everywhere concave downward. Hence, the
minimum reflux ratio is established by the tie line which when extended passes
through F.
Mark the point F in H–x, y diagram. xF = 0.3763
HF = 325.2 × 17.6237 = 5731.23 kJ/kmol
Find the tie line which when extended, passes through point F. One method is,
draw the two nearest tie lines from the given equilibrium data. Here, these two
tie lines are (i) x = 0.35 and y = 0.9456, and (ii) x = 0.4 and y = 0.969. Then draw
the line having average slope of two tie lines and passes through point F. Extend
this tie line and allow it to intersect x = xD line. Intersection point is ΔDm.
From Fig. 8.6 at ΔDm, Hm = 31 818 kJ/kmol.

Rm =

(a) Enthalpy-Concentration Diagram for Aqueous Ammonia Solution

 (b) Theoretical Stage Calculations for Distillation of Aqueous Ammonia Solution
Fig. 8.6 Separation of Ammonia from Aqueous Solution

Rm = = 0.213 44

Actual R is 0.3137. Thus actual reflux ratio is 1.47 times the minimum reflux
ratio.
For R = 0.3137
=

0.3137 =

H = 33 690 kJ/kmol
Mark ΔD on x = xD line corresponding to the value of H.
FHF = DH + WH ′
426.88 × 5731.23 = 41.0176 × 33 690 + 385.8624 H ′
H ′ = 2759.18 kJ/kmol
Mark ΔW on x = xW line corresponding to value of H′. ΔD, F and ΔW are on the
same line. In Fig. 8.6(a), ΔD at xD = 1, H = 33 690 kJ/kmol and ΔW at xW = 0.31,
H ′ = 2759.18 kJ/kmol are plotted. Auxiliary random lines from the Δ points are
drawn and allowed to intersect the saturated vapour and saturated liquid curves
at values of y and x, respectively. These (x, y) points are the points of operating
curves. Operating curves are plotted in Fig. 8.6(b). Stagewise construction from
xD to xW between equilibrium curve and operating curve gives number of
theoretical stages required for the desired separation.
Table 8.12 x, y Values for Enriching Section Operating Curve and Stripping Section Operating Curve
Enriching Stripping
x y x y

1.00 1.000 0.381 0.942

0.95 0.990 0.373 0.900
0.90 0.982 0.360 0.800
0.80 0.970 0.350 0.700
0.70 0.962 0.339 0.600
0.60 0.955 0.329 0.500
0.50 0.949 0.324 0.400
0.381 0.942 0.310 0.310

Here stagewise construction is started from xD = 0.999 72 to xW = 0.31. Total

number of theoretical stages required (as shown in Fig. 8.6(b)) is 5.36.
N = 5.36.
H = HD + = 9159.6 +

ϕC = 1006 178.1 kJ/h ≡ 279.493 kW - condenser duty

H′ = HW – = 7347.86 –

ϕB = 1770 599.1 kJ/h ≡ 491.83 kW - reboiler duty

Methods Valid for Multicomponent Distillation To determine by number of
theoretical stages or equilibrium stages required for desired separation in
multicomponent distillation many methods are available. None of them is totally
reliable. Hence, for a new case of distillation, it is recommended to use more
than one method for the determination of theoretical stages. Also it is better to
design the distillation column, first for pilot plant and then modify the design
based on the results of pilot plant column rather than designing the column
directly for the commercial scale plant. Methods available for multicomponent
distillation can be broadly divided into two groups:
1. Short cut methods: FUG method, FUE method, Smith–Brinkley method
2. Rigorous methods: Lewis Matheson method, Thiele–Geddes method,
Relaxation methods, Tridiagonal Matrix method, Linear Algebra methods
FUG Method (Fenskey–Underwood–Gilliland’s Method)8 It is the most
widely used short-cut method for determining number of equilibrium stages for
multicomponent distillation.
In this method, first minimum reflux ratio Rm is determined by Underwood’s
method, which was explained in the Section 8.4.4.
Then Fenskey’s equation [Eq. (8.26)] is used to find the minimum number of
theoretical stages required for the desired separation.

Nm =   (8.26)

where, αLK = Average relative volatility of light key with respect to heavy key
(xLK, xHK)d = Mole fractions of light key and heavy keys in distillate
(xLK, xHK)b = Mole fractions of light key and heavy keys in residue
If there is a wide difference in the values of αLK at the topmost temperature and
at bottommost temperature, then use Eq. (8.27). Fenskey’s equation in such
cases gives lesser value of Nm than actual Nm required, if αLK is determined by
Eq. (8.8).
αLK = (αLK, top × αLK, bottom)1/2  (8.8)
αLK = (αLK, top × αLK, feed × αLK, bottom)1/3   (8.27)
where, αLK, feed = relative volatility of light key with respect to heavy key at the
bubble point of feed.
After finding the values of Rm and Nm, number of theoretical stages required or
number of equilibrium stages required for desired separation can be determined
by Gilliland’s correlation. It was originally presented in terms of graph. Many
attempts are made to represent Gilliland’s correlation empirically. One of the
empirical equations for Gilliland’s correlation is developed by Malokanov8.

f (N) = = 1 – exp   (8.28)

where, ψ =   (8.29)

Figure 8.7 is the plot representing Gilliland correlation.

Fig. 8.7 Gilliland Correlation

(Reproduced from Perry’s Chemical Engineeris’ Handbook, 6th Ed., McGraw-Hill, USA)

Determine the number of theoretical stages required for desired

separation by FUG method for the case given in Example 8.4.
Solution
Minimum reflux ratio determined by Underwood’s method in the Example 8.4.
Rm = 1.4509
Using Fenskey’s equation to find the minimum number of theoretical stages
required for desired separation, Nm for total reflux.

Nm =   (8.26)

Nm =

Nm = 4.117
Using Eqs.   (8.28) and   (8.29), Table 8.13 is prepared.
Table 8.13 Values of N for Different Values of R
R ψ f(N) N

1.7 0.092 26 0.5616 10.67

2.0 0.183 03 0.4752 8.75
2.5 0.299 74 0.3811 7.27
3.0 0.387 30 0.3198 6.52
3.5 0.455 36 0.2761 6.07
4.0 0.509 82 0.2433 5.76

Above the value of reflux ratio R = 3, there is a tapering in values of N. Hence,

the optimum reflux ratio is nearer to R = 3.
Using Eq. (8.29), ψ = 0.3873. Substituting ψ in Eq. (8.28),
Number of theoretical stages required for the desired separation N = 6.52.
For ψ = 0.3873;
Read f(N) = = 0.32 from Fig. 8.7

N = 6.525
A fatty acid mixture contains 11% palmitic acid, 4% stearic
acid, 28.5% oleic acid and rest linoleic acid (by mass). It is to be distilled to
separate in pure components in distillation columns operating at 2 Torr absolute
at top.
 Fig. 8.8(a) Separation of Fatty Acids
Table 8.14 Pressure vs Boiling Point Data of Fatty Acids
Absolute Pressure Boiling Point, °C

kPa Torr Palmitic acid Stearic acid Oleic acid Linoleic acid

0.133 1 167.4 183.6 177.6 178.5

0.267 2 179.0 195.9 189.5 190.1
0.533 4 192.2 209.2 202.6 202.8
1.07 8 206.1 224.1 217.0 216.9
2.13 16 221.5 240.0 232.9 232.6
4.27 32 238.4 257.1 250.6 250.0
8.53 64 257.1 276.8 270.3 269.7
17.1 128 278.7 299.7 292.5 291.9

34.1 256 303.6 324.8 317.7 317.2

68.3 512 332.6 355.2 346.5 346.5
 101.325 760 351.5 376.1 364.9 365.2

(Table 2 of Ref. 15)

Table 8.15 Chemical Formulas and Molar Masses of Fatty Acids
Fatty acid Chemical formula Molar mass Melting point, °C

Palmitic acid C16H32O2 256.42 63.1

Stearic acid C18H36O2 284.48 69.6

Oleic acid C18H34O2 282.46 16.3

Linoleic acid C18H32O2 280.45 –5.0

Feed is a saturated liquid. Determine the number of theoretical stages required

for the desired separation.
Solution
Boiling points of oleic acid and linoleic acid are very close. Hence, they are near
impossible or very difficult to separate by distillation. Study of melting points of
fatty acid (Table 8.15) suggests that fractional crystallization can be explored for
separation of oleic and linoleic acids. It is planned to separate the following three
products from the feed mixture.
Products Purity in % by mole
1. Palmitic acid 99%
2. Mixture of linoleic and oleic acids 97% (Combinedly)
3. Stearic acid 95%
There are three (N′) products. Hence, two distillation columns (N′ – 1) are
required as discussed in Section 8.4.1. First column will separate 99% by mole
pure palmitic acid as top product and remaining as bottom product. Second
column will give 97% by mole pure (combinedly) mixture of linoleic and oleic
acid as the top product and 95% by mole stearic acid as the bottom product. Both
columns will be operated under vacuum and operating pressure at the top of both
columns will be 2 Torr a.
Determination of number of theoretical stages for first distillation column
This column operates as a multicomponent (3) distillation column. Top product
is nearly pure palmitic acid. At 179°C vapour pressure of palmitic acid is 2 Torr
(Table 8.14). Hence, top temperature of column will be very close to 179°C. Let
operating pressure at the base of first column be 30 Torr a. (This is decided
based on trial calculations). Bottommost temperature of distillation column is
bubble point temperature of residue. To carry out the material balance and to
find the feed composition in mole %, let the feed flow rate be 1000 kg/h (basis).
Mass flow rate of feed, ṁF = 1000 kg/h
 Table 8.16 Feed Composition
Component mass % Molar mass kmol/h mole %

Palmitic acid 11 256.42 0.428 98 11.94

Stearic acid 4 284.48 0.140 61 3.91
Oleic acid 28.5 282.46 1.008 99 28.08
Linoleic acid 56.5 280.45 2.014 62 56.07

Total 100.0 3.593 20 100.00

Average molar mass of feed, MF = 1000/3.5932

= 278.3
Table 8.17 Composition of Distillate and Residue of First Column
Distillate, D Bottom, W
Component
kmol/h kg/h mole % kmol/h kg/h mole %

Palmitic acid 0.3971 101.824 99.0 0.031 88 8.175 1.00

Stearic acid — — — 0.140 61 40.001 4.40
Oleic acid 0.002 0.565 0.5 1.006 99 284.434 31.55
Linoleic acid 0.002 0.561 0.5 2.012 62 564.439 63.05
Total 0.4011 102.950 100.0 3.192 10 897.049 100.00

Vapour pressures of oleic acid and linoleic acid are very close. Hencte, in
distillation calculations they are considered as one product. For first column
palmitic acid is the light key and mixture of oleic acid-linoleic acid is the heavy
key.
Component balance around first distillation column: Palmitic acid balance.
0.1194 F = 0.99 D + 0.01 W
Let mole % of palmitic acid in residue = 1%. (refer to Table 8.17)
0.1194 × 3.5932 = 0.99 D + 0.01 (3.5932 – D)
D = 0.4011 kmol/h, W = 3.1921 kmol/h
At bubble point temperature of residue,
Σxwi pvi = pt = 30 Torr
Trial temperature = 250°C
Table 8.18 Vapour Pressures as Read from Fig. 8.8(b)
Vapour pressure, Torr
Component
At 179°C At 250°C

Palmitic acid 2 51.71

Stearic acid 0.714 23.29
Linoleic acid/Oleic acid 0.903 31.43
 Fig. 8.8(b) Vapour Pressures of Fatty Acids based on Table 8.14

Using Fig. 8.8(b) and Table 8.17,

0.01 × 51.71 + 0.044 × 23.29 + (0.3155 + 0.6305) × 31.43 = 31.27 Torr ≅ 30
Torr
Hence, bottommost temperature of first distillation column ≅ 250°C
Table 8.19 Relative Volatility Calculations
Component αtop αbottom αav = (αtop · αbottom)0.5

Palmitic acid (LK) 2.2148 1.6452 1.9089

Stearic acid 0.7907 0.741 0.7654
Linoleic acid/Oleic acid (HK) 1 1 1

Here FUG method is used to find the number of theoretical stages required for
desired separation.
Minimum number of theoretical stages required by using Fenskey’s method, Nm:

Nm =   (8.26)

= = 14.144
Underwood’s method

= 1 – q  (8.15)

Feed is saturated liquid. Hence, q = 1

= 1 – 1 = 0

Solving by Mathcad,
For ϑ = 1.719 32, f (ϑ) = 0.001 ≅ 0
Also, ϑ ∈ {1, ..., 1.9089}

= Rm + 1

= Rm + 1

Rm = 8.9545
Gilliland’s correlations; Eqs.   (8.28) and   (8.29):
Table 8.20 Evaluation of Parameters
R ψ f(N) N
10 0.095 0.5586 33.31
12 0.234 0.4318 25.65
15 0.378 0.326 21.47
17 0.447 0.2813 20.07

Let, R = 15, for which N = 21.47 ≅ 22

For ψ = = 0.378

From Fig. 8.7 = 0.3062 or N = 20.83

FUG method does not give information about feed tray location. Feed tray
location can be determined by using following equation (Kirkbride equation)10:

log =   (8.30)

where, Nr = Number of stages above the feed

Ns = Number of stages below the feed

log = 0.206 log

= 2.292

Nr + Ns = 22
3.292 Ns = 22
Ns = 6.683 ≅ 7, Nr = 15
Hence, seventh stage from bottom will be the feed stage.
Second column acts as a binary distillation column. For second distillation
column W = F ′ (as shown in Fig. 8.7(a)), where F ′ is molar flow rate of feed to
second distillation column.
Average molar mass of feed, MF′ = 897.049/3.1921
= 281.02
Linoleic-Oleic acids balance around second column:
F ′ × (0.3155 + 0.6305) = 0.97 D ′ + 0.05 W ′
3.1921 × 0.946 = 0.97 D ′ + 0.05 (3.1921 – D ′)
2.860 12 = 0.92 D ′
D ′ = 3.1088 kmol/h, W ′ = 0.0833 kmol/h
Table 8.21 Composition of Distillate and Residue of Second Column
Distillate, D′ Bottom, W′
Component
kmol/h kg/h mole % kmol/h kg/h mole %

Palmitic acid 0.0319 8.180 1.026 — – —

Stearic acid 0.0615 17.496 1.978 0.0791
 22.502 95.00
Oleic acid 1.0056 284.042 32.347 0.0014 0.395 1.68
Linoleic acid 2.0098 563.648 64.649 0.0028 0.785 3.36

Total 3.1088 873.366 100.000 0.0833 23.682 100.00

Operating pressure at the top of second column = 2 Torr a.

Topmost temperature of distillation column is dew point temperature of
overhead vapour.
At dew point temperature
= 1

Let topmost temperature = 190.1°C (Trial) yi = xDi

At 190.1°C temperature, from Fig. 8.7(b) vapour pressure of components are as
follows.
Table 8.22 Vapour Pressures as Read from Fig. 8.8(b)
Vapour pressure, Torr
Component
At 190°C At 268.2°C

Linoleic/oleic acid 2 63.5

Stearic acid 1.64 50

Table 8.23 Relative Volatilities Calculations

Component αtop αbottom αav = (αtop · αbottom)0.5

Linoleic/oleic acid 1.22 1.27 1.245

Stearic acid 1 1 1

For the calculations of number of theoretical stages required for desired

separation feed to second column can be approximated as binary system.
F′ = 3.1921 , D′ = 3.1088 , W ′ = 0.0833 ,

zF = 0.946, xD = 0.97, xW = 0.05

Vapour-liquid equilibrium data can be determined by using following equation:
y =

Table 8.24 VLE Data for Second Column

x y

0 0
0.2 0.2374
0.4 0.4536
 0.6 0.6513
0.8 0.8328
1.0 1.0000

Feed is saturated liquid; q = 1.

Fig. 8.8(c) McCabe–Thiele Diagram for Second Column

From Fig. 8.7(c)

= 0.4301

Rm = 1.2553
Let R = 3 × Rm = 3.7659

= = 0.203 53

For R = 3.7659, number of theoretical stages required for desired separation, N =

33.
Feed tray location = 4th equilibrium stage from top
Rigorous Methods Rigorous or detailed methods available for multicomponent
distillation for finding or verifying number of theoretical stages required for
desired separation are as follows.
(a) Lewis–Matheson method
(b) Theile–Geddes method
(c) Relaxation methods
(d) Equation tearing procedures using tridiagonal matrix algorithm.
Among these methods, only Lewis–Matheson method calculates the number of
theoretical stages. In other methods, number of equilibrium stages required
above and below the feed point must be specified by the user. In other methods,
for the given number of theoretical stages, product compositions are determined.
(a) Lewis–Matheson Method13
In this method, constant molar overflow is assumed. In other words, it is
assumed that for each stage, molar latent heat of vaporization of liquid phase is
equal to latent heat of condensation of vapour phase. Hence, moles of liquid
vaporized over each theoretical stage is equal to moles of vapour condensed on
that stage. This implies that, molar flow rates of liquid and vapour in enriching
section (L and V) and also in stripping section (L– and V–) remain constant. To
use this method following information must be specified:
(i) Feed flow rate, conditions and compositions
(ii) One product flow (D or W)
(iii) Product compositions or distribution of key and non-key components
(iv) Reflux ratio
(v) Column pressure
Stepwise Lewis–Matheson method is as given below:
• Assuming that total condenser is used. L = RD and V = (R + 1)D. Determine
the dew point temperature of overhead vapour by using the equation,
= 1 (For non-ideal system)  (8.3)

or = 1 (For ideal vapour-liquid equilibrium)  (8.1)

where, yNi = xDi = xoi

yNi = Mole fraction of component i in vapour phase leaving the Nth stage
xDi = Mole fraction of component i in distillate
xoi = Mole fraction of component i in reflux
pt = Operating pressure at top of distillation column, kPa a
pvi = Vapour pressure of component i at dew point temperature of overhead
vapour, kPa a
φi = Fugacity coefficient of component i in vapour phase
γi = Activity coefficient of component i in liquid phase
Same calculations (dew point temperature calculation) also give the value of xNi

xNi = (For ideal VLE) – Raoult’s Law  (8.1)

where xNi = mole fraction of component i in liquid phase leaving the Nth tray.
Composition of vapour leaving the (N – 1)th tray can be determined by a material
balance around the top section as shown in Fig.   (8.9).
V y(N – 1)i = LxNi + DxDi  (8.31)

Fig. 8.9(a) Material Balance of Top Tray

 Fig. 8.9(b) Material Balance for (N–1)th Tray

By using above equation y(N – 1)i is determined. After finding y(N – 1)i, i.e.,
composition of vapour leaving the (N – 1)th tray, find the dew point of this
vapour.
= 1 (For ideal VLE)  (8.32)

Same calculations also give the values of x(N – 1)i, composition of liquid phase
leaving the (N – 1)th tray.
Material balance around (N – 1)th tray gives composition of vapour leaving the
(N – 2)th tray.
Vy(N – 2)i = Vy(N – 1)i + Lx(N – 1)i – LxNi   (8.33)
These calculations (i.e., dew point temperature calculations and material balance
calculations) are repeated up to the feed tray.
• Feed Tray Identification To identify the feed tray following conditions must be
satisfied.
If feed is liquid then mole fraction ratio or molar ratio of light key to heavy key
in the feed should lie in between the same ratios of the liquid leaving the feed
tray and of the liquid leaving the tray, next above the feed tray.

or ≤   (8.34)
If the feed is a mixture of liquid and vapour then also the above condition is
valid. However, the difference is only liquid phase molar ratio is to be
considered for satisfying this condition.
If feed is vapour then condition is

or   (8.35)

In a distillation column, any tray from top to bottom can be the feed tray. Hence,
in tray-to-tray calculations from top, this condition must be checked for each
tray.
• For Stripping Section:
Molar flow rate of liquid, L– = L + qF
Molar flow rate of vapour, V– = L– – W
where, W = Molar flow rate of residue
Calculate the bubble point temperature of residue using either Eq. (8.3) or Eq.
(8.1).
Same calculations also give the values of yWi composition of vapour leaving the
reboiler.

yWi =   (8.36)

The material balance around the reboiler gives

L–x1i = V–yWi + WxWi  (8.37)
Calculate the bubble point temperature of liquid leaving the first tray
= 1 (For ideal VLE)  (8.38)

where, pt′ = Pressure in reboiler

= pt + Δp
Same calculations also give y1i, composition of vapour leaving the first tray.
Then material balance around 1st tray gives x2i.
L–x2i = L–x1i + V–y1i – V–yWi  (8.39)
x2i is the composition of liquid leaving the second tray. These iterative
calculations are repeated from bottom to feed tray. For the liquid leaving the
feed tray xLK /xHK ratios, one obtained from enriching section calculations and
other obtained from stripping section calculations, must be very close.

Fig. 8.10 Material Balance around Reboiler

• Normally, in the first trial calculations, it happens that on getting close values
of the ratios xLK /xHK, one from enriching section and other from stripping
section, mole fractions of the components of liquid leaving the feed tray, one
obtained from top section and other from bottom section do not match. If they
are different then it is called “mismatch”.
If composition of liquid phase, leaving the feed tray, obtained from top section
calculations and the same from bottom section calculations are different, then to
match them one has to change product distribution and product rates. In other
words one has to change xDi, xWi, D and W. For the new values, entire
calculations must be repeated until the entire composition of liquid leaving the
feed tray calculated from top and bottom closely match.
In the first trial calculations, if composition of liquid leaving the feed tray from
top to down calculations and the same from bottom to up calculations are not
matching, then Smith suggests the following method to make the change in
product composition for second trial calculations12:
Change in moles of component i in distillate for second trial calculations, Δdi is
calculated by following equation:
Δdi =   (8.40)

where, xFBi = Mole fraction of component i in the liquid from feed tray as found
from the bottom to up calculations
xFDi = Mole fraction of component i in the liquid from feed tray as found from
the top to down calculations
di, bi = Moles of component i in top and bottom products in last iteration
Then corresponding changes required in bottom product for component i is Δbi =
– Δdi. For new distribution, entire calculation is repeated. These calculations are
continued until compositions of liquid, leaving the feed tray, obtained from top–
down calculations and bottom–up calculations are approximately same.
For more exact calculations, enthalpy balance should also be done for each tray
to find the variation in liquid rate and vapour rate on each tray. But these
calculations require detailed and exact enthalpy data. Also, if one considers it,
then calculations become more laborious and complicated. Use of a computer is
then essential.
A saturated liquid, consisting of phenol and cresols with some
xylenols, is fractioned to give a top product of 95.3 mole % phenol. Metacresol
is heavy key and phenol is light key component. Total condenser is used. The
composition of the top product and of the phenols in the bottoms are given in
Table 8.25.
(a) Complete the material balance over the still for a feed rate of 100 kmol/h.
(b) Calculate the minimum reflux ratio by Underwood’s method.
(c) For R = 3Rm, calculate the composition of vapour entering to the topmost tray
by Lewis–Matheson method.
Table 8.25 Distillation Column Data and Relative Volatilities
Component αav Feed, mole % Top product, mole % Bottom product, mole %

Phenol 1.98 35 95.30 5.24

o–Cresol 1.59 15 4.55 ?
m–Cresol 1.00 30 0.15 ?
Xylenol(s) 0.59 20 — ?

Total 100 100.00

Solution
(a) Overall material balance
F = D + W
100 = D + W(1)
Phenol balance:
F × 0.35 = D × 0.953 + W × 0.0524(2)
From Eqs. (1) and (2), D = 33 kmol/h, W = 67 kmol/h.
Component balance gives the composition of residue.
o–Cresol balance:
F × 0.15 = 0.0455 × D + xWo × W
100 × 0.15 = 0.0455 × 33 + xWo × 67
xWo = 0.2015
Similarly, xWm = 0.447, xWx = 0.2985
where, xWm = mole fraction of m–cresol in residue
xWx = mole fraction of xylenol in residue
(b) Rm by Underwood’s method

= Rm + 1(3)

q = 1.0 (saturated liquid feed)

= 1 – 1 = 0(4)

By trial-and-error calculations, or by Mathcad, determine ϑ.

Also ϑ ∈ {1, … 1.98}.
ϑ = 1.22
On substituting this value of ϑ in Eq. (3) gives
Rm = 1.67
(c) R = 3 Rm = 3 × 1.67 = 5.01
L = RD, V = (R + 1) D
L = 5.01 × 33 = 165.33 kmol/h
V = (5.01 + 1) × 33 = 198.33 kmol/h
Refer to Fig. 8.9(a).
Component balance around top segment
Vy(N–1)i = LxNi + DxDi
Composition of vapour entering to topmost tray
 y(N – 1)i = xNi + xDi (5)

At dew point of overhead vapour

= ΣxNi = 1 (6)

But here operating pressure pt is not specified. So xNi cannot be determined by

Eq. (6). xNi can be determined based on the data of average relative volatility,
αavi.
ΣxNi = xNP + xNO + xNm + xNx = 1
where, xNi = Mole fraction of component i in the liquid leaving the Nth tray
xNP = Mole fraction of phenol in the same
xNo = Mole fraction of o–cresol in the same
xNm = Mole fraction of m–cresol in the same
xNx = Mole fraction of xylenol in the same

Average relative volatility of phenol with respect to heavy key component,

αavP =

= (7)

= (8)

From Eqs. (7) and (8):

xNP =

xNP = 0.9411
Since total condenser is used, yNi = xDi
Similarly,

xNo = = 0.055 95

xNm = 0.002 95
On substituting the values of xNi in Eq. (5)

y(N – 1)i = × xNi + × xDi

y(N – 1)P = × 0.9411 + × 0.953

y(N – 1)P = 0.8336 × 0.9411 + 0.1664 × 0.953

= 0.943
y(N – 1)o = 0.8336 × 0.055 95 + 0.1664 × 0.0455 = 0.0542
y(N – 1)m = 0.8336 × 0.002 95 + 0.1664 × 0.0015 = 0.0027
Find the number of theoretical stages required for the desired
separation by Lewis–Matheson method for the following feed mixture.
Table 8.26 Distillation Column Details
Component Feed, mole % Distillate, mole % Residue, mole %

n-Butane 37 95.0 16.3

i-Pentane 32 5.0 41.6
n-Pentane 21 0.0 28.5
n-Hexane 10 0.0 13.6
Total 100 100.0 100.0

In Table 8.26, the distillate and residue compositions are specified for the first
trial calculations. These are not the final values.
1. F = 4750 mol/h, D = 1250 mol/h, W = 3500 mol/h and q = 1
2. n–Butane is light key component and i–Pentane is heavy key component.
3. Operating pressure pt = 2 atm a
4. Reflux ratio R = 3
5. Vapour pressures of components can be calculated using Antoine equation.
ln pv = A – , pv in Torr and T in K(2.1)

Table 8.27 Antoine Constants10

Component A B C

n-Butane 15.6782 2154.9 – 34.42

i-Pentane 15.6338 2348.67 – 40.05
n-Pentane 15.8333 2477.07 – 39.94
n-Hexane 15.8366 2697.55 – 48.78

Solution
Lewis–Matheson method
Molar flow rates of liquid and vapour in enriching section
L = RD, V = (R + 1)D
L = 3 × 1250 = 3750 mol/h, V = (3 + 1) × 1250 = 5000 mol/h
For total condenser,
yNi = xDi = xoi
To find the dew point of overhead vapour
= 1

For T = 295 K (i.e., t = 22°C), pvB = 1650 Torr, pviP = 615 Torr

≅ 1

or 0.875 15 + 0.1236 = 0.998 75 ≅ 1

Hence, dew point temperature of overhead vapour, tDP = 22°C.
The same calculations also give
xNB = 0.875 15
xNiP = 0.1236
Material balance around the top segment, shown in Fig. 8.9(a).
Vy(N – 1)i = LxNi + DxDi

y(N – 1)i = xNi + xDi

y(N – 1)B = × xNB + × xDB

y(N – 1)B = 0.75 × 0.875 15 + 0.25 × 0.95 = 0.8939

y(N – 1)iP = 0.75 × 0.1236 + 0.25 × 0.05 = 0.1052
Find the dew point temperature of vapour leaving the (N – 1)th tray by trial-and-
error calculations.
At t = 24°C or T = 297 K
pviB = 1757.3 Torr, pviP = 660.65 Torr

= = 1.012 ≈ 1

Dew point temperature of vapour leaving the (N – 1)th tray is 24°C.

The same calculations also give
x(N –1)B = 0.77 and x(N –1)iP = 0.24
Here Σx(N –1)i = 1.01. To make this total = 1.000, normalised values of x(N – 1)i are
determined as follows.
x(N –1)B = = 0.7624

x(N –1) iP = = 0.2376

Check whether the (N – 1)th tray is feed tray or not.

= (= 1.156)

Hence, (N – 1)th tray is not the feed tray.

Component balance around (N – 1)th tray
Refer to Fig. 8.9(b):
Vy(N – 2)i = Lx(N – 1)i + Vy(N – 1)i – LxNi
y(N – 2)i = x(N – 1)i + y(N – 1)i – xNi

y(N – 2)B = 0.75 × 0.7624 + 0.8911 – 0.75 × 0.875 15

y(N – 2)B = 0.8065, y(N – 2)iP = 1– y(N – 2)B = 0.1935
Find the dew point temperature of vapour leaving the (N – 2)th tray.
At, T = 301 K or t = 28° C, pvB = 1987.6 Torr, pviP = 760 Torr

= 0.6168 + 0.387 = 1.0038 ≈ 1

The same calculations also give
x(N – 2)B = 0.6168 and x(N – 2)iP = 0.387
Normalised values of x(N – 2)i
x(N – 2)B = 0.6182, x(N – 2)iP = 0.3818
Check whether (N – 2)th tray is feed tray or not.

(N – 2)th tray is not the feed tray.

Component balance around (N – 2)th tray
Vy(N – 3)i = Lx(N – 2)i + Vy(N – 2)i – Lx(N – 1)i

y(N – 3)i = x(N – 2)i + y(N – 2)i – x(N – 1)i

y(N – 3)B = 0.75 × 0.6182 + 0.8065 – 0.75 × 0.7624 = 0.6984

y(N – 3)iP = 0.3016
Find the dew point of vapour, leaving the (N – 3)th tray.
At T = 305 K or t = 32°C, pvB = 2239.9 Torr, pviP = 870.63 Torr.

= = 1.00

= 0.4789 + 0.5266 = 1.0005 ≈ 1

Dew point temperature of vapour leaving the (N – 3)th tray, tDP = 32°C
The same calculations also give
x(N – 3)B = 0.47 and x(N – 3)iP = 0.53
Check (N – 3)th tray is feed tray or not.

= = 0.887 < 1.156

(N – 3)th tray or 4th tray from the top is the feed tray.
Molar flow rate of liquid and vapour in stripping or exhausting section:
L– = L + Fq = 3750 + 4750 = 8500 kmol/h
V– = L–– W = 8500 – 3500 = 5000 kmol/h
Correct mole fractions in liquid leaving the feed tray
xBF L– = xB L + x ′B F

xBF = 0.47 × = 0.414

(where x ′B is mole fraction of n–Butane in feed)

xiPF = 0.53 × = 0.413

xnPF = = 0.1173

xnHF = = 0.056

Composition of the liquid, leaving the feed tray obtained from top to down
calculations.
xBF = 0.412, xiPF = 0.413, xnPF = 0.1173, xnHF = 0.056

≅ 1 < 1.156

Bottom to up calculations:
Assume Δp ≈ 0, i.e., column operates at nearly same pressure from top to
bottom.
Find the bubble point of residue via trial-and-error calculations:

At t = 48°C or T = 321K
=

= 1.01 ≅ 1
Hence, bubble point temperature, tBP = 48°C
The same calculations also give the vapour composition, leaving the reboiler.
yWB = = 0.375
yWiP = 0.395, yWnP = 0.2098, yWnH = 0.033
Component balance around reboiler gives (Ref. Fig. 8.10):
L–x1i = V–yWi + WxWi
x1i =

= 0.5882 yWi + 0.4118 xWi

Composition of liquid, leaving the first tray
x1B = 0.5882 × 0.375 + 0.4118 × 0.163 = 0.2877
x1iP = 0.5882 × 0.395 + 0.4118 × 0.416 = 0.4036
x1nP = 0.5882 × 0.2098 + 0.4118 × 0.285 = 0.24
x1nH = 0.5882 × 0.033 + 0.4118 × 0.136 = 0.075

= = 0.7128, = 1.0036

Hence, and first tray (from bottom) is not the feed tray.

Find the bubble point of liquid leaving the first tray by trial-and-error
calculations.
At t = 41°C or T = 314 K
pvB = 2894.5 Torr, pviP = 1164.97 Torr, pvnP = 893.26 Torr
pvnH = 288.76 Torr, (by using Antoine constants)

= 1.012 ≈ 1 = y1i

Bubble point of liquid leaving the first tray, tBP = 41°C. Same calculations also
give the composition of vapour, leaving the first tray.
y1B = = 0.548

y1iP = 0.309, y1nP = 0.141, y1nH = 0.014

Normalised values of y1i
y1B = = 0.5415

y1iP = = 0.305

y1nP = = 0.139

y1nH = = 0.0138

Component balance around the first tray gives

L–x2i = V–y1i + L–x1i – V–ywi
x2i = y1i + x1i – ywi

x2B = × 0.5415 + 0.2877 – × 0.375 = 0.3856

x2iP = × 0.305 + 0.4036 – × 0.395 = 0.3506

x2nP = × 0.139 + 0.24 – × 0.2098 = 0.1984

x2nH = × 0.0138 + 0.075 – × 0.033 = 0.0637

= (= 1.025) – close enough

Second tray from bottom can be called the feed tray. Ideally values of this ratio
obtained from bottom to up calculations should be same as that obtained

from top to down calculation. Here composition of liquid leaving the feed tray
obtained from top to down calculations and the same obtained from bottom to up
calculations are not matching.
Table 8.28 Summary of First Trial Calculations of Feed Tray Liquid
Calculated mole fractions
Component
Composition based on top–down calculation Composition based on bottom–up calculation
n-Butane 0.412 0.3856
i-Pentane 0.413 0.3506
n-Pentane 0.1173 0.1984
 n-Hexane 0.056 0.0637

To match this composition and to get the exact value of number of theoretical
stages required for the desired separation of key components, second set of
calculations are required. Based on first trial calculations number of theoretical
stages required for the desired separation of key components is 6 (Reboiler, 1st ,
2nd = N – 3 = Feed tray, N – 2, N – 1, N).
How to start the second trial calculations?
To start the second trial calculations, product composition xDi and xWi (hence, D
and W also) must be changed. Based on the Smith’s suggestion following
equation can be used to make the change in product composition for the 2nd trial
calculations.
Change in moles of component in distillate.

Δdi =   (8.40)

For n–Butane or for nC4

ΔdnC4 = = –25.81 mol/h

nC4 in distillate for the second trial calculation = 1250 × 0.95 – 25.81 = 1161.69
mol/h
nC4 in residue = 0.163 × 3500 + 25.81 = 596.31 mol/h
Note that total of molar flow rates of each component should be constant.
For iso–Pentane or iC5:

ΔdiC5 = = –9.1 mol/h

iC5 in distillate = 1250 × 0.05 – 9.1 = 53.4 mol/h

iC5 in residue = 3500 × 0.416 + 9.1 = 1465.1 mol/h
For n-Pentane or nC5:
ΔdnC5 = = 0.6913 mol/h

In first trial calculation dnC5 = 0. But to use above equation one cannot put dnC5 =
0. Let dnC5 = 1 mol/h (negligible value)
nC5 in distillate = 1 + 0.6913 = 1.7 mol/h
nC5 in residue = 3500 × 0.285 – 1.7 = 995.8 mol/h
For n-Hexane or nC6:

ΔdnC6 =

Let dnC6 = 1 mol/h (negligible)

ΔdnC6 = 0.137 mol/h
nC6 in distillate = 1.137 mol/h
nC6 in residue = 3500 × 0.137 – 1.137 = 474.86 mol/h
Table 8.30 Product Composition for Second Trial Calculations
Distillate, D Residue, W
Component
mol/h mole fraction mol/h mole fraction

nC4 1161.69 0.954 596.31 0.169

iC5 53.4 0.044 1465.1 0.415

nC5 1.7 0.001 995.8 0.282

nC6 1.137 0.001 474.86 0.134

Total 1217.93 1.000 3532.07 1.000

Based on this new product distribution one can start the second set of
calculations. These trial calculations are repeated until the composition of liquid
leaving the feed tray obtained from top–down calculations match with the same
obtained from bottom–up calculations. Results obtained at the end of first trial
calculations are matching with the same given by FUG method.
(b) Thiele Geddes Method1:
Original method of Thiele and Geddes was developed for manual calculations in
1933. Thereafter computer programs are developed for this method to reduce the
labour of calculations. This method is a rating method in which distribution of
components between distillate and bottoms is predicted for a specified number
of stages.
In this method, following variables are to be specified as input data:
(i) Number of equilibrium stages above and below the feed.
(ii) Column pressure
(iii) Temperatures (stagewise) – Dew point temperature of vapour leaving the
equilibrium stage is equal to bubble point temperature of liquid leaving the
same stage and this temperature is called the stage temperature.
(iv) Reflux flow rate
(v) Distillate rate
(vi) Feed flow rate, feed condition, and feed composition
(vii) Liquid and vapour flow rates, leaving each stage
Calculation procedure is as follows.

Fig. 8.11 Material Balance around Overhead Condenser

Material balance for any component around the condenser is

G1 y1 = Lo xo + DzD

= + 1

For a total condenser, y1 = xo = zD = xD

and = R

= + 1 = R + 1 = Ao + 1

where, Ao = = R (for total condenser)

For a partial condenser, zD = yD = Ko xo

where, Ko = equilibrium constant for partial condenser

= = Ao + 1

where, Ao = (for partial condenser)

Component balance up to tray 1 gives

G2 y2 = L1 x1 + DzD

= + 1  (8.41)

Fig. 8.12 Material Balance around Top Section up to First Tray

For Tray 1, absorption factor

A1 =   (8.42)

L1 x1 = G1 y1 A1
From Eqs.   (8.41) and   (8.42),

= + 1

= (Ao + 1) Al + 1 = A1 Ao + A1 + 1  (8.41a)
Similarly, for the vapour coming from any nth tray of enriching section

= = A0 A1 A2 … An – 1
+ A1 A2 … An – 1 + … + An – 1 + 1  (8.43)
For the vapour coming from feed tray
= A0 A1 A2 … AF – 1 + A1 A2 … AF – 1 + … + AF – 1 + 1  (8.44)

For stripping or exhausting section:

Component balance around reboiler
= yN + 1 + WxW

= + 1

= SW + 1  (8.45)

where, SW is stripping factor of reboiler.

For kettle-type reboiler (yN + 1 = Kb xW) stripping factor

SW =   (8.46)

where, Kb = equilibrium constant for kettle-type reboiler

For thermosyphon reboiler yN + 1 ≅ xW

Fig. 8.13 Material Balance around Reboiler

Stripping factor for thermosyphon reboiler

Sw =   (8.47)
Component balance around bottom section up to the last tray.

Fig. 8.14 Material Balance around Bottom Section up to Final Tray

L–N – 1 xN – 1 = G–N yN + WxW

= + 1  (8.48)

Stripping factor for N th tray

SN =

where, KN = Equilibrium constant for Nth tray

G–N yN = L–N xN SN   (8.49)
From the Eqs.   (8.48) and   (8.49),

= SN + 1  (8.50)

From Eqs.   (8.45) and   (8.50),

= (SW + 1) SN + 1 = Sw SN + SN + 1  (8.51)

For any mth tray of stripping section

= Sm + 1 . Sm + 2 … SN – 1 · SN · Sw + Sm + 1 · Sm + 2 … SN – 1 SN

+ Sm + 1 · Sm + 2 … SN – 1 + … + Sm + 1 + 1  (8.52)
For the liquid leaving the feed tray

= SF + 1 · SF + 2 … SN · Sw + SF + 1 · SF + 2 … SN

+ SF + 1 … SN – 1 + … + SF + 1 + 1  (8.53)
Absorption factor for feed tray

AF =   (8.54)

where, KF = Equilibrium constant for feed tray

=   (8.55)

Symbol zD is mole fraction of distillate which can be liquid or vapour.

In calculations, GF yF /DzD ratio is determined by Eq. (8.44). (L–F xF )/(WxW)
ratio is determined by Eq. (8.53).
Absorption factor for feed tray is determined by Eq. (8.54). The values are
substituted in Eq. (8.55) and the ratio Wxw /DzD is determined.
Overall component balance
FzF = WxW + DzD  (8.56)

= + 1

DzD =   (8.57)

Equation (8.57) gives the value of DzD, then Eq. (8.56) gives WxW. Thus finally
one can get the product composition.
This product composition is correct, if all input data are correct. Normally, in the
first trial calculations temperatures at the trays are assumed to calculate the
equilibrium constants; Ki. To verify whether assumed values of temperatures are
correct or not, find yn by following equation:

yn =   (8.58)
where, is determined by Eq. (8.43).

For any tray n, yns for all components must be determined by Eq. (8.58). Then
check whether Σyni = 1 or not.
If Σyni = 1, it means assumed value of temperature for tray n is correct. If Σyni ≠
1, find the normalized values of yni and find new dew point temperature. Same
calculations should be done for all the trays.
Similarly, in stripping section for each tray (m) find

xm =   (8.59)

and check whether Σxmi = 1 or not.

Here is determined by Eq. (8.52).

If Σxmi ≠ 1, then find the normalized values of xmi and find new bubble point.
For new temperatures, carry out the material balance and enthalpy balance
simultaneously and find the new values of molar liquid flow rates and molar
vapour flow rates (new values of L′s, G′s, L–′s, G–′s ). Then repeat the
calculations.
Find the product composition by Thiele–Geddes method based
on the following data (same as Table 8.26):
(i) Table 8.31 Feed Composition
Component mole %

n–Butane 37
i–Pentane 32
n–Pentane 21
n–Hexane 10

n–Butane is light key component and i–Pentane is heavy key component.

(ii) Feed flow rate F = 4750 mol/h. Feed is saturated liquid at its bubble point.
(iii) Reflux ratio, R = 3
(iv) D = 1250 mol/h, W = 3500 mol/h
(v) Operating pressure, pt = 2 atm a
(vi) Assume constant molar overflow. Total condenser is used.
 Table 8.32 Equilibrium Constant Data (Ki s)
Tray number Temperature KC4 KiC5 KnC5 KC6

1 22°C 1.086 0.4046 0.3000 0.0870

2 24°C 1.156 0.4346 0.3230
 0.0950
3 28°C 1.308 0.5000 0.3747 0.1125
4 = feed 32°C 1.474 0.5730 0.4324 0.1330
5 41°C 1.904 0.7660 0.5877 0.1900
Reboiler 48°C 2.300 0.9500 0.7360 0.2500

(Read from Fig. 6.20)

Solution
Molar flow rates of liquid and vapour in enriching section.
L = R D = 3 × 1250 = 3750 mol/h
G = (R + 1) D = 4 × 1250 = 5000 mol/h
× AF

= A0 A1 A2 … AF– 1 + … + AF– 1 + 1  (8.44)

Here fourth tray is feed tray.

= A0 A1 A2 A3 + A1 A2 A3 + A2 A3 + A3 + 1

for n–Butane
A0 = R = 3

A1 = = 0.6906

A2 = = 0.6488

A3 = = 0.5734

= 3 × 0.6906 × 0.6488 × 0.5734 + 0.6906 × 0.6488 × 0.5734

+ 0.6488 × 0.5734 + 0.5734 + 1 = 2.973
AF =   (8.54)

L– = L + Fq = 3750 + 4750 × 1 = 8500 mol/h

KF = KC4 at feed tray temperature (equilibrium constant for n–Butane)
KF = 1.474
AF = = 1.1533

= SF + 1 · SF + 2 … SN · SW + SF + 1 … SN + … + SF + 1 + 1  (8.53)

G– = G + F(q – 1) or G– = L– – W = 8500 – 3500 = G = 5000 mol/h

= S5 Sw + S5 + 1

S5 = = 1.12

SW = = 3.286

= 1.12 × 3.286 + 1.12 + 1 = 5.8003

=   (8.55)

DzD = = 1104.58 mol/h  (8.57)

Wxw = 652.92 mol/h

zD = 0.8837, xW = 0.1865
For iso–Pentane:
A0 = R = 3, A1 =

A1 = = 1.8537

A2 =
A3 = = 1.5

= 3 × 1.8537 × 1.7257 × 1.5 + 1.8537 × 1.7257 × 1.5 + 1.7257 × 1.5 + 1.5

+ 1
= 24.282
Absorption factor for feed tray
AF = = 2.967

= S5SW + S5 + 1

S5 =

SW =

= 0.45 × 1.357 + 0.45 + 1 = 2.060 65

= × 2.967 = 34.96

DxD = = 42.27 mol/h

WxW = 1477.73 mol/h

xD = 0.0338, xW = 0.4222
For i-pentane, similarly one can calculate, xD = 0.005 58, xW = 0.283 and for n-
hexane xD ≅ 0 and xW = 0.1357
Table 8.33 Resulting Product Compositions
Component xD xDnor xW xWnor

n-Butane 0.8837 0.9571 0.1865 0.1815

i-Pentane 0.0338 0.0366 0.4222 0.4110
n-Pentane 0.0059 0.0063 0.2829 0.2754
n-Hexane — — 0.1357 0.1321
 Total 0.9234 1.0000 1.0273 1.0000

xDnor = Normalised values of xDi

xWnor = Normalised values of xWi
Note: Results of Tables 8.30 and 8.33 are close enough.
(c) Equation–Tearing Procedures Using Tridiagonal Matrix Algorithm8:
This method require the help of a computer. This method is more general and
flexible, compared to Lewis–Matheson method and Thiele–Geddes method.
Consider a general continuous flow steady state, multicomponent multistage
distillation operation. Assume that phase equilibrium between an existing vapour
phase and existing liquid phase is achieved at each stage, that no chemical
reaction takes place, and that neither of the existing phases entrains the other
phase. Consider the general equilibrium stage j as shown in Fig. 8.15.

Fig. 8.15 General Equilibrium Stage

Here Fj represent molar flow rate of feed to a stage j at temperature TFj (in K)
and pressure pfj and with overall composition in mole fractions zij. If stage j is
not the feed stage then in calculation Fj = 0. Heat is transferred from stage j (+ve
) or to stage j (–ve) at rate ϕj. If there is no external heat exchanger, attached with
stage j, then ϕj = 0. Equilibrium vapour and liquid phases leave the stage j at
temperature (in K) and pressure pj and with mole fraction yij and xij, respectively.
The vapour may be partially withdrawn from the column as side stream at a
molar flow rate Wj and the remainder Vj is sent to adjacent stage j – 1 above.
Similarly, existing liquid may be split into a side stream at a molar flow rate of
Uj, with the remainder Lj, sent to adjacent stage j as follows:
For each stage j, following independent equations can be written:
(i) Component material balance equations (M)
Lj–1 xi j–1 + Vj+1 yi j + 1 + Fj zij – (Lj + Uj) xij – (Vj + Wj) yij = 0  (8.60)
(ii) Phase equilibrium equation (E)
yij – Kij xij = 0   (8.61)
(iii) Mole fraction summation equation (S)
Σyi j – 1.0 = 0
Σxij – 1.0 = 0   (8.62)
(iv) Energy balance or enthalpy balance equation (H)
Lj – 1 HLj–1 + Vj+1 Hvj+1 + Fj HFj – (Lj + Uj) HLj – (Vj + Wj) Hvj – ϕj = 0  
(8.63)
These equations are known as MESH equations. For evaluating different
variables namely; flow rates, compositions, temperatures and pressures, number
of MESH equations are necessary. For this purpose, degree of freedom dictates
that for any stage j total 2C + 3 MESH (= C No. of material balance equations +
C No. of phase equilibrium equations + 2 No. of mole fraction summation
equations + 1 No. of energy balance equation) equations (where, C = number of
components) can be written. For total N number of equilibrium stages, N (2C +
3) MESH equations can be written. Total number of variables are N (3C + 10) +
1. It includes (C – 1) numbers of zij, yij and xij, flow rates; Fj, Lj, Vj, Uj, Wj,
enthalpies; HFj, HVj, HLj, pressures; pFj, pj, temperatures; tFj, tj and heat transfer
rates; ϕj as stagewise variables and equilibrium stages ‘N’ as variable.
Total number of variables = [3(C – 1) + 5 + 3 + 2 + 2 + 1) N + 1]
= (3C + 10) N + 1
Difference between total number of variables and number of equations is N (C +
7) + 1. Hence, N (C + 7) + 1 variables must be specified.
On combining MESH equations to eliminate yis and Ljs, following N · C
equations are obtained:
Aj xi j – 1 + Bi j xij + Cij xi, j + 1 = Dij  (8.64a)
where, Aj = Vj + (Fm – Wm – Um) – Vi 2 ≤ j ≤ N

Bij = – 1 ≤ j ≤ N

Cij = Vj + 1 Ki j + 1 1 ≤ j ≤ N – 1
Dij = – Fj Zij 1 ≤ j ≤ N
Equation (8.64a) is a general representation of total N · C equations. Each set of
N equations is a tridiagonal matrix equation. For example, if N = 5, for each
component following equations can be written in tridiagonal matrix form:

  (8.64b)

Each set of above equations give the mole factions of one component in the
liquids, leaving all the trays (equilibrium stages). Substitution of xij values in the
phase equilibrium equations give the values of yij. Like that all unknown
variables are determined. For solving such equations, a mathematical software
(such as Mathcad®) is useful.
8.4.6 Step 6: Selection of Type of Tower
Basically two types of distillation columns are in use: packed towers and tray
columns. Criteria for selection of the type of column were discussed in Section
8.3.
Packed column is widely used in distillation operation. Various packed column
designs were discussed in Chapter 7. Figure 7.16 gives schematics for towers,
packed with different types of packings. Characteristics of various packings are
listed in Table 9.1 Number of theoretical stages are calculated as per general
procedure outlined in the Section 8.4.5. Height Equivalent to Theoretical Stage
(Plate) – HETP is obtained from a packing supplier. Both these values gives
required height of a packed bed for the service. Refer to Table 8.36 for the
recommended values of HETP. Packed tower internals for distillation service are
different from those used for liquid-liquid extraction and are similar to those
discussed in Chapter 9.
In tray towers, distillation trays provide hold-up for a portion of liquid through
which vapour flows to accomplish the mass transfer required for separation.
Variety of tray tower designs are available for distillation service. Trays are
essentially used for distillation service when
(a) towers are very large in diameter (multi-pass trays),
(b) liquids contain solids or foulants,
(c) there are many internal transitions, and
(d) liquid loads are high.
Most commonly used tray tower is a cross-flow tray tower which is less
expensive. Sieve trays, bubble cap trays and valve trays fall in this category.
Modified sieve trays (e.g., Linde trays) and other high capacity trays are also
developed for variety of applications.
Sieve trays A typical conventional cross-flow type sieve tray is shown in Fig.
8.16. This tray has become a dominate tower internal for mass transfer columns
owing to its simple shape and ease of manufacture. Liquid travels horizontally
across the tray and into a vertical downcomes (also called a downspout) while
vapour flows up through the holes across the flow of liquid.

Fig. 8.16 Typical Sieve Tray

A detailed sketch of the sieve tray arrangement is shown in Fig. 8.26 for a
specific design. Space occupied by downcomer vary from 8 to 12% of total tray
area. Perforated holes occupy 5 to 15% of the active tray area which excludes
two downcomer areas from total cross-sectional area. Hole diameter (DH) varies
from 4 to 8 mm. Distance between two trays normally varies from 150 to 600
mm in general but it can be more (up to 900 mm) for foamy liquids. Typical weir
height (hw) varies from 30 to 90 mm. Clear liquid depth (hL = hW + how) also
decides liquid inventory on the tray. It can be safely assumed that liquid head on
the tray will be equal to hL or higher (Refer to Eq. (8.70)). When hL/DH exceeds
about 2, vapour momentum is exchanged with significant quantity of liquid,
two-phase mixture is continuous in liquid and behaves as froth, offering good
mass transfer.
There are several other flow patterns offered by different design engineers which
are shown in Fig. 8.21. Reverse flow pattern, shown in Fig. 8.21(b), is typically
developed by Linde, Germany and these trays are known as Linde trays. Double
pass and four pass trays are selected when tower diameter is large and it is feared
that liquid gradient (Δ) could be high in a single pass cross-flow tray. Linde trays
are claimed to offer at least 10% more efficiency than that obtained with single
cross-flow trays.
High capacity sieve (also known as dual-flow) trays are developed without
downcomers. As a result tray diameter reduces considerable. In these type of
trays, hole diameter varies from 12.5 to 25 mm. Holes are drilled in the entire
active area of the tray. Dual-flow trays have a higher capacity and lower pressure
drop, compared to conventional distillation trays. However, dual-flow trays have
a very narrow efficient operating range. Vapour and liquid flow counter currently
through the same holes. The majority of commercial applications of high
capacity trays are in small to moderate sized columns. Alternately, slots are
provided on the trays for countercurrent flow of both fluids.
Bubble cap trays Bubble cap trays were in use until the late 1950s when it was
superseded by sieve and valve trays. It consists of a flat perforated deck in which
the holes are enclosed with vapour risers and caps in the form of inverted cups
mounted on top of the risers. The caps can either be equipped with slots or holes,
through which vapour escapes. Vapour is forced into the surrounding cross-
flowing liquid such that aerated liquid is trapped on the tray floor to a depth at
least equal to the weir height. This gives the bubble cap tray the advantage to
operate at extremely low liquid and vapour rates. Figure 8.17 shows a bubble
cap tray along with enlarged view of a sample cap design and three other designs
of bubble caps.
Bubble cap trays, generally, have the ability to handle wide ranges of liquid and
vapour flow rates satisfactorily due to their leakproof performance. Standard cap
sizes of 75, 100 and 150 mm are available. Typical distance between two bubble
caps is kept 1.25 to 1.4 times of bubble cap size. Weir height (hw) is kept more
than bubble cap height such that it is immersed in liquid during operation.
Downcomers are similar to those for cross-flow sieve trays.

Fig. 8.17 Typical Bubble Cap Tray

Valve trays Valve trays are essentially flat perforated trays with moveable or
fixed valve units with or without a cage structure covering the holes. Moveable
valves are disk-shaped-type devices which are enclosed within a cage structure
or contain legs formed out of the valve disc. Fixed valves are units with integral
legs formed out of the tray deck. Figure 8.18 represents a valve tray and Fig.
8.19 shows with different valves that are in common use.

Fig. 8.18 Typical Floating Cage Valve Tray

 Fig. 8.19 Different Types of Valves

The open or vertical curtain area of the valve through which vapour escapes in a
horizontal direction is defined by the restrictive legs integral with the valve unit
or the leg rise of the cage structure attached to the tray deck. As the vapour rate
is increased, the movable valve units rise and the upper limit of opening is
controlled by the valve leg height. Valve trays perform over a wider operating
range and has a higher turndown compared with the sieve trays.
Variations in valve diameter (dv) from 38 to 50 mm, valve lift from 8 to 20 mm
and valve slot area from 12 to 30% of active tray area are common in designs.
Minimum and maximum distance between two valves is maintained at 15 mm
and 30 mm, respectively. Weir height (hw = 20 to 40 mm) for valve tray is
slightly less than that for cross-flow sieve tray. There are several other design
variations of valve trays offered by the manufacturers. For example in one
design, alternate rows of valves can have different weight (up to double),
claiming better vapour distribution on the tray.
Apart from the above three types of trays, there are many other tray designs
offered by various suppliers. It is not possible to cover all designs in the book.
Selection of tray type
Different types of trays used in tray tower are (i) sieve tray, (ii) bubble cap tray,
(iii) valve tray, (iv) Linde trays and other modified sieve trays.
Various factors that must be considered for the selection of tray type are: (a)
cost, (b) capacity, (c) operating range, (d) efficiency, and (e) pressure drop.
(a) Cost: Bubble cap tray is the costliest type of tray. With mild steel as a
material of construction, the ratios of cost
Bubble cap: Valve: Sieve trays ≅ 3:1.5:1 (general guideline)
In other words, cost of bubble cap tray is three times the cost of sieve tray and
double the cost of valve tray with mild steel as material. With different
materials, these ratios could be somewhat different.
(b) Capacity: If tower diameter required for given flow rates is more it means
capacity of tray is less. The ranking for capacity is in the order of modified
sieve tray, sieve tray, valve tray and bubble cap tray. In other words, for the
given flow rates, tower diameter required is maximum with bubble cap trays
and minimum with modified sieve tray. However, the difference in the values
of tower diameters is not large.
(c) Operating range: It is an important factor for the selection of tray. Operating
range of tray means range of liquid and vapour flow rates over which tray
works satisfactorily. Bubble cap tray provides the maximum operating range
while sieve tray provides the minimum operating range. Operating process
plants are designed and operated with some maximum capacity and also for
some minimum capacity. Hence, distillation columns are operated in the
range of some maximum feed flow rate to minimum feed flow rate. The ratio
 of design flow rate to operating flow rate is called turn-down ratio. Some
process plants are frequently operated at very high turndown ratio to meet the
product’s demand and to avoid the shut down and start-up conditions. For
higher turn-down ratio, bubble cap trays are preferred. Valve trays also offer
more flexibility or higher turn down ratio than the sieve trays. However, in
actual plant operation, movable float valves get stuck-up or fly-away if
operating conditions are frequently changed. In such a case, valve trays loose
efficiency.
(d) Efficiency: This factor is not generally considered for the selection of tray
type. Efficiency of trays are almost same, if they are operated at design flow
rates, or with minimum turn down. However, bubble cap columns give good
efficiency over a wider range of turn down in general.
(e) Pressure drop: Pressure drop over the tray and for entire column is important
for vacuum distillation for which a packed tower is the first choice. But for
the low vacuum (up to 700 Torr) operation, tray tower can be selected to get
other advantages. For example, to get the better vapour-liquid contact in large
diameter column or to facilitate the side draws. Among the trays, sieve tray
provides the lowest pressure drop, followed by valve tray, while bubble cap
provides the highest pressure drop in general.
Sieve trays and modified sieve trays are the cheapest and are satisfactory for the
majority of the applications. Nowadays, majority of chemical plants are
having automatic and precise control system and hence, flow rates are not
much fluctuating. However, to meet the fluctuations in the demand of
product, some plants are intentionally operated with different flow rates or
capacities. To handle the higher turn-down ratio, valve trays or bubble cap
trays are considered. Valve tray perforations are covered with liftable caps.
Vapour flow lifts the cap. Thus lifting cap directs the flow of vapour
horizontally into the liquid which provides intense mixing as compared to a
sieve tray. Valve trays are cheaper than bubble cap trays but they are not
suitable for very low vapour rate. Bubble cap tray provides higher efficiency
even with low vapour flow rate.
8.4.7 Step 7: Process Design of Tray Towers
Process design of a packed tower for absorber duty is detailed in Sec. 9.3.2 of
Chapter 9. In analogous manner, a packed tower can be designed for distillation
system. Process design of a sieve tray tower is detailed below.
Determining the tower diameter of sieve tray tower Flooding in sieve tray tower
fixes the upper limit of vapour velocity through the tower. Velocity of vapour at
which flooding in sieve tray tower occurs is called flooding velocity; vF. Tower
diameter of sieve tray column is decided based on the value of vF. Actual
velocity or design velocity through distillation column should be in the range of
70 to 90% of flooding velocity. Ideally it should be in between 80 and 85%.
Flooding is an inoperative condition. Flooding phenomena or flooding velocity
are defined differently for different equipments. For example, in case of knock
back condenser, flooding velocity is the velocity of vapour at which entrainment
of liquid droplets in vapour just starts or the velocity of vapour at which %
entrainment of liquid droplets in vapour becomes greater than zero. While in
distillation column entrainment of liquid droplets in vapour up to 10% is
normally accepted.
Flooding in tray tower of distillation column can occur or observed in either of
the following ways16:
1. Jet flooding: In distillation operation froth of liquid-vapour mixture forms on
each tray from which nearly clear vapour is separated and rises upward to
meet the liquid on the next above tray. When froth of liquid-vapour mixture
touches the next above tray (i.e. unstable operation) it is called jet flooding or
operational flooding. Actually the vapour flows through perforations of tray
forms a free flowing jet after leaving the orifice. Liquid droplets are entrained
in these free flowing jets. These free flowing jets combinedly form the froth.
2. Downcomer flooding: In distillation column, liquid flows in downward
direction by gravitational force but it flows against the pressure. When liquid
flows from one tray to next below tray, it flows from lower pressure to higher
pressure. Hence, to compensate that, it elevates certain level inside the
downcomer. When the liquid level in downcomer on any tray rises above the
weir, it is called downcomer flooding or hydraulic flooding.
Nowadays, it is possible to find tray or trays which restrict the capacity of
existing tower or which decide the maximum capacity of existing tower. To find
these trays of existing tower, one can intentionally operate the tower with
maximum flow rates or even in flooding condition for a short period of time.
During that time photograph of each section of tower can be taken with γ-rays.
In these photographs, one can clearly see the froth zone above the trays and
liquid level inside the downcomers. Based on the observation one can find a tray
or more than one tray which decide the maximum capacity of tower. If these
sieve trays, which restrict the capacity of sieve tray tower, are replaced by
modified or high capacity sieve trays, then it is possible to increase the capacity
of existing tower (may be up to 20%).
In calculations, tower diameter is decided based on jet flooding velocity.
Downcomer flooding conditions are checked in a calculation of liquid back-up
in downcomer (Refer Eq. (8.83)). Flooding velocity through sieve tray tower can
be determined by the following equation:

vF =   (8.65)

where, vF = Flooding vapour velocity based on net cross-sectional area, m/s

Cf = Flooding constant which can be obtained from Fig. 8.20.
σ = Surface tension of liquid, N/m
ρv = Density of vapour, kg/m3
ρL = Density of liquid kg/m3

FLV = liquid-vapour flow factor =   (8.66)

ṁL = mass flow rate of liquid, kg/s

ṁV = mass flow rate of vapour, kg/s
Figure   (8.20) gives flooding velocity with ±10% error, subject to the
following restrictions:
(a) System is non-foaming.
(b) Weir height is less than 15% of plate spacing.
(c) Sieve plate perforations are 13 mm or less in diameter.
(d) Value of Cf obtained from Fig.   (8.20) is valid only if Ah /Aa ≥ 0.1. If Ah
/Aa < 0.1 then Cf value obtained from Fig.   (8.20) must be corrected by
multiplying it with following correction factor.
Ah /Aa Correction Factor
0.08 0.9
0.06 0.8
where, Ah = Total area of all active holes
Aa = Active or bubbling area of tray
An = Net area available for vapour-liquid disengagement
For a single pass tray, An = Ac – Ad and Aa = Ac – 2Ad (max.)
where, Ac = Total column cross-sectional area
Ad = Cross-sectional area of downcomer. Downcomer area Ad usually ranges
from 8 to 12 % of Ac
Actual active area excludes calming area, area occupied by support devices, etc.,
and is less than Aa.
For a double pass tray, Aa = Ac – 4 Ad (max.)
In sieve trays, hole layout can be triangular pitch (similar to tube layout on a
tubesheet of shell and tube heat exchanger) or could have equal diametrical
pitch, depending on An/Ac ratio.

Fig. 8.20 Flooding Velocity (Sieve Plates)8 (Reproduced with the permission of McGraw-Hill Education,
USA)

Selection of liquid flow pattern After finding the tower diameter, liquid flow
pattern over sieve tray is decided. Common liquid flow patterns of cross-flow
sieve tray are shown in Fig. 8.21. In most of the cases single pass (cross flow,
(Fig. 8.21(a)) pattern is selected. Liquid flow pattern depends on the liquid flow
rate and tower diameter. Table 8.34 can be used for the selection of liquid flow
pattern.

Fig. 8.21 Common Liquid Flow Patterns in Cross-Flow Plates8

(Reproduced with the permission of McGraw-Hill Education, USA)
Table 8.34 Selection Guidelines for Liquid Flow Pattern
Range of liquid flow rate, m3/s

Column diameter, m Cross flow

Reverse flow Cascade double pass
Single pass Double pass

0.9 0 – 1.90 × 10–3 1.90 × 10–3 – 0.0126

1.2 0 – 2.50 × 10–3 2.50 × 10–3 – 0.0190
1.8 0 – 3.15 × 10–3 3.15 × 10–3 – 0.0250 0.025 – 0.0440
2.4 0 – 3.15 × 10–3 3.15 × 10–3 – 0.0315 0.0315 – 0.050
3 0 – 3.15 × 10–3 3.15 × 10–3 – 0.0315 0.0315 – 0.057 0.057 – 0.0880
3.66 0 – 3.15 × 10–3 3.15 × 10–3 – 0.0315 0.0315 – 0.063 0.063 – 0.1000
4.6 0 – 3.15 × 10–3 3.15 × 10–3 – 0.0315 0.0315 – 0.070 0.070 – 0.1136
 6 0 – 3.15 × 10–3 3.15 × 10–3 – 0.0315 0.0315 – 0.070 0.070 – 0.1260

Tower diameter also depends on tray spacing. Value of flooding constant

depends on the value of tray spacing (Fig. 8.20). Tray spacing ranges from 0.15
m to 1 m. At design stage, if the tray spacing is increased, required tower
diameter decreases, tray efficiency increases and hence, actual number of trays
required for desired separation decreases but overall height for the tower
increases. Ideally one should find the optimum tray spacing or the tray spacing
for which total cost of tower is minimum. If column diameter is less than 1 m,
tray spacing is kept 0.2 to 0.3 m. For columns above 1 m diameter, tray spacing
of 0.3 to 0.6 m is used. For the first trial calculations one can start with tray
spacing equal to 0.3 m but if column diameter is calculated to be more than 1 m
then for the second trial calculations, tray spacing can the taken equal to 0.4 to
0.5 m. Even with smaller tray spacing of 0.15 to 0.3 m, larger spacing is needed
between certain plates to accommodate nozzles for feed and side streams, for
manholes, etc.
Checking of liquid entrainment As a rough guide, the % of liquid entrainment
should be less than 10%.
% liquid entrainment = × 100  (8.67)

or % liquid entrainment = 100ψ

where, ψ = Fractional liquid entrainment
ψ =

Value of ψ as a function of liquid-vapour factor (FLV) with % of flood as a

parameter is given by Fair17. The same is reproduced in Fig. 8.22. ψ should be
less than 0.1 to get higher tray efficiency. Optimum value of ψ can be greater
than 0.1.
% of flood = × 100

= × 100(8.68)

An elaborate empirical approach for entrainment is given by Bennett et al in Ref.

16.
Weir design The height of weir decides the height of liquid over the tray. In most
cases, weirs are straight and rectangular plates. With bubble cap trays, to
compensate the fluctuations in the flow rate, V–notch-type weirs are also used.
Increase in weir height increases depth of liquid pool (hL) and contact time.
Hence, it increases tray efficiency of the tray with which it is attached. However,
increase in weir height increases liquid entrainment in vapour, leaving the tray
and hence decreases tray efficiency of the next above tray. Therefore, effect of
increase of weir height on the tray efficiency cannot be predicted. However,
increase in weir height increases pressure drop. For a distillation column
operating above atmospheric pressure, weir height is kept in at 30 mm to 90 mm;
while 30 mm to 50 mm is more common. For vacuum operation, to control the
 Fig. 8.22 Entrainment Correlation for Sieve Plates8
(Reproduced with the permission of Chemical Engineering by Access Intelligence, USA)

tray pressure drop, weir height is reduced and it is kept in between 6 and 12 mm.
In old designs, inlet weir was also used, but in recent designs inlet weir is not
recommended as it results in hydraulic jump.
Weir length fixes area of segmental downcomer. Weir length is normally kept in
between 0.6 to 0.85 times column diameter. For the first trial calculations, weir
length can be taken as 0.77Di and equivalent down comer area is 12%. Decrease
in weir length increases the net area of tray and decreases the tower diameter, but
it decreases the downcomer area, increases the pressure drop in downcomer and
also decreases the volume of downcomer. It also elevates the liquid level in
downcomer. Increase in weir length increases the tower diameter. Hence,
balanced weir length and weir height are required. Relation between weir length
and downcomer area is given in Table 8.35.
Table 8.35 Relationship of Weir Length and Downcomer Area

0.600 5.25
0.650 6.80
0.700 8.80
0.705 9.00
0.715 9.50
0.727 10.00
0.745 11.00
0.770 12.00
0.780 13.00
0.800 14.00
0.815 15.00
0.826 16.00
0.840 17.00
0.850 18.00

Clear liquid depth over the tray is equal to height of the weir, hw, plus the depth
of the crest of liquid over the weir, how. The height of the liquid crest over the
weir can be estimated by using Francis formula (Ref. 9).

how = 750   (8.69)

where, how = Weir crest of liquid, mm

ṁL = Mass flow rate of liquid, kg/s
lw = Weir length, m
ρL = Liquid density, kg/m3
hL = hw + how  (8.70)
Checking for weeping At a very low value of vapour velocity, liquid rains down
through the perforations. This phenomenon is known as weeping. It is again an
inoperative condition. Minimum vapour velocity required to avoid the weeping
is given by Eduljee’s equation:

vh min =   (8.71)

where, vh min = Minimum vapour velocity through the holes, m/s

dh = Hole diameter, mm
ρV = Density of vapour, kg/m3
K = Constant, a function of depth of clear liquid (hL) on the plate, which can be
obtained from Fig.   (8.23).

Fig. 8.23 Weep-Point Correlation (Eduljee, 1959)10 (Reproduced with the permission of Elsevier, UK)

Another empirical approach for checking weeping is developed by Bennett et al.

(Ref. 16) based on liquid Froude number.
FrL =   (8.72)

where, ρV = Vapour density, kg/m3

vh = Vapour velocity through holes, m/s
ρL = Liquid density, kg/m3
g = Gravitational constant = 9.81 m/s2
hL = Liquid head over the plate, m
Froude number should be greater than 0.5 to ensure no weeping.
Basically Froude number (FrL) specifies the ratio of inertia force to gravity force
on an element of fluid. Hence, this equation is more appropriate in defining the
weeping condition.
Total pressure drop in sieve tray tower Total pressure drop in sieve tray tower,
ΔpT is given by the following equation:
ΔpT = Pressure drop in inlet nozzle + Pressure drop/tray × No. of trays +
Pressure drop in outlet nozzle
Major pressure drop in vapour is provided by trays.
ΔpT ≅ N × Δpt  (8.73)
where, N = Total number of sieve trays of tower
Δpt = Tray pressure drop, Pa
Tray pressure drop in Pa or N/m2 can be determined from the tray pressure drop
in mm of liquid column by following equation:
Δpt = 9.8 × 10–3 ht ρL   (8.74)
where, Δpt = Tray pressure drop, N/m2 or Pa
ht = Tray pressure drop, mm of liquid column (LC)
ρL = Density of liquid, kg/m3
To find the tray pressure drop simple additive model is used.
ht = hd + (hw + how) + hr  (8.75)
where, hd = Pressure drop through dry plate, mm LC
hw = Height of weir, mm
how = Height of liquid crest, mm
hr = Residual pressure drop, mm LC
Tray pressure drop (ht) is the sum of pressure drops which occurs as vapour rises
through the perforations of dry plate (hd), pressure drop which occurs as vapour
rises through the pool of clear liquid (hw + how) and pressure drop which occurs
in vapour-liquid disengagement space (hr). Due to the high vapour velocity, froth
formation takes place over the tray. Some liquid particles are entrained in vapour
phase. Hence, some additional energy loss is taking place in carrying the liquid
particles. These liquid particles are disengaged from the vapour after travelling a
certain distance. This additional energy loss is counted by residual pressure drop,
hr.
Pressure drop through dry plate is given by following equation:

hd = 51   (8.76)

where, hd = Dry plate pressure drop, mm LC

vh = Velocity of gas through the holes, m/s

=   (8.77)

ρV, ρL = Densities of vapour phase and liquid phases, respectively, kg/m3

Co = Orifice coefficient which is a function of ratio of plate thickness to hole
diameter and % perforated area of plate. It can be determined by using
Fig. 8.24.

% Perforated area = × 100  (8.78)

where, Ah = Total area of all active holes of tray, m2

Aa = Active area of tray = Ac – 2 Ad, m2 for single pass
Ac = Inside area of column, m2
Ad = Downcomer area of tray, m2
 Fig. 8.24 Discharge Coefficient (Sieve Plates)8 (Reproduced with the permission of McGraw-Hill
Education, USA)

Residual pressure drop can be estimated by the following equation:

hr =   (8.79)

where, hr = Residual pressure drop, mm LC

ρL = Density of liquid, kg/m3
Design of downcomers There are two functions of a downcomer:
1. It provides the passage for the flow of liquid.
2. Foamy or frothy liquid enters into downcomer. Hence, downcomer also acts
as a vapour-liquid separator. To facilitate the vapour-liquid disengagement,
certain minimum residence time (at least 3 seconds) is required in the
downcomer.
Following types of downcomers are used:
(a) Straight segmental downcomer: It is the simplest and cheapest and is
preferred for most applications. In this type, downcomer channel is formed by
a flat plate. The portion of the flat plate which extends down below the plate
from the outlet weir is called apron. With this type of downcomer, apron is
straight and vertical.
(b) Inclined segmental downcomer: With this type of segmental downcomer,
apron is inclined towards the wall. It is preferred with foamy liquid, with
which vapour-liquid disengagement is difficult. Inclined or sloppy apron
provides more resistance to flow and makes the phase separation easier. It
also increases tray area available for perforations. But with this type, liquid
back-up in downcomer is higher compared to straight segmental downcomer.
(c) Circular downcomers or pipes: They are used for small liquid flow rates in a
few cases.
Liquid is falling from tray to tray by passing through the downcomer. Due to the
pressure drop across the tray, liquid is flowing from lower pressure to higher
pressure. To overcome the pressure difference liquid elevates its level in the
downcomer. This is called liquid back-up in the downcomer. Also some pressure
drop occurs when liquid flows through the downcomer area. Downcomer
pressure drop increases the liquid back-up further. Head loss or pressure drop in
the downcomer is estimated by following equation:

hdc = 166   (8.80)

where, hdc = Downcomer pressure drop, mm LC

ṁd = Liquid flow rate in downcomer, kg/s
ρL = Density of liquid, kg/m3
Am = Ad or Aap whichever is smaller, m2
Ad = Downcomer area, m2
Aap = Clearance area under downcomer apron, m2
Aap = hap lw/106   (8.81)
where, hap = Height of the bottom edge of the apron above the plate, mm
lw = Length of weir, mm
hap = hw – (5 to 10) mm   (8.82)
where, hw = Height of weir, mm
Liquid back-up in downcomer in terms of clear liquid is given by following
equation:
hb = (hw + how) + ht + hdc   (8.83)
= Downcomer back-up measured from plate surface, mm (i.e., elevation of clear
liquid in downcomer)
Due to the formation of froth and foam, actual height or foamy or frothy liquid
in downcomer is quite great, nearly double than hb.
Density of froth or “aerated” liquid is in between 0.4 to 0.7 times that of the
clear liquid. For design purpose, it is taken as 0.5 times the density of liquid.
Hence, to avoid the downcomer flooding.
hb >|   (8.84)

where, hb = Liquid back-up in the downcomer measured from plate surface, mm

It = Tray spacing, mm
hw = Height of weir, mm
Certain minimum residence time must be allowed in the downcomer to facilitate
the disengagement of vapour and liquid phases. Residence time of at least 3
seconds is recommended. Residence time in downcomer is given by following
equation:
θr =   (8.85)

where, Ad = Downcomer area, m2

hbc = Liquid back-up in downcomer in terms of clear liquid, m
ρL = Density of liquid, kg/m3
ṁd = Liquid flow rate in downcomer, kg/s
Design of bubble cap tray column and valve tray column Design of bubble cap
columns can be made in accordance with the procedure described in Refs. 2, 8,
20 to 23. Hydraulic design of valve tray column is quite complicated and most
data for different valve systems are proprietary. Reference 23 gives some date
for valve tray columns. A preliminary design can be made on the basis of these
data and it can be checked with the supplier of the valve tray based on his
hydraulic data.
8.4.8 Step 8: Tray Efficiency
Like VLE data, in case of tray efficiency also, it is better to predict the tray
efficiency based on actual data on the same system or similar system of
commercial scale plant or of pilot plant. There is no entirely satisfactory method
available for finding the tray efficiency. Various correlations are available for
finding tray efficiency, but they are used only if the reliable actual data on the
same or similar system are not available. Overall efficiency of a tray tower is
given by the following equation:

ηoc =   (8.86)

where, Nt = Number of theoretical stages required for the desired separation.

Na = Actual number of trays required for the desired separation.
Tray efficiency depends on (i) system composition and properties, (ii) rate of
throughput, and (iii) geometry of tray tower (geometry of tray, tray spacing,
etc.).
Two methods are given here to find the tray efficiency:
1. Van Winkle’s method
2. AIChE method.
If actual data for predicting tray efficiency for the given system are not available,
then it can be predicted approximately by using these two methods. For the safe
design calculation, it is suggested that if the value of tray efficiency obtained
based on any of these methods is more than 0.5 then consider the value of
overall efficiency as 0.5.
1. Van Winkle’s Correlation or Method19 Van Winkle’s correlation can be
used to predict the tray efficiency of a binary system of distillation. This
correlation can be used for both bubble-cap and sieve trays.
η = 0.07 Dg0.14 ScL0.25 Re0.08  (8.87)
where, Dg = Surface tension number = σL/(μL · uV)   (8.88)
σL = Liquid surface tension, N/m
μL = Liquid viscosity, N · s/m2 or kg/(m · s)
uv = Superficial vapour velocity, m/s
ScL = Liquid Schmidt number = μL /(ρL · DLK)  (8.89)
DLK = Liquid diffusivity of light key component, m2/s

Re = Reynolds number =   (8.90)

where, hw = Weir height, mm

ρV = Vapour density, kg/m3
fA = Fractional area = Ah /Ac
2. AIChE Method10 The point efficiency η̇ is related to number of transfer units
by the following equation:

  (8.91)

where, η̇ = Point efficiency

NG = Number of gas phase transfer units
m = Slope of equilibrium curve at the point in consideration
V, L = Molar flow rates of vapour and liquid phases, respectively at the point in
consideration
NL = Number of liquid phase transfer units
The number of gas phase transfer units is given by the following equation:

NG =   (8.92)

where, hw = Weir height, mm

FV =
va = Vapour velocity based on active tray area, m/s

va =
Lp = Volumetric liquid flow rate across the plate divided by average width of
plate, m3/(s · m)
Average width of plate =   (8.93)

ZL = Length of liquid path across the tray, m

μV = Viscosity of vapour, N · s/m2
ρV = Density of vapour, kg/m3
DV = Vapour diffusivity, m2/s
The number of liquid phase transfer unit is given by the following equation:
NL = (4.13 × 108 DL)0.5 (0.21 FV + 0.15)θL  (8.94)
where, DL = Liquid diffusivity, m2/s
θL = Liquid contact time, s
Liquid contact time θL can be determined by the following equation:

θL =   (8.95)

where, ZL = Length of liquid path from inlet downcomer to outlet weir, m

Zc = Liquid hold-up on the plate, m3 per m2 active area
For sieve trays,
Zc = 0.006 + 0.73 × 10–3 hw – 0.24 × 10–3 FV hw + 1.22 Lp  (8.96)
Tray efficiency η is equal to the point efficiency η̇ , if the liquid on tray is
perfectly mixed. For the real tray, this may not be true.
Tray efficiency η can be determined from point efficiency by using Fig.   
(8.25). In this figure, graph of vs is drawn for different values of Peclet
number. For a tray, Peclet number is given by the following equation:
Pe =   (8.97)

where, De = Eddy diffusivity, m2/s

For sieve trays, the Eddy diffusivity can be estimated by the following equation.
De = (0.0038 + 0.017 va + 3.86 Lp + 0.18 × 10–3 hw)2  (8.98)
In actual operation liquid droplets get entrained in vapour phase. This reduces
the value of tray efficiency η. Actual tray efficiency ηa considering the liquid
entrainment can be calculated by following equation.

ηa =   (8.99)

where, ψ = Fractional liquid entrainment.

In Sec. 8.4.7, various parameters relating to the process design of a distillation
tower were discussed in general. However, detailed design of the tower internals
are well covered in the book by Billet20. For preparation of the specifications of
a tray tower, useful guidelines can be derived from the article by Mukherjee21.

Fig. 8.25 Relationship between Plate and Point Efficiencies10 (Reproduced with the permission of McGraw-
Hill Education, USA)

8.4.9 HETP of Packings

For the distillation in packed tower, height of packed bed (Z), required for the
desired separation, is determined by following equation:
Z = Nt × HETP  (8.100)
where, Nt = Number of theoretical stages required for the desired separation
HETP = Height equivalent to a theoretical plate.
The height of an equivalent equilibrium stage or theoretical plate (HETP) is the
height of packing that gives the same separation as an equilibrium stage. HETP
for the given type and size of random packing is nearly constant and is nearly
independent of the system physical properties. HETP values available in
literature are valid only for good liquid distribution and for the reasonable
pressure drop.
HETP values for various packing can be obtained from Table 8.36.
Table 8.36 Recommended HETP for Various Packings
Sr. No. Packing types Packing size, mm HETP, m

1. Pall ring, random 25 0.40 – 0.50

(for Δp > 17 mm WC 38 0.60 – 0.75
per m of packing) 50 0.75 – 1.00
2. Berl saddle and 25 0.40 – 0.50
Intalox sadle, random 38 0.60 – 0.75
(for Δp > 29 mm WC 50 0.75 – 1.00
per m of packing)
3. Raschig rings 25 to 50 1.0
4. Structured packings* ≤ 0.5

*HETP of a structured packing depends on a factor known as FV which is

defined as
FV = vg   (8.101)
where, vg = Superficial vapour velocity, m/s
ρV = Vapour density, kg/m3
HETP equal to around 0.5 m is reported for FV factor, not exceeding 3, beyond
which pressure drop increases steeply.
Bennett (Ref. 16) has given empirical equations for estimation of HETP for
different types of packings.
Design a crossflow-type sieve tray tower for the distillation of
acetic acid-water system, specified in Example 8.5. Maximum feed flow rate is
12 000 kg/h and minimum feed flow rate is 8400 kg/h (70% of the maximum).
Solution
Data obtained from Example 8.5:
(i) Feed composition: 80% acetic acid and 20% water (by mass) or
54.55% acetic acid and 45.45% water (by mole)
(ii) Mole fractions of water: In distillate xD = 0.9302, In residue xW = 1.666 ×
10–4
(iii) q = 1.272 for liquid feed at 30°C
(iv) Reflux ratio, R = 4.2
(v) Number of theoretical stages required for desired separation, N = 24
Flow rates
Average molar mass of feed,
Mav = Σxi Mi = 0.5455 × 60 + 0.4545 × 18 = 40.911 kg/ kmol
Molar flow rate of feed, F =

F = D + W
FxF = DxD + WxW
D + W = 293.22
0.9302 D + 1.666 × 10–4 W = 0.4545 × 293.32 = 133.314
0.9302 D + 1.666 × 10–4 (293.32 – D) = 133.314
D = 143.29 kmol/h
W = 293.32 – 143.29 = 150.03 kmol/h
Molar flow rates of vapour and liquid at the top in enriching section:
L = RD = 4.2 × 143.29 = 601.818 kmol/h
V = (R + 1) D = 5.2 × 143.29 = 745.108 kmol/h
Molar flow rates of vapour and liquid in stipping section:
L– = L + Fq = 601.818 + 293.32 × 1.272 = 974.92 kmol/h
V– = F (q – 1) + V = 293.32 (1.272 – 1) + 745.108 = 824.89 kmol/h
Tower diameter calculation
(a) Tower diameter required at top
Operating pressure at top of column = 1 atm = 101.325 kPa
Here total condenser is used, hence, composition of vapour leaving the topmost
tray and composition of liquid entering to the topmost tray (reflux) are equal.
=
where, m.L, m.V = Mass flow rates of liquid and vapour at the top, kg/s
Density of vapour, ρv

ρv =

Average molar mass of vapour at top

Mav = Σxi Mi = (1 – 0.9302) × 60 + 0.9302 × 18 = 20.9316 kg/ kmol

ρv =

To find the temperature of vapour at top, in t–x–y data, put y = xD = 0.9302 and
find the corresponding value of temperature from Table 8.6.
y = xD = 0.9302 t = tT = 100.64°C
ρv = 0.6823 kg/m3
Density of liquid at top, ρL

ρL = (6.55)

Density of pure water at 100.64°C, ρw = 958 kg/m3

Density of pure acetic acid at 100.64°C, ρAA = 1010 kg/m3
Mass fraction of water in distillate = 0.8
Mass fraction of acetic acid in distillate = 0.2

ρL = = 967.97 kg/m3

Liquid–vapour flow factor at top,

FLV =   (8.66)

FLV = = 0.021 444

For the first trial calculations, tray spacing = 0.3 m (assumed)

From Fig. 8.20,
Cf = 0.0635
Flooding velocity,
σ = Surface tension of liquid, N/m
σ1/4 = Σσi1/4 xi = (1 – 0.9302) σAA1/4 + 0.9302 σW1/4
where, σW = Surface tension of pure water, N/m = 58 × 10–3 N/m
σAA = Surface tension of pure acetic acid, N/m
σAA at 100.64°C:
σAA1/4 = (6.93)
σAA = Surface tension of acetic acid, dyn/cm
ρ′L = Density of liquid acetic acid, mol/cm3

ρ′L =

ρ′V = Density of acetic acid vapour, mol/cm3

ρ′V =

ρ′V = mol/cm3

[P] = Parachor can be calculated from Table 3–343, of Ref. 14.

[P] = 55.5 (Contribution of CH3 –) + 73.8 (Contribution of – COOH)
[P] = 129.3
σAA1/4 = 129.3 (0.016 833 – 3.26 × 10–5) = 2.1723
Surface tension of mixture,
σAA = 22.268 dyn/cm ≡ 22.268 × 10–3 N/m
σ1/4 = (1 – 0.9302) × (22.268 × 10–3)1/4 + 0.9302 × (58 × 10–3)1/4
σ = 55 × 10–3 N/m
Equation (8.65):

vF = 0.0635

= 2.93 m/s
Let actual vapour velocity through tower, v = 0.85 vF = 0.85 × 2.93
= 2.49 m/s
Volumetric flow rate of vapour at top
qV = = 22 858.4 m3/h ≡ 6.3496 m3/s

Net area required at top

An =

Let downcomer area, Ad = 0.12 Ac

An = Ac – Ad = Ac – 0.12 Ac = 0.88 Ac
where, Ac = Inside cross-sectional area of tower
0.88 Ac = 2.55
Ac = = 2.8977 m2

Inside diameter of column required at top

Di = = 1.92 m say 1.95 m at the top

(b) Tower diameter required at bottom

Operating pressure at the base of column = Operating pressure at top + ΔpT
where, ΔpT = Total pressure drop in sieve tray tower
ht = 120 mm WC per tray
ΔpT ≅ Actual no. of trays × ρL g ht
Assuming tray efficiency = 0.5
Actual number of trays, Na =

ΔpT = 48 × 1000 × 9.81 × 120 × 10–3

= 56 506 Pa ≡ 56.5 kPa (assumed)
Operating pressure at base
p′t = pt + ΔpT = 101.325 + 56.5 = 157.825 kPa ≡ 1.558 atm
Bottom product is nearly pure acetic acid. Antoine equation14 for acetic acid
ln pv =
When pv = p′t, T = Tb = Base temperature of column

ln (157.825) = 15.8667 –

T = 406.74 K t = 133.6°C (Estimated)

Molar flow rates of vapour and liquid at bottom
V– = 824.89 kmol/h, L– = 974.92 kmol/h
At the base of column,
≅

where , = Mass flow rates of vapour and liquid at bottom, kg/s

Density of vapour at bottom
ρV =

ρV = 2.8368 kg/m3
Density of liquid (nearly pure acetic acid) at bottom
ρL = Density of acetic acid at 134.26°C and at 160.185 kPa
= 1000 kg/m3 – saturated liquid
Liquid–vapour flow factor at bottom; Eq. (8.66):
FLV = = 0.062 96

For tray spacing = 0.3 m, from Fig. 8.20;

Cf = 0.0616
Flooding velocity
Surface tension of pure acetic at 134.26°C temperature:
=
–3 × = 0.016 67 mol/cm3
= 1000 × 10

= 2.8368 × = 4.728 × 10–5 mol/cm3

= 129.3 (0.016 67 – 4.728 × 10–5) = 2.1493

σAA = 21.34 dyn/cm ≡ 21.34 × 10–3 N/m
vF = 0.0616

vF = 1.17 m/s
Actual vapour velocity, v = 0.85 × 1.17 = 0.9945 m/s
Volumetric flow rate of vapour at bottom
Q̇ = = 4.8464 m3/s

Net area required at bottom

An = = 4.8732 m2

Inside area of column

Ac = = 5.538 m2

Inside diameter of column required at base

Di = = 2.655 m

Up to 1 m diameter 0.3 m tray spacing is sufficient. Here, tower diameters are

greater than 1 m. Hence, for the 2nd trial calculations, increase the tray spacing to
0.45 m.
For tray spacing = 0.45 m
(a) Tower diameter required at top
For FLV = 0.021 44 and tray spacing = 0.45 m
From Fig. 8.20, Cf = 0.0814

vF = = 3.756 m/s

Actual vapour velocity, v = 0.85 vF = 3.1926 m/s

Net area required
An = = 1.989 m2

Ac = = 2.26 m2

Inside diameter at top

Di = = 1.7 m

(b) Tower diameter required at bottom

For FLV = 0.062 96 and tray spacing = 0.45 m
From Fig. 8.20, Cf = 0.0794
vF = = 1.5 m/s

Actual vapour velocity, v = 0.85 vF = 1.275 m/s

Net area required, An = = 3.8 m2

Inside diameter of column required at bottom

Ac = m2

Inside diameter of column required at bottom

Di = = 2.345 m say 2.36 m

Provide different diameters in two different sections.

Table 8.37 Summary of Calculations
Parameter Enriching Section Stripping Section

Inside diameter of column (Di), m 1.7 2.36

Inside area of column (Ac), m2 2.27 4.374

Net column area (An), m2 1.998 3.85

Downcomer area, (Ad ), m2 0.2724 0.525

Maximum active area, (Aa = 0.76 Ac), m2 1.7252 3.3242

Hole area, (Ah), m2 0.172 52 0.332 42

For the first trial, hole area Ah is 10% of active area, Aa.
Volumetric flow rate of liquids
In enriching section, qL =

= 13.0138 m3/h ≡ 3.6 × 10–3 m3/s

In stripping section,

qL = = 54.495 m3/h ≡ 16.25 × 10–3 m3/s

Referring Table 8.34, single pass tray can be used for both sections.
Check for weeping
Minimum vapour velocity through holes to avoid the weeping is given by the
following equation:
vh min =   (8.71)

where, K = constant can be obtained from Fig. 8.23 as a function of (hw + how)
Take
Weir height, hw = 50 mm (For both sections)
Hole diameter, dh = 5 mm (For both sections)
Plate thickness, t = 5 mm (For both sections for MS trays)
= 3 mm (For both sections for SS trays)
(a) For enriching section
Height of liquid crest over the weir

how =   (8.69)

For the checking of weeping conditions minimum value of how at 70% turn down
must be determined. Minimum value of
ṁL = 0.7 LMav = 0.7 × 601.818 × 20.9316
= 8817.9 kg/h ≡ 2.45 kg/s (minimum)
ρL = 967.97 kg/m3
Referring to Table 8.35, for Ad /Ac = 0.12, lw /Di = 0.77
lw = Length of weir = 0.77 Di
lw = 0.77 × 1.7 = 1.309 m

Minimum how = = 11.64 mm

At minimum rate, hL = hw + how = 50 + 11.64 = 61.64 mm

From Fig. 8.23, K = 30.38
vh min = m/s

Actual vapour velocity through holes at minimum vapour flow rate,

vha = m/s

Since, vha > vh min, there will be no weeping.

A check by Eq. (8.72) would also confirm no weeping.
FrL =

= 0.88 > 0.5

Thus in the enriching section, minimum operating rate is well above weep point.
(b) For stripping section
Minimum value of = 0.7 × ṁL = 0.7 × 974.92 × 60
Minimum = 40 946.64 kg/h ≡ 11.374 kg/s
ρL = 1000 kg/m3
lw = 0.77 × 2.36 = 1.8172 m

Minimum how = 750 = 25.47 mm

At minimum rate, hL = hw + how = 50 + 25.47 = 75.47 mm

From Fig. 8.23, K = 30.69
vh min = m/s

Actual vapour velocity through holes at minimum vapour flow rate

vha = m/s

In this case also vha > vh min. Hence, no weeping will take place in the stripping
section.

Eq. (8.72): FrL =

= 0.6317 > 0.5

Note: FrL of the stripping section is close to the limit which indicates that
stripping section is more sensitive to weeping than enriching section. Since
column’s operational stability depends on both the sections, overall stability
will be governed by the performance of the stripping section in this case.
Tray pressure drop
(a) Tray pressure drop for enriching section
Dry plate pressure drop
hd = 51   (8.76)

vh = m/s (maximum)

ρV = 0.6823 kg/m3, ρL = 967.97 kg/m3

From Fig. 8.24, for = 1 and

Ap = Perforated area is slightly less than active aera, Aa, which excludes some
calming zone.
However, for calculation purpose, Ap = Aa

hd = 51 mm LC

hw = 50 mm LC for ρL = 967.97 kg/m3

Maximum height of liquid crest over the weir

Maximum how = 750 = 14.765 mm LC

Residual pressure drop,

hr =   (8.79)

= = 12.9136 mm LC

Total tray pressure drop

ht = hd + (hw + how) + hr  (8.75)
ht = 68.64 + (50 + 14.765) + 12.9136
= 146.32 mm LC for ρL = 967.97 kg/m3
Note: This is equivalent to 142 mm WC. Tray pressure drop was assumed
equal to 120 mm WC to estimate the base pressure. The calculations can be
repeated with a revised value of tray pressure drop but small changes in
physical properties will have little effect on the tray design.
(b) Tray pressure drop for stripping section
Dry plate pressure drop, hd

vh = = 14.58 m/s

ρV = 2.8368 kg/m3,
ρL = 1000 kg/m3,
Co = 0.8422 (From Fig. 8.20)

hd = 51

hd = 43.36 mm LC
hw = 50 mm
Maximum height of liquid crest over the weir
how = 750 = 32.31 mm LC

Residual pressure drop

hr = = 12.5 mm LC

Total tray pressure drop

ht = 43.36 + (50 + 32.31) + 12.5 = 138.17 mm LC or WC
Note: In this section also ht is higher than 120 mm WC. Calculations with
revised ht may not significantly effect the tray design. However, total pressure
drop with correct ht values in both sections should be worked out and base
temperature at the bottom should be recalculated which is important for
reboiler design. Considering ht = 150 mm WC for tray, bottom temperature can
be recalculated as 136.63°C (409.78 K).
Checking of downcomer design
Types of downcomer: Straight and segmental downcomer area, Ad = 0.12 Ac (For
both sections).
(a) For enriching section
Pressure drop in downcomer is given by the following equation:

hdc = 166   (8.80)

ṁd = Liquid flow rate through downcomer, kg/s
= LMav
= 601.818 × 20.9316 ×

= 3.5 kg/s
ρL = 967.97 kg/m3
Am = Ad or Aap, whichever is smaller, m2
Ad = 0.2724 m2 (Table 8.37)
Aap = hap × lw
lw = 1.309,
hap = hw – 10
= 50 – 10 = 40 mm ≡ 0.04 m
Aap = 0.04 × 1.309 = 0.052 36 m2
Aap << Ad. Therefore, take Am = Aap.
Am = 0.052 36 m2

hdc = 166

= 0.79 mm, say 1 mm

Liquid back-up in downcomer
hb = hw + how + ht + hdc  (8.83)
= 50 + 14.765 + 146.32 + 1 = 212 mm
= = 250 mm

hb = 212 < which satisfies Eq. (8.84).

Downcomer area and tray spacing are acceptable.

Check residence time:
θr =   (8.85)

= 15.97 s > 3 s – satisfactory

(b) For stripping section
Pressure drop in downcomer

hdc = 166

Downcomer area, Ad = 0.525 m2

Aap = hap · lw
hap = 50 – 10 = 40 mm, lw = 1.8172 m
Aap = 0.04 × 1.8172 = 0.0727 m2
Am = 0.0727 m2
hdc = 8.29 mm
Liquid back-up in downcomer
hb = hw + how + ht + hdc
hb = 50 + 32.31 + 140 + 8.29 = 230.6 mm
= = 250 mm

hb = 230.6 < which satisfies Eq. (8.84).

Here also downcomer area and tray spacing are acceptable.

Residence time in downcomer
θr = =

= 7.45 s > 3 s
Hence, it is satisfactory.
Checking for entrainment
(a) For enriching section
Vapour velocity based on net area, vn =

vn = m/s

Flooding =
FLV = 0.021 444, ψ = 0.21 (from Fig. 8.22)
Entrainment = 21% > 10% – somewhat high
If ψ is less, then higher tray efficiency is obtained. But to decrease the value of
ψ, tower diameter must be increased further. Ideally, optimum value of ψ must
be determined for the given case. Optimum value of ψ can be greater than 0.1 or
10%.
(b) For stripping section
% Flooding = 85 %, FLV = 0.062 96
From Fig. 8.22, ψ = 0.084
% entrainment = 8.4% < 10%
It will provide higher tray efficiency.
Tray efficiency by AIChE method
(a) For enriching section
Physical properties of liquid and vapour for the same tray are:
ρL = 967.97 kg/m3, σL = 55 × 10–3 N/m
μL = Viscosity of reflux or distillate (water–acetic acid solution at
100.64°C)
ln μL = w1ln μ1 + w2ln μ2(6.56)
w1 = 0.8 = mass fraction of water
w2 = 0.2 = mass fraction of acetic acid
μw = Viscosity of pure water at 100.64°C = 0.28 mPa · s or cP
μAA = Viscosity of pure acetic acid at 100.64°C = 0.48 mPa · s or cP
(Fig. 3–43 of Ref. 8)
μL = 0.312 mPa · s or cP = 0.312 × 10–3 (N · s)/m2
Diffusivity of acetic acid in water at 100.64°C:
DL = 1.24 × 10–5 cm2/s at 25°C
(Table 3.319 of Ref. 8)
To find DL at 100.64°C,

= constant

where, μ = Viscosity of solvent (water)

μw at 25°C = 0.95 mPa · s or cP
DL =

DL = 5.275 × 10–5 cm2/s ≡ 5.275 × 10–9 m2/s

For vapour:
ρV = 0.6823 kg/m3
μv = Viscosity of vapour mixture

μv = (6.77)

Θij =

and Θji =

[(Eq. (3–87) to (3–89) of Ref. 8)]

Viscosity of acetic acid vapour at 100.64°C:
μAA = 1000 × 10–7 poise = 0.01 cP or mPa · s (i for acetic acid)
Viscosity of water vapour at 100.64°C
μW = 1250 × 10–7 poise = 1250 × 10–5 cP or mPa · s (j for water)

Θij = = 0.4691

Θji =

yi = 1 – 0.9302 = 0.0698, yj = 0.9302

μv =

μv = 0.0215 cP or mPa · s ≡ 0.0215 × 10–3 (N · s)/m2

DV = Diffusivity of acetic acid vapour in water vapour.

DV =   (8.102)

where, DV is in cm2/s. T in K, p is operating pressure in atmosphere, MA and MW

are molar masses and ΣV is atomic diffusion volume.
(Equation 3–133 of Ref. 8. ΣV can be determined by using Table–342 of the
same reference.)
For acetic acid (CH3COOH)
ΣVAA = 2 × 16.5 + 4 × 1.98 + 2 × 5.48 = 51.88
For water (H2O),
ΣVW = 2 × 1.98 + 5.48 × 1 = 9.44
T = 273 + 100.64 = 373.64 K
DV =

DV = 0.25 cm2/s = 0.25 × 10–4 m2/s

Number of gas phase transfer units is calculated by the following equation:

NG =   (8.92)

hw = weir height = 50 mm
FV = va   (8.84)

va = Vapour velocity based on active tray area =

= = 2.5764 m/s
FV = 2.5764 = 2.128

Lp =

Lp =   (8.103)

where, ZL = Length of liquid path

= Horizontal distance between weir and downcomer apron
=   (8.104)

Using Mathematical Table 1.19a of Ref. 8.

Chord h is given by
= 0.362

h = = 0.181 Di

ZL = Di – 2h = Di – 2 × 0.181 Di = Di (1 – 2 × 0.181)
= 0.638 × Di = 0.638 × 1.7 = 1.0846 m
Lp = 1.5843 × 10–3 m2/s

= = 1.26

NG =

= 0.588
Liquid hold-up on the plate, m3 per m2 active area, is given by the following
equation:
Zc = 0.006 + 0.73 × 10–3 hw – 0.24 × 10–3 Fv hw + 1.22 Lp  (8.96)
= 0.006 + (0.73 × 10–3 × 50) – (0.24 × 10–3 × 2.128 × 50)
+ (1.22 × 1.5846 × 10–3) = 0.0189 m3/m2
Liquid contact time
θL =   (8.95)

= = 12.94 s

Number of liquid phase transfer unit

NL = (4.13 × 108 DL)0.5 (0.21 FV + 0.15) θL  (8.94)
= (4.13 × 108 × 5.275 × 10–9)0.5 (0.21 × 2.128 + 0.15) × 12.94 = 11.4
Eddy diffusivity De for sieve tray can be estimated by the following equation:
De = (0.0038 + 0.017 va + 3.86 Lp + 0.18 × 10–3 hw )2
= (0.0038 + 0.017 × 2.5764 + 3.86 × 1.5843 × 10–3 + 0.18 × 10–3 × 50)2
= 0.003 93 m2/s

Peclet number Pe = = = 23.13  (8.96)

Point efficiency, η̇

=   (8.91)

Slope of equilibrium curve of top tray

m =

(Value of y and x are obtained from vapour-liquid equilibrium data, given in

Example 8.5.)
m = (1.0136 × 1.0301 × 1.0454)1/3 = 1.0296
= = 1.2747

Note: At 70 % turndown, individual values of V and L will change, but value

of ratio V/L will remain constant.

η̇ = 0.424
 = 1.2747 × 0.424 = 0.54 and Pe = 23.13

From Fig.   (8.25) = 1.297

Efficiency of sieve tray η = 1.297 × 0.424 = 0.55

Actual tray efficiency ηa, considering liquid entrainment

ηa =   (8.99)

ηa = = 0.4798

(b) For stripping section

To find the tray efficiency of the bottommost tray of stripping section, physical
properties of liquid and vapour of the same tray are required.
Liquid
ρL = 1000 kg/m3
σL = 21.334 × 10–3 N/m
Viscosity of acetic acid liquid at 134.26°C
μL = 0.34 mPa · s or cP = 0.34 × 10–3 kg/(m · s)
Diffusivity DL at 134.26°C

DL =

= 5.75 × 10–5 cm2/s ≡ 5.75 × 10–9 m2/s

Vapour
ρV = 2.8368 kg/m3
μV = 0.009 mPa · s or cP

DV =
= 0.1839 cm2/s ≡ 0.1839 × 10–4 m2/s
Number of gas phase transfer unit
Vapour velocity based on active area
va = = 1.02 m/s

FV = 1.02 = 1.718

Lp =

ZL = 0.638 Di = 0.638 × 2.36 = 1.5 m

Lp = 5.1328 × 10–3 m2 /s

= = 0.1725

NG =

= 2.7234
Liquid hold-up on the plate,
Zc = 0.006 + (0.73 × 10–3 × 50) – (0.24 × 10–3 × 1.718 × 50)
+ (1.22 × 5.1328 × 10–3) = 0.028 m3/m2
Liquid contact time

θL = = 8.18 s

Number of liquid phase transfer units,

NL = (4.13 × 108 × 5.75 × 10–9)0.5 (0.21 × 1.718 + 0.15) × 8.18
= 6.44
Point efficiency, η̇

=   (8.91)

η̇ = 0.7625
 = 2.5 × 0.7625 × = 1.613

Eddy diffusivity
De = (0.0038 + 0.017 va + 3.86 Lp + 0.18 × 10–3 hw)2  (8.98)
= (0.0038 + 0.017 × 1.02 + 3.86 × 5.1328 × 10–3 + 0.18 × 10–3 × 50)2
= 0.0025 m2/s
Peclet number
Pe = = 110  (8.97)

From Fig. 8.25,

= 2.46

η = 2.46 × 0.7625 = 1.876

Efficiency cannot be greater than 1.
Hence η = 1
Actual tray efficiency ηa considering liquid entrainment

ηa = = 0.916

Tray efficiency by Van Winkle’s correlation19

For the topmost tray
η = 0.07 Dg0.14 ScL0.25 Re0.08  (8.87)
where, Dg = Surface tension number =   (8.88)

σL = 55 × 10–3 N/m, μL = 0.312 × 10–3 (N · s)/m2

uV = Superficial vapour velocity

uV = = = 2.797 m/s
Dg = = 63.6

fA = Fractional area =

Liquid Schmidt number, Eq. (8.89):

ρL = 967.97 kg/m3, DLK = 5.275 × 10–9 m2/s

ScL = = 61.1

Reynolds number,

Re = × 10–3 = = 3962.6  (8.90)

η = 0.07 (63.6)0.14 (61.1)0.25 (3962.6)0.08

= 0.679,

ηa = = 0.5752

For the bottommost tray

Surface tension number

Dg =   (8.88)

σL = 21.334 × 10–3 N/m, μL = 0.34 × 10–3 kg/(m · s)

uV = Superficial vapour velocity = = 1.108 m/s

Dg = = 56.63

ρL = 1000 kg/m3, DL = 5.75 × 10–9 m2/s

ScL = = = 59.13
fA = = = 0.076

Re = =

= 6082
η = 0.07 (56.63)0.14 (59.31)0.25 (6082)0.08
= 0.6863

ηa = = 0.6457

Table 8.37 Resulting Data of Tray Efficiency

Tray Efficiency
Method
Topmost Tray Bottommost Tray

AIChE Method 0.4798 0.916

Van Winkle’s Method 0.5752 0.6457

Let tray efficiency for enriching section = 0.45

Tray efficiency for stripping section = 0.5
Actual number of trays in enriching section = ≅ 16

Actual number of trays in stripping section = 17/0.5 = 34

Actual number of trays = 18 + 34 = 52, assume 55.
Feed nozzles can be provided at 32nd, 36th and 40th trays. More number of feed
nozzles are provided to facilitate the flexibility in the operation. For example, in
actual operation if the feed is introduced on 36th tray, purity of distillate could be
more than desired but bottom product (in this case it is pure acetic acid) may not
be of desired purity. In such a case to get desired result, feed can be introduced
on 32nd tray.
Pressure drop
Pressure drop in the topmost tray = 146.32 mm LC
Pressure drop in the bottommost tray = 138.17 mm LC
Assume tray pressure drop for the entire column = 145 mm WC
ρL (at top) = 967.97 kg/m3, ρL (at bottom) = 1000 kg/m3
ΔpT = 55 × 145 = 7975 mm WC
ΔpT = 7975 × 10–3 × 1000 × 9.81 = 78 235 N/m2 ≡ 78.2 kPa
Value of 78.2 kPa is higher than that assumed value of 56.5 kPa (p. 522). With
the correct ΔpT, temperature of boiling liquid in the reboiler is to be reworked.
Refer to Fig. 8.26 for crossflow sieve trays’ details for enriching and stripping
sections.
 Fig. 8.26 Crossflow Sieve Tray Details for Example 8.13

8.5 BATCH DISTILLATION

It is a very common unit operation used in the manufacturing of chemicals
which are produced in batch processes. Batch distillation is carried out with or
without using a rectification column (tray tower or packed tower). Thus, it is
divided in two types: (i) simple distillation, and (ii) batch distillation with
rectification.
8.5.1 Simple Distillation
This type of batch distillation consists of a heated vessel (coiled vessel or
jacketed vessel or kettle-type reboiler), a condenser and one or more receiving
tanks. Feed mixture is charged into the vessel and brought to boiling. Vapours
are condensed and condensate is collected in the receiver. No reflux is returned
(Fig. 8.27). Simple batch distillation provides only one theoretical plate of
separation. In some batch reactors of small plants, at the end of reactions,
volatile components are separated from high boiling liquid solution by simple
distillation. Following equation is applicable to the simple batch distillation.

Fig. 8.27 Simple Batch Distillation

  (8.105)

where, nF = Moles of liquid in initial condition

nW = Moles of liquid in final condition
xF = Mole fraction of more volatile component in liquid phase in initial
conditions
xB = Mole fraction of more volatile component in liquid phase in final conditions
x, y = Mole fraction of more volatile component in liquid and vapour phases at a
given time
If the liquid mixture is a binary mixture for which relative volatility α is
constant, then above Eq. (8.105) can be converted into the following equation:

ln = ln + ln   (8.106)

For a binary system and for simple batch distillation, the following equation is
also applicable.
  (8.107)

where, nAf , nAi = Moles of component A in the pot after and before distillation,
respectively
nBi, nBf = Moles of component B in the pot before and after distillation,
respectively.
In analogue manner, for simple or differential batch distillation of a multi-
component mixture, the following equation is applicable:

  (8.108)

where, nif, nii = Moles of component i in the residue and in the feed in the pot
after and before distillation, respectively
nrf, nri = Moles of reference component (r) in the residue and in the feed,
respectively
αi = Relative volatility of component i with respect to reference component r
1000 kg of feed, containing 30% by mass ethylene glycol and
remaining water, is charged to a batch still. Simple batch distillation is carried
out at 30.4 kPa absolute pressure to get the residue which must contain 95% by
mass ethylene glycol. Find the amount of residue (final product).
 Table 8.39 Vapour – liquid Equilibrium Data of Ethylene Glycol–Water System at 30.4 kPa a8
Temperature, °C x y

69.5 1.00 1.000

76.1 0.77 0.998
78.9 0.69 0.997
83.1 0.60 0.990
89.6 0.46 0.980
103.1 0.27 0.940
118.4 0.15 0.870
128.0 0.10 0.780
134.7 0.07 0.700
145.0 0.03 0.530
160.7 0.00 0.000

Solution
Feed contains 300 kg ethylene glycol and 700 kg water.
Molar mass of water = 18
Molar mass of ethylene glycol = 62
nF = = 43.7276 kmol

Mole fraction of water in feed, xF = = 0.8893

Mole fraction of water in residue, xB = = 0.153 47

ln =

To solve this integration, Table 8.40 is prepared based on the given VLE data as
well as based on the equilibrium curve drawn.
Table 8.40 Batch Distillation Data

x y y – x

xB = 0.153 47 0.875 0.721 53 1.385 94

0.270 00 0.940 0.670 00 1.492 54

0.350 00 0.970 0.620 00 1.612 90
0.460 00 0.980 0.520 00 1.923 08
0.600 00 0.990 0.390 00 2.564 10
0.690 00 0.997 0.307 00 3.257 33
0.770 00 0.998 0.228 00 4.385 96
 xF = 0.889 30 0.999 0.109 70 9.115 77

Draw the graph of vs x (Fig. 8.28).

= area under curve from xB to xF = 2.0643 (integral value)

ln = 2.0643

nW = 5.5493 kmol
Batch distillation data, presented in Table 8.40, were fitted (using Excel*) in an
equation form.
F(x) = (138.6652x4 – 232.9768x3 + 142.0108x2 – 33.8861x + 4.0347) (R2 =
0.9987)
 Fig. 8.28 Evaluation of Integral for Batch Distillation

Integral = 2.116 27

Thus integral value can be obtained by fitting an equation without drawing the
graph. The value differs by 2.51% from the integral value, obtained from the
graph. Advantage of the equation form is that it can be integrated in any desired
range easily.
Recovery of ethylene glycol in final product (nW) =

= 97.05%
A mixture of hydrocarbons at 5.44 atm a is to be differentially
distilled until the moles of propane is reduced to 10 moles in the residue. Total
moles of feed is 100 mol.
Feed composition
Component mole %
Ethane (C2) 10
Propane (C3) 25
n-butane (n – C4) 35
i-Butane (i – C4) 30
Solution
To find the average relative volatilities, bubble point of feed and bubble point of
final residue are required.
At 10°C and 5.44 atm a for feed solution
KC2 = 4.5, KC3 = 1.18, KnC4 = 0.33, KiC4 = 0.48(Ref. Fig. 6.20)
ΣKixi = 4.5 × 0.1 + 1.18 × 0.25 + 0.33 × 0.35 + 0.48 × 0.3 = 1
Hence, 10°C is the bubble point of feed.
Bubble point of Residue:
1st trial calculations
Assume that bubble point of final residue is 40°C.
At 40°C and 5.44 atm a,
KC2 = 7.2, KC3 = 2.2, KnC4 = 0.75, KiB = 1.0(Ref. Fig. 6.20)
Here propane is reference component.
Average relative volatilities with respect to C3

αC2 =

αC3 = 1, αnC4 = 0.309, αiC4 = 0.43

Moles of components in residue
nC3f = 10 moles

nC2f = nC2i

= 10 = 0.3938 mol

nC4f = 35

niC4f = 30

Total moles of the residue = 0.3938 + 10 + 26.37 + 20.23 = 56.9938 mole

Calculated composition of residue
xC2 = 0.0069, xC3 = 0.175, xnC4 = 0.463, xiC4 = 0.355 at 40°C
ΣKixi = 7.2 × 0.0069 + 0.175 × 2.2 + 0.463 × 0.75 + 1 × 0.355 = 1.13 > 1
Hence, actual bubble point is less than 40°C.
At 35°C
KC2 = 6.7, KC3 = 2, KnC4 = 0.67, KiC4 = 0.92
αC2 = 3.58, αC3 = 1, αnC4 = 0.307, αiC4 = 0.432
nC2f = 0.378 mol, nC3f = 10 mol
nnC4f = 26.4 mol, niC4f = 20.2 mol
Total moles of residue = 56.97 mol
Calculated composition of residue
xC2 = 0.0066, xC3 = 0.175, xnC4 = 0.464, xiC4 = 0.354
ΣKixi = 6.7 × 0.0066 + 2 × 0.175 + 0.67 × 0.464 + 0.92 × 0.354 = 1.03 ≈ 1.
Thus residue has 56.97 mol with bubble point of 35°C.
8.5.2 Batch Distillation with Rectification
Simple batch distillation is a single stage separation hence it does not give good
separation unless the relative volatility is very high. In many cases, a
rectification column which may be a tray tower or a packed tower with reflux is
used to improve the performance of batch distillation. If the column is not too
large, then it is mounted on top of the still or heated vessel (Fig. 8.29).

Fig. 8.29 Batch Distillation with Rectification Column

This type of batch distillation consists of a still (heated vessel or reboiler), a

rectification column, a condenser, some means of splitting off a portion of the
condensed vapour (distillate) as reflux and one or more top product receivers. In
operation, a batch of liquid feed is charged to the reboiler or still. Then first,
batch distillation is carried out under total reflux. After achieving steady state,
portion of the overhead condensate is continuously withdrawn in accordance
with the established reflux policy. To get the constant overhead composition, the
amount of reflux returned to the column must be continuously increased
throughout the run.
8.5.2.1 Process design of binary batch rectification with constant overhead
composition8 Since compositions within column are continuously changing
during the batch distillation, rigorous calculation methods are extremely
complex. For a binary system, approximate but useful method is based on
McCabe–Thiele graphical method. In addition to the assumptions of McCabe–
Thiele method (negligible heat of solution and equimolar overflow on the trays),
this approximate method assumes negligible hold-up of liquid on the trays in the
column and in the condenser. Stepwise procedure of this approximate method is
as follows:
1. Determine the minimum reflux ratio by using the following equation:
Rmin =   (8.109)

where, = Minimum internal reflux ratio

=   (8.110)

where, xD = yD = Mole fraction of more volatile component in distillate, if total

condenser is used.
For partial condenser,
yD = Mole fraction of more volatile component in distillate vapour
xD = Mole fraction of more volatile component in reflux
xpi = xF = Mole fraction of more volatile component in liquid feed
ypi = Value of y which is in equilibrium with xpi which can be determined from
the vapour-liquid equilibrium curve or data or equation.
xpi and ypi defined above are valid only for normal equilibrium curve, as shown
in Fig. 8.30.
For the equilibrium curve with inflection as shown in Fig. 8.31, xpi and ypi are
equilibrium compositions corresponding to the inflection point, “P”.
2. Actual reflux ratio is 1.5 to 10 times minimum. Decide the actual reflux ratio
 R for starting of batch distillation. Determine the value of internal reflux ratio
L /V required for starting condition.
3. Draw the operating line for starting of batch distillation. Slope of operating
line is L /V (for starting condition). Determine the number of theoretical
stages required for the desired separation at the start of batch distillation. To
find this, start the staircase construction from xD to xF in between equilibrium
curve and operating line.

Fig. 8.30 Determination of Minimum Reflux for Equilibrium Curve without Inflection
 Fig. 8.31 Determination of Minimum Reflux for Equilibrium Curve with Inflection

4. Select the slightly higher value of R and L /V than the same of starting
condition. For this new value of L /V draw the operating line from xD. Draw
the same number of steps (or equilibrium stages) from xD which was obtained
for starting condition. Find the composition of liquid in reboiler (xWi)
corresponding to new value of R and L /V.
5. Select again slightly greater value of R and L /V. Draw the operating line for
the new value of L /V. Find the composition of liquid in reboiler or still
(residue); xWi, by drawing the same number of steps from xD which was
required for starting condition. Continue these calculations until, new value of
xWi obtained, is equal to xB, where xB is equal to mole fraction of more
volatile component required in bottom product.
6. Time required for batch distillation rectification and with constant overhead
composition can be calculated by Bogart equation8:
θ =   (8.111)

where, θ = Total time required in batch distillation with rectification, s

nF = Total moles of feed-charged, kmol
V = Molar flow rate of vapour phase, kmol/s
L = Molar flow rate of liquid phase, kmol/s
xD = Mole fraction of more volatile component in distillate
xB = Mole fraction of more volatile component in bottom product or final
residue.
xW = Mole fraction of more volatile component in liquid inside the still or
reboiler
xF = Mole fraction of more volatile component in feed
Above equation can also be written as

θi =   (8.112)

where, xWi = Intermediate composition (mole fraction of more volatile

component) of liquid in still or reboiler.

Define F′(xW) =   (8.112a)

By using Eq. (8.112), for the different values of θ, L /V and R can be determined.
Based on these data, the graph of R vs θ can be developed. This graph or
curve is also useful in actual operation, giving information about how to
increase the value of reflux ratio R with respect to time to maintain the
constant composition of overheard (top) product.
7. The quantity of distillate at any time θ can be determined by using the
following equation:
nF – nWi =   (8.113)

where, nWi = Moles of liquid in still at any time θi, kmol

At the end of distillation, quantity of material distilled
nD = nF – nB =   (8.114)
where, nB = Moles of final bottom product, kmol
8. Carry out the material of construction of packed tower or tray tower.
9. Select the type of tower; tray tower or packed tower.
10. Determine the tower diameter and pressure drop per unit height or tray.
11. Determine the tray efficiency of tray tower or HETP in case of packed tower.
12. Determine the actual number of trays required in tray tower or height of
packing required in packed tower.
13. Determine the actual total pressure drop.
14. Design the condenser, reboiler, etc.
1000 kg of feed, containing 40% by mass methanol and
balance water, is to be separated in batch distillation with a rectification column.
Top product must contain 99.9% by mass of methanol. Composition of distillate
should remain constant throughout the batch distillation. Desired composition of
residue is 95% by mass water.
(a) Design the rectification column.
(b) Determine the operating (actual) reflux ratio vs. time data.
Operating pressure in rectification column is atmospheric.
Table 8.41 Equilibrium Data for Methanol-Water Data at 101.3 kPa
Temperature, °C Mole fraction of methanol Relative volatility αi = Y/X
x X = x/(1 – x) y Y = y/(1 – y)

100.0 0.0 0 –
96.4 0.02 0.0204 0.134 0.1547 α1 = 7.5820
93.5 0.04 0.0417 0.230 0.2987 α2 = 7.1688
91.2 0.06 0.0638 0.304 0.4368 α3 = 6.8429
89.3 0.08 0.0870 0.365 0.5748 α4 = 6.6102
87.7 0.100 0.1111 0.418 0.7182 α5 = 6.4639
84.4 0.150 0.1765 0.517 1.0704 α6 = 6.0656
81.7 0.200 0.2500 0.579 1.3753 α7 = 5.5012
78.0 0.300 0.4286 0.665 1.9851 α8 = 4.6318
75.3 0.400 0.6667 0.729 2.6900 α9 = 4.0351
73.1 0.500 1.0000 0.779 3.5249 α10 = 3.5249
71.2 0.600 1.50000 0.825 4.7143 α11 = 3.1429
69.3 0.700 2.3333 0.870 6.6923 α12 = 2.8681
67.5 0.800 4.0000 0.915 10.7647 α13 = 2.6912
66.0 0.900 9.0000 0.958 22.8095 α14 = 2.5344
65.0 0.950 19.0000 0.979 46.6190 α15 = 2.4536
64.5 1.000 1.000 –

(Table 13-1 of Ref. 8)

Average volatility, αavg = = 4.448

Solution
(a) Graphical Method:
Basis: 1000 kg feed consisting 40% methanol (by mass)
nF = = 45.833 kmol

Mole fraction of methanol in feed, xF = = 0.2727

Mole fraction of methanol in distillate, xD = = 0.9982

Mole fraction of methanol in residue,

xB = = 0.028 75

nD =   (8.114)

= 11.5333 kmol
nB = 45.833 – 11.5333 = 34.2997 kmol
Draw the equilibrium curve and find the minimum reflux ratio, Rm.

Rm =   (8.109)

where, (L/V)min =   (8.110)

Figure 8.32 shows equilibrium curve for methanol-water system at 101.325 kPa.
xD = yD = 0.9982
xpi = xF = 0.2727
Corresponding equilibrium value of y of xpi
ypi = 0.6415
(L/V)min = = 0.491 66

Rm = = 0.9672

Let the reflux ration R at the starting of batch distillation (at θ = 0) is equal to
2Rm.
R1 = 2 × 0.9672 = 1.9344
Slope of operating line
L /V = = 0.6592

Point on y-axis R xB

1.9344 0.3400 0.6354

 3.0 0.249 55 0.2296

3.5 0.2218 0.1427

4.0 0.2000 0.1036

4.5 0.1815 0.0774

5.0 0.1664 0.0616

6.0 0.1426 0.0447

8.0 0.1109 0.0294

9.0 0.1000 0.0288

Fig. 8.32 McCabe–Thiele Diagram for Batch Rectification of Methanol-Water Solution

Intercept of this operating line

xD/(R1 + 1) = 0.9982/2.9344 = 0.34
Figure 8.32 shows that for R1 = 1.9344 it is not possible to get the bottom
composition of xB = 0.028 75. Hence, it is planned to increase the reflux ratio
gradually from R1 = 1.9344 to R9 = 9.
For reflux ratio R9 = 9 and intercept = 0.1, approximately 10 number of
theoretical stages required to get xB = 0.028 75 as inferred from Fig. 8.32. Based
on this the value of number of theoretical stages N = 10 is fixed for all values of
reflux ratio. For each value of reflux ratio, value of bottom composition xWi is
determined for N = 10.
Table 8.42 Batch Distillation Data

R L/V xW

R1 = 1.9344 0.65921 0.34017 0.63540 —

R2 = 3 0.75000 0.24955 0.27270 7.5995

R2 = 3 0.75000 0.24955 0.22960 6.7711

R3 = 3.5 0.77778 0.22182 0.14270 6.1485

R4 = 4 0.80000 0.19964 0.10360 6.2476

R5 = 4.5 0.81818 0.18149 0.07740 6.4868

R6 = 5 0.83333 0.16637 0.06160 6.8398

R7 = 6 0.85714 0.14260 0.04470 7.6994

R8 = 8 0.88889 0.110 91 0.02940 9.5890

R9 = 9 0.90000 0.09982 0.02875 10.6402

 Note: Table 8.42 shows that for R1 = 1.9344 and N = 10, xw = 0.6354 > xF. It
means that for R1 = 1.9344, more than 10 number of equilibrium stages are
required to get the desired separation. In actual operation for R = 1.9344 and N
= 10, distillate composition obtained will be less than xD even at the starting of
batch distillation. It is calculated to be 0.9917 from the graph (not shown in
Fig. 8.32). For R2 = 3 and N = 10, xw = 0.2296 < xF. Hence, it is decided to
keep R = 3 (min) at the start of batch distillation.
Time required for batch distillation and with constant overhead composition is
given by Bogart equation [Eq. (8.111)].

θ =

Draw the graph of F′(xW) = vs xW and find the area under the
curve.

Fig. 8.33 Evaluation of Integral for Batch Distillation

From Fig. 8.33, area under the curve from xW = 0.028 75 to xW = 0.2727 is equal
to 1.637 47 (integral value).
Batch rectification data, presented in Table 8.42, were fitted (using Excel®) in an
equation form.
F′(xW) = 4500 637.98x6W – 3944 492x5W + 1356 915.1x4W – 234 196.7x3W
+ 21 563.33x2W – 1024.75xW + 26.56 (R2 = 0.9794)

Integral = dxW = 1.649 81

θ =

Vθ = 54.449 kmol
If the total time of batch distillation is fixed at θ = 6 h,
V = = 9.0748 kmol/h

Total time for batch distillation in any batch process is fixed based on the time
required by the equipment which is governed by the batch process. In most of
the cases, time required by the reaction in batch reactor decides the time of batch
process and time of other separation operations like batch distillation.
Total distillation time, θ = 6 h
Here reflux ratio R at the starting of batch distillation is 3. In operation R can be
kept equal to 3 from θ = 0 to θ = θ1.

θ1 =

θ1 = × 6 = 1.122 h

For R3 = 3.5, area of segment from xW = 0.2296 to xW = 0.1427 is 0.558 89

(determined from Fig. 8.33)
Hence, θ2 = × 6 = 2.048 h

Similarly,
For R4 = 4, θ3 = × 6 = 0.8883 h

For R5 = 4.5, θ4 = × 6 = 0.5509 h

For R6 = 5, θ5 = × 6 = 0.4485 h

For R7 = 6, θ6 = × 6 = 0.4468 h

For R8 = 8, θ8 = × 6 = 0.473 h

For R9 = 9, θ9 = × 6 = 0.0225 h

Table 8.43 Reflux Ratio vs Time

R θ, h xW

3 0 to 1.122 0.2727 to 0.2296

3.5 1.122 to 3.17 0.2296 to 0.1427
4 3.17 to 4.0583 0.1427 to 0.1036
4.5 4.0583 to 4.6092 0.1036 to 0.0774
5 4.6092 to 5.0577 0.0774 to 0.0616
6 5.0577 to 5.5045 0.0616 to 0.0447
8 5.5045 to 5.9775 0.0447 to 0.0294
9 5.9775 to 6 0.0294 to 0.02875

(b) Numerical Integration Method

Alternate method of obtaining the integral value is given by Koppel24.
The following equation can be derived by combining material balance equation
with a vapour-liquid equilibrium equation:

yn–1 =   (8.115)

and yn–1 =   (8.116)

Combining the above equations,

xn–1 =   (8.117)

For total condensation in the overhead condenser, xD = y1. Also for constant
molar flow rates on each tray, L/D = R and V/L = (R + 1/R). Substitution of these
ratios in terms of reflux ratio, Eq. (8.117) can be solved.
By selecting R and xD, xn–1 can be calculated from Eq. (8.117) which represents
the mole fraction of light component on tray 2 from top. Similarly, for the next
tray (i.e., number 3), xn–2 can be calculated from xn–1. In this manner xn can be
calculated for all theoretical trays; the last one will represent xW. Tray-to-tray
calculations are summarised in Table 8.44.
For xD = 0.9982, N = 10 and α = 4.448,
Table 8.44 Tray-to-Tray Calculations for Varying Reflux Ratios
R = 3 R = 4
Tray No.
xn xn–1 xn xn–1

1 xD = 0.9982 0.9920 xD = 0.9982 0.9920

2 0.9920 0.9721 0.9920 0.9708

3 0.9721 0.9114 0.9708 0.9024
4 0.9114 0.7581 0.9024 0.7253
5 0.7581 0.5029 0.7253 0.4433
6 0.5029 0.2740 0.4433 0.2185
7 0.2740 0.1581 0.2185 0.1186
8 0.1581 0.1158 0.1186 0.0858
9 0.1158 0.1023 0.0858 0.0762
10 0.1023 0.0982 = xW 0.0762 0.0734 = xW

Similar calculations cab be made for different reflux ratios. These are
summarised in Table 8.45.
Table 8.45 Computation of Integral between xW and xF
F′(xW)
R xW 1/(xD – xW) R/(R + 1) Integral value F′(xW)avg · ΔxW
Eq. (8.112a)

3 xF = 0.2727 1.3784 0.7500 7.5995

3 0.0981 1.1110 0.7500 4.9373 1.094 41

3.5 0.0840 1.0938 0.7778 5.3842 0.072 84
4 0.0731 1.0813 0.8000 5.8463 0.059 45
4.5 0.0652 1.0718 0.8182 6.3183 0.049 90
5 0.0587 1.0643 0.8333 6.7970 0.042 94
6 0.0489 1.0534 0.8571 7.7670 0.071 34
7 0.0419 1.0457 0.8750 8.7474 0.057 65
8 0.0366 1.0400 0.8889 9.7341 0.048 32
9 0.0326 1.0356 0.9000 10.7249 0.041 56
10 0.0293 1.0321 0.9091 11.7184 0.036 45
10.5 0.0279 1.0307 0.9130 12.2159 0.016 65 = xW

Total 1.591 50

Compare integral value 1.637 47 (to attain xW = 0.028 75) from graph (Fig. 8.33)
with the value 1.5915 (to attain xW = 0.016 65), obtained in Table 8.45. Both
values differ by 2.81%. This difference can be attributed to the use of average
volatility (αavg) for numerical calculations.
Design of distillation column
Type of tower = Packed tower (because of small capacity)
Type of packing = Stainless Steel Pall rings
Size of packing = 16 mm
Void space = 92%
Packing factor Fp = 230 m–1 (Table 9.1)
Tower diameter required at bottom:

FLG = (9.15)

Molar flow rate of vapour, V = 9.0748 kmol/h

V = L + D and L /D = R
Maximum value of actual or operating reflux ratio in the entire operation is 9.
Lmax = Rmax D = 9 D
V = L + D = 9D + D = 10 D
D = = 0.907 48 kmol/h

L = 8.1673 kmol/h
Composition and average molar mass of vapour and liquid phases at bottom are
changing with time. Average molar mass is maximum at the starting of
distillation.
At the starting of distillation
Mav = 0.2727 × 32 + (1 – 0.2727) × 18 = 21.8178 kg/kmol

= = = 0.9

Bubble point temperature at x = 0.2727 is 79oC (from t–x–y data).

ρG = = 0.7554 kg/m3
ρL =

ρL = 870.9 kg/m3

FLG = 0.9 = 0.0265

From Fig. 9.5, KF = 0.2

Let actual velocity of vapour through packed tower = 70% of flooding velocity.

× 100 = 70

K = (0.70)2 or KF = 0.098
Corresponding Δp = 83.3 mm WC/m of packing (From Fig. 9.5).

GW = (9.18)

K = 0.098, ρG = 0.7554 kg/m3, ρL = 870.9 kg/m3

Fp = 230 m–1
(Density of water is from Table 3.28, of Ref. 14)
ψ = = = 1.1166

ψ is inverse of specific gravity of liquid.

μL = viscosity of liquid solution
Viscosity of 40% methanol solution (by mass) at 79°C,
μL = 0.45 cP = 0.45 mPa · s
(Fig. 3.43 of Ref. 8)

GW =

= 1.7 kg/(m2 · s)
Mass flow rate of vapour, ṁv = 9.0748 × 21.8178 = 198 kg/h ≡ 0.055 kg/s

Area required at bottom = = 0.032 35 m2

A = 0.032 35 =

Di = 0.203 m
Tower diameter required at top
= = 0.9

Mav = 0.9982 × 32 + (1 – 0.9982) × 18 = 31.9748 kg/kmol

ρG = = = 1.1546 kg/m3

Top temperature is the dew point temperature of overhead vapour. It can be

determined from t-x-y data. Find the value of t for y = xD = 0.9982.
ρL = Density of methanol at 64.5 °C = 760 kg/m3

FLG =

From Fig. 9.3, KF = 0.18

Let actual velocity of vapour is equal to 70% of flooding velocity.
K = (0.7)2 KF = 0.0882
Corresponding Δp = 90 mm WC/m of packing
μL = Viscosity of methanol at 64.5°C = 0.34 cP = 0.34 mPa · s

GW =

= 1.7819 kg/(m2 · s)
Mass flow rate of vapour, ṁv = 9.0748 × 31.9748 = 290.165 kg/h ≡ 0.0806 kg/s
Area required at top = = 0.045 23 m2

Di = = 0.24 m
Select higher value of the tower diameter (Di) of two values. Take Di = 0.24 m
for the entire tower.
HETP (height equivalent theoretical plate) for Pall rings = 0.5 m
Total height of packing = HETP × N = 0.5 × 10 = 5 m
Let the total height of packing = 5.1 m
Number of packing sections = 3 each with 1.7 m packed height
Type of liquid distributor = Pipe type
Number of holes required in pipe-type reflux distributor,

Fig. 8.34 Pipe-Type Liquid Distributor

nh =

For each 30 in2 (194 cm2) area, one distribution point is required.
nh = = 2.33

Let nh = 4
However, the pipe of 25 mm NB is proposed for distribution.
Let the velocity of reflux through pipe = 2 m/s
Volumetric flow rate of reflux
=

= 9.545 × 10–5 m3/s

Inside diameter of pipe:

di = = 0.007 79 m

Let di = 8 mm
Let the velocity of solvent through holes = 3 m/s
Let dh = diameter of hole, mm

4 × = = 3.181 67 × 10–5

dh = 0.0032 m
Let dh = 3.5 mm
Packing support
Packing support should be selected such that flow area provided by packing
support for the flow of gas should be greater than flow area provided by packing
material.
Void space for 16 mm size SS Pall rings = 92% (Table 9.1)
Hence with Pall rings, any gas-injection-type packing support must be used. Let
the type of packing support is cap-type packing support.
Actual outer diameter of packing support is greater than 240 mm as some
portion of packing support is sandwiched between two flanges.
 Fig. 8.35 Cap-Type Packing Support

Let h = Height of slot in riser, mm

a = Width of slot in riser, mm
dr = Diameter of riser, mm
n = Number of risers, mm
Di = Inside diameter of column, mm
Let dr = = 40 mm

Number of risers, n = 9
Let ns = number of slots per riser
Total area of slots of risers = ns n h a where, distance between two successive
slots is a mm.
Let a = 5 mm
ns = = = 12.566

Let ns = 12
Hence, total area of slots of risers = 12 × 9 × h × 5 ≥ 0.92 × (240)2

h ≥ 77 mm
Let h = 80 mm
8.5.2.2 Process design of a binary batch rectification to attain definite
minimum concentration of distillate In many cases of batch distillation, it is not
required to attain high purity of distillate as discussed in the previous section.
Since the recovered distillate is reused in the plant, purity requirement can be
lower but a certain minimum concentration of more volatile component is
desired. For design of such a system, at times reflux ratio is kept constant and
distillation is carried out for a known period. For such designs, graphical2 and
numerical procedures24 differ from the procedures described in the previous
section. Reader is advised to see the original references for the procedures.
8.5.3 Process Design of Multicomponent Batch
Rectification
Stepwise procedure for process design for multicomponent batch rectification2
with constant overhead composition is as under.
1. Select light key and heavy key components.
2. Fix the distribution of light key and heavy key component in distillate and
final residue. Calculate the distribution of non-key components in distillate
and residue and complete the material balance using Eq. (8.108).
3. Calculate dewpoint of distillate, bubble point of feed and bubble point of final
residue.
4. Calculate the average relative volatilities of all components with respect to
heavy key component.
5. Calculate the minimum number of equilibrium stages (Nm) required for the
desired separation by Fenskey’s equation Eq. (8.26).
6. Fix the number of equilibrium stages required for the desired separation. N
can be 1.5 to 3 times Nm.
7. Keep mole fraction of light key component in distillate (xLD) and mole
 fraction of heavy key component in distillate (xHD) as constant. Change the
values of mole fraction of light key in residue (xLW) from xLF to xLB, where,
xLF and xLB are mole fractions of light key in feed and in final residue,
respectively. For any intermediate value of xLW, corresponding value of xHW
can be determined by material balance.
nF = nDi + nWi (i = intermediate value)  (8.118)
nF · xLF = nDi · xLD + nWi · xLW  (8.119)
From equations   (8.118) and   (8.119), calculate nDi and nWi. Then find
xHW by using following material balance equations.
nF · xHF = nDi · xHD + nWi · xHW  (8.120)
8. Find Rm by following equation:

Rm =   (8.121)

where α is average relative volatility of light key with respect to heavy key.
9. Find the actual reflux ratio R based on the values of Rm, Nm and N by using
Gilliland’s correlation, Eq. (8.28) and Eq. (8.29) or Fig. 8.7.
10. Calculate by the following equation:

  (8.122)

Thus for the different values of xLW from xLF to xLB, corresponding values of
xHW, Rm, R and L/V can be determined.
11. Time required for multi component batch distillation with rectification and
with constant overhead composition can be calculated by Bogart equation:

θ =   (8.123)

12. Develop R vs θ curve.

Determine the reflux ratio vs time data for the batch distillation
with rectification for the following mixture.
Feed: 200 kmol with 50% benzene, 25% toluene and 25% ortho xylene (by
mole)
Distillate: 99% pure benzene (by mole)
Distillate composition should remain constant throughout the course of
distillation.
Distillation is to be continued until the moles of benzene is reduced to 10 kmol
in the residue.
Data: Antoine equation:

Log10 pv = (2.1)

where, pv is vapour pressure in bar and T is the temperature in K.

Table 8.46 Antoine Constants
Component A B C

Benzene 4.725 83 1660.652 –1.461

Toluene 4.078 27 1343.943 –53.773

o-Xylene 4.129 28 1478.244 –59.076

Solution
Here benzene is light key and toluene are heavy key components.
Assume that remaining 1% of distillate is toluene.
Table 8.47 Material Balance
Feed Distillate Residue
Component
kmol xF kmol xiD
 kmol xiB

Benzene 100 0.5 90 0.99 10 0.0917

Toluene 50 0.25 0.91 0.01 49.09 0.4500

o-Xylene 50 0.25 – – 50 0.4583

Total 200 1.00 90.91 1.00 109.09 1.0000

Bubble point temperature of feed

Operating pressure = 1 atm = 1.013 25 bar
At 96°C (369 K)
pvB = 1.6126 bar, pvT = 0.6529 bar, pvX = 0.2289 bar
Σpvi · xFi = 1.6126 × 0.5 + 0.6529 × 0.25 + 0.2289 × 0.25 = 1.027 bar
≅ 1 atm
Hence bubble point of feed = 96°C
Dewpoint of distillate
At dewpoint temperature,

At 81°C (354 K),

Hence, dewpoint of distillate = 81°C (354 K)

Bubble point of final residue
At 117°C (390 K)
pvB = 2.8297 bar, pvT = 1.2054 bar, pvX = 0.4595 bar
ΣxBi pvi = 0.0917 × 2.8297 + 0.45 × 1.2054 + 0.4583 × 0.4595 = 1.0125 bar
≡ 1.013 25
Hence bubble point of final residue = 117°C (390 K)
Average relative volatilities of components with respect to heavy key

αBi =
αBf =

αB = (2.529 × 2.4655)1/2 = 2.497

αT = 1
Vapour pressure of o-xylene at 81°C,
pvX = 0.1309 bar

αXi =

αXf =

αX = 0.3459
Minimum number of theoretical stages required for the desired separation

Nm =

Nm =

= 6.7598
Let actual number of theoretical stayes required for the desired separation, N =
12
=

Using Eq. (8.28) and Eq. (8.29),

= 0.2615

From Fig. 8.7, = 0.2678

For xLW = xLF = 0.5

Rm =

Rm =

Rm = 1.256

= 0.2615

R = 2.055,

For xLW = 0.4

Rm =

nF = nD + nW = 200
= nDi + nWi
nF · xLF = nDi · xLD + nWi · xLW
200 × 0.5 = nDi × 0.99 + nWi × 0.4
= nDi × 0.99 + (200 – nDi) × 0.4
nDi = 33.9 kmol, nWi = 166.1 kmol
nFxHF = nDixHD + nWixHW
200 × 0.25 = 33.9 × 0.01 + 166.1 × xHW
xHW = 0.299
Rm = 1.5975

= 0.2615

R = 2.5173
 = = 10.1046

Similarly, following table can be prepared:

Table 8.48 Multicomponent Batch Rectification Data

xLW xHW Rm R L/V

0.5 0.25 1.256 2.055 0.6727 12.7239

0.4 0.299 1.5975 2.5173 0.7157 10.1043

0.3 0.348 2.156 3.2735 0.7660 8.9761

0.2 0.3969 3.2646 4.7747 0.8268 9.2528

0.15 0.4214 4.369 6.2701 0.8625 10.3034

0.0917 0.45 7.1747 10.0693 0.9097 13.7176

Figure   (8.36) is the plot of xLW vs F″(xLW) = .

 Fig. 8.36 Evaluation of Integral for Batch Distillation with Rectification

Total area under the curve = 4.1211 (integral value)

θ =

θ = × Area under curve from xLW = 0.0917 to 0.5

Vθ = 200 (0.99 – 0.5) × 4.1211 = 403.87 kmol

Batch rectification data, presented in Table 8.48, were fitted (using Excel®) in an
equation form.
F″(xLW) = 1204.42xLW4 – 1626.18xLW3 + 861.85xLW2 – 201.72xLW + 26.12
(R2 = 1.00)

Integral =

Let θ = 6 h (total time period for batch distillation)

V = 403.87/6 = 67.31 kmol/h
Area of segment [from xW = xF = 0.5 to xW = 0.4]
For R = 2.5173
= 1.1197
θ1 = = 1.6031 h

Area of segment [from xW = xF = 0.5 to xW = 0.3] = 2.0665

For R = 3.2735
θ2 = = 3.0087 h

Area of segment [from xW = xF = 0.5 to xW = 0.2] = 2.9069

For R = 4.7747
θ3 =

Area of segment [from xW = xF = 0.5 to xW = 0.15] = 3.4466

For R = 6.2710
Table 8.49 Calculated R vs θ
θ, h R

0 2.055
1 2.35
2 2.691
3 3.267
4 4.28
5 6.226
6 10.07

θ4 =

Using Ri vs θi data, Fig. 8.37 is prepared. For integral values of θi, Ri values are
read from Fig. 8.37 which are tabulated in Table 8.49.

Fig. 8.37 Evaluation of Integral for Batch Distillation with Rectification

8.6 SHORT PATH DISTILLATION25,26

Short path distillation is very high vacuum distillation. Operating pressure at
distillation surface in short path distillation unit (SPDU) could be as low as 10–6
bar (1 μbar). Short path distillation provides the distillation at the minimum
possible temperature which is desirable for heat sensitive products.
In distillation, maximum boiling point is of the bottom product. If the normal
boiling point of any product of distillation is higher than the decomposition
temperature or cracking temperature, distillation cannot be carried out at
atmospheric pressure. Reduction in absolute pressure results in the reduction in
distillation temperature or boiling temperature. In lower pressure range, more
reduction in boiling temperature of pure component is observed for a small
reduction in operating pressure. For example, boiling point of stearic acid is
reduced by 0.2°C when operating pressure is reduced from 251 mbar to 250
mbar. But, its boiling point is reduced by 3.8°C when operating pressure is
reduced from 5 mbar to 4 mbar. In both cases, reduction in operating pressure is
1 mbar but in lower pressure range reduction in boiling point is quite high.
Hence, reduction of pressure in lower pressure range is important as it creates
considerable reduction in boiling point. For a food grade product (for example,
orange juice) and for a highly heat sensitive product (for example, vitamin E,
i.e., α–tocopherol) reduction in boiling point means better recovery. Maximum
recovery is obtained at the minimum boiling temperature.
Decreases in pressure results in the increase of vapour volume per unit mass and
in turn it results in increase in vapour velocity. In the lower pressure range, even
low vapour flow rate or non-condensable flow rate results in relatively higher
value of pressure drop.
It will be interesting to note change in properties of ice and water vapour from
Table 8.50 at low, high and ultra high vacuum.
Table 8.50 Properties of Water, Ice and Water Vapour at Low Temperatures27
Melting point
Saturation vapour pressure, psDensity of ice, ρsSpecific volume of saturated vapour, V Specific enthalpy, kJ/kg
T t
K °C Pa bar kg/m3 m3/kg Ice SublimationVapour

273.15 0 611.65 0.611 × 10–2 916.708 205.99 –333.4 2834.3 2500.9

270 –3.15 470.06 0.470 × 10–2 917.170 264.96 –340.0 2835.1 2495.1

265 –8.15 305.92 0.306 × 10–2 917.886 399.64 –350.2 2836.2 2486.0

260 –13.15 195.81 0.196 × 10–2 918.594 612.65 –360.3 2837.1 2476.8

255 –18.15 123.15 0.123 × 10–2 919.295 955.48 –370.2 2837.8 2467.6

250 –23.15 76.02 0.760 × 10–3 919.980 1517.5 –380.0 2838.3 2458.3

245 –28.15 46.01 0.460 × 10–3 920.658 2457.3 –389.5 2838.6 2449.1

240 –33.15 27.27 0.273 × 10–3 921.319 4061.4 –398.9 2838.7 2439.8

235 –38.15 15.81 0.158 × 10–3 921.973 6860.7 –408.0 2838.6 2430.6

230 –43.15 8.95 0.895 × 10–4 922.611 11 861 –417.0 2838.4 2421.4

225 –48.15 4.94 0.494 × 10–4 923.233 21 021 –425.9 2838.0 2412.1

220 –53.15 2.65 0.265 × 10–4 923.839 38 249 –434.5 2837.3 2402.8

215 –58.15 1.39 0.139 × 10–4 924.437 71 588 –443.0 2836.6 2393.6

210 –63.15 0.702 0.702 × 10–5 925.018 1.3809 × 105 –451.3 2835.6 2384.3

205 –68.15 0.344 0.344 × 10–5 925.583 2.7514 × 105 –459.4 2834.4 2375.0

200 –73.15 0.1626 1.626 × 10–6 926.132 5.6759 × 105 –467.3 2833.1 2365.8

195 –78.15 0.7404 0.740 × 10–6 926.655 1.2155 × 106 –475.1 2831.6 2356.5

Reference State:
Specific enthalpy of saturated liquid water at 273.16 K (0.01°C) = 0 kJ/kg

Specific volume of saturated water vapour at 0.1 bar (boiling point = 45.83°C),
1.232 × 10–3 bar (or 1.232 mbar) and 1.626 × 10–6 bar (or 1.626 μbar) is 14.675
m3/kg, 955.48 m3/kg and 5.6759 × 105 m3/kg, respectively. Thus several fold
increase in vapour volume will have to be handled at ultra high vacuum. This
will call for big size pipes (ducts) for transporting vapours at ultra high vacuum
so that low pressure drops are achieved.
Hence, in all vacuum distillation columns, minimum possible value of absolute
pressure at vaporisation surface of boiling liquid in reboiler is decided by
pressure drop of vapour flowing from vaporisation surface of liquid in reboiler to
the heat transfer surface of external condenser. In conventional packed tower-
type vacuum distillation column, vapour travels a long path from reboiler to
external condenser. Also, in conventional design, no extra care is generally taken
to reduce the flow of non-condensables.
To break this physical barrier of pressure drop of vapour and non-condensables
travelling very long path from vapourising surface to external condenser, in short
path distillation, an internal condenser is used and because of the same, vapour
travels through a very short distance, in order of few centimetres, from
vaporisation surface of liquid film over heating surface to the surface of internal
condenser. Because of the shortening of the path length of vapour and non-
condensables, this new unit is known as short path distillation unit (SPDU).
There is a confusion in between the words “Molecular Distillation” and “Short
Path Distillation”. In the latter, if the distance travelled by vapour molecules
from the vaporisation surface to the surface of internal condenser is less than
mean free path of molecules between two successive collisions, then short path
distillation is called molecular distillation. If in this path length, there is no
collision between vapour molecules then pressure drop in vapour flow is the
minimum. Decrease in pressure increases the mean free path of molecules
between two successive collisions. For commercial applications, it is not
necessary to achieve the condition of molecular distillation for each application,
hence, only a few short path distillation units are operated as molecular
distillation units.
8.6.1 Design and Working of Short Path Distillation
Unit
A short path distillation unit (Fig. 8.38) consists of a vertically mounted double
walled vacuum chamber with a central inner condenser and roller wiper system.
 Fig. 8.38 Typical Short Path Distillation Unit

Feed is continuously fed on a rotating distributor plate. At the bottom of rotating

distribution plate, 3 to 12 guiding rods are mounted. These guiding rods contain
the rollers which are passed over the rods. Distribution plate with rods and
rollers is rotated by an electric motor. This entire rotating system is known as
Roller–Wiper system. Feed material is thrown by rotating distributor to wall or
to heat transfer surface. Then material flows down by gravity over the heat
transfer surface. Rollers do not touch the heat transfer surface but they create the
pressure and turbulence in liquid film. Rollers are rotated about the central axis
of the unit and are also rotated about their own axis. Friction between liquid film
and rollers create the automatic rotation in rollers about their own axis. This
roller wiper system creates effectively mixed, uniform and thin film over the
heating surface. Heating is provided by thermic fluid, saturated steam or hot
water. Liquid film is partially vaporized. Vapour travel through the shortest path
from liquid film surface to the surface of inner central condenser. Cooling
medium may be cooling water, chilled water, oil or brine.
Residue and condensate (distillate) both are falling down by gravity. Partition is
provided to avoid the possibility of the mixing of residue and distillate.
Vacuum nozzle located at the side bottom of vessel, through which non-
condensables (and some vapours) are drawn out by vacuum system. Vapours are
condensed in a cold trap. Vacuum measuring instrument is also provided in the
bottom part.
Unit is kept in the perfect vertical position and hence, residence time of liquid on
the heating surface is in the order of few seconds. The distillation takes place in
“single pass” or in “once through”. In this operation, in the short residence time
and the lowest, vaporisation temperature reduces the possibility of thermal
degradation of product. Thus SPDU provides the most gentle conditions for the
distillation of heat sensitive products.
Sometimes it is advisable to have a degassing stage in front of short path
distillation unit to remove the dissolved noncondensable gases like oxygen,
nitrogen, etc. In addition to this maximum tightness of the system, the
prevention of gaseous cracked products and the lack of stripping gas reduce size
of vacuum system required and also reduces the operating cost of the vacuum
system.
Standard short path distillation units with heating surface area range from 0.01
m2 to 42 m2, heating temperature up to 450°C and product throughput range
from a few g/h to several t/h (multiple units) are available.
Theoretical treatment of agitated thin film evaporator is given by Salden28 in
which heat transfer surface area requirement can be calculated (i.e., estimated)
by knowing properties of the fluid to be processed in the evaporator and the
heating medium. This area correspondences to the jacket heat transfer surface.
However, most recognised suppliers of such equipments provide trial facility on
a pilot model. It is recommended to conduct some trials on the pilot plant and the
model suitable for the service is selected.
8.6.2 Applications of SPDU25,26
The use of short path distillation to modify the traditional processes or for the
development of a new process requires special knowledge because many
classical rules and laws of traditional vacuum distillation cannot apply to short
path distillation.
Short path distillation finds its application in pharmaceutical industries, fine
chemicals industry, polymer industries, food industries, oil industries and
refineries. Some important applications are:
1. Separation of Vitamin E: The short path distillation is applied for
synthetically produced vitamin-E as well as for natural vitamin-E. The
synthesised vitamin-E exists in the final stage as acetic acid ester (as vitamin E-
acetate). Typically 100 kg crude vitamin-E acetate contains 78.75 kg pure
vitamin-E acetate, 13.75 kg residue and 7.5 kg low boiling impurities. In two-
stage short path distillation, pure vitamin E acetate is obtained as light yellow
pure oil.
Deodoriser distillate of an edible oil refinery is a source of natural vitamin-E. Its
concentration varies from 3 to 10% (by mass) in the distillate of soybean oil,
depending on the operating conditions of the deodoriser. Concentration up to
40% (by mass) can be carried out in SPDU as shown in Fig. 8.39.

Fig. 8.39 Distillation of Vitamin-E

2. Separation of Monoglyceride:
Monoglycerides are mono fatty acid esters of glycerol.
Mono fatty acid + Glycerol → Monoglyceride + Water
Monoglycerides are used as emulsifiers in many of food products. The modern
food product technology demands the concentration of monoglyceride more than
90%. It is a heat sensitive product. More than 90% concentration of
monoglycerides is achieved by two-stage short path distillation (Fig. 8.40).
3. Carotene: It is a pro-vitamin which is used as a supplement in many food
products. Red palmoil ester contains about 0.06% carotene. Three-stage short
path distillation plant (Fig. 8.41) is required to increase the concentration up to
40%.
4. Dimeric Fatty Acids: They are basic materials for polyester or alkyd resins.
They are produced by dimerisation of oleic acid. After reaction a mixture of
monomer, dimer and oligomer fatty acids are formed. The fractionation of this
mixture takes place in two or there stage short path distillation plant (Fig. 8.42).
5. To recover more middle distillate from the residuum, coming from vacuum
distillation unit of refinery, currently vacuum distillation is carried out in
conventional distillation equipment in which operating pressure at the base of
column or in reboiler is decided by pressure drop in the column. The same
operating pressure also decides the certain boiling range of residuum. Currently,
considerable amount of middle distillate remains in the residuum. With
conventional vacuum distillation unit, one can

Fig. 8.40 Manufacturing of Monoglyceride

 Fig. 8.41 Concentration of Cartone in Palm Oil Ester

increase the % recovery of middle distillate by increasing base temperature but

the same cannot be permitted because increase in the base temperature above the
certain limiting value results in considerable thermal cracking of the residuum.
In short path distillation, the same residuum can be boiled or distilled at much
lower temperature range. Reduction in operating pressure in SPDU creates the
sizeable reduction in boiling temperatures. Hence, with SPDU, it is possible to
recover the middle distillate further without causing thermal cracking.
6. Oleoresin: Oleoresin is a red colour natural pigment separated from paprika.
It is used as a natural colour additive in food products. The oleoresin is extracted
by hexane. From the extracted product, which contains 96% oleoresins and 4%
hexane (by mass), hexane can be completely removed from oleoresin (down to
10 ppm) by two stage short path distillation (Fig. 8.43).
 Fig. 8.42 Separation of Dimeric Fatty Acids

Fig. 8.43 Manufacturing of Oleoresins

7. Deacidification of Ricebran Oil and Palm Oil: Ricebran oil and palm oil are
preferred to be used as cooking mediums because of their high nutritional
values. But because of considerable high content of free fatty acids,
deacidification by conventional distillation is difficult. With short path
distillation, these oils can be gently deacidified to the concentrations of less than
0.1% without loss of nutritional values.
8. Separation of Monomers from Prepolymers: Polyurethanes are polymers
with different compositions and wide range of characteristics. They are produced
by polyadduct-type polycondensation reaction from toluene diisocyanate and
polyols. In the first step, so-called prepolymers are formed. These prepolymers
still contain a certain amount of non-reacted diisocyanate monomers. It is
necessary to reduce monomer content before selling to end manufacturers
because these monomers are toxic in nature. It can be reduced by short path
distillation.
8.6.3 Economics
It is a general impression that cost of short path distillation unit is quite high.
The reason for this is historical. The first industrial short path distillation plants
were operated with very low throughputs to maintain the very low value of
operating pressure at distillation surface. Vacuum technology was not advanced
at that time and this was the limiting factor for the throughput. Also in initial
stages, there were high overdimensioning of the unit. These all reasons created
the general impression that short path distillation is very expensive operation and
the distillation cost can only be borne by very expensive products. But now,
reality is different.
Improved vacuum technology, considerately higher throughput quantities,
increased availability of units from different manufacturers and low personnel
requirements with automated units have led to a substantial reduction in cost of
distillation. Results have shown that on increasing product throughput from 150
kg/h to 3600 kg/h, total cost of short-path distillation in ₹/kg of product reduces
by more than 85%.
Short-path distillation unit requires only one week shutdown time per year.
Hence, it requires relatively less maintenance.

8.7 REACTIVE AND CATALYTIC

DISTILLATION29,30,35
In most of the chemical plants, one can find the reaction followed by separation.
Usually reaction and separation are performed one by one. First reactants are
allowed to react in the reactor, then in a separate equipment product mixture is
separated from unconverted reactants and/or from byproducts and inerts. It is
rarely seen that both reaction and separation combined and carried out
simultaneously in one equipment.
However, by combining reaction and separation of product in one equipment one
can get two advantages:
1. Removal of the product from the reaction mass at the reaction conditions
increases equilibrium conversion of reactant or increases the extent of
reaction. As per Le Chatelier’s principle, for the reaction in equilibrium, if
any change in any process variable like pressure, temperature or composition
of reaction mass is made, reaction or process is moved in such a direction
which will nullify the effect of the change in the variable. In reactive
distillation, change in the composition of reaction mass is made by
continuously removing the product from the reacting mass. Hence, to nullify
the effect of this change, reaction proceeds in the forward direction and
produces more and more products and so higher conversion is achieved.
2. Separate equipment for the separation is not required. Also separate piping,
instruments, etc., are not required.
There are four commonly used techniques to combine reaction and separation:
(a) Reactive distillation (Reaction + Distillation)
(b) Membrane reaction (Reaction + Membrane separation)
(c) Extractive reaction (Reaction + Extraction)
(d) Reaction with absorption (Reaction + Absorption)
In reactive distillation, reaction and distillation both are carried out
simultaneously in one equipment. If reactive distillation is carried out in
presence of heterogeneous solid catalyst, then it is called catalytic reactive
distillation.
The concept of reactive distillation is not new. This technique was first applied
in 1920 to esterification process using homogeneous liquid phase catalyst. In
manufacturing of ethyl acetate, reaction between ethanol and acetic acid and
separation of crude ethyl acetate from reaction mass by distillation were carried
out simultaneously in one equipment. Ethanol was used as excess reactant.
Overhead vapour from reactive distillation column is a ternary azeotrope having
composition 82.16% ethyl acetate, 8.4% ethanol and 9% water (by mass). On
condensation heteroazeotrope liquid is converted into two layers, ester rich layer
and ethanol-water solution layer. Part of ester rich layer is taken out as top
product (distillate) and remaining is recycled back to column as reflux. Sulphuric
acid was added in feed as catalyst. Waste water (water + sulphuric acid) is
continuously taken out as bottom product. Alcohol-water stream is taken out as
an intermediate product.
In manufacturing of butyl acetate overhead vapour is a ternary heteroazeotrope
of butyl acetate, butanol and water. On condensation, it is converted into two
immiscible liquid layers.
Table 8.49 Ternary Azeotrope Separation at 90°C31
Layer % by mass % by mass

Ester Butanol Water

Upper 75 80.5 13.4 6.1

Lower 25 1.92 1.69 97.39

(Table 12.7 of Ref. 31)

Lower aqueous layer is continuously taken out from decanter for recovery of
butanol (separately) while upper layer or ester rich layer is recycled back to
column as a source of reflux. Bottom product is crude butyl acetate. It is sent for
further purification. Thus in both cases, concept of reactive distillation, invented
in 1920, was applied. But after that for a long period of time very little work was
done in this direction. Almost after 50 years, an important research work was
done by Sermewald in 197132, that applying reactive distillation with
heterogeneous catalyst is far more useful than homogeneous catalyst. Nowadays,
this technique is being successfully applied to some processes like synthesis of
methyl tertiary butyl ether (MTBE), synthesis of ethyl tertiary butyl ether
(ETBE), selective hydrogenation of butadine of C4 stream, synthesis of ethyl
propionate, methyl acetate, etc.
If a solid catayst is used, then the same catalyst can also act as packing material
of packed tower. In that case solid catalyst plays double role to enhance the rate
of chemical reaction as well as to enhance the rate of mass transfer for
distillation. This special type of reactive distillation which uses solid catalyst is
known as catalytic distillation.
Even though reactive distillation has two major advantages over conventional
reactor followed by distillation, it finds its place only in few processes. Reason
is, it also has one major disadvantage.
Disadvantage
In majority of the processes which involve reaction followed by distillation,
optimum conditions for reaction and optimum conditions for distillation are
different. Hence, to carry out both symultaneously in one equipment, one has to
compromise in reaction or in distillation. In most of the cases, the compromise is
in the distillation. In a few cases like in the process of MTBE, optimum
conditions for the reaction and the same for distillation are more or less same.
Hence, one can apply the concept of catalytic distillation in the process of
MTBE. But in many cases, this combination is not economically feasible. For
example, in the manufacturing process of acetic acid by carbonylation of
methanol, first reaction is carried out in presence of homogeneous phase catalyst
at 50 atm pressure and at 180oC temperature with conversion of methanol up to
50%. After the reaction, reactants, promoters and catalyst are separated from
acetic acid by distillation at atmospheric or near atmospheric pressure. Hence, in
this case one cannot apply the concept of reactive distillation because optimum
conditions for the reaction and for the distillation are very much different.
8.7.1 Design Methods for Reaction Distillation System
Complex process design of reactive distillation system has received significant
attention from many research scientists. As such the system involves
simultaneous occurrence of distillation and reaction, high non-ideality of the
mixtures in question and the formation of azeotropes. In developing a model of
the reactive distillation system, problems are experienced due to simultaneous
chemical and physical equilibriums and the types of variables involved in
defining the hypothesis, general composition variables in molar units and the
extents of reactions are the variables that are considered in modelling.
There are a number of methods, reported in literature for modelling reactive
distillation system. A short method, presented by M. Carrera-Rodriguez et
al.33,34 is based on tray-to-tray calculations with a strategy to find reflux ratio
and to locate the feed tray that minimises the total number of stages. The method
is demonstrated for modelling synthesis of MTBE from i-butene and methanol in
presence of an inert; n-butane.
Another method is based on Ponchon-Savarit diagram for design of a binary
reactive distillation column. Developed by J. W. Lee et al.4 it is more reliable as
enthalpies are used in preparation of the diagram. Enthalpy values are reworked
to take care of the heat of reaction. It is assumed that reaction takes place in the
rectifying section and hence the enthalpy ordinate for the rectifying section only
is modified. With the assumption of no reaction in the stripping section, its
enthalpy ordinate is not required to be modified.
Reader is advised to refer original articles for more details.
Kai Sundmacher and Achim Kienle have authored a book35 on reactive
distillation. It is an exhaustive volume on the subject covering design and
practical aspects
8.7.2 Comparison of Reactive Distillation with
Conventional Technology
Conventional technology involves two steps (Fig. 8.44).

Fig. 8.44 Typical Conventional Process

In the conventional process, first step is the reaction step. In this step, reactants
are brought into contact with solid or liquid catalyst at appropriate temperature
and pressure.
In second step, stream from the reactor is sent to separation section, where
desired product is separated from unconverted reactants, catalyst, byproducts
and/or inerts. In most of the cases, this separation is done by distillation.
When concept of reactive distillation (or catalytic distillation, if catalyst is in
solid phase) is applied to such type of processes, flow sheet is changed. Typical
flow sheet of catalytic distillation column is shown in Fig. 8.45.
 Fig. 8.45 Typical Flow Sheet of Catalytic Distillation

In the middle section of catalytic distillation column, reaction and distillation

take place simultaneously while upper and lower sections represent enriching
and stripping sections of distillation or they are utilised only for separation.
The advantages of reactive distillation can be described best by considering its
advantages over reaction step and distillation step of conventional process.
Advantages over reaction step
1. For equilibrium limiting reaction, continuous removal of one of the products
at process conditions increases equilibrium conversion. Thus reactive
 distillation provides high conversion compared to conventional reaction step.
2. For exothermic or endothermic reaction, separate cooling or heating system is
required for reactor in conventional process. In reactive distillation, separate
heating or cooling system is not required for reaction zone. In reactive
distillation, heat of reaction is utilised in distillation either for condensation or
for boiling, thus achieving energy conservation.
Advantages over distillation step
Advantages related to reduction in capital cost
1. In conventional technology, separate equipment is required for distillation.
Reactive distillation eliminates use of major process equipment and
associated pump, piping and instrumentation.
2. Heat of reaction reduces either condensation duty or reboiling duty. Hence, it
either reduces condenser size or reboiler size.
3. In reactive distillation, very less amount of uncoverted reactants are to be
separated from products compared to distillation in conventional technology.
Hence, reactive distillation requires lesser number of stages for same value of
reflux or less amount of reflux compared to distillation of conventional
technology for the same value of reflux ratio or it requires lesser size of
distillation column compared to the same required for the distillation in
conventional technology for the given production rate. Less amount of reflux
also decreases sizes of condenser and reboiler.
Advantages related to reduction in operating cost
1. Heat of reaction reduces either condensation duty or reboiling heat duty.
Hence, it reduces either load of cooling system or the same of heating system.
Consequently, it reduces operating cost.
2. Operating cost required for recycling of unreacted reactants is reduced
considerably.
3. In reactive distillation, amount of distillate is reduced considerably. Hence, for
the same reflux ratio, lesser amount of reflux is required in reactive
distillation compared to distillation column of conventional technology. It
decreases condenser heat duty, reboiler heat duty and power required for
pumping the reflux.
8.7.3 Different Ways by Which Reactive Distillation
Can be Applied
There are three different ways by which Reactive Distillation can be applied:
1. Reactive distillation alone
2. Reactive distillation plus non-reactive distillation
3. Reaction alone plus reactive distillation plus non-reactive distillation
1. Reactive Distillation Alone: In this case, throughout the whole column,
reaction and distillation take place simultaneously. At one of the ends, purified
product is obtained. This is possible only for the systems for which pure product
point lies within reaction space.
2. Reactive Distillation plus Non-reactive Distillation: There are many cases
for which pure product point does not lie within reaction space. For such cases,
non-reactive distillation section (either rectifying or stripping or both) is also
required to achieve pure product and pure reactants.
3. Reaction alone plus Reactive Distillation Plus Non-reactive Distillation
(Optional): If the reaction space required is considerably greater than distillation
space, then this method is applied.
Sometimes it is not recommended to shift entire reaction space to reactive
distillation unit. The reason may be one of the following:
(a) If the optimum reaction conditions and optimum distillation conditions are
different, then it is better to improve only equilibrium conversion in reactive
distillation, while major part of reaction (up to equilibrium conversion) can be
carried out separately in another reactor.
(b) If the existing plant of conventional technology have more number of
reactors in series than one may shift only last or a few of the last reactors to
reactive distillation unit.
8.7.4 Various Contact Devices Used for Catalytic
Distillation
In catalytic distillation, contact devices are required to improve mass transfer
rate and to improve catalyst contact. Various types of contact devices used for
catalytic distillation are as follows:
1. Tray column in which bed of catalyst is placed either in downcomers or above
the trays.
2. Packed tower using random packing elements which contain the catalyst (with
special coating) or which may be made of the catalyst.
3. Packed tower using ordered packing elements which contain catalyst particles.
In orderly packing element, catalyst particles are sandwiched between two
supporting materials. Two types of ordered pickings are used:
(a) Bale packing
(b) Structured packing
(a) Bale Packing: In bale packing (Fig. 8.46), catalyst particles are enclosed
within several pockets in a fibre glass cloth belt.
The pockets are normally 2.5 to 5 cm wide with 0.625 to 1.25 cm web between
them. Open ends of pockets are sewn closed. Belts are rolled up with alternating
layers of stainless steel mesh which increases structural strength of cloth belt.
Bales constructed for commercial use are roughly cylindrical shape with 20 to 35
cm diameter and 50 cm height. Several bales then packed tightly into a column
such that pockets remain vertical. Flow channels facilitate the contact between
vapour and liquid. Ideally, contact between liquid reactants and catalyst is taking
place in pockets and vapour-liquid contact takes place in flow channel. HETP as
well as pressure drop relations for bale packings could be obtained from the
supplier.
(b) Structured packing: KATMAX structural packing is developed by Koch
Engineering Company, USA. From outside this packing looks like a brick with
corrugated surface. In such type of packing, solid catalyst is held in a screen
envelope.
 Fig. 8.46 Bale Packing

The catalyst is sandwiched between the layers of screen. It is easily accessed by

the liquid. Between two layers of catalyst (surrounded by screen), gaps are
provided, known as flow channels which facilitate the flow of vapour. Inside the
flow channel vapour makes the contact with liquid film.
8.7.5 Industrial Applications of Reactive Distillation
1. Manufacturing of ethyl acetate from ethanol
2. Manufacturing of methyl acetate from methanol
3. Manufacturing of butyl acetate from butanol
4. Manufacturing of MTBE (methyl tertiary butyl ether) and ETBE (ethyl
 tertiary butyl ether)
5. Selective hydrogenation of butadine of C4 stream
6. Selective hydrogenation of pentadine of C5 stream
7. Selective hydrogenation of hexadine of C6 stream
8. Manufacturing of benzyl chloride from toluene
9. In benzene separation of gasoline
10. Manufacture of ethyl propionate from ethanol
11. Manufacture of amino resins (refer to Exercise 8.16)
12. Manufacture of monochloro acetic acid
Among these applications, manufacturing of ETBE by catalytic distillation is
discussed in this chapter.
8.7.6 Manufacturing of ETBE
Ethyl tertiary butyl ether is used as octane number improver like MTBE.
Main relation
60 – 90°C
(CH3)2C = CH2 + C2H5OH (CH3)3COC2H5
9 atm
i–Butylene Ethanol ETBE
Side reactions
(1) (CH3)2C = CH2 + (CH3)2C = CH2 (CH3)2C = (CH2)2
i-Butylene i-Butylene Diisobutylene
(2) (CH3)2C = CH2 + H2O (CH3)3COH
i–Butylene Water i–Butyl alcohol
Conventional Process Refer to Fig. 8.47. Isobutylene rich stream is water
washed in wet scrubber to remove trace contaminants that can deactivate the
catalyst. Then this stream is mixed with rectified spirit and mixture is sent to
isothermal, shell and tube type plug flow reactor. Ethanol is used as excess
reactant. Second reactor is operated adiabatically. Product from second reactor is
sent to distillation column. In first distillation column, the mixture of ethanol,
water and isobutylene is separated from ETBE. Ethanol cannot be separated
from the distillate of first column by common distillation because ethanol forms
azeotrope with isobutylene as well as with water.
Further addition of water as solvent separates ethanol and water mixture from
isobutylene in liquid-liquid extractor. Isobutylene rich stream is coming out from
the top of extractor as light product or as raffinate. Ethanol-water stream or
extract stream leaves from the bottom of the extractor. From the extract phase,
azeotrope (95.6% ethanol and 4.4% water by mass) is separated by distillation
and it is recycled back to the reactor.
Bottom product of first distillation column is ETBE-ethanol mixture from which
ethanol is separated from ETBE by distillation.
Reactive Distillation Process36 In reactive (catalytic) distillation technology
(Fig. 8.48) only equilibrium conversion is increased in reactive distillation unit.
Up to 80% conversion of isobutylene is achieved in a (conventional) separate
isothermal shell and tube heat exchanger-type plug flow reactor. Hence, feed to
reactive distillation column is rich in ETBE but still contains ethanol and
isobutylene. Instead of rectified spirit, absolute alcohol is used as raw material
with isobutylene.
Overhead product from reactive distillation column is non-reactive hydrocarbons
(n-butane, n-butylene, etc.), isobutylene and ethanol, respectively.
Fig. 8.47 Conventional Process of ETBE
 Fig. 8.48 Reactive Distillation Process of ETBE36

Reactive Distillation column has total ten packing sections. First and second
packing sections use simple structured packing, act as rectification or enriching
section. Third, fourth and fifth packing sections use orderly packing and act as
reactive distillation zones. Sixth to tenth packing sections used simple structural
packing which act as stripping section.
Typical data of this process are as follows.
1. Feed composition: 29.1% ETBE, 9.1% ethanol,
73% isobutylene, 54.5% n–butylene (by mole)
2. Isobutylene conversion: 97.4%
3. Operating pressure = 900 kPa a
4. Reflux ratio = 5.0
5. Bottom composition = 96.5% ETBE, 2.8% ethanol
0.7% butylene, 0.04% diisobutylene (by mole)
6. Distillate composition = 96.5% n-butylene, 1.7% isobutylene,
0.7% ethanol (by mole)
8.7.7 Effect of Various Parameters on Reactive
Distillation of ETBE
1. Effect of Feed Composition: Increase in isobutylene percentage in feed
(isobutylene rich stream) results in
(a) Decrease in energy cost per kg of ETBE produced. As lesser energy is
“wasted” in heating and cooling of inert compounds.
(b) Decrease in size and fixed cost of reactive distillation unit and associated
equipments for the same production rate of ETBE.
(c) Increase in the reactant concentrations in reaction zone, increases reaction
rate as well as equilibrium conversion.

f ( Xe ) =   (8.124)

In this reaction, Δn = change in moles during reaction = 2 – 1 = 1. Hence,

decrease in total number of moles of reaction system increases the value of
equilibrium conversion Xe.
(d) Increase the temperature of reaction zone due to decrease in stabilising effect
of inert components. Reaction is exothermic.
2. Effects of Excess Ethanol: Excess of ethanol is required to suppress rate of
side reactions and to increase yield of ETBE. But, increase in excess of ethanol
decreases the purity of ETBE as a bottom product of reactive distillation. It has
been found that 4 to 7% excess ethanol is sufficient to produce high yield of
ETBE without affecting the purity of ETBE as a bottom product from the
reactive distillation unit.
3. Effect of Column Pressure: In conventional distillation, operating pressure is
decided based on use of cooling water or cheaper cooling medium in overhead
condenser. Sometimes, to avoid the thermal cracking of product or to alter the
vapour-liquid equilibrium or to facilitate the use of cheaper heating medium,
etc., conventional distillation is carried out in vacuum.
However, in reactive distillation, choice of operating pressure is more
complicated. In reactive distillation pressure affects equilibrium conversion of
reaction also.
In ETBE, synthesis moles are decreasing during the reaction, hence higher
pressure is favourable as it increases equilibrium conversion. But, this is not
always true, as increase in pressure increases temperature of reaction zone of
reactive distillation unit. Since ETBE synthesis is exothermic reaction, increase
in temperature decreases equilibrium conversion. It is found that best conversion
is achieved at 400 to 500 kPa a. But considering distillation aspect, operating
pressure is kept 700 to 900 kPa a.
4. Effect of Reactive Distillation Stages: Increase in length of reactor zone or
increase in number of stages of reactive distillation zone improves the
conversion, but at the same time decomposition of product starts in the lower
stages, if the number of stages increased above the optimum number of stages.
5. Effect of Separation Stages: Ideally the rectification or enriching section of
reactive distillation column for ETBE synthesis should (a) remove light inerts
from reaction zone, (b) prevent loss of ether or alcohol in distillate, and (c)
recycle unreacted reactants to the reaction zone. In actual practice, this is almost
impossible to achieve as favourable conditions for the reaction cannot be
changed. However, the loss of ETBE in distillate can be minimised without
rejecting isobutylene from the column. More enriching stages are required to
prevent loss of ethanol with distillate.
Ideally, stripping section should perform the following functions:
(a) Remove ETBE from the reaction zone to maintain favourable reaction
conditions.
(b) Purify the ether (ETBE) product.
(c) Prevent loss of reactive isobutylene with ether (ETBE) product.
(d) Minimise ethanol loss with the ether product.
These functions are easily achievable at temperature and pressure of reaction
conditions.
6. Effect of Reflux Ratio: Effect of increasing reflux ratio are as follows:
(a) The concentration of reactants in the distillate is reduced.
(b) The reaction zone temperature reduces which is favourable for exothermic
 reaction to increase equilibrium conversion.
(c) The concentration of ether in the reaction zone reduces. It reduces catalytic
cracking of ether (ETBE).
All these three effects increase the conversion. However, higher reflux ratio
increases tower diameter, condenser and reboiler sizes and energy requirements.
7. Effects of Feed Point: In ETBE synthesis, optimum feed point for reactive
distillation column is just below the reactive distillation zone. In reaction zone,
ETBE can decompose or it can react by cracking reaction in presence of catalyst.
So to avoid the decomposition of ETBE in the feed (i.e., ETBE rich stream), it is
fed just at the bottom of reaction zone. Feeding too far below the reaction zone
reduces the stripping potential and increases energy requirement for the same
height of stripping section.
8. Effects of Reboiler Duty: At some optimum value of reboiler duty, higher
conversion (more than 95%) and higher purity (~90%) is achieved. If reboiler
duty is too high or too low, conversion and product purity both are reduced.
Manufacture of ETBE was discussed as a case study of reactive distillation so
that a process engineer can learn to analyse a system before actual process
design is taken up.
8.7.8 Manufacture of Ethyl Propionate
Ethyl propionate can be synthesised by esterification of ethyl alcohol with
propionic acid in a reaction distillation unit using cation exchange resin as
catalyst. Bhanvase et al37 have given pilot plant results for scale up and
developing a process flow sheet for commercial scale plant.

8.8 AZEOTROPES AND SEPARATION THEREOF

In the previous sections, distillation of liquid mixtures was dealt which are
amicable to ordinary fractional distillation. In these cases, VLE plot or an
equilibrium curve of a binary is on one (upper) side of the diagonal line.
However, when the equilibrium curve crosses the diagonal line, an azeotrope
occurs. Thus at a given pressure, a mixture of specific composition boils at a
constant temperature which produces vapour of the same composition. When
this happens, the mixture cannot be separated and it is known as azeotropic
mixture. When the vapour of the azeotropic mixture is condensed, liquid mixture
so obtained decides the type of azeotrope. If the condensed liquid phase is
homogeneous, it is called a homogeneous azeotrope. Ethanol-water forms
homogeneous azeotrope at 101.325 kPa and 78.1°C having 95.6% ethanol (by
mass). Figures 8.49 (a) and (b) and Figs. 8.50(a) and (b) refer to homogeneous
azeotropes.
A heterogeneous azeotrope is one for which the overall liquid mixture forms two
liquid phases. Mixture of n-butanol–water forms an azeotrope at 101.325 kPa
and 92.2°C having 57.5% (by mass) or 24.8% (by mole) n-butanol in the overall
liquid mixture. It separates in two layers. At 40°C, top layer contains 78.51% n-
butanol while the bottom layer contains 6.60% n-butanol (by mass). Refluxing a
specific layer provides a method to separate the components of the mixture.
Figures 8.51 (a) and (b) refer to a heterogeneous azeotrope.
The temperature at which an azeotrope is distilled is important. A minimum
boiling azeotrope is one for which the boiling temperature is less than that of
pure components. Figures 8.49(a) and (b) and Figs. 8.51(a) and (b) are t–x–y and
y–x diagrams of minimum boiling azeotropes. For example, at 101.325 kPa, n-
butanol and water boil at 117.7°C and 100°C, respectively but their azeotrope
boils at 92.2°C. In case of maximum boiling azeotrope, the boiling point of the
azeotrope is higher than that of the pure component. Chloroform and acetone
boil at 61.2°C and 56.4°C, respectively at 101.325 kPa. The mixture of these two
compounds forms a maximum boiling azeotrope at 64.5°C, containing 65.5%
(by mole) chloroform. Figures 8.50(a) and (b) are t-x-y and y-x diagrams of the
system, comprising chloroform and acetone. Maximum boiling azeotropes are
usually homogeneous. About 90% of the known azeotropes are of the minimum
variety. Heterogeneous azeotropes are minimum boiling azeotropes. However,
many homogeneous azeotropes are also minimum boiling azeotropes. Minimum
boiling azeotropes have positive deviation from ideality. In other words, their
partial pressures in vapour phases are greater than the same given by Raoult’s
law. Maximum boiling azeotropes have negative deviation from ideality.
Fig. 8.49 Homogeneous Minimum Boiling Azeotrope of Toluene-n-Butanol System38 (Reproduced with the
permission of Chemical Engineering by Access Intelligence, USA)

Azeotropic data are well tabulated in literature. Perry’s Chemical Engineers’

Handbook8 tabulate data for binary and ternary azeotropes of common interest.
Extensive 3-volume compilation by J. Gmehling (Ref. 9) cover pratically all
known azeotropes.

Fig. 8.50 Homogeneous Maximum Boiling Azeotrope of Chloroform-Acetone System38 (Reproduced with
the permission of Chemical Engineering by Access Intelligence, USA)

Property of some organic compounds to form azeotropes with water is

advantageously used in many chemical reactions. Manufacture of butylated
amino resins involve etherification reactions. Water is a product of reaction. One
of the reactants is n–butanol which forms a heterogeneous azeotrope with water.
While a batch of reaction is in progress, water is removed from the mass as an
azeotrope. The vapour of azeotropic composition is condensed which separates
into two liquid layers. Top n-butanol rich layer is recycled to the reactor and
bottom aqueous layer is removed. This way one of the products of reaction is
removed from the reaction mass, thereby reaction proceeds in forward direction.
 Fig. 8.51 Heterogeneous Minimum Boiling Azeotrope of n-Butanol-Water System38 (Reproduced with the
permission of Chemical Engineering by Access Intelligence, USA)

In manufacture of alkyd resins also, water is a product of reaction. Toluene or

xylene is added to the reaction mass and water is removed from the mass in form
of a heterogeneous azeotrope. On condensation, two layers separate. Water is
sparingly soluble in the hydrocarbon. Top hydrocarbon layer is recycled to the
reactor.
In all esterification reactions, water is a byproduct. It is a common practice to
remove water in the form of an azeotrope from top in a specially designed
distillation column.
Removal of water in the form of an azeotrope has also expedited other
operations, such as nitration of benzene and toluene. In this case, water formed
in the nitration of the hydrocarbon is removed continuously, utilising
hydrocarbon itself as an entrainer. Contrary to semi-continuous operation, mixed
acid, consisting of nitric acid and sulphuric acid, is used for nitration in a batch
reactor. This batch process suffers from a drawback of disposal of spent acid or
recovery of acids from the spent acid is a problem. Bottom still acts as a reaction
vessel in which nitric acid and excess hydrocarbon are initially charged. Nitrated
hydrocarbon has higher boiling point and thus remains in the still. Azeotrope of
hydrocarbon and water goes to the bottom of the dehydrating column. Optimum
temperatures and concentrations can readily be maintained in the system with
simple controls. Parameters, normally required to be controlled are feed rates of
nitric acid and hydrocarbon, rate of supply of heat to respective heating units and
rate of withdrawal of water at the top of the dehydrating column. Product is thus
formed in single pass with almost 100% yield on both; nitric acid and
hydrocarbon.
Similar azeotropic process is continuous production of acetamide. Continuous
streams of ammonia and acetic acid are fed to a dehydrating column.
Ammonium acetate, so formed is continuously dehydrated in azeotropic
distillation system to give water and acetamide. Water, formed in both the
reactions, goes overhead with the help of an entrainer. Acetamide with excess
acetic acid flow out from the bottom of the column. This mixture is sent to
second distillation column from which pure acetamide is withdrawn from the
bottom. Acetic acid vapour from the column is recycled back to reaction-cum-
azeotropic distillation column.
In general, following are the methods, used for separating homogeneous
azeotropic mixtures:
1. As mentioned earlier, azeotrope is dependent on the operating pressure in
most cases. If pressure is altered, it may be possible to avoid formation of the
azeotrope thereby fractional distillation can be carried out. Table 8.50 gives
azeotropic compositions of ethanol-water mixtures at different operating
pressures.
 Table 8.50 Effect of Pressures on Ethanol-Water Azeotropes39
Absolute Pressure Concentration of Ethanol Temperature °C

Torr kPa mole fraction mass fraction

80 10.666 1.000 1.000 30.30

95 12.666 0.987 0.995 33.35
130 17.332 0.967 0.987 39.30
198 26.398 0.934 0.973 47.63
300 39.997 0.925 0.969 56.50

405 53.996 0.909 0.962 63.04

760 101.325 0.895 0.956 78.15
1075 143.322 0.889 0.953 87.10
1451 193.451 0.887 0.953 96.40
2585 344.638 0.882 0.950 112.60
5171 689.410 0.874 0.947 136.70

It can be seen from Table 8.50 that at pressure lower than 80 Torr, there does not
exists an azeotrope. Similarly, for methanol-benzene mixture, mole fractions
of methanol in azeotropic mixture at 200 Torr and at 760 Torr absolute
pressures are 0.557 and 0.613, respectively.
A few azeotropes, however, do not change composition with the pressure. For
example, n-hexane forms an azeotrope with ethanol38, containing 21%
ethanol (by mass) which boils at 58.7°C at 101.325 kPa. However, the
composition of this azeotropic mixture is unchanged over the pressure range
of 20 Torr to 10 000 Torr.
2. A popular method is to add specially chosen chemical to shift the azeotropic
point to a more favourable position. When this chemical (also known as
entrainer) appears in appreciable amounts at the top of the column, the
operation is called azeotropic distillation. When the additional component
(called solvent) appears mostly at the bottom of the column, the operation is
called extractive distillation.
3. In distillation using ionic salt, the salt dissociates in the liquid mixture and
alters the relative volatilities sufficiently that the separation becomes possible
like fractional distillation.
4. Pressure-swing distillation is yet another method where a series of columns,
operating at different pressures, are used to separate binary azeotropes which
change appreciably in composition over a moderate pressure range or where a
separating agent which forms a pressure-sensitive azeotrope is added.
5. Use other techniques like pervaporation or adsorption for separating a
component (such as water) from the azeotrope. These techniques are
 discussed in Sec. 8.9.3.
6. Reactive distillation can be considered where the separating agent reacts
preferentially and reversibly with one of the azeotropic constituents. The
reaction product is then distilled from the non-reacting components. Dilute
alcohol solutions or dilute aqueous acetic acid solution can be treated by this
method to form esters in reactive distillation units.
7. Fixed adsorptive distillation method is yet another technique, found useful in
separation of an azeotropic mixture. Mujiburohman et al40 have developed a
system consisting of two distillation columns with an interbed filled with an
adsorbent, is installed. In this way the azeotropic point is avoided and purer
products are obtained from two columns.
8.8.1 Azeotropic Distillation
Heterogeneous azeotropic distillation is a widely practised process for
dehydration of number of compounds such as acetic acid, chloroform, ethanol,
acetonitrile, etc. The technique involves addition of a third component in a
homogeneous azeotrope, called an entrainer to form a minimum boiling
heteroazeotrope with at least one of the components of the feed carrying water
overhead and leaves near dry (pure) product in the bottom. The overhead
vapours are condensed to two liquid phases; entrainer rich phase is refluxed
while the aqueous phase is decanted.
Dehydration of ethanol-water mixture is carried out with the help of benzene as
an entrainer. Ethanol and water forms azeotrope having 95.6% ethanol (by mass)
at 78.15°C and 101.325 kPa. When benzene is added to the azeotrope (about
0.427 kg/kg azeotrope) it comes as ternary heterogeneous azeotrope at the top
containing 74.1% benzene and 18.5% ethanol (by mass). Bottom product from
the column is pure dry ethanol. Top azeotrope at 65°C is condensed which forms
two layers. Top layer contains 90.9% benzene by mass (and practically no
water). It is refluxed to the column. Bottom aqueous layer contains 53% ethanol
and 11% benzene (by mass). Since this layer contains appreciable content of
benzene, it can be distilled in another column to recover it in the form of ternary
azeotrope. Water is removed from the bottom of the third column. Thus three
columns are normally required for this distillation.
Process design of an azeotropic distillation system for separation of ethanol-
water using trichloroethylene is given by Coates38 and using benzene is given by
Norman41.
Another entrainer for ehtanol dehydration is n-pentane which is eco-friendly.
The tower is operated at near 3.45 bar a. Total condenser operates at 3 bar a and
a ternary azeotrope, containing 82.64% n-pentane and 12.33% ethanol (by mole)
is formed. Condensate separates in two layers, top layer contains 87.9% n-
pentane and 11.6% ethanol (by mole) and is refluxed to the column. Bottom
aqueous layer contains 23.8% ethanol (by mole) and practically no n-pentane.
This layer is distilled in second column to recover ethanol as azeotrope with
water as top product at atmospheric pressure. In this case only two columns are
required.
Another example of separation of homogeneous azeotrope is that of iso-
propanol–water. It forms azeotrope at 80.4°C and 101.325 kPa. For their
separation, cyclohexane or benzene is used as entrainer. Cho and Jeon42 have
given process design of heterogeneous azeotropic distillation system for
dehydration of iso-propyl alcohol using benzene as the entrainer.
When the binary azeotrope is hetergeneous, separation of the components is
possible without addition of an entrainer. Design of n-butanol–water distillation
system is given by McKetta and Cunningham43 in detail.
8.8.2 Extractive Distillation
Extractive distillation can be defined as distillation in the presence of a miscible,
high boiling relatively non-volatile component (called the solvent) that forms no
azeotropes with other components of the mixture. Solvent is continuously added
near the top of the column so that an appreciable amount is present in the liquid
phase throughout the column. The component having the greater volatility (not
necessarily the lowest boiling point component) is taken overhead as a relatively
pure distillate. The other component leaves with the solvent via the column
bottoms. The solvent is then separated from the bottoms in second distillation
column. Selection of an extractive distillation solvent is the most important step
in developing a successful extractive distillation sequence. Important selection
criteria are given below.
1. It should enhance the relative volatility of the key component significantly.
2. Required solvent quantity to be added to the azeotropic mixture should not be
excessive.
3. It should be soluble in the feed components.
4. It should be easily separable from the bottom product.
5. It should be relatively inexpensive and easily available.
6. It should be stable at the operating conditions of both the columns.
7. It should be non-reactive with the feed components.
8. It should have low latent heat of vaporisation.
9. It should be non-corrosive and non-toxic.
10. It should not form immiscible liquid mixtures at any point in the column.
Perrys’ Chemical Engineers’ Handbook8 serves as a good reference for the
solvent selection.
Cyclohexane is manufactured from benzene. The mixture from the reactor
comes out with the desired product (cyclohexane) and also a significant amount
of unreacted benzene which is to be recycled back to the reactor for economic
reasons. Product mixture contains nearly 45% cyclohexane and balance benzene
(by mole). It is desired to operate the distillation column at 150 kPa a as dictated
by the process flow sheet. At this pressure, cyclohexane and benzene boil at
94.3°C and 93.5°C, respectively. Binary homogeneous azeotrope at this pressure
is formed at 91°C, containing 45.5% cyclohexane (by mole) which is nearly the
same composition of reactor exit product mix. In this case propylene glycol is
selected as a solvent as benzene is highly soluble in the solvent. Further the
solvent has high boiling point (200.4°C) at 150 kPa a and no new azeotropes are
formed. Distillate from the column is 99.3% pure cyclohexane. Bottoms from
the first column is fed to second column from which distillate is 85% pure
benzene which is recycled back to the reactor. Bottoms from the second column
is nearly pure solvent which is recycled to the first column.
Apart from organic solvents, inorganic salts such as potassium and sodium
acetates, calcium chloride, calcium nitrate, caustic soda, etc., can also be used to
break the azeotropes.
Another well-known example of extractive distillation is dehydration of nitric
acid-water azeotrope (containing 64.1% HNO3 by mass at 101.325 kPa a and
122°C). In this case sulphuric acid is used as solvent. Refer to Example 8.21.
Apart from above examples, extractive distillation finds many applications in the
petroleum refinery. Pyrolysis gasoline byproduct from naphtha steam cracking
contains a wealth of aromatics. Separation of aromatics, such as xylenes, is
carried out by extractive distillation. Similarly, extractive distillation can be used
to selectively extract the sulphur species from cracked gasoline to very low
levels without losing its octane value.
Advantages of extractive distillation over azeotropic distillation are:
1. Choice of third component to be added is greater as selection of third
component does not depend upon an incident of azeotrope formation.
2. Energy requirement is less as the third component or solvent is not completely
vaporised in principle column or in auxiliary column. Only small fraction of
 solvent is vaporized in both columns as the same is obtained as bottom
products. While entire entrainer is to be vaporized in azeotropic distillation in
both columns as the same is coming out as one of the component of overhead
vapour from both the columns. However, this advantage of extractive
distillation is lost if volatilised component or the component to be taken out
as top product along with entrainer which is the minor component of feed.
Hence, in dehydration of ethanol from the azeotropic mixture (containing
4.4% by mass water), energy consumption of azeotropic distillation using n-
pentane as entrainer is less as compared to the same by extractive distillation
using ethylene glycol as solvent.
8.8.3 Pressure Swing Distillation
Although known from good old days, only now pressure swing distillation
technique is gaining popularity for separation of a binary azeotrope over
heterogeneous azeotropic distillation and extractive distillation. The chief reason
for not finding many industrial applications till date was the lack of experimental
VLE data for binary systems at different pressures. Two main advantages of
pressure swing distillation are; energy saving by heat integration of the system
and no need to add an entrainer or a solvent for the separation.
The technique is based on the principle that the composition of almost all
azeotropes varies with the operating pressure. In industrially important water-
solvent mixtures, the water content of the azeotrope increases with increasing
pressure. This variation provides a technique for separation of the components.
A shift in composition of azeotropic mixture with respect to pressure can be well
understood by studying Figs. 8.52(a) and (b).
Acetonitrile—ACN (also called methyl cyanide; formula: C2H3N) forms a
minimum boiling azeotrope with water at 1 atm and 77°C containing 83.5% by
mass or 69.0% by mole ACN. This azeotrope composition shifts to 60.0% by
mole ACN at 3.02 bar a which boils at 112.5°C. Using these data a two-column
system can be designed to operate at two different pressures to separate ACN
and water from the mixture. Feed with lower concentration than 69.0 mole %
ACN can be fed to the first column operating at 1 atm. Distillate from the
column with near azeotropic composition is the feed to the second column
operating at 3.02 bar a. Bottom products from the first and second column are
water and ACN, respectively. Distillate from the second column is recycled to
the first column along with the fresh feed.
An important point to note is that in heterogeneous azeotropic distillation,
bottom product from the first column is the solvent (low boiling component)
while in pressure swing distillation, bottom product is water (high boiling
component).
In pressure swing distillation since both the columns operate at different
pressures, their heat exchangers can be readily thermally integrated. This will
result in substantial savings in energy for distillation. However, such an
integration results in complexity of the system which requires adequate process
control and instrumentation. Also hazards of handling flammable (and toxic)
solvents at high pressure must be given due considerations while designing such
a system.
 Fig. 8.52 Homogeneous Minimum Boiling Azeotropes Acetonitrile-Water System44 (Reproduced with the
Permission of Dr. Repke)

Both batch and continuous systems are developed. Jen-Uwe Repke et al.45,46
have developed mathematical models and process control systems for ACN-
water mixture.
Knapp and Doherty47 have listed more than 35 pressure sensitive binaries which
can be separated by pressure swing distillation technique.
Ethanol-water is also a pressure sensitive azeotrope (Ref. Table 8.50) but its
separation by pressure swing distillation method is considered an uneconomical
proposition. For separation of this azeotrope, Knapp and Doherty47, 48 have
suggested to add small amount of acetone in the dilute alcohol stream from a
fermentor. The entrainer does not form a ternary azeotrope at atmospheric
pressure but forms a ternary azeotrope at high pressure (say at 10 atm). On this
basis a scheme is presented to operate first column at 1 atm and second column
at 10 atm. Resulting pressure sequence yields dry ethanol from the bottom of the
second (HP) column. Also thermal integration of both the columns result in
nearly 53% energy saving over the conventional heterogeneous azeotropic
distillation sequence.
Exercise 8.15 is yet another example of separation of tetrahydrofuran (THF)
from aqueous THF solution using pressure swing distillation technique.
Isopropyl alcohol (IPA) forms a minimum boiling azeotrope
with water at 80.4°C and 101.325 kPa, having a composition of 31.47 mole %
water. Vapour pressure of IPA can be calculated using Antoine equation14.
log pv = 6.8651 –

where, pv = Vapour pressure, kPa

T = Saturation temperature, K
(a) A feed consisting of 50 mass % IPA under saturated liquid condition is to be
distilled to give 67.5 mole % IPA as distillate (near azeotrope) and 0.1 mole
% (0.0033 mass %) IPA in the bottoms. Using McCabe–Thiele technique and
for R = 3, find the number of theoretical stages, required for the separation at
1 atmosphere and locate the feed introduction stage.
(b) Make tray-to-tray calculations for the desired separation using van Laar
constants.
Solution
(a) Azeotrope contains 0.3147 mole fraction (0.121 mass fraction) water. To find
the non-ideal behaviour of the IPA-water system, use of van Laar equation can
be made. Assuming vapour fugacity (i.e., vapour escaping tendency) to be unity
at low pressure.
Modified Raoult’s law:
p.yi = x iγ i.pvi  (8.3)
where, p = Total pressure of the system = 1 atm = 760 Torr
yi = Mole fraction of ith component in vapour
xi = Mole fraction of ith component in liquid
γ i = Activity coefficient of ith component
pvi = Vapour pressure of the ith component at the boiling point, Torr
If x1 and x2 are mole fractions of the two components in a binary liquid mixture,

ln γ 1 = A   (8.125)

ln γ 2 = B   (8.126)

Equations   (8.125) and   (8.126) are called van Laar equations and A and
B are called van Laar constants. Let 1 refer to IPA, the more volatile component
and, 2 refer to water.
Boiling point of the azeotrope, t = 80.4°C or T = 353.55 K.
log pv1 = 6.8651 – = 1. 973 05

pv1 = 93.984 kPa (vapour pressure of IPA at 80.4°C) ≡ 704.94 Torr

From steam tables, at t = 80.4°C, pv2 = 48.14 kPa = 361.08 Torr
At the azeotropic point,
γ1 = 760/704.94 = 1.0781
γ2 = 760/361.08 = 2.1048

ln (1.0781) = A

ln (2.1048) = B
Solving the equations by Mathcad,
A = 2.311 87 and B = 1.107 77
These constants are themselves functions of the liquid compositions and
temperature but their variations can be ignored for all practical purposes (for low
heats of mixing) at a given total pressure and varying temperature conditions.
Using these constants, van Laar equations can be used to develop the vapour-
liquid equilibrium curve.
Consider x1 = 0.6 (of IPA) and x2 = 0.4 (of water) and calculate activity
coefficients.
ln γ1 = 0.135 51 or γ1 = 1.145 12
ln γ2 = 0.636 31 or γ2 = 1.889 49
Assume boiling point of the mixture to be 81°C (or 354.15 K),
pv1 = 96.292 kPa = 722.25 Torr
pv2 = 49.311 kPa = 369.86 Torr
p = Σxi · γi · pvi
= (0.6 × 722.25 × 1.145 12) + (0.4 × 369.86 × 1.889 49)
= 496.24 + 279.54 = 775.78 Torr ≠ 760 Torr
Since total pressure does not equal to 760 Torr, assume new temperature equal to
80.5°C (353.65 K).
pv1 = 94.365 kPa = 707.8 Torr
pv2 = 48.336 kPa = 362.55 Torr
p = (0.6 × 707.8 × 1.145 12) + (0.4 × 362.55 × 1.889 49)
= 486.31 + 274.0 = 760.31
= 760 Torr
Mole fraction of IPA in vapour (corresponding to x1), y1 = x1 · γ1 · pv1/p
= 0.6 × 707.8 × 1.145 12/760
= 0.64
In this manner, x-y data can be generated for the entire range which is tabulated
in Table 8.51.
Table 8.51 Vapour-Liquid Equilibrium of IPA-Water at 1 atm (Based on van Laar Constants)
Mole fraction of IPA
Boiling point (t), °C
in Liquid, x in Vapour, y

0 0 100.0
0.02 0.2495 92.65
 0.05 0.3993 87.4
0.075 0.4561 85.15
0.1 0.489 83.85
0.2 0.5371 82.0
0.3 0.5545 81.5
0.4 0.573 81.3
0.5 0.6 80.8
0.6 0.64 80.5
0.6853 0.6853 80.4 (azeotrope)
0.8 0.7682 80.6
0.9 0.87 81.15
0.95 0.93 81.65
1.0 1.0 82.5

Figure 8.53 is the plot of the vapour-liquid equilibrium of IPA-water system.

Feed (zF) is marked at x1 = 0.2306 (equivalent to 0.5 mass fraction) of IPA.
Desired purities of distillate and bottoms are marked at xD = 0.675 and xW =
0.001. q-line is vertical as the feed is saturated liquid.

= 0.3537

R m = 0.9084
In case of azeotropic distillation, other methods are available in literature41 for
calculation of minimum reflux ratio which are more reliable. Also actual reflux
ratio is kept much higher than the minimum.
 Fig. 8.53 Azeotropic Distillation of Aqueous Isopropyl Alcohol

Let R = 3 (3.3 times Rm)

= = 0.1688

For R = 3, from Fig. 8.53, total 10 theoretical stages are required for the
separation and feed will be introduced at third stage from bottom.
(b) Tray-to tray Calculations:
Azeotropic distillation column design of IPA-water system can be carried out by
numerical method, i.e., tray-to-tray calculations. Average latent heats of
vaporisation of IPA and water are 40 093 and 41 081 kJ/kmol, respectively. Also
column is assumed to operate at atmospheric pressure from top to bottom.
Activity coefficients for each tray liquid mixture will be calculated using same
van Laar constants, used for developing Table 8.51.
Basis: 100 kmol/h of feed consisting of 0.2306 mole fraction IPA.
With distillate, containing 0.675 mole fraction IPA and bottoms, containing
0.001 mole fraction IPA, material balance yields
D = 34.065 kmol/h and
W = 65.935 kmol/h
Reflux R = L = 3D = 3 × 65.935 = 102.195 kmol/h
Calculations relate to stripping section.
Liquid flow in column L– = L + F · q = L + F (q = 1)
= 102.195 + 100
= 202.195 kmol/h
Vapour from the bottommost tray (No. 1),
V–1 = L– – W
= 202.195 – 65.935
= 136.26 kmol/h
For x1 = 0.001 (IPA) and x2 = 0.999 (water)
γ2 = 9.996 57 and γ2 = 1.000 00
At t1 = 99.46°C, y1 = 0.0191 and y2 = 0.9809 (Refer Table 8.51)
Latent heat of vapour leaving Tray-1
λ1 = 0.0191 × 40 093 + 0.9809 × 41 081
= 41 062.3 kJ/kmol
Table 8.52 Material Balance of Tray-1 (from Bottom)
Vapour, V–1 Liquid flow from Tray-2, L–2
Component Bottom product, W kmol/h
mole fr. kmol/h kmol/h mole fr.

IPA 0.066 0.0191 2.6 2.666 0.0132

Water 65.869 0.9809 133.66 199.529 0.9868

Total 65.965 1.0000 136.26 202.195 1.0000

With x1 = 0.0132 and x2 = 0.9868

γ 1 = 8.917 39 and γ2 = 1.000 82
At t2 = 94.83°C, y1 = 0.1904 and y2 = 0.8096
λ 2 = 0.1904 × 40 093 + 0.8096 × 41 081
= 40 901 kJ/kmol
Adjust vapour flow from tray-2 with molar latent heats of vapour from tray-1
and tray-2.
V–2 = 136.26 × 41 062.3/40 901
= 136.797 kmol/h
 Table 8.53 Material Balance of Tray-2 (from Bottom)
L–2 V–2 V–1 L–3
Component
kmol/h mole fr. kmol/h kmol/h kmol/h mole fr.

IPA 2.666 0.1906 26.046 2.60 26.112 0.1288

Water 199.529 0.5696 110.751 133.66 176.520 0.8712

Total 202.195 1.0000 136.797 136.26 202.732 1.0000

With x1 = 0.1288 and x2 = 0.8712

γ 1 = 3.858 09 and γ 2 = 1.063 53
At t3 = 83.02, y1 = 0.5121 and y2 = 0.4879
λ 3 = 40 591.2 kJ/kmol
Adjust vapour flow from tray-3 with molar latent heats of vapour from tray-2
and tray-3.
V–3 = 136.797 × 40 901/40 591.2
= 137.841 kmol/h
Table 8.54 Material Balance of Tray-3 (without Fresh Feed)
L–3 V–3 V–2 L–4
Component
kmol/h mole fr. kmol/h kmol/h kmol/h mole fr.

IPA 26.112 0.5121 70.588 26.046 65.654 0.3394

Water 176.629 0.4879 67.253 110.251 133.631 0.6606
Total 202.732 1.0000 137.841 136.797 202.285 1.0000

Composition of liquid mixture on tray-4 is 0.3394 mole fraction IPA which is

higher than mole fraction 0.2306 of IPA in the fresh feed. Hence, feed is to be
introduced on 3rd tray.
Table 8.55 Liquid Composition on Tray-4 after Addition of Fresh Feed on Tray–3
L–4 F L4
Component
kmol/h kmol/h kmol/h mole fr.

IPA 68.654 23.06 45.594 0.4457

Water 133.631 76.94 56.691 0.5543

Total 202.285 100.00 102.285 1.0000

Tray-to-tray calculations for the rectification section can be carried out in

analogous manner to the stripping section. These are summarised in Table 8.57.
Sample calculation for rectification section:
Table 8.56 Material Balance of Tray-4 (from Bottom)
L4 V4 V3 L5
Component
kmol/h mole fr. kmol/h kmol/h kmol/h mole fr.
 IPA 45.594 0.5837 80.628 70.588 55.634 0.5424
Water 56.691 0.4163 57.505 67.253 46.943 0.4536
Total 102.285 1.0000 138.133 137.841 102.577 1.0000

Further tray calculations of rectification section are summarised in Table 8.57.

Total number of theoretical tray are calculated to be 23 with feed on 3rd tray
from bottom.
Table 8.57 Summary of Tray-to-Tray Calculations for IPA-Water System at 101.325 kPa
Tray No. Liquid Activity Vapour
Temperature, Liquid flow Vapour flow Average molar latent heat composition coefficients composition
(from
°C rate, kmol/h rate, kmol/h of vaporisation λ, kJ/kmol
bottom) x1(IPA) x2(H2O) γ1(IPA) γ2(H2O) y1(IPA) y2(H2O)

1 2 3 4 5 6 7 8 9 10 11

Stripping Section:
1 99.46 65.965 136.260 41 062.3 0.0010 0.9990 9.99657 1.00000 0.0191 0.9809
2 94.83 202.195 136.797 40 926.5 0.0103 0.9897 9.15797 1.00005 0.1565 0.8435
Saturated Fresh Feed 100.000 0.2306 0.7694
Rectification Section
3 83.02 202.732 137.841 40 591.2 0.1288 0.8712 3.85809 1.06353 0.5121 0.4871
4 80.96 102.285 138.133 40 505.3 0.4457 0.5543 1.38036 1.54487 0.5837 0.4163
5 80.62 102.577 138.480 40 462.8 0.5424 0.4576 1.21118 1.75378 0.6148 0.3849
6 80.52 102.924 138.483 40 461.9 0.5843 0.4157 1.16117 1.85170 0.6324 0.3678
7 80.47 102.927 138.526 40 449.2 0.608 0.392 1.13745 1.90897 0.6433 0.3568
8 80.44 102.970 138.585 40 432.1 0.622 0.378 1.12478 1.94343 0.6500 0.3499
9 80.44 103.029 138.465 40 449.8 0.6317 0.3683 1.11654 1.96757 0.6553 0.3451
10 80.422 102.909 138.527 40 431.2 0.6388 0.3612 1.11078 1.98537 0.6587 0.3413
11 80.415 102.971 138.546 40 425.8 0.6434 0.3566 1.10717 1.99697 0.6611 0.3388
12 80.413 102.990 138.545 40 426.2 0.6467 0.3533 1.10463 2.00532 0.6630 0.3371
13 80.41 103.089 138.555 40 423.3 0.6486 0.3514 1.10318 2.01014 0.6640 0.3360
14 80.395 103.099 138.554 40 423.7 0.6499 0.3501 1.10318 2.01344 0.6649 0.3351
15 80.38 103.098 138.559 40 422.2 0.6511 0.3489 1.10318 2.01649 0.6657 0.3343
16 80.367 103.103 138.560 40 421.8 0.6522 0.3478 1.10318 2.01929 0.6665 0.3335
17 80.354 103.104 138.560 40 421.3 0.6533 0.3467 1.10318 2.02209 0.6672 0.3327
18 80.343 103.104 138.564 40 420.3 0.6542 0.3458 1.10318 2.02439 0.6679 0.3321
19 80.332 103.108 138.562 40 421.0 0.6552 0.3448 1.10318 2.02694 0.6686 0.3314
20 80.32 103.106 138.570 40 418.2 0.6561 0.3439 1.10318 2.02924 0.6692 0.3308
21 80.312 103.114 138.564 40 419.9 0.6569 0.3431 1.10318 2.03128 0.6698 0.3302
22 80.301 103.108 138.575 40 416.7 0.6577 0.3423 1.10318 2.03333 0.6703 0.3297
23 80.294 103.119 138.570 40 418.1 0.6584 0.3416 1.10318 2.03512 0.6708 0.3292

Note: It may be noted that in this method number of theoretical trays are worked out to be 23 (to achieve
67.08 mole % IPA in vapour) with fresh feed entry on third tray from the bottom. This number is much
greater than that obtained by McCabe–Thiele diagram. It is therefore advisable to carry out tray to tray
calculations for design of an azeotropic column with non-ideal behaviour. Further it is recommended to
adjust pressure drop on each tray with the help of anticipated pressure drop across the tray. This will
change total operating pressure on each tray.
With reflux flow of 102.195 kmol/h, difference in calculated in tray 23 is 0.924 kmol/h, i.e., 0.9%. This is
quite low. However, if energy balance of each tray is also calculated, this could have been negligible.
This type of repetitive calculations can be conveniently carried out with a spreadsheet (such as Excell®) on
the computer.
 Refer to Table 8.51 in Example 8.18. It can be seen that boiling
point of 10 mole % IPA mixture is 83.85°C. This indicates that boiling point
variation of only 1.35°C takes place from 0.1 mole fraction to the azeotropic
point. Thus if the column is operated to separate up to 0.1 mole fraction, the
column operation is nearly isothermal. In the overhead condenser, cooling
medium at 72°C (max.) can be used while in the reboiler, heating medium at
95°C should be adequate.
For the above reason it is decided to operate the main column in the range of 0.1
mole fraction to 0.675 mole fraction IPA. It will be provided its own reboiler as
shown in Fig. 8.54.
 Fig. 8.54 Azeotropic Distillation of Aqueous Isopropyl Alcohol (Conventional System)

Liquid mixture, containing 0.1 mole fraction, is fed to another column in which
open steam will be used for stripping to achieve 0.1 mole % (0.0033 mass %)
IPA in the bottoms. Vapour from this column will be fed to the main column to
supplement the heat.
For the new arrangement, rework the number of theoretical stages, required for
both the columns for R = 3.
Solution
Note: It can be noted that cooling medium; water for vapour at 72°C is
required in the overhead condenser while heating medium at 95°C is required
in the reboiler. This suggests water as an auxiliary heat transfer medium in the
heat pump (refer to para 6 of Sec. 8.9.1). Saturation pressure of water at 72°C
is 0.34 bar a and at 95°C is 0.845 bar a. Compression ratio across the thermo-
compressor will be just 2.485.
In the second column, since live steam is to be injected, its temperature can be
just above 100°C. This permits use of flash steam at low pressure (may be at
105°C or 1.21 bar a) as live steam.
Use of live steam at low pressure and water as the auxiliary heat transfer
medium in the heat pump will make distillation quite energy efficient. Thermal
energy requirement will be considerably reduced with this arrangement.
From Fig. 8.55 number of theoretical stages required for the separation up to 10
mole %. IPA from 23.06% mole IPA (feed), N = 8.
Second column is an isothermal steam stripper. In stripping liquid, solution is
contacted with vapour or gas in a countercurrent manner to strip off the most
volatile component of liquid solution. Most volatile component of liquid solution
is transferred to gas or vapour. Stripping is a reverse of gas absorption. For
stripping, low pressure and high temperature are favourable.
Industrial strippers are of two types:
1. In first type of stripper, hot gas/steam (can be called stripping agent) is
introduced at the bottom of stripper and the same is obtained from external
source.
2. In second type, hot vapour is generated by partial vaporization of liquid
solution in reboiler at bottom. This type of stripper is like a bottom section
(stripping section) of distillation column. Feed solution is introduced from top
and reboiler is provided at the bottom. Reflux is not provided at the top. If the
component to be separated from feed is having very high relative volatility,
 then this type of stripper can be selected.
In this example, first type of stripper is selected. Stripping can be assumed as
isothermal at 100°C.
Let, Gs = Molar flow rate of steam, kmol/h
Ls = Molar flow rate of water of feed, kmol/h
Y =

X =

Equilibrium data can be approximated from Table 8.51.

x y X Y
0 0 0 0
0.02 0.2495 0.0204 0.3324
0.05 0.3993 0.0526 0.6647
0.075 0.4561 0.081 0.8385
0.1 0.489 0.1111 0.957
Equilibrium curve is plotted in Fig. 8.55.
 Fig. 8.55 Stripping of Dilute IPA with Live Steam

Minimum amount of steam required for desired separation can be obtained by

following equation:
Gsm(Y1m – Y2) = Ls(X1 – X2)   (8.127)
where, Y1 and X1 represent the molar ratios at top of the stripper and Y2, X2
represent the molar ratios at the bottom of stripper.
Here equilibrium curve is concave upward, hence Y1m is equilibrium value of Y
corresponding to the value of XF.

xF = 0.1, XF = = 0.1111

Y1m = 0.957 (From Fig. 8.55)

Y2 = 0, X1 = XF = 0.1111, X2 = ≅ 0.001

Gsm =

or = = 0.115

Let actual amount of steam, Gs = 1.5 Gsm

= 1.5 × 0.115 = 0.1725 =

0.1725 =

Y1 = 0.6383
Draw the operating line starting from (X2, Y2) to (X1, Y1). Staircase construction
between equilibrium curve and operating line, starting from (X1, Y1) to (X2, Y2)
gives number of theoretical stages required for the desired separation. From Fig.
8.55, number of theoretical stages required in the steam stripper; N = 6.
Ethanol forms a minimum boiling azeotrope at 78.15°C and
101.325 kPa, with water having composition of 10.57 mole % water. Vapour
pressure of ethanol can be calculated using Antoine equation14.
log pv = 7.2371 –

where, pv = Vapour pressure, kPa

T = Saturation temperature, K
Extractive distillation technique is known for separation of an azeotrope using a
solvent. In an extractive process, aqueous inorganic salt solution is used as a
solvent to break the azeotrope. Lowering of vapour pressure is experienced when
an inorganic salt is dissolved in water. Vapour pressures of aqueous calcium
chloride solutions are given in Table 8.58.
Table 8.58 Vapour Pressure of Water over Aqueous Calcium Chloride Solution49
Calculated vapour pressure of water, Torr

Temperature, °C Strength of CaCl2 solution, mass %

0 20 25 35

40 55.3 47.9 41.5 30.1

70 233.7 197.7 179.6 136.4
90 525.9 448.4 408.7 318.0
100 760.0 660.6 593.9 466.1
110 1074.6 905.9 841.9 629.5
120 1489.2 — — 886.6

It is proposed to use aqueous 25% CaCl2 solution (by mass) as a solvent for
dehydration of aqueous ethanol. Proposed flow diagram is shown in Fig. 8.55.
Fig. 8.56 Extractive Distillation of Aqueous Ethanol with Aqueous Calcium Chloride Solution (25 mass %)

Plot x-y diagrams of ethanol-water system at 101.325 kPa, with and without the
solvent. Superimpose both the diagrams and study the effect of addition of
aqueous CaCl2 solution on the separation.
(a) A feed consisting of 50 mass % ethanol under saturated liquid conditions is to
be distilled to give 99.8 mole % (99.92 mass %) ethanol as distillate and 0.1
mole % ethanol in bottoms in the extractive distillation column. Using
McCabe–Thiele technique and for R = 2 Rm, find the number of theoretical
stages required for the separation at 101.325 kPa and locate the feed stage.
(b) If the distillate quality is restricted to 96.0 mole % (98.4 mass %) ethanol for
the case (a), find the number of theoretical stages required for the separation
and locate the feed stage.
Solution
Azeotrope contains 0.1057 mole fraction (0.0442 mass fraction) ethanol. Let
suffix 1 represent ethanol and suffix 2 represent water.
At the boiling point of the azeotrope, t = 78.15°C or T = 351.3 K,

log pv1 = 7.2371 – = 2.0033

pv1 = 100.763 kPa = 755.78 Torr

From steam tables, pv2 = 43.93 kPa = 329.50 Torr
γ 1 = 760/755.78 = 1.0056
γ 2 = 760/329.50 = 2.3065
van Laar equations:

ln (1.0056) = A

ln (2.3065) = B

Solving by Mathcad,
A = 1.950 34 and B = 0.9329
Using values of A and B, t-x-y data can be generated in an analogous manner to
the IPA-water system (Example 8.18). When use of aqueous CaCl2 solution (25
mass %) is considered, pv2 will be replaced by p′v2, i.e., vapour pressure of water
over the aqueous solution. For this purpose a graph of p′v2 vs temperature can be
plotted and p′v2 values are read at the required temperature.
Sample calculations for x1 = 0.9 and x2 = 0.1:
γ1 = 1.004 98 and γ2 = 2.318 96
Assume t = 78.2°C or T = 351.35 K.
pv1 = 757.22 Torr
p′v2 = 254.6 Torr against pv2 = 330.15 Torr
p = Σxi · γi · pvi
= (0.9 × 757.22 × 1.004 98) + (0.1 × 254.6 × 2.318 96)
= 684.89 + 59.04 = 743.93 Torr ≠ 760 Torr
Revise t = 78.8°C or T = 351.9 K.
pv1 = 775.49 Torr
p′v2 = 260.8 Torr
p = 761.9 Torr ≈ 760 Torr
y1 = (0.9 × 775.49 × 1.004 98)/760 = 0.923
Similarly t-x-y data are generated for ethanol-water-solvent system. Both t-x-y
data are presented in Table 8.59 and Fig. 8.57(a).
Table 8.59 Vapour-Liquid Equilibrium for Ethanol-Water System at 1 atm
Without solvent, mole fraction ethanol With solvent, mole fraction ethanol

In Liquid In Vapour Boiling point, °C In Liquid In Vapour Boiling point, °C

x y t x y t

0 0 100.0 0 0 107.0
0.025 0.2503 92.8 0.025 0.298 97.7
0.05 0.365 88.8 0.05 0.4234 92.95
0.1 0.468 85.0 0.1 0.531 88.35
0.2 0.542 82.3 0.2 0.6026 85.1
0.3 0.578 81.2 0.3 0.638 83.75
0.4 0.6096 80.35 0.4 0.668 82.7
0.5 0.647 79.65 0.5 0.703 81.8
0.6 0.693 79.05 0.6 0.745 80.9
0.7 0.75 78.6 0.7 0.7945 80.05
0.8 0.817 78.25 0.8 0.8513 79.3
0.8943 0.8943 78.15 – azeotrope — — —
0.96 0.953 78.1 0.9 0.921 78.8
1.0 1.0 78.1 1.0 1.0 78.1

Case (a)
Refer to Fig. 8.57(b), xD = 0.998, zF = 0.2811 (or 0.5 mass fraction) and xW =
0.001

= 0.2005

R m = 3.98
R = 2 R m = 2 × 3.98 = 7.96
 = 0.1114

Total 29 theoretical stages are required for the separation and the feed will be
introduced on 14th stage.

Fig. 8.57(a) Vapour Liquid Equilibrium Curves for Ethanol Water System with and without Aqueous
Calcium Chloride Solution (25 mass %)

Case (b)
For xD = 0.96, the graph [similar to Fig. 8.56(c)] was redrawn and following data
were collected:
= 0.2701

R m = 2.554
R = 2 × 2.554 = 5.108
= 0.1572

With these values of R, the number of theoretical stages were calculated to be 17

and the feed will have to be introduced on 14th stage.
Required number of theoretical stages for both the cases were found using
McCabe–Thiele diagrams [Figs. 8.56(b) and (c)]. However, it is recommended to
recalculate the requirement by tray-to-tray calculations for actual process design.

Fig. 8.57(b) Extractive Distillation with Aqueous Calcium Chloride Solution (25 mass %)

Note: It may be noted that single stage extractive distillation (ED) column is
adequate to produce ethanol with 99.8 mole % purity while in most ED
systems, two columns are required for the same degree of separation. Boiling
point elevation at atmospheric pressure (at the bottom of the column) is just
7°C. Also this ED system calls for 50% or less energy than the conventional
ED columns.
Presence of high chlorides means corrosive nature of the aqueous system. The
column, reboiler and concentrator should be built with material of construction,
resistant to chlorides. Special stainless steel, such as Ti stabilised SS 316-L or
high nickel alloy, could be used for such corrosive service. Circulation pumps
of Alloy-20 could be used. However, cost of such materials of construction is
higher than that of normally used materials. Nevertheless, overall operating
cost of the ED system with aqueous CaCl2 solution will be less than the
conventional ED system.

Fig. 8.57(c) Extractive Distillation with Aqueous Calcium Chloride Solution (25 mass %)

Results of Case (a) and Case (b) indicate that the ED column will be more than
70% taller in Case (a). Also reflux ratio of Case (a) is nearly 55% more than
Case (b). Both these factors call for higher investment and higher energy
consumption for Case (a).
Distillate of Case (b) contains 4 mole % (1.6 mass % or 16 032 ppm or 12 650
mg/L) water. To dehydrate ethanol further, three options are shown in Fig. 8.56.
In a conventional manner (Option I), distillate with 4 mole % water will be
distilled in another column to yield the azeotrope at the top (to be recycled) and
anhydrous ethanol at the bottom. In Option II, part of the vapour from the ED
column will be fed to a pervaporation unit. Here again, contaminated water
(with some ethanol), removed in the membrane unit, can be recycled to the
feed of the ED column. Drying of the distillate from ED column by molecular
sieves is shown as Option III. All three options are well proven in the industry
and the selection should be made on the basis of minimum total operating cost.
Please refer Section 8.9.3 for detailed discussions on various options.
Nitric acid water mixture forms maximum boiling azeotrope at
1 atm a and 122°C, with a composition of 64.1% by mass of HNO3. Distillate
contains 99.9% HNO3 (by mass). Separation is achieved from the azeotropic
mixture by extractive distillation using concentrated Sulphuric acid (99% by
mass) as solvent. Residue of this extractive distillation column is 80% by mass
H2SO4 and remaining water. Extractive distillation is carried out in glass packed
tower type distillation column at 180 Torr a pressure. Glass Raschig rings are
used as a packing material. Determine height of packing required for the desired
separation in this extractive distillation column.
Solution
Mole fraction of HNO3 in feed,

xF = = 0.3378

xD =

Let xW = 0.001
Here H2SO4 is not vaporised in distillation column and in reboiler. Hence, this
distillation can be approximated as binary distillation and MeCabe–Thiele
method can be applied. However, effect of addition of H2SO4 on vapour-liquid
equilibrium data or on relative volatility of HNO3–H2O mixture must be
considered.
Topmost temperature of column = Dew point temperature of overhead vapour
≅ Condensation temperature of pure HNO3 vapour
From Table 3.16 of Ref. 8,
At 180 Torr a, condensation temperature of 100% HNO3 is 48°C.
At 48°C, vapour pressure of water vapour over 80% H2SO4 (by mass) is 0.106 ×
10–2 bar. (From Table 3.13 of Ref. 8).
Relative volatility at topmost temperature
α =

pvA = Vapour pressure of 100% HNO3 = 180 Torr

pvB = Vapour pressure of water over 80% H2SO4
= 0.106 × 10–2 bar = 0.795 Torr
α1 =

Bottommost temperature = Bubble point of 80% H2SO4 ≅ 160°C

(From Table 3.15 of Ref. 8)
But at 160°C vapour pressure of HNO3 is not available. Hence, here at
intermediate temperature or at 90°C second value of α is determined.
At 90°C pvA = 820 Torr (From Table 3.16 of Ref. 8)
pvB = 0.0115 bar = 8.6257 Torr (From Table 3.13 of Ref. 8)

αav =

Considering this value as αav, Table 8.60 is prepared.

Table 8.60 Vapour-Liquid Equilibrium Data for HNO3–H2SO4 System

x y = x y =

0.02 0.66 0.4 0.9845

0.04 0.7984 0.5 0.9896
0.06 0.8585 0.6 0.993
0.08 0.892 0.7 0.9955
0.1 0.9135 0.8 0.9973
0.2 0.9596 0.9 0.999
0.3 0.976 1.0 1.0
Equilibrium curve of y vs. x is shown in Fig. 8.58.

Fig. 8.58 Extractive Distillation of Aqueous Nitric Acid

Equation of q-line.
y =

q =

Let feed is a liquid at atmospheric temperature = 30°C.

q =

Assume that bubble point of the feed mixture at 180 Torr a = 85°C
At 85°C λN = 39 000 kJ/kmol (Ref. 50)
λW = 41 326 kJ/kmol (from steam tables)
λF = Σ λi xi = 39 000 × 0.3378 + 41 326 × 0.6622
= 40 540.3 kJ/kmol
At 57.5°C CLN = 110.88 kJ/(kmol · °C)
CLW = 75.36 kJ/(kmol · °C)
CLF = 110.88 × 0.3378 + 75.36 × 0.6622
= 87.36 kJ/(kmol · K)
q =

y =

= 9.439 x – 2.85
For y = 0.7, x = 0.376
= 0.98 (From Fig. 8.58)

= 0.98

Rm = 0.02
Let R = 0.5
=

From Fig. 8.58 N′ = 3 for enrichment from x′r to xF

N = 3 + Ns
where, Ns = Nos. of theoretical stages required from 0.01 to 0.001
xW = 0.001, x′r = 0.01

Ns =   (8.21)

K′ = slope of equilibrium curve =

S′ = = slope of stripping section operating line

Ns =

Total number of theoretical stages required,

N = 3 + 1.73 = 4.73, say 5
HETP for glass Raschig rings = 1 m
Total height of packing required = 5 m
Note: Although Rm = 0.02 only, R was selected 0.5, i.e., 25 times. This is
because the system is highly exothermic and non-ideal. McCabe–Thiele
method is not really suitable for such a non-ideal system.

8.9 ENERGY CONSERVATION IN DISTILLATION

It was stated earlier that distillation is the most widely used separation technique
in chemical process industries but energy consumption in distillation is relatively
high. Hence, saving of energy in distillation operation is important for lowering
overall cost of process plant. This aspect is also becoming important in
biotechnology routes.
For saving the energy in distillation systems there are two options:
1. Make the distillation more energy efficient; either by design or by operating
the column more efficiently.
2. Replace the distillation partially or completely by a new separation technique.
8.9.1 Energy Efficient Distillation by Design
To decrease the energy consumption of distillation various design options can be
considered. To name a few
1. Optimum design of the system
2. Heat integration
3. Use of high efficiency trays or high efficiency packings
4. Advanced process control
5. Use of thermally coupled distillation columns
6. Use of heat pumps (ejector/thermocompressor).
1. Optimum design of system: Process engineers have a general tendency to
overdesign the system but at the expense of energy. For example, very high
purities are specified for the distillate or bottom product which may not be
warranted. At times, solvents are recycled in the system and very high purity
may not be required. By designing the system for the desired purity only, reflux
ratio can be trimmed thereby energy consumption can be reduced. In Sec. 8.8.2,
a case of extractive distillation of cyclohexane and benzene mixture was
discussed. As cyclohexane is a product, its purity of 99.3% was specified as a
distillate from the first column. However, since benzene is a reactant which is to
be recycled for further conversion to cyclohexane, its purity of 85% was
specified as a distillate from the second column. If benzene purity of over 99% is
specified, second column will be much taller with high reflux ratio, resulting in
high energy consumption in the reboiler. Take another example of ethanol-water
mixture. When ethanol is a desired product for mixing with petrol, very high
purity of ethanol will be needed.
For heterogeneous azeotropic distillation, first column is used as a concentrator
to produce distillate of near azeotropic composition. Subsequently, two more
columns are required for separation of the components. Cho and Jeon42 have
proved after rigorous simulation by varying design parameters that while
separating aqueous IPA solution, the first column (the concentrator) should be
designed to produce distillate containing 38.7 mole % IPA (and not the
azeotrope) which should be subsequently processed in a two-column system to
dehydrate IPA. This design is claimed to consume lowest energy for the entire
distillation system.
Knapp and Doharty47, 48 have proposed a two-column system (and not three
column system) for concentrating dilute aqueous ethanol solution (containing
just 4.2 mole % ethanol) from fermentor by adding approximately 0.043 kmol
acetone per kmol feed and applying pressure swing distillation technique. This
system is claimed to consume much less energy than conventional system.
Several such distillation case studies can be cited from the literature in which it
has been demonstrated that by careful selection of process parameters at the
design stage and the right technique, substantial energy savings can be achieved
in operation. However, most energy efficient design could be expensive from
equipment cost point of view. Nevertheless, energy saving outweighs capital cost
in most cases.
2. Heat integration: Heat integration is low risk technology and it results in
substantial energy saving. Reboiler of distillation column can act as a cooler for
other unit. For example, in acetaldehyde plant, exhothermicity of the reaction is
removed by circulating saturated liquid leaving from the bottom of distillation
column in tubeside of shell and tube-type reactor. Liquid is vaporised and vapour
is sent back to the distillation column. Thus reactor act as a reboiler for the
distillation column. An additional reboiler is provided for the distillation column
for supplying the balance heat (vapour) to the column.
Similarly, overhead condenser can act as a heater or reboiler for other unit. For
example, in cryogenic distillation of air, overhead condenser of higher pressure
distillation column also acts as a reboiler for the lower pressure distillation
column. In such a heat exchanger, nitrogen condenses on one side and on other
side, liquid oxygen boils.
Overhead vapour of distillation column can be partially used as heating medium
of reboiler of other distillation column. This can be achieved conveniently with
an ejector as discussed in Sec. 8.9.1 (f).
In Sec. 8.8.3, it was seen that pressure swing distillation technique offers an
opportunity for thermal integration of condensers and reboilers of two columns
as they operate at different temperatures.
Bottom product cooler can also act as feed preheater. Such an energy integration
decreases the heat requirement of the preheater and also reduces the cooling load
of the product cooler. It decreases heat required for reboiler and improves
(NPSH)A of bottom product transfer pump.
3. Use of high efficiency trays or high efficiency packings: Number of
equilibrium stages of existing distillation column can be increased by replacing
sieve trays or bubble cap trays with high efficiency packings or trays. For the
fixed degree of separation, increase in number of equilibrium stages reduces
reflux ratio and consequently reduces reboiler energy. Other advantage of using
high efficiency packings (such as structured packings) or high efficiency trays
(such as modified sieve trays) is to increase in capacity of the distillation
column. Disadvantage of this replacement is higher cost of structured packings
or of high efficiency trays. Generally savings in energy cost outweigh the cost of
high efficiency packings/trays.
Example51: Replacement of sieve trays of 400 mm diameter distillation column
of styrene plant by a packed column, packed with structured packings, resulted
in saving of energy consumption by 17.7%. This column was operating at its
capacity of 24 500 kg/h of styrene. Its feed composition was 54% ethyl benzene
and 46% styrene (by mole), distillate composition was 96.9% ethyl benzene and
residue composition was 99.7% styrene (by mole). After replacement of sieve
trays by a column with structured packings, reflux ratio was reduced from 7.3 to
5.8 and reboiler energy was reduced from 1605 kJ/kg of styrene to 1321 kJ/kg of
styrene. In this modification, feed flow rate and purity of products were kept
same.
Similarly replacement of conventional bubble cap trays by structured packing
increases the capacity of the column or saves energy by decreasing the reflux
ratio. However, structured packing cannot work efficiently at higher turndown
ratio.
4. Advanced process control: Plant operators normally operate the distillation
columns with a sufficient safety margin by achieving actual compositions of
distillation products which are better than specifications. Operating a distillation
column in this manner requires extra reflux and energy, but some margin is
necessary so that the plant operator can cope with process upsets and other
problems. The ‘safety’ margin is smaller in the plants which are properly
designed and operated.
Installation of advanced process control systems based on accurate analytical
sensors, reliable control loop hardware and instrumentation and good process
models increases confidence in operators. In many instances, this added
confidence translated to a reduced margin of ‘safety’, reflux ratio and reboiler
energy. Nowadays, instrumentation and controls are given due weightage at the
design stage itself. In fact data logging system provides useful information on
optimising the column performance.
Example51: A conventional distillation column was used for separating a binary
mixture of propane and proplyene. Feed composition was 73% propylene and
27% propane (by mole). Required purity of top product was 99.7 mole %
propylene for which reflux ratio was 13.5. With a sufficient margin,
conventional distillation column was operated with excess reflux ratio of 15.5 to
get the purity of propylene of 99.9 mole % instead of the required purity of 99.7
mole %. After installation of distributed control system (DCS) on this column,
more accurate control of the purity of propylene at 99.7 mole % could be
achieved which reduced energy consumption up to 13.5%. However, from a
practical standpoint, operators cannot control product purity precisely at required
composition in the conventional system.
5. Use of thermally coupled distillation52: In the separation of multicomponent
mixture (more than 2), thermally coupled distillation columns require less energy
and fixed cost as compared to conventional distillation of multicomponent
mixture. To understand this, consider the separation of a ternary mixture having
components A, B and C by distillation. A is the most volatile component and C
is the least volatile component. Component B is having intermediate volatility.
Conventional methods for separating the ternary mixture by distillation is either
via direct sequence or indirect sequence, as shown in Figs. 8.59(a) and (b).

Fig. 8.59 Multicomponent Distillation in Sequence

Fig. 8.60 Thermally Coupled Distillation Columns

For the given ternary mixture, selection between these two conventional
arrangement of distillation column depends on composition of the ternary
mixture, relative volatilities, latent heats of vaporisation, boiling points, heat
sensitivities, etc. Compared to these conventional arrangements of distillation
columns thermally coupled distillation columns require less energy consumption
and fixed capital investment. For the separation of the same ternary mixture of
A, B and C by thermally coupled distillation, two types of arrangement are
possible, as shown in Figs. 8.60(a) and (b).
In thermally coupled side column rectifier, vapour stream is withdrawn from the
stripping section of main column and is sent for rectification in side column
where an overhead condenser is provided. Liquid stream from the bottom of side
column is sent back to the main column.
In thermally coupled side column stripper, liquid stream from the enriching
section the main column is withdrawn and is sent to side column for stripping
where reboiler is provided at bottom. Vapour stream from side column is sent
back to the main column.
Selection between these two arrangements depend on concentration profile of
intermediate component B. If concentration of component B is maximum at any
point of stripping section of the main or upstream column, then side column
rectifier is selected. By withdrawing a side stream from the point at which the
concentration of B is maximised, the distillation of B from either C or A is made
much easier.
Side rectifiers and side strippers can be cost effective on any ternary separation
where less than half of the feed leaves in middle product or where relatively low
purity of middle product is acceptable. Compared to conventional column
arrangements side stripper or side rectifier can reduce the energy consumption
up to 50%. Saving in fixed cost is also achieved because thermally coupled
distillation columns avoid the use of one condenser or reboiler and require
reduced column diameter or fewer stages for at least one column. Concept of
thermally coupled distillation is also applicable to any pseudoternary systems,
i.e., multicomponent mixtures requiring separation into three products. Hence, it
is applicable to the distillation of any mixture of more than two components.
Refer to Example 8.9 in which separation of fatty acids from a mixture is dealt.
Such a system can be considered for thermal coupling. In this case thermally
coupled side column is recommended.
In a petroleum refinery, this technique is extensively practised to produce
different cuts (products like kerosene, naphtha, petrol, etc.) from crude by having
number of side-stream rectifiers and strippers.
As an alternate to the thermally coupled columns, a novel idea in design of tray
type distillation column is to have partitioned distillation column (Ref. 53) (also
called a divided-wall column or Petyluk column) for multicomponent system. In
such a column, a vertical baffle is placed (Fig. 8.61), which separates the feed
location from that at which the intermediate boiling product is withdrawn. On
the feed side of the partition, separation is achieved between the light (A) and
heavy (C) fractions while the intermediate boiling fraction (B) is allowed to
migrate to top and bottom sides of the partition. On the other side of the
partition, the light fraction is washed down, thereby intermediate boiling fraction
(B) of desired purity is withdrawn. Such a column is larger in diameter than the
otherwise required two columns but overall capital and operating costs could be
significantly (up to 30%) lower. Process design of such columns is quite
complex and is not available in open literature.

Fig. 8.61 Partitioned Distillation Column

6. Use of heat pumps in distillation: Steam ejectors have been used in limited
distillation applications as inexpensive heat pumps. An ejector uses high
pressure fluid to entrain low pressure fluid. Steam ejectors can be operated in
different ways.
Simplest way to use an ejector is to operate with live steam. Figure 8.62 shows
the proposed arrangement. Part of the top vapour is compressed in an ejector
utilising live (fresh) steam as motive fluid. Two or more ejectors can be used in
parallel to improve flexibility. Use of such an ejector system requires that (i) top
product is water and (ii) temperature difference between the bottom and top of
the column is low. Hydrogen peroxide-water distillation column can be fitted
with a live steam ejector. This ejector system reduces steam and cooling water
requirements for the distillation unit. It also reduces the load of vacuum pump.
Refer to Example 11.1 for use of a thermocompressor in a triple-effect
evapourator system. Similar arrangement is also recommended in distillation.
Figure 8.63 is a capacity ratio curve for the thermocompressor (ejector),
operating with saturated steam at 20.7 bar g.
In Fig. 8.64, a system is shown in which top vapour product is used as a motive
as well as entrained fluid. In this case, it should be possible to heat the top
product to a substantially higher temperature and pressure than the column’s
operating conditions. Also the temperature difference between top and bottom of
the column should be low. Ethyl benzene/styrene splitter can be fitted with such
an ejector system. A 225 × 106 t/a styrene monomer plant when operated with
such an ejector system reduced steam and cooling water requirements by 25%
and 32%, respectively. Capital investment for the modifications is expected to be
repaid in less than 18 months. Acetic acid-water, ethylene glycol-water systems
can also be operated in this manner.

Fig. 8.62 Ejector System with Live Steam as the Driving Medium54 (Reproduced with the Permission of Gul
Publishing Co., USA)
Fig. 8.63 Capacity Ratios of Steam Jet Thermocompressors Operating with Saturated 20.7 bar g Motive
Steam
(Courtesy: Mazda Ltd. Ahemedabad)
Fig. 8.64 Ejector System with Use Top Product as the Driving Medium54 (Reproduced with the Permission
of Gul Publishing Co., USA)

In a third alternative, bottom product can be used to generate motive fluid for the
ejector (Fig. 8.65). Part of the top vapour is condensed in a heat exchanger with
bottom product which operated under reduced pressure (say subatmospheric).
High pressure bottom product vapour provides the driving medium for the
ejector. For this system to operate satisfactorily, three conditions need be
satisfied: (i) It should be possible to heat the bottom product to a substantially
higher temperature and pressure than the operating conditions of the column, (ii)
Top product is not suitable for compression with the ejector, and (iii)
Temperature difference between bottom and top of the column is low. Alcohol
(methanol or ethanol)-water distillation column can be fitted with such an ejector
system.
Innovative flow sheets were presented as Figs. 8.62 8.64 and 8.65 for use of an
ejector for energy conservation. Process engineer should therefore always look
for innovative designs which are technically feasible and select the best option
based on all aspects, including operational convenience.
It may be noted that in all the applications, it is necessary to have low
temperature difference between top and bottom products of the column. Also the
ejectors should be large in size so that with low differential pressure between
motive and entrained fluids is permitted. In Figs. 8.62, 8.64 and 8.65, number of
peripheral equipments are shown which require substantial capital expenditure.
Because of these reasons, ejector installation is of lesser interest for the smaller
columns. However, when large distillation column and substantial energy input
are required, use of ejector(s) could prove economically viable (Ref. 54).
In Fig. 8.62, live steam is used for compressing the overhead vapours in an
ejector to the reboiler for its use as heat transfer medium. As an alternate to the
ejector, a vapour compressor (Ref. 55) can be used to compress the overhead
vapours (Fig. 8.66). The compressor could be a twin lobe blower, a radial or an
axial–flow turbocompressor

Fig. 8.65 Ejector System with Bottom Product as the Driving Medium54 (Reproduced with the Permission of
Gul Publishing Co., USA)
 Fig. 8.66 Distillation Columns with Direct Vapour Compressor55 (Reproduced with the Permission of
Chemical Engineering by Access Intelligence, USA)

or a screw compressor, driven directly or via gear box by an electric motor or a

steam turbine. For overhead vapour having low molar mass (such as water),
compression is rather restricted up to 10 bar. For high molar mass vapours,
multistage compressors can be used to compress them up to 80 bar. For partial
load operation, the compressor should have a variable speed drive or kick-back
arrangement, performance of Vapour Recompression System (VRC) is measured
by calculating the ratio of reboiler duty (kW) to the power consumption (kW) of
the compressor. This ratio is called Coefficient of Performance (COP) and
typical value ranges from 6 to 15. Although VRC system can be used for any
distillation column on technical ground, it is favoured in the following cases due
to the economic factors:
1. When reboiler duty is high (> 2 MW), VRC system is promising.
2. Lower temperature rise across the compressor is preferred.
3. When chilled water or brine is used to condense overhead vapours in the
condenser, use of VRC system is advantageous.
4. Columns that separate products with low separation factors require larger
number of stages and high reflux ratio. Such columns are most suitable for
adoption of VRC system. Separation of propylene-propane, ethylene-ethane,
ethyl benzene-xylene, iso-propanol-water (refer to Example 8.18), etc., are
the systems which can benefit significantly by adoption of the VRC system.
In a typical case, 60% (by mass) iso-propanol feed (balance water) is distilled to
achieve azeotrope (87.5% iso-propanol) with bottom product containing 15%
iso-propanol. When overhead azeotropic mixture is compressed and used as a
heating medium in the reboiler, coefficient of performance in excess of 10 could
be achieved.
Where low cost heat or free waste energy is available for use in the reboiler,
VRC system is not recommended. Also when the column is operated under high
vacuum (>710 Torr), VRC is not economical because compression ratio
becomes high.
One more option for reducing energy consumption in distillation is to adopt heat
pump with an auxiliary heat transfer medium (Ref. 56). In Fig. 8.67, distillation
column with auxiliary heat transfer medium is shown. In the overhead
condenser, vapours are condensed by evapourating the auxiliary medium. A
compressor compresses auxiliary medium vapours. In the reboiler, auxiliary
medium vapours are condensed and liquefied medium is recycled back to the
overhead condenser through an expansion valve, thus completing the cycle.

Fig. 8.67 Distillation Column Heat Pump with Auxiliary Heat Transfer Medium56 (Reproduced with the
Permission of Chemical Engineering by Access Intelligence, USA)

Typical auxiliary heat transfer media are refrigerants. Water (R*-718), ammonia
(R–717), R–134a, R–22, etc., can be used for the purpose. Among these, water
(R–718) has many advantages over others. Water is chemically and thermally
stable. No negative environmental effect in case of leakage, no special material
of construction is required, cheap, high heat transfer co-efficients and high latent
heat of vaporization are chief factors that favour its selection. Performance of
heat pump with auxiliary medium is measured by Rankin coefficient of
performance. Generally, COP of water is higher than that of other refrigerants.
Considering capital investment for the auxiliary heat–transfer medium recycle
system and larger condenser and reboiler (as compared to conventional system),
heat pump system with auxiliary medium is cost effective when reboiler duty is
high (say > 2 MW). In this case also when close boiling components are to be
separated, the heat pump is economical. Separation of p-xylene and ethyl
benzene mixture and styrene and ethyl benzene mixture can be good examples of
the system. For a 225 000 t/a styrene monomer plant, COP of heat pump with
water as auxiliary medium could be in excess of 5.5.
A distillation column is to be designed for aromatics
separation. Material balance around the column is fixed and compositions of
various streams are given below.
Table 8.61 Composition of Distillation Column Streams
Composition, mole %
Component
Feed Distillate Bottoms

Benzene 2.2 22.8 0.

Toluene 7.4 72.2 0.5
Ethyl benzene 43.4 5.0 47.5
Styrene 47.0 0 52.0

Feed flow rate is 70 kmol/h as saturated liquid. Column must be operated under
vacuum to avoid thermal cracking and polymerisation.
It is planned to use a heat pump (similar to one shown in Fig. 8.67) for the
system. Refrigerant R–134a (1,1,1,2–tetrafluoroethane) is to be used as an
auxiliary heat transfer medium. Bubble point of the distillate will be 12°C in the
overhead condenser and accordingly operating pressure of the condenser will be
fixed.
Determine the required number of theoretical stages for the desired separation.
Also find the location of the feed stage. Based on number of theoretical stages,
fix the height of packed bed, packed with fibreglass reinforced plastic (FRP) Pall
rings of 25 mm size. Assume pressure drop in the column to be 0.04 atm and
calculate the operating pressure of the reboiler. In the reboiler, compressed R–
134a vapour will be condensed.
Assume approach of 10°C in the condenser and the reboiler for effective heat
transfer. Calculate heat duties of the condenser and reboiler and the circulation
rate of the refrigerant. Consider ideal behaviour of all liquid and gaseous
mixtures.
Solution
Bubble point of distillate: 12°C
Considering ideal VLE,
Σ xi pvi = pt(A)
xDB = 0.228, xDT = 0.722 and xDEB = 0.05
Table 8.62 Antoine Constants14
Antoine Constants
Component
A B C

Benzene 14.1603 2948.78 –44.5633

Toluene 14.2515 3242.38 –47.1806
Ethyl benzene 6.0821 1424.3 –59.95
Styrene 14.3284 3516.43 –56.1529

ln pv = A – for benzene, toluene and styrene(B)

and log pv = A – for ethyl benzene(C)

where, pv is in kPa and T is in K.

At T = 12 + 273 = 285 K,
pvB = 6.659 kPa, pvT = 1.8547 kPa and pvEB = 0.567 kPa
Operating pressure of the condenser,
pt = 0.228 × 6.659 + 0.722 × 1.8547 + 0.05 × 0.567 = 2.8857 kPa
Pressure drop in the column = 0.04 atm = 4.053 kPa
Operating pressure at the bottom of the column, p′t = 2.8857 + 4.053 = 6.9387
kPa
This pressure ( pt′ ) will fix the bubble point of the liquid in the reboiler.
Σ xwi pvi = pt′(D)
By trial and error, bubble point is to be fixed to satisfy Eq. (D).
At tB = 62.5°C, TB = 335.5 K
pvB = 55.9708 kPa,
pvT = 20.2005 kPa, pvEB = 8.1878 kPa and pvS = 5.7 kPa
ΣxWi pvi = 0.005 × 20.2005 + 0.475 × 8.1878 + 0.52 × 5.7
= 6.954, close enough to 6.9387 kPa.
In this example, toluene is the light key component and ethyl benzene is the
heavy key component.
Table 8.63 Relative Volatilities

Component αtop = αbottom = αav = (atop · abottom)0.5

Benzene 11.744 6.836 8.96

Toluene (LK) 3.271 2.467 2.8407
Ethyl benzene (HK) 1 1 1
Styrene 0.625 0.6962 0.6596

FUG method
Minimum number of theoretical stages can be found, using Fenskey’s equation:

Nm =   (8.26)

Nm = = 6.919, say 7

Underwood method
= 1 – q  (8.15)

= 0

Solving by Mathcad,
ϑ = 2.4052
Also ϑ ∈ {1, …, 2.8407}
= Rm + 1  (8.14)
 = Rm + 1

For ϑ = 2.4052, Rm = 3.9856

Gilliland correlation
f (N ) = = 1 – exp.   (8.28)

where, ψ =   (8.29)

Table 8.64 Evaluation of Parameters

R ψ f(N) N

4.5 0.093 53 0.5603 17

6 0.287 8 0.39 12
7 0.376 8 0.3288 10.76
8 0.446 0.2819 10
9 0.501 44 0.2483 9.53

Referring to Table 8.64, R = 8 is considered optimum and required number of

theoretical stages are 10.
Feed tray location can be determined by Kirkbride equation,

log = 0.206 log   (8.30)

For determining W and D, material balance calculations are to be carried out.

Overall material balance:
F = D + W
70 = D + W (E)
Benzene balance:
0.022 × F = 0.228 × D + 0 × W
0.022 F = 0.228 D (F)
Solving Eqs. (E) and (F),
D = 6.7544 kmol/h and W = 63.2456 kmol/h

log = 0.206 log

 = 0.8839

Nr + Ns = 10
Ns = 5.308. Let Ns = 6
Therefore, sixth stage from bottom will be the feed stage. Take HETP = 0.5 m
for FRP Pall rings. Thus total packed bed height works out to 5 m.
Pressure drop per stage is calculated to be 0.405 kPa which is reasonable.
Heat duty of condenser, ϕC
ϕC = (R + 1) D λav
Using Watson equation, latent heats of vaporisation for the components of
distillate are calculated as under at 12°C.
λ B = 34 252.6 kJ/kmol, λ T = 38 452.8 kJ/kmol and
λ EB = 42 497.0 kJ/kmol
λ av = Σλi · yi = 0.228 × 34 252.6 + 0.722 × 38 452.8 + 0.05 × 42 497.0
= 37 697.3 kJ/kmol distillate
ϕ C = (8 + 1) × 6.7544 × 37 697.3 = 2291 603.8 kJ/h ≡ 636.56 kW
Let t = 12°C be the reference temperature
HD = 0, HD · D = 0
Refer to Table 6.29 of Chapter 6.
At T = 335.5 K
HW = –17.712 (335.5 – 285) + 1054.193 × 10–3

– 1838.075 × 10–6

+ 1694.332 × 10–9

= 9240 kJ/kmol mixture

ϕW = HW · W = 9240 × 63.2456 = 584 389.3 kJ/h ≡ 162.33 kW
Feed is saturated and hence enthalpy of feed at its bubble point is to be
calculated.
At bubble point,
 = Σyi = 1

Since feed is nearly in the middle of the column, assume average pressure at the
feed point
p′′t = = 4.9122 kPa

0.022 pvB + 0.074 pvT + 0.434 pvEB + 0.47 pvS = 4.9122

This equation can be solved by trial-and-error calculations. At t = 49°C (322 K),
pvB = 34.18, pvT = 11.6264, pvEB = 4.435, pvS = 3, all in kPa
Σxi pvi = 4.947, close enough to 4.9122 kPa
Refer to Table 6.30 of Chapter 6.
HF = –16.023 (322 – 285) + 1030.422 × 10–3

– 1809.476 × 10–6

+ 1693.124 × 10–9

= 6561.5 kJ/kmol feed

ϕF = HF · F = 6561.5 × 70 = 459 305 kJ/h ≡ 127.6 kW
Assume 5% heat loss to surrounding and no heat of mixing.
ϕ B = (ϕC + HD · D + HW · W – HF · F) 1.05
= (636.56 + 0 + 162.33 – 127.6) 1.05
= 704.85 kW ≡ 2537 460 kJ/h
Approach for heat transfer in the overhead condenser and reboiler is 10°C. Since
distillate bubble point is 12°C, auxiliary heat transfer medium (R–134a) should
evaporate at 2°C. Similarly in reboiler, boiling takes place at 62.5°C and hence
R–134a vapours are to be condensed at 72.5°C.
Table 8.65 Properties of R–134a27
Auxiliary heat transfer medium enthalpy, kJ/kg Vapour,
Temperature °C Pressure, bar a
Cp /Cv
Liquid Latent Vapour

2.0 3.150 (satd.) 202.68 197.07 399.75 1.182

72.5 22.385 (satd.) 308.57 120.42 428.99 1.653
130.0 23.0 (sup.) — — 500.00 —
(Reference state: Enthalpy of saturated R–134a liquid at 0°C = 200 kJ/kg)
In the condenser, R–134a liquid from the reboiler (at 72.5°C) will be flashed to
2°C and vapours will be produced.
Heat, taken up in condenser,
H1 = 399.75 – 308.57 = 91 kJ/kg

Evaporation rate of R–134a, ṁR =

= 2291 603.8/91 = 25 182.5 kg/h

These vapours will be superheated by 10°C before taking to the compressor to
avoid any liquid carryover. This can be done by exchanging heat with liquid
from reboiler. However, for the present, this difference is neglected. Saturated
vapour at 3.15 bar a will be compressed to 23 bar a.
Overall compression ratio = 23/3.15 = 7.302
Since this is high, it will be compressed in a two-stage reciprocating compressor.
Compression ratio in each stage = = 2.702
Vapour will be compressed from 3.15 bar a to 8.511 bar a in the first stage,
cooled suitably and compressed to 23.0 bar a in the second stage. Its temperature
is expected to be 130°C.
Heat, given up in the reboiler, H2 = 500 – 308.57
= 191.43 kJ/kg
Heat available, ϕ ′B = 25 182.5 × 191.43 = 4820 686 kJ/h ≡ 1339.1 kW
Since ϕ ′B is much higher (by 90%) than ϕB, a trim condenser will be needed to
take care of the excess heat load. This heat (low level) can be utilised to heat the
feed or some other stream in the plant. In addition, during start-up, overhead
condenser will not have any heat-load but reboiler will have to be
commissioned. Therefore, an additional auxiliary reboiler will be needed with
some other heat source, say with low pressure steam for the start-up. It can be
seen that adoption of heat pump system will thus call for a number of additional
equipments.
Notes:
(i) Since required number of equipments are many more than for the
conventional distillation system, initial investment will be substantially
high.
(ii) Operating pressure of the reboiler shell will be 23 bar a. This will call for
 thick shell, calling for higher investment.
(iii) Adequate instrumentation and controls (I&C) will be needed for the
refrigerant circulation system in addition to the I&C for distillation system,
calling for additional investment.
(iv) Cost of Refrigerant R–134a is much higher than conventional utilities.
Enough care is required to prevent its leakage from the circulation system.
Fixed and variable costs for driving the compressor and extra cost of the heat
pump system should be carefully weighed against savings in normal utilities,
used in the conventional distillation system. Also skilled personnel will be
required for operating the integrated system.
Heat pumps save energy in distillation but at a definite cost. Careful economic
evaluation is therefore necessary before its implementation.
8.9.2 Efficient Operation of Distillation Columns
Existing distillation columns can be operated in such a manner that substantial
energy can be saved.
1. Many distillation columns are operated to excessively purify the products
which might call for high reflux ratio and hence high reboiler duty. This
should be avoided by having adequate I&C. Refer to example given in para 4
of Sec. 8.9.1.
2. Damaged or worn column internals can be costly. Over a period of operation,
the column internals may get damaged. For example, in bubble cap columns,
top nut may loosen and cap may open up. In floating valve trays, valves may
fly off, in packed columns ceramic packings may crumble, scaling on trays
may take place over a period of operation, etc., Due to similar reasons, the
existing columns should be opened for inspection periodically and preventive
maintenance should be carried out. Such preventive maintenance practice will
pay rich dividends in terms of energy saving in operation.
3. Overhead condenser needs periodic cleaning. This will ensure condensation of
overhead vapour at the right pressure and loss, if any, of vapour can be
avoided.
8.9.3 Replace the Distillation Partially or Completely
by New Separation Technique
Replacment of distillation by supercritical extraction or by reactive distillation
can result in significant energy savings. The same is discussed in Secs. 7.7 and
8.7 of this text. Similarly, energy savings can be achieved by replacing
distillation partially or completely by new separation techniques like membrane
separation techniques (reverse osmosis, pervaporation, membrane distillation,
membrane reactor, etc.), molecular sieve adsorption, crystallisation, fixed
adsorptive distillation, etc.
Examples: 1. Conventionally absolute alcohol (ethanol) is separated from
azeotropic mixture of ethanol-water (95.6% ethanol, 4.4% water by mass) by
azeotropic distillation using benzene or cyclohexane as an entrainer. This
conventional method uses three distillation columns and requires very high
energy. One of the new methods for separating the same mixture is by Pressure
Swing Adsorption (PSA) which utilise 3 Å molecular sieves51. Molecular sieve
adsorption consumes less energy and eliminates the use of the entrainer. Benzene
is a proven carcinogen and therefore an air pollutant. For the production of 3400
kg/h of fuel–grade ethanol, installation of molecular sieve—pressure swing
adsorption system instead of azeotropic distillation unit allows a reduction in
energy consumption of boiler from 2257 to 147 kW. But this new technique
requires additional electrical energy in the amount of about 100 kW to drive the
vacuum pumps and other equipments for the PSA unit. The use of benzene or
cyclohexane as entrainer is eliminated.
For the separation of same azeotropic mixture, pervaporation can also be used.
For the production of fuel grade ethanol with the same rate; 3400 kg/h,
installation of pervaporation allows a reduction in energy consumption of boiler
from 2257 to 264 kW. But fresh electrical energy, equivalent to 150 kW is
required to drive the vacuum pumps and other equipments for the pervaporation
unit.
2. Iso-propyl–alcohol (IPA) is yet another application for which pervaporation
can be used. IPA forms an azeotrope with water having composition of 87.9%
(by mass) IPA. Dehydration of IPA azeotrope with zeolite membrane is
commercially carried out.
Table 8.65 Comparison of Pervaporation and Adsorption with Distillation
Parameter Distillation Pervaporation Adsorption

Capacity, m3/d 3 to 30 0.5 to 30 10 to 30

Feed water content, % 5–95% 0.5 to 30 0.05 to 5
Product water content, mg/L < 2000 < 1000 < 100
Operating cost High Low Medium
Capital cost Low Medium Medium
Floor space requirement Large Small Large
Safety Medium High Medium
Environment friendliness Low Excellent Medium
 Ease of installation Medium Easy Medium
Operation Skilled Easy Skilled
Turndown capability Medium Excellent Poor
Start-up and shut-down time Long Short Long
Maintenance requirements High Low High
Availability/Uptime Medium High Medium
Flexibility Limited Excellent Medium

3. Iso-propyl alcohol (IPA) can be concentrated nearly pure IPA by dixed

adsorptive distillation system40.
4. Freeze crystallization is yet another technique which can be used for recovery
of organic species, such as acetic acid, acrylic acid, etc., from very dilute
solutions with 75 to 90% energy reduction as compared to conventional
distillation. Heist57 has given a process flow sheet for 95% recovery of acetic
acid from 1% aqueous solution (by mass) by freeze crystallization. Normally,
1% solution could be considered as effluent with high non-biodegradable COD
and its disposal will be a serious threat to the industry. Recovery will turn
effluent disposal problem into an economical proposal by adoption of freeze
crystallization.

EXERCISES
8.1 Fill the gaps in the following crossword puzzle:
 Fig. 8.68 Crossword Puzzle58 (Reproduced with the Permission of Dr. Ming Tham of School of Chemical
Engineering and Advanced Materials, University of Newcastle upon Tyne, UK)

KEYS:
Down:
1. A conduit that directs liquid from one tray to another.
4. Liquid carried by vapour up to the tray above.
6. This is determined by the amount of material going through the column.
9. Brought about by excessive vapour flow.
14. An alternative to the use of trays or plates.
15. Made-up of the two components.
Across:
2. A liquid mixture which when vaporised produces the same composition as the
liquid.
3. A type of tray used in distillation/absorption.
5. An equipment used to cool vapours, coming off the top of the column.
7. Distillation operations require lots of this.
8. Operation performed by the top section of the column.
10. Product stream taken off the top of the column.
11. Liquid that is fed back to the top of the column.
12. An equipment used to raise vapour.
13. A mixture that is to be separated.
16. Expansion of liquid due to vapour or gas.
17. Operation performed by the bottom section of the column.
18. Ensures that there is sufficient liquid on a distillation tray.
19. Caused by low vapour flow.
20. Product stream taken off bottom of the column.
8.2 Refer to Examples 8.5 and 8.13. Considering the mixture to be an ideal one,
find separate average volatilities for rectification section (α1avg) and for
stripping section (α2avg) for the mixtures. Using these average volatility
values, reflux ratio (R) as 4.2, same feed conditions, calculate number of
theoretical tray requirement for achieving xW = 0.02 by tray to tray
calculations. Specify the feed tray position.
Hint: Use Eqs.   (8.17) and   (8.18) for evaluating liquid and vapour
compositions on trays of both sections, considering yD = 0.9302 and xW =
0.02.
8.3 Distillation column is used to separate aqueous solution of dimethyl
formamide (DMF). Determine the following based on given data:
(a) Minimum reflux ratio.
(b) Number of theoretical stages required for desired separation for R = 2 Rm.
(c) Overhead condenser duty.
(d) Reboiler duty.
Data:
(i) Feed composition: 20% by mass DMF, Molar mass of DMF = 73.1 kg/kmol
Feed is saturated liquid. Feed flow rate = 100 kmol/h
(ii) Distillate composition: 99% water (by mass).
(iii) Residue composition: 99% DMF (by mass).
(iv) Operating pressure at top: 650 Torr vacuum or 110 Torr a.
(v) For vapour pressure data refer Table 8.67.
(vi) Properties of DMF
Latent heat of vaporisation λ = 576.85 kJ/kg at the topmost temperature of
distillation column
Specific heat CL = 2 kJ/(kg · °C)
Table 8.67 Vapour Pressure Data of DMF and Water
Temperature °C pv of DMF Torr pv of water Torr

53.7 18 110
60 26 149.38
70 42 233.7
80 65 355.1
90 100 525.76
92 110 567
100 165 760

8.4 NRTL parameters for dimethyl formamide (1) and water (2) are as follows:
τ12 = b12/RT, τ21 = b21/RT, R = 8.314 kJ/(kmol · K)
b12 = –267.667, b21 = 470.31

ln γ1 = x 22   (8.127)

In γ2 = x12   (8.128)

where, G12 = exp (–α τ12), G21 = exp (–α τ21)

α = 0.277
Solve Exercise 8.3 considering non-ideality of DMF-water system. To find the
VLE data following equation can be used:
y = x  (8.3)

y, x = Mole fraction of more volatile component (water) in vapour and liquid

phase, respectively.
γ = Activity coefficient of water in solution.
pt = Operating pressure = 110 Torr a
8.5 In the production of chloromethanes, a saturated vapour mixture consisting
of 60% methyl chloride, 28% methylene dichloride, 9% chloroform and 3%
carbon tetra- chloride (by mass), is sent to distillation section for the
separation of pure products. In first distillation column 99.9% (by mass) pure
methyl chloride is separated as top product. Residue contains 0.2 kmol/h
 methyl chloride. Determine the following for this distillation column:
(a) Operating pressure of distillation column.
(b) Minimum reflux ratio by Underwood’s method.
(c) For R = 3 Rm calculate the number of theoretical stages required for desired
separation.
(d) Determine the tower diameter of distillation column, selecting sieve tray
tower for the same.
Feed is a saturated vapour mixture and its flow rate is 5000 kg/h.
8.6 Refer to Example 8.16. Consider that total distillate at the end of distillation
should contain 99% by mole (min.) methanol. Considering a constant reflux
ratio (R) of 4, same feed conditions and 6 hours of distillation period,
calculate the quantities of distillate and residue with their compositions. Make
calculations for N = 2 and 10.
8.7 Determine the minimum reflux ratio and number of theoretical stages
required for reflux ratio, R = 2 Rm, for the following given separation by
(a) Ponchon-Saverit method based on given H-x-y data and equilibrium x vs y
data, given in Table 8.68.
(b) McCabe–Thiele method using phase equilibrium data x vs y, given in Table
8.68.
(c) McCabe–Thiele method by using vapour–liquid equilibrium data x vs y,
which can be derived by using modified Raoult’s law.
yi pt = γi pvi xi  (8.3)
where, γi = Activity coefficient can be determined by van Laar equations. van
Laar constants A = 1.978, B = 1.401.
In Table 8.68, “1” stands for ethanol and “2” for benzene.
Data:
(i) Feed: 100 kmol/h contains 10% ethanol and 90% benzene (by mole)
Feed is saturated liquid.
(ii) Distillate: It contains 44% ethanol (by mole).
(iii) Residue: 99.9% benzene (by mole)
Table 8.68 VLE Data and Enthalpy Data for Ethanol-Benzene System at 760 Torr
Mole fraction of ethanol Temperature Enthalpy of mixture, kJ/kmol
x1 y1 * t, °C HL HG

0 0 79.7 –15 870 19 090

0.01 0.075 78.0 –15 910 18 380
0.023 0.150 76.0 –16 120 17 580
0.045 0.217 74.0 –16 410 17 170
 0.095 0.305 71.0 –17 170 16 540
0.200 0.395 69.0 –18 630 15 910
0.300 0.425 68.4 –20 050 15 740
0.447 0.447 68.2 –22 150 15 570

(Reference states: Enthalpies of ethanol and benzene in ideal gas state at 0 K = 0 kJ/kmol)

8.8 1000 kg of feed, containing 30% by mass ethylene glycol and 70% by mass
water, is to be separated in batch distillation with rectification column.
Bottom product must contain 98.8% by mass of ethylene glycol. Composition
of distillate should remain constant throughout the batch distillation. Desired
composition of distillate is 98% by mass water.
(a) Design the rectification column.
(b) Determine the operating (actual) reflux ratio vs time data.
(c) Compare the resulting data of this example (batch distillation with
rectification with the resulting data of Example 8.14 (simple batch
distillation).
Operating pressure in rectification column is 30.4 kPa a (maximum). Vapour
liquid equilibrium data for ethylene glycol-water system at 30.4 kPa a are
given in Example 8.14.
8.9 Refer to Exercise 10.5 relating to chlorination of benzene. At the end of
reaction, the product mixture is expected to be 10.63 kmol and will have 13%
benzene, 71.3% monochlorobenzene and 15.7% dischlorobenzene. It is
desired to separate 99% pure benzene as distillate by batch distillation with
rectification. Final residue will contain 1% benzene. Total time considered for
distillation is 4 h. Determine the reflux ratio (R) vs time data (θ). All
concentrations are mole %
Table 8.69 Antoine Equation Constants
Antoine Constants
Component Normal boiling point at 101.325 kPa, °C
A B C

Benzene 4.725 83 1660.652 –1.641 80.15

Monochlorobenzene 4.110.83 1435.675 –55.124 131.75
Dichlorobenzene 4.122 90 1575.110 –64.637 174.06

Antoine equation: log pv = A –

where, pv = Vapour pressure, bar

T = Saturation temperature, K
8.10 Chloroform and methanol form a minimum boiling azeotrope at 101.325
kPa and 53.5°C containing 65.17 mole % chloroform. Using van Laar
 equations, generate t-x-y relationship and plot y vs. x diagram.
Table 8.70 Antoine Equation Constants14
Antoine constants Normal boiling point at 101.325 kPa, °C
Component
A B C

Chloroform 14.5014 2938.55 –36.9972 61.15

Methanol 16.4948 3593.39 –35.2249 64.65

Antoine equation: ln pv = A –

where, pv = Vapour pressure, kPa

T = Saturation temperature, K
8.11 Formic acid and water form a maximum boiling azeotrope at 101.325 kPa
and 107.1°C, containing 43.3 mole % water. Using van Laar equations
generate t-x-y relationship and plot y vs. x diagram.
Table 8.71 Antoine Equation Constants14
Antoine Constants
Component Normal boiling point at 101.325 kPa, °C
A B C

Formic acid 6.7067 1699.20 –12.45 100.8

Water 7.0436 1636.90 – 48.25 100.0

Antoine equation: log pv = A –

where, pv = Vapour pressure, kPa

T = Saturation temperature, K
8.12 In Example 8.23, refrigerant R–134a was selected. Consider use of
refrigerant R–22 (chlorodifluoromethane) or refrigerant R–717 (ammonia) as
alternate auxiliary heat transfer mediums. Properties of R–22 and R–717 are
given below.
Table 8.72 Properties of R–2227
Enthalpy, kJ/kg
Temperature °C Pressure, bar a Vapour Cp/Cv
Liquid Latent Vapour

2.0 5.315 (satd.) 202.36 202.65 405.01 1.296

72.5 31.51 (satd.) 296.44 118.15 414.59 2.147
130.0 32.0 — — — 478.7 —

(Reference state: Enthalpy of saturated liquid at 0°C = 200 kJ/kg)

Assume 10°C approach in the overhead condenser and reboiler for R–22.
 Table 8.73 Properties of R–71727
Enthalpy, kJ/kg Vapour
Temperature °C Pressure, bar a
Cp/Cv
Liquid Latent Vapour

7.0 5.54 (satd.) 232.55 1236.31 1468.86 1.415

72.5 35.06 (satd.) 558.85 923.10 1481.95 1.834
130.0 35.50 — — — 1692.34 —

(Reference state: Enthalpy of saturated ammonia liquid at 0°C = 200 kJ/kg)

Ammonia evaporator is normally a flooded one. Therefore, assume approach of
5°C in the overhead condenser and approach of 10°C in the reboiler.
Calculate required circulation rates of R–22 and R–717 for the same heat duties.
Also review the performances of all three auxiliary heat transfer mediums and
critically evaluate system requirements, including environment friendliness.
8.13 Extractive distillation of 50 mass % aqueous isopropyl alcohol (IPA) is to
be carried out with the help of aqueous 35 mass % CaCl2 solution. Vapour
pressure data of aqueous calcium chloride solution are given in Table 8.58.
Calculate t-x-y data for the system and plot the equilibrium (x, y) curve.
8.14 In a refinery, a gas stream, having composition 6.84% ethane, 40.81%
propane, 9.70% i-butane and 42.65% n-butane (by mole), is to be fractionated
at a rate of 300 kmol/h in a column to separate LPG fraction59. Feed gas is
compressed, cooled and flashed in the fractionating column at 9.7 bar a and
45°C. Overhead vapours from the column are taken to partial condenser from
which C2 and C3-fractions are separated in gaseous form. Composition of the
gaseous stream from the condenser is 28.25% ethane, 68.55% propane, 1.6%
i-butane and 1.6% n-butane (by mole). Flow rate of overhead vapour is 68.52
kmol/h. It leaves the condenser at 9.4 bar a and 20°C. Saturated liquid from
the condenser at 20°C is recycled as reflux to the column at the rate of 1.5
kmol/kmol vapours separated. Pressure drop in the overhead condenser may
be considered negligible (< 10 kPa) for calculation purpose. Bottom residue
(LPG fraction) is removed from the reboiler at 10 bar a. Use Fig. 6.20 for dew
point and bubble point calculations.
(a) Make complete material balance calculations of the column.
(b) Determine minimum reflux ratio.
(c) Determine number of theoretical stages for the reflux ratio 1.5 kmol/kmol
and the stage at which the feed should be introduced by FUG method.
(d) In Tables 8.74 and 8.75, enthalpies of liquid and gaseous hydrocarbons are
given. Using these values, calculate the heat duties of the overhead
 condenser and the reboiler.
(e) Considering low temperature of the vapours (20°C), leaving the overhead
condenser, use of ammonia (R-717) is proposed as an auxiliary heat
transfer medium in the overhead condenser and the reboiler. Evaporation of
ammonia at 15°C in the overhead condenser and condensation of ammonia
at 65°C in the reboiler are fixed as design parameters. Consider conditions
of ammonia at the compressor discharge as 30 bar a and 125°C. Properties
of ammonia are given in Table 8.76. Calculate ammonia circulation rates,
required in the overhead condenser and the reboiler. Comment on the
results.
Table 8.74 Enthalpy of Liquid Hydrocarbons at Saturated Conditions60 (in kJ/kmol)
Temperature, K Ethane Propane i-Butane n-Butane

290 304 1918 2521 2525

300 1840 3112 3939 3941
305.34c 4154
310 — 4353 5397 5391
320 — 5650 6896 6878
330 — 7015 8442 8404

cCritical temperature of ethane

Reference state: Enthalpy of saturated liquid hydrocarbon at 0°C = 0 kJ/kmol

Table 8.75 Enthalpy of Gaseous Hydrocarbons60 (in kJ/kmol)

Hydrocarbon Temp., K Pressure, bar a

Ethane 1 2 3

280 11 700 11 630 11 560

300 12 730 12 680 12 620
320 13 820 13 770 13 720
330 14 380 14 330 14 280

Propane 5 6 8

290 17 785 17 615 —

300 18 585 18 425 18 105
320 20 205 20 085 19 825
340 21 885 21 785 21 565

i-Butane 0.1 0.5 1.0

280 21 935 21 835 21 695

300 23 825 23 745 23 635
320 25 825 25 755 25 665
340 27 935 27 875 27 805

n-Butane 0.1 0.5 1.0 2.0 3.0

 280 23 545 23 425 22 845 — —
300 25 475 25 385 25 255 24 505 —
305 — — — — 25 245
320 27 515 27 435 27 325 27 105 26 875
340 29 655 29 575 29 485 29 305 29 105

(Reference state: Enthalpy of saturated liquid hydrocarbon at 0°C = 0 kJ/kmol)

Table 8.76 Properties of Ammonia (R-717)26
Enthalpy, kJ/kg
Temperature, °C Pressure bar a Condition Cp/Cv
Liquid Vapour

15 7.286 Saturated 270.2 1475.9 1.438

65 29.476 Saturated 518.6 1487.1 1.748
125 30.0 Superheated — 1695.0 —

(Reference state: Enthalpy of saturated liquid ammonia at 0°C = 200 kJ/kg)

8.15 Tetrahydrofuran (also known as 1, 4-epoxy butane; formula: C4H8O) is
widely used as a solvent and also as a reaction medium in chemical industry.
Typical THF recovery method involves adsorption on activated carbon,
followed by steam stripping of the bed. Resulting THF-water mixture is a
dilute aqueous stream of THF which is required to be separated in pure
components for reuse of THF.
THF forms a minimum boiling azeotrope at 63.5°C, containing 5.3% by mass of
water at atmospheric pressure. For the separation of THF from the mixture,
pressure swing distillation process61 is used. Basic flow sheet of the process is
shown in Fig. 8.69.
In the first column, THF-water mixture, containing 0.06 mole fraction THF is
fed. Distillate D1 from the top has near azeotropic composition (0.805 mole
fraction THF). Reflux ratio of 0.26 kmol/kmol distillate is maintained in this
column. Bottom product of the first column is pure water (in which THF
content is 1 × 10–6 mole fraction).
Distillate D1 is fed to the second column, operating at 7.908 bar a. At this
pressure, THF-water mixture forms a minimum boiling azeotrope at 136°C,
containing 12% by mass water. Distillate D2 from the top has near azeotropic
composition (0.65 mole fraction THF) and is recycled to the first column.
Bottom product from the second column is pure THF (in which water content
is 1 × 10–5 mole fraction). Reflux ratio in the second column is maintained at
0.65 kmol/kmol distillate.
 Fig. 8.69 Pressure Swing Distillation of THF - Water

Note: All the concentrations are expressed as mole fractions.

Assume that van Laar correlations represent non-ideality of THF-water mixtures
in both the columns. Carry out tray-to-tray calculations for both the columns
and fix feed locations of F and D1 for fresh feed rate of 1000 kmol/h.
Table 8.77 Properties of THF and Water
THF Water

Normal boiling point, °C 66 100

Freezing point, °C –108.5 0
Density, kg/L 0.889 1.0
Temperature range, °C
Antoine constants
60–150 60–70 70–80 80–90 90–100 100–150

A 6.995 15 7.1365 7.1095 7.0883 7.0733 7.0436

B 1202.29 1695.2 1678.9 1665.9 1656.4 1636.9
C 226.25 – 42.75 – 44.15 – 45.35 – 46.25 – 48.25
Average latent heat of vaporisation, λ, kJ/kmol 27 901 41 453

Antoine equations: log pv = A – for THF

log pv ′ = A – for water

where, pv = Vapour pressure of THF, Torr

pv ′ = Vapour pressure water, kPa
t = Saturation temperature, °C
T = Saturation temperature, K
8.16 A plant employs a batch reactor with refluxing arrangement (with cooling
water) to manufacture butylated amino resins. One of the products is
melamine-formaldehyde (MF) resin. Melamine, aqueous formaldehyde (37%
by mass) and n-butanol are the chief reactants. Phosphoric acid (98% pure,
0.5% by mass) is used as a catalyst.
In a typical batch of 4 kL, average consumptions for a specific grade of MF resin
are reported to be 0.6 t n-butanol, 0.4 t melamine and 0.3 t formaldehyde
(100%) per t MF (100% strength) product. Etherification and condensation
reactions take place in the reactor at atmospheric pressure and in the
temperature range of 90 to 105°C. Stoichiometry of above reaction
mechanism reveals that one mole of water is produced per one mole of
melamine condensed (reacted) and one mole of n-butanol reacted.
Product water and water of aqueous formaldehyde solution are removed from
the reactor in the form of an azeotrope with n-butanol. Condensed azeotrope
separates in two layers; compositions of both the layers depend on the
 temperature. Butanol rich layer is recycled and aqueous layer at 40°C is
collected in a separate tank for recovery of n-butanol by distillation. In the
batch, 100% excess n-butanol is used for completion of reaction. At the end
of reaction, excess n-butanol is recovered by distillation. Batch is normally
completed in 8 to 10 h.
Product contains up to 1.25% formaldehyde and 2 to 2.5% n-butanol (by mass).
Since the reaction takes longer time, substantial crosslinking takes place
which is more than desired.
 Fig. 8.70 Typical Polymer Chain

Molar mass of polymer = 324.3n + 30.03a + 56.16b + 18.02

Total water Produced = (3n + b –1) 18.02
It is decided to adopt a continuous system for production of MF resin as the
product demand has shot up. For this purpose, process research was carried
out in the laboratory. Use of strong cation exchange resin (up to 4% by mass),
suitable for continuous operation at 140°C, was made as a catalyst instead of
phosphoric acid. Results of the 20-litre flask showed that while consumptions
of the reactants were nearly same as that of batch process (for the same
grade), better quality product was produced in 3 h. MF resin so produced had
less crosslinking and low formaldehyde (< 0.5%) content. Further
experiments with a glass column, packed with glass Raschig rings, confirmed
that 3 theoretical stages for stripping are required for achieving bottom
product (MF resin) with less than 1% n-butanol (by mass). Data on n-butanol-
water azeotrope are covered in Sec. 8.8. Heterogeneous liquid from overhead
condenser separates in two layers. Mutual solubility data are given in Table
8.78.
Table 8.78 Mutual Solubility of Water-n-Butanol mixture62
Temperature in separator, °C mass % n-Butanol

Upper layer Lower layer

20 79.93 7.81
30 79.38 7.08
40 78.51 6.60
50 77.58 6.46
60 76.38 6.52
80 73.53 6.89

Consider: (i) Excess n-butanol will be just sufficient to remove water in the form
of the azeotrope, and (ii) Residence time of reaction mass in the distillation
column in contact with the resin will be 3 h.
Based on above information, develop a process flow diagram for the reactive
distillation system for continuous production of MF resin along with desired
instrument and control system with technical reasons. Consider whether pilot
plant studies are required to implement the project commercially. Also study
the advantages and disadvantages/limitations of proposed reactive distillation
system over the batch system.

REFERENCES
1. Treybal, R. E., Mass Transfer Operations, 3rd Ed., McGraw-Hill, USA, 1980.
2. Ludwig, E. E., Applied Process Design for Chemical and Petrochemical
Plants, Vol. 2, 3rd Ed., Gulf Publishing Co., USA, 2002.
3. Smith, B. D., Design of Equilibrium Stage Processes, McGraw-Hill, USA,
1963.
4. Gmehling and Onken, DECHEMA Chemistry Data Series, DECHEMA,
Germany Vol.1, 1977.
5. Chu, J. C. Wang, Levy, S. L. and Pual, R., Vapour—Liquid Equilibrium Data,
Edwards Inc., USA, 1956.
6. Hala, E., Wichterle, I. and Lineks, J., Vapour—Liquid Equilibrium Data
Bibliography, Elsevier, The Netherlands, 1976.
7. Hala, E., Wichterle, I., Polak, J. and Boublik, T., Vapour-Liquid Equilibrium
Data at Normal Pressure, Pergamon, UK, 1968.
8. Perry, R. H. and Green D., Perry’s Chemical Engineer’s Handbook, 7th Ed.,
McGraw-Hill, USA, 1998.
9. Gmehling, J., Azeotropic Data, 3 Vol., Wiley-VCH, Germany, 2004.
10. Sinnott, R. K. Coulson and Richardson’s Chemical Engineering, Vol. 6,
Revised 2nd Ed., Asian Books Pvt. Ltd., 1988.
11. McCade, W. L., Smith, J. C., Harriott, P., Unit Operation of Chemical
Engineering, 7th Ed., McGraw-Hill, USA, 2005.
12. King, C. J., Separation Processes, 2nd Ed., McGraw-Hill, USA, 1980.
13. Deshpande, P. B., Distillation Dynamics and Control, Arnold, USA, 1985.
14. Bhatt, B. I. and S. M. Vora, Stoichiometry, 4th Ed., McGraw-Hill Education
(India), 2004.
15. Kroschwitz, J. I. (Ed.) Kirk and Othmer’s Encyclopedia of Chemical
Technology, 4th Ed., Vol. 5, John Wiley, USA, 1993, p. 175.
16. Bennett, D. L. and K. W. Kovak, Chem. Engg. Progress, 96 (5), 2000, p. 20.
17. Kister, H. Z., Distillation Design, McGraw-Hill, USA, 1992.
18. Fair J. R., Chem. Engg., 70 (14), July 8, 1963, p. 119.
19. Van Winkle, M., Distillation, McGraw-Hill, 1967.
20. Billet, R., Distillation Engineering, Chemical Publishing Co., USA, 1979.
21. Mukherjee, S., Chem. Engg., 112 (9), 2005, p. 52.
22. AIChE Bubble Cap Design Manual, American Institute of Chemical
Engineers, USA, 1958.
23. Lockett, M. J., Distillation Tray Fundamentals, Cambridge University Press,
UK, 1986.
24. Koppel, P. M. Chem. Engg. 79(23), 1972, p. 102.
25. Fischer, W., Seminar on “Short Path Distillation”, Mumbai, 1996.
26. Kukla, N., Chemical Plant Processing, 1996, p. 74.
27. Bhatt, B. I., Design Databook: Properties of Steam, Selected Refrigerants’ n-
Hexane and Brines, CBS Publishers and Distributors, New Delhi, 2007.
28. Salden, D. M., The Chemical Engineer Supplement, 1987, p. 17.
29. Degarmo J. L., V. N. Parwleker, V. Pinjala, Chem. Engg., Progr., 86 (3),
1992, p. 45.
30. Mahajani S. M., A. K. Kolah, Ind. Engg. Chem. Res., 35, 1996, p. 4587.
31. Groggins, P.H., Unit Process in Organic Synthesis, 5th Ed., McGraw-Hill,
USA, 1958, p. 731.
32. Jacobs R., R. Krishna., Ind. Engg. Chem. Res., 32, 1993, p. 1706.
33. Rodriguez, M. C., J. G. S. Harnandez and A. B. Petriciolet, Proceedings of
the 5th International Symposium on Design, Operation and Control of
Chemical Processes, 2010, PSE ASIA Conference, p. 1083.
34. Lee, J. W., S. Hauan, K. M Lien and A. W. Westerberg, Proceedings:
Mathematical, Physical and Engineering Sciences, Vol, 456, 2000, Part I on
p. 1953 and Part II on p. 1965.
35. Sundmacher, K. And A. Kienle, Reactive Distillation, Wiley-VCH,
Germany, 2006.
36. Snessby M. G., Tade M. J., Datta R., Smith T. N., Ind. Engg. Chem. Res., 36,
1997 p. 1855.
37. Bhanvase, B. A., Y. R. Soman and R. V. Naik, Chem. Engg. World, 41 (2), p.
43, 2006.
38. Coates, J., Chem. Engg., 67 (10), May 16, 1960, p. 121.
39. Ruether, J. A. and B. C. -Y. Lu, Can. J. Chem. Engg., 50, April 1972, p. 266.
40. Mujiburhman, M., W. B. Sediawan and H. Sulistyo, Separation and
Purification Technology, 48, 2006, p. 85.
41. Norman, W. S., Absorption, Distillation and Cooling Towers, Longmans,
USA, 1961.
42. Cho, J. and Jong-Ki Jeon, Korean J. Chem. Engg., 23 (1), 2006, p. 1.
43. Mcketta, J. J. and W.A. Cunningham, Encyclopedia of Chemical Processing
and Design, Vol. 5, Marcel Dekker, Inc., USA, 1977, p. 257.
44. Private Communication with Dr. Jen-Uwe Repke, TU Berlin, Germany.
45. Repke, Jen-Uwe and A. Klein, Proceedings of the 15th European Symposium
on Computer Aided Process Engineering, (Ed. L. Puigianer and A. Espuna),
Elsevier Science B.V., 2005.
46. Repke, Jen-Uwe, F. Forner and A. Klein, Chem. Engg. & Tech., 28 (10),
 2005, p. 1151.
47. Knapp, P. J. and M. F. Doherty, Ind. Engg. Chem., Res., 31, 1992, p. 346.
48. Levy, S. G., D. B. Van Dongen and M. F. Doherty, Ind. Engg. Chem.
Fundam., 24, 1985, p. 463.
49. International Critical Tables, 1st Ed., Vol. 3, McGraw-Hill, USA, 1928, p.
292.
50. Kroschwitz, J. I. (Ed.) Kirk and Othmer’s Encyclopedia of Chemical
Technology, Vol. 17, 4th Edition, John Wiley, USA, 1992, p. 39.
51. Humphrey, J. L. and Seibert, A. F., Chem. Engg. Progr., 88 (3), March, 1992,
p. 32.
52. Finn, A. J., Chem. Engg. Progr., 89 (10), October, 1993, p. 41.
53. H. Becker, S. Godorr, H. Kreis and J. Vaughan, Chem. Engg., 108 (1),
January 2001, p. 68.
54. Meszaros, I. and A. Meili, Hydrocarbon Processing, 81 (3), March 2002, p.
51,
55. Meili, A. and A. Stuecheli, Chem. Engg., 94 (2), February 16, 1987, p. 133.
56. Meili, A. and A. Stuecheli, Chem. Engg. Progr., 89 (5), 1993, p. 49.
57. Heist, J. A., Chem. Engng., 86 (10), May 7, 1979, p. 72.
58. Private Communication with Dr. Ming Tham, University of Newcastle upon
Tyne, UK.
59. Kaiser V. and J. P. Gourlia, Chem. Engg., 92 (17), August 19, 1985, p. 45.
60. Younglove, B. A. and J. K. Ely, J. Phy. Chem. Ref. Data, 16 (4), 1987, p.
577.
61. Samir, I., Abu-Elshah and W. L. Luyber, Ind. Engg. Chem., Process Des.
Dev., 24 (1), 1985, p. 132.
62. International Critical Tables, Vol. 3, McGraw-Hill, USA, 1928, p. 388.
 9
Process Design of Absorbers
9.1 INTRODUCTION
Absorption is an important unit operation of chemical process industries. In
absorption, a gas mixture (or a gas-vapour mixture) is contacted with a suitable
liquid, called solvent, to separate one or more components of the gas mixture by
dissolving them in the liquid. Hence, after the absorption, components of gas
mixture are transferred to the solution. Dissolved component in the solution can
be separated by distillation or stripping (reverse of absorption). For example, in
the manufacturing of synthesis gas for ammonia from the gas mixture containing
nitrogen, hydrogen, carbon dioxide and traces of other components, carbon
dioxide is removed or separated by absorption in ethanol amine solution. Then
carbon dioxide from the resulting solution is separated by stripping. “Absorber”
and “scrubber” are sometimes used as synonyms. Truly speaking, if a component
of the gas mixture, transferred to liquid, is a gas at operating conditions of mass
transfer, then it is called absorption. If a component transferred is vapour at
operating conditions then it is called scrubbing. Below critical temperature, a gas
is known as vapour (which could be saturated or superheated). For example,
separation of acetone vapour from air. Acetone vapour mixture in contact with
water as solvent is called scrubbing. Operating conditions of this scrubber is
atmospheric pressure and ambient temperature. Critical temperature of acetone
(Tc = 508.1 K) is well above the ambient temperature. Absorption or scrubbing
is chiefly carried out in a packed tower, although plate (tray) columns are not
uncommon. Various types of equipment used as absorbers or scrubbers are:
1. Packed tower
2. Tray tower
3. Falling film absorber
4. Spray chamber
5. Venturi scrubber
6. Stirred tank

9.2 CRITERIA OF SELECTION

9.2.1 Use of Packed Tower
Packed towers are used for all mass transfer operations. In Chapter 7, utility of a
packed tower for extraction was discussed and clarified that in extraction
columns the mass transfer area is almost independent of the packing surface, but
is directly linked to the drop phase hold-up in the tower. Contrary to this flow
path process, a packed tower provides surface for mass transfer for distillation
and absorption operations. Both random and structured packings are used for the
latter operations. However, internals are some what different than those for
extraction service. For example, disperser and coalescer play important role in
extraction but they are not needed in other two operations. Instead, a demister
pad at the top to catch entrainment is an important internal for two operations. In
majority of absorption applications, packed towers are preferred possibly for the
following reasons:
1. Tower diameter required for the majority of the applications is less than 0.75
m. Packed tower is generally cheaper than tray tower when tower diameter is
less than 0.75 m.
2. Many absorption or scrubbing systems are corrosive systems. For corrosive
systems plastic packed towers (for example FRP packed towers) with plastic
packing are cheaper than metallic tray towers. For corrosive systems, use of
graphite heat exchangers is increasingly considered in which gas and solvent
flow in co-current fashion.
3. In majority of absorption with chemical reaction systems, overall rate of gas-
liquid reaction is totally controlled by rate of mass transfer (i.e., rate of
chemical reaction > > rate of mass transfer). Packed tower provides higher
rate of mass transfer due to higher interfacial area of contact compared to
other equipments.
4. For the liquids having high foaming tendency and also for a heat sensitive
liquid, packed tower is a better choice than tray tower.
5. For removing obnoxious gases from the vent gases or from the flue gases by
absorption, available pressure of feed gas, fed to absorber or scrubber, is
usually very low; close to atmospheric pressure. Hence, for such a case,
pressure drop in absorber is the main criteria for selecting the equipment.
Packed tower provides less pressure drop compared to a tray tower. Hence,
for such applications, packed towers are preferred.
9.2.2 Use of Spray Tower
For removing sulphur dioxide (SO2) from boiler flue gases in large scale plants,
available pressure of flue gases is very low; close to atmospheric pressure. For
this case, even packed tower is not suitable because pressure drop offered by
packed tower is normally greater than 100 mm WC. Normally for such an
application spray tower or spray chamber is used. Boiler flue gases are also
contaminated by fine carbon particles, if coal is used as a fuel. This favours the
selection of spray tower over packed tower. Alternatively, a venturi scrubber can
be used which induces vacuum in the suction.
9.2.3 Use of Tray Tower
Three different types of tray tower were discussed in Section 8.4.6. These tray
towers can also be used for absorption service.
Tray tower is selected as absorber in the following cases:
1. In case of exothermic absorption, it can be improved by providing external
cooling during absorption. Cooling coils can be more readily built into tray
towers and liquid can readily be removed from the trays to be passed through
coolers and returned, than from packed towers. Take, for example, absorption
of acetaldehyde from acetaldehyde vapour-gas mixture by using ethanol-
water solution as solvent. This absorption is exothermic and hence tray tower
with intermediate coolers is used as absorber. Similarly, absorption of
formaldehyde in water is exothermic which can be conveniently carried out in
a plate column. Refer to Example 9.6 and Figs. 4.2 and 4.8. Absorption of
carbon dioxide in ammonia plant is preferably carried out in a tray column
which is exothermic in nature, although packed columns are also in use.
2. For the absorption systems involving low liquid flow rate, tray towers are
preferred to get the better gas-liquid contacting efficiency. Low liquid rate
may lead to incomplete wetting of column packing in case of packed tower
and can give lower contacting efficiency.
3. If required tower diameter is large, say more than 1.5 m, then uniform liquid
distribution is difficult in a packed tower. Hence, in such a case, tray tower is
preferred.
4. If solids are present in liquid or gas to a limited extent then tray towers are
preferred because they permit easier cleaning. In practice, solids are getting
accumulated in packed tower and result in increasing the pressure drop. For
example, absorption of chlorine in aqueous caustic soda solution, fibre glass
reinforced plastic (FRP) tray tower is preferred over FRP packed tower. Use
of packed tower for this application restricts the conversion of NaOH to NaCl
and NaOCl. Once the concentration of salts in the circulating solution
increases above the limiting value then fine suspended solids may appear in
 the solution. Hence, with the packed tower, solution bled from the circulating
system contains unconverted NaOH. Use of tray tower permits complete
conversion of NaOH in the circulating solution.
High Speed Mass Transfer Trays For absorption service under high pressure,
high speed mass transfer trays are developed by many suppliers. For example,
dehydration of natural gas with ethylene glycol solution, sweetening of gases,
rectification of hydrocarbon mixtures, removal of noxious gases by scrubbing,
etc., require intense mass transfer with high liquid loading and high level of
flexibility. For such applications, high speed trays are found economical as the
tray column becomes compact in terms of diameter and height.
One such example of high speed tray is shown in Fig. 9.1.
 Fig. 9.1 High Speed Tray
(Courtesy: HAT International Ltd., U. K.)

In this design, swirl-type cyclone elements are fixed into the opening of the
plate. The element consists of a vertical cylinder with annular row of holes in the
lower section. The axial swirler, firmly attached to the inside of the cylinder has
inclined vanes in upper part. Gas from bottom induces liquid from the holes,
travels through the swirl element and enters the cyclone separator where liquid
separates from the froth mixture and falls back on the tray. Turbulence created in
the swirl element can result in high mass transfer efficiency, reducing the tower
diameter and column height. Needless to say, such trays offer high pressure drop
as compared to conventional trays. Efficient separation of liquid also helps in
reducing entrainment. Also sensitivity to weeping is reduced.
9.2.4 Use of Falling Film Absorber
For highly exothermic absorption, e.g., absorption of hydrogen chloride gas in
water and absorption of ammonia gas in water, more efficient heat removal
system is required to get the higher efficiency and higher concentration of
solution. For such applications, falling film absorbers are used. This could be a
shell and tube heat exchanger, in which absorption is carried out inside the tubes
while cooling medium is circulated on shell side. In a heat exchanger, made of
graphite blocks, absorption is carried out inside the vertical tube (path) while
cooling medium is passed through horizontal holes (perpendicular to vertical
tube path). Design of a falling film absorber is dealt later in this chapter
(Example 9.5).
9.2.5 Use of Venturi Scrubbers
Venturi scrubbers are normally preferred for removing particulate matter from a
gas stream as opposed to absorbing soluble vapour from vapour-gas mixture.
Venturi scrubber provides higher gas-liquid contact or higher absorption
efficiency but at the expense of relatively large liquid side pressure drop and
consequently it requires higher power consumption.
9.2.6 Use of Stirred Tanks and Sparged Towers
Stirred tanks and sparged towers are preferred for slow gas-liquid reactions.
Normally, in cases of absorption with chemical reaction, solvents or liquid
reactants are selected in such a way that they offer fast or very fast chemical
reaction with the solute gas. Hence, these equipments are normally not used as
absorbers.

9.3 DESIGN OF PACKED TOWER-TYPE

ABSORBER
9.3.1 Packings
Before attempting the process design of the packed tower, characteristics of
various packings be understood. Packings can be divided into two major groups;
random (or dumped) packing and structured packing. Random packings are
either cylindrical or half cylinder (saddle) shaped. Some known packing shapes
are shown in Fig. 9.2.

Fig. 9.2 Random Packings

Both rings and saddles provide surface area within the structure, resist the
deformation and prevent from nesting when dumped. Within the packing types,
packing elements are available in different sizes and material of construction.
Elements fabricated from a given material look essentially same except their
sizes. This implies that they are geometrically similar and differ only by a
characteristic dimension; known as nominal packing size (δp). Mass transfer of a
packing is significantly impacted by the specific area of the packing (at) which is
inversely proportional to δp. Data on various random packings are given in Table
9.1. Packing factor Fp is the key to the pressure drop calculations and is
determined experimental. For example, Fp for metal Raschig rings of 25 mm
nominal size (δp) is 450 m–1 while that of 25 mm metal Pall rings is 157 m–1,
less than half. This is attributable to the openness of the Pall rings compared to
Raschig rings. This can also be inferred from % void (ε) which is 86% for
Raschig rings and 94% Pall rings. Correlations are reported in literature4 for
calculating at and Fp, knowing δp, but one should rely more on published data of
the supplier or generalised data (similar to those given in Table 9.1).
Pall rings are the preferred and commonly used random packings but their cost
per unit volume is high. Pall rings require minimum diameter and minimum
height for the given absorption or distillation duty as compared to other types of
random packings. Pall rings are available in metals and plastics. For high
temperature services polyethylene or polypropylene Pall rings cannot be used.
Plastic packings are also attacked by some organic solvents. For corrosive
services ceramic Berl saddles are cheaper than metallic Pall rings. For example,
in distillation of the mixture of phosphorous trichloride and phosphorous
oxychloride, packed tower with ceramic packing is used. Ceramics are brittle at
high temperature and have poor strength. For very high temperature and
corrosive services glass packed columns with glass Raschig rings are preferred
against packed column of special alloy. For example, in vacuum distillation (for
extractive distillation of concentrated nitric acid using concentrated sulphuric
acid as solvent) it is carried out in glass packed tower with glass Raschig rings.
Raschig rings are the cheapest packing material but are less efficient. Raschig
rings require the maximum diameter and height for the given absorption or
distillation
Table 9.1 Characteristics of Dumped (Randomly Packed) Tower Packings3
Nominal Wall Outside Approximate Approximate Percent Packing
Approximate No.
Packing type size δp, thickness, diameter & mass per m3, surface area, at, void factor Fp,
of elements per m3
mm mm length, mm kg m2/m3 space, ε m–1
Berl saddles, 6 — — 3.78 × 106 900 900 60 2950
ceramic 13 — — 590 000 865 465 62 790
25 — — 77 000 720 250 68 360
38 — — 22 800 640 150 71 215
50 — — 8 800 625 105 72 150
Intalox saddles, 6 — — 4.15 × 106 865 984 75 2380
ceramic 13 — — 730 000 720 625 78 660
25 — — 84 000 705 255 77 320
38 — — 25 000 670 195 80 170
50 — — 9 400 760 118 79 130
75 — — 1 870 590 92 80 70
Intalox saddles, (No. 25) — — 168 400 — — 97 135
metal (No. 40) — — 50 100 — — 97 82
(No. 50) — — 14 700 — — 98 52
(No. 70) — — 4 630 98 43
Intalox saddles, 25 — — 55 800 76 206 91 105
plastic 50 — — 7 760 64 108 93 69
(polypropylene) 75 — — 1 520 60 88 94 50
Pall rings, 16 26 gauge 16 — — — 92 230
metal 25 24 gauge 25 49 600 480 205 94 157
38 22 gauge 38 13 000 415 130 95 92
50 20 gauge 50 6 040 385 115 96 66
90 90 1170 270 92 97 53
Pall rings, 16 — 16 214 000 116 340 87 310
plastic 25 — 25 50 100 88 205 90 170
(polypropylene) 38 — 38 13 600 76 130 91 105
50 — 50 6 360 72 100 92 82
90 — 90 1 170 68 85 92 52
Raschig rings, 6 1.6 6 3.02 × 106 960 710 62 5250
ceramic 13 2.4 13 378 000 880 370 64 2000
19 2.4 19 109 000 800 240 72 840
25 3.2 25 47 700 670 190 74 510
38 6.4 38 13 500 740 120 68 310
50 6.4 50 5 800 660 92 74 215
75 9.5 75 1 700 590 62 75 120
100 9.5 100 700 580 46 80
Raschig rings, 19 1.6 19 111 000 1500 245 80 730
Steel 25 1.6 25 46 300 1140 185 86 450
38 1.6 38 14 100 785 130 90 270
50 1.6 50 5 900 590 95 92 187
75 1.6 75 1 800 400 66 95 105
Hy-Pac, steel (No.1) 30 30 000 300 177 96 141
(No.2) 60 3 780 225 95 97 59
Leva packing (No.1) 34 000 270 164
(No.2) 10 500 210 118
Low-density 25 39 700 160 250 83
polyethylene
Tellerettes

(Reproduced with the permission of McGraw-Hill Education, USA)

duty compared to other types of random packings. Hence, total fixed cost of
packed column is higher. Diameter of tower is recommended to be more than 10
times the packing size. Increase in packing size decreases required tower
diameter but increases height of packing. As increase in packing size reduces
pressure drop per unit height of packing but decreases surface area of packing
per unit volume and hence, decreases mass transfer area and rate of mass
transfer. Increase in packing size increases the value of HETP or HOG. In general
increase in packing size reduces the cost of packing per unit volume.
The second category of packing refers to structured packing. It can be made of
woven or solid material, usually thin metal sheet, which is typically corrugated
and bundled into segments that are placed layer by layer into the column shell.
In most cases, three types of structured packing are used. Type B (also known as
sheet metal packing) is fabricated from corrugated sheets as explained above.
Most B-type packing have a corrugation angle of 45°and a fold angle of 90°.
Second category is Type BSH packing which combines the essential features and
characteristics of sheet metal and wire gauze packing. In the BSH-packing
surface structure consists of rhombic perforations with alternating burred-up.
Third category is Type A which is basically Wire Gauze Packing. It is
constructed of multiple layers of corrugated and pierced wire gauze sheets.
9.3.2 Process Design
Process design of packed tower type absorbers involve two major steps: (i)
determination of tower diameter and (ii) determination of height of packing.
Method for determining tower diameter is same for all types of packed tower
absorbers. But methods for finding the height of packing are different for the
following cases:
1. Physical absorption from dilute gas mixture in isothermal conditions.
2. Physical absorption from concentrated gas mixture in isothermal conditions.
3. Physical absorption in non-isothermal condition, for example, absorption of
HCl or NH3 in water.
4. Absorption with chemical reaction, for example, absorption of chlorine in
sodium hydroxide solution.
Packed Tower Type Absorber for Physical Absorption Example of an absorber
for physical absorption from dilute gas mixture in isothermal condition is the
absorption of acetone vapour in water from air-acetone vapour mixture
containing less acetone (say up to 6%). Step-wise design for this case is as
follows.
1. From the given gas mixture data, find G1, Gs, y1 and Y1.
where, G1 = Molar flow rate of gas mixture, kmol/h
Gs = Molar flow rate of non-solute gas mixture, kmol/h
y1 = Mole fraction of solute in feed gas
Gs = G1 (1 – y1)   (9.1)

Y1 =

Y1 is the ratio of the moles of solute gas to the same of non-solute gas for
incoming gas,

Fig. 9.3 Variables Used for Material Balance and Design

G1 = Gs(1 + Y1)   (9.2)

Decide the degree of absorption to be carried out in absorber, if it is not
specified. Find y2, Y2 and G2.
where, y2 = Mole faction of solute in outgoing gas mixture

Y2 = , is the ratio of the moles of solute gas to non-solute gas for outgoing
gas mixture
G2 = Gs (1 + Y2)   (9.3)
where, G2 = Molar flow rate of outgoing gas mixture, kmol/h
If absorber is a pollution control tower then value of y2 or Y2 is decided by
pollution control norms.
2. Select a suitable solvent. In the selection of solvent, the important properties
to be considered are solubility of solute gas in the solvent, volatility, cost,
corrosiveness, etc1. Solvent which reacts with the solute by chemical reaction
provides high solubility. If recovery of solvent or solute gas is essential then
chemical reaction should be reversible. For example, for the absorption of CO2
gas, caustic soda solution provides higher solubility or absorptivity than ethanol-
amine solution. But in actual application, ethanolamine solution is selected as
solvent rather than caustic soda solution because CO2 gas can be easily separated
from ethanolamine solution at higher temperature and lower pressure in a
stripping column while caustic soda solution does not easily release CO2 gas in
stripping operation.
3. Find the gas-liquid equilibrium data. Draw the equilibrium curve in terms of Y
vs X.
4. Find the minimum amount of solvent required for the desired separation.
Minimum amount of solvent is the maximum amount of solvent for which
infinite number of equilibrium stages are required for the desired separation. In
other words, it is the maximum amount of solvent for which desired degree of
absorption is not possible. Minimum amount of solvent, in physical absorption is
determined by graphical method.
(a) In this method, first draw the equilibrium curve Y vs X.
(b) Mark (X2, Y2) point, where, X2 = ratio of the moles of solute to moles of
solvent for incoming solvent, Y2 = ratio of the moles of solute to non-solute
gas for outgoing gas mixture.
(c) Draw the tangent starting from point (X2, Y2) and allow it to intersect with
horizontal line from Y1, if the equilibrium curve is concave upward [Fig.
 9.4(a)]. But if the curve is concave downward as shown in Fig. 9.4(b), draw
the line from (X2, Y2) and join it with the point of intersection of equilibrium
curve and horizontal line from Y1.
In any case, intersection point is represented as (X1m , Y1), where X1m is the value
of X1 corresponding to minimum amount of solvent and X1 is the ratio of the
moles of solute to moles of solvent in outgoing solution.

Fig. 9.4 Minimum Amount of Solvent

(d) Write the overall solute balance equation

Lsm(X1m – X2) = Gs (Y1 – Y2)  (9.4)
substitute the value of X1m obtained from the graph and find the value of Lsm
from the above equation.
5. Decide the actual molar flow rate of solvent; Ls. Ideally it should be optimum.
Optimum amount of solvent is the amount of solvent which gives minimum total
cost of absorption. Total cost includes fixed cost, operating cost, interest on fixed
capital, etc. At design stage, if it is not possible to find the optimum solvent flow
rate, then actual flow rate Ls should be the greatest of the following values:
(a) Ls = 1.5 Lsm  (9.5a)

(b) Ls =   (9.5b)

where, MW = Molar mass of solvent

Certain minimum flow rate is required to facilitate the use of a centrifugal pump.
(c) Ls ≥ MWR*
*Minimum wetting rate (MWR) is the liquid rate, required to form a liquid film
over all the packing. No 100% reliable method is available to find MWR. Some
designers2 consider minimum liquid rate for any tower packing as 7340 kg/(h ·
m2 of tower cross section). Packing wetting rates are related to packing material
surface
Table 9.2 Minimum Wetting Rate for Different Packings
Surface Minimum wetting rate (MWR), m3/(m2 · h)

Unglazed ceramic 0.5

Glazed ceramic 2.0
Glass 2.5
Stainless steel 3.0
Carbon steel 0.7
Polypropylene 4.0
Polyvinyl chloride 3.5
Polytetrafluoroethylene (PTFE) 4.0

Morris and Jackson equation gives minimum solvent rate required in m3/(h · m2
of cross section of tower) by following equation2, 3:
MWR = R · at  (9.5c)
where, at = Packing surface area per unit volume, m2/m3.
Values of R in m2/h for rings less than 75 mm size is 0.08 and for packing more
than 75 mm size it is 0.12.
Kister3,4 has suggested values of MWR in m3/(m2 · h), 1.26 to 5.4 for random
packings and 0.252 to 0.504 for structured packing.
(d) For exothermic absorption LS should be such that it gives per pass rise in
temperature of liquid less than 10°C.
(e) Find the value of X2 based on overall solute balance.
Ls(X1 – X2) = Gs(Y1 – Y2)   (9.6)
6. Find NOG or NOL.
where, NOG = Number of overall gas phase transfer units
NOL = Number of overall liquid phase transfer units
Height of packing required for the given absorption duty is determined by using
one of the following equations:
Z = HOG · NOG   (9.7)
or Z = HOL · NOL   (9.8)
If overall resistance to mass transfer is controlled by the gas film, then Eq. (9.7)
is used to find the packing height Z for the desired absorption duty. If the gas is
highly soluble in liquid (solvent), then major resistance to mass transfer is
provided by the gas film. For example, HCl (hydrogen chloride gas) is highly
soluble in water. Hence, in this absorption, overall resistance to mass transfer is
governed by hydrogen chloride gas film resistance. Another example of gas-film
controlled absorption is absorption of acetone vapour in water from air-acetone
vapour mixture. In the same case, if the concentration of solute gas in gas
mixture is less than 10% (by mole) then molar flow rate of gas can be assumed
as constant and height of packing (Z), is given by following equation:

Z = HOG · NOG =   (9.9)

where, HOG =   (9.10)

and NOG =   (9.11)

where, Gm = Molar flow rate of gas per unit cross-sectional area, kmol/(m2 · s)
KG = Overall gas phase mass transfer coefficient, kmol/(atm · m2 · s)
a = Interfacial surface area per unit volume, m2/m3
p = Operating pressure, atm
HOG = Height of an overall gas phase transfer units, m
NOG = Number of an overall gas phase transfer units
Practically, it is not possible to find individual value of a. Hence majority of
correlations are developed for KGa.
NOG can be determined by Eq. (9.11), for which area under the curve of 1/(y – ye)
vs y within limits from y2 to y1 is determined. In the special case, if equilibrium
curve and operating line, both can be assumed as straight lines, NOG can be
calculated by the following equation5:

NOG =   (9.12)
where, Gm = Molar gas velocity, kmol/(m2 · s)
Lm = Molar liquid velocity, kmol/(m2 · s)
and m is the slope of the equilibrium line and Gm/Lm is the slope of the operating
line.
If the overall resistance to mass transfer is controlled by liquid film then Eq.
(9.8) is used to find the packing height. If the solubility of gas in liquid phase is
very low then overall resistance to mass transfer is governed by liquid film
resistance. For example, (i) absorption of carbon dioxide gas (CO2) in water
from CO2 rich gas mixture and (ii) absorption of dimethyl ether in water. In such
cases, normally a chemical reagent is added into liquid phase which reacts with
solute gas and thereby increases the rate of mass transfer dramatically. For
example, for the absorption of CO2, potassium carbonate (K2CO3) or ethanol
amine is added into water to improve the rate of the absorption of carbon dioxide
(CO2). In the same case, if the concentration of solute gas in gas mixture fed to
absorber is less than 10% (by mole) or for the dilute system, molar flow rate of
liquid can be assumed as constant and height of packing required can be
calculated by the following equations:
Z = HOL · NOL  (9.8)

where, HOL =   (9.13)

and NOL =   (9.14)

where, Lm = Molar liquid flow rate per unit cross-sectional area, kmol/(m2 · s)
NOL = Number of overall liquid phase transfer units
HOL = Height of an overall liquid phase transfer unit, m
KLC = Overall liquid-phase mass transfer coefficient, m/s
a = Interfacial surface area per unit volume, m2/m3
Ct = Total molar concentration, kmol/m3
x1 = Mole fraction of solute in the liquid phase at the outlet from absorber
x2 = Mole fraction of solute in the liquid phase at the inlet to the absorber
xe = Equilibrium mole fraction of solute in liquid phase
Units of mass transfer coefficient depend on units of driving force. KLC
represents mass transfer based on concentration gradient.
Here again, a cannot measured experimentally and hence, value of KLCa are
reported.
NOL can be determined by Eq. (9.14), for which area under the curve of vs
x within limits x1 to x2 should be determined.
Determination of Tower Diameter In a packed column as the throughput
increases beyond the stable operating region, pressure drop begins to rise much
more quickly and liquid entrainment increases from higher liquid holdup. As
random packing is relatively open to cross-flow, liquid and vapour/gas tend to
redistribute themselves. At very high throughput, there is sharp fallout in mass
transfer. This point is the flood point.
For packed tower-type absorber, tower diameter is determined based on
flooding velocity. Actual velocity of gas through packed tower is kept about 60
to 70% of flooding velocity. Recommended range of pressure drop for packed
tower-type absorber is 15 to 50 mm water column per m of packing height
depending on the application. Flooding velocity of gas for packed tower is
defined as a velocity of gas through the packed tower at which pressure drop
across the tower does not remain stable, even though liquid flow and gas flow
through the tower are kept constant. It goes on increasing and tends to infinite.
At the time of flooding, one of two conditions may occur; (i) Liquid phase may
occupy the entire cross section of tower. Continuous phase of liquid body rises in
the column. The change in pressure drop is very high with only a slight change
in gas rate. (ii) Phase inversion occurs and gas bubbles through the liquid.
Pressure drop rises rapidly as phase inversion occurs. To find the tower diameter,
at first the following factor (FLG) is determined:

FLG =   (9.15)

where, Lw = Mass velocity of liquid, kg/(m2 · s)

Gw = Mass velocity of gas, kg/(m2 · s)
ρG = Density of gas, kg/m3
ρL = Density of liquid, kg/m3
Using Fig. 9.5, corresponding value of K at flooding (KF) is determined. Actual
value of K is determined as follows. This graph is known as Sherwood-Leva-
Eckert correlation chart. Actual velocity of gas through the tower is kept in the
range of 50 to 70% of the flooding velocity. Let the actual velocity of gas
through tower be 66% of flooding velocity.
K = (0.66)2 KF   (9.16)
Corresponding value of pressure drop per m of packed height can be determined
by using Fig. 9.5. Recommended range of pressure drop in atmospheric packed
tower type absorber is 15 to 50 mm WC per m of packing. This chart (having
log-log scales) is a good approximation of the pressure drop in packed beds but
the values are correct within ±20%. This graph has limitations that it does not
take into account of the packing geometry and is based on limited data source.
Kister and Gill4 later developed different charts for different packings on
semilog scale. Y-axis is a linear scale in new charts which make interpolation
easier. While X-axis factor is kept same as that given in Fig. 9.5. Y-axis factor
includes packing factor. Massive data base was used to prepare the charts which
give more accurate values of pressure drop and flood line is omitted from the
graphs. Reader is advised to refer original article for these charts.
Billet and Schultes6 have presented in a concise manner updated equations for
calculating gas and liquid side mass transfer coefficients, pressure drops of dry
and irrigated random and structured packings, their loading and flooding points
as well as liquid hold-ups. Since their presentation is based on physical models
with large experimental data bases, they are highly reliable in designing packed
columns for mass transfer operations. The methods do not present graphical
presentations such as that of Sherwood–Leva–Eckert but present empirical
equations to be solved. However, the method involves elaborate and iterative
calculations. Use of a computer is recommended for these repetitive calculations.
Reader is advised to refer original article for more details.
K is a function of Gw.

K =   (9.17)

where, K = Function of velocity of gas through tower

ρG, ρL = Density of gas and liquid, respectively, kg/m3
Gw = Mass velocity of gas, kg/(m2 · s)
Gw =   (9.18)

g = 9.81 m/s2
Fp = Packing factor, m–1 (Ref: Table 9.2)
ψ = Ratio of density of water to density of liquid (inverse of specific gravity)
μL = Viscosity of liquid, mPa · s (= cP)

Fig. 9.5 Generalised Flooding and Pressure-Drop Correlation Chart for Packings3
(Sherwood-Leva-Eckert Correlation Chart)
(Reproduced with the Permission of McGraw-Hill Education, USA)

Determination of HOG or HOL Many different correlations are available to find

out the height of transfer unit or to find out the mass transfer coefficients. But,
these correlations are not as reliable as the same for finding the heat transfer
coefficients. Hence, it is suggested to find out the height of transfer unit of mass
transfer coefficient by using the correlation which is derived for the same or a
similar system. But, if the correlation for the same or a similar system is not
available, then, use more than one generalised correlations for finding the height
of transfer unit (HOG or HOL) and the maximum value obtained from the
different correlation is considered for finding the height of packing. Two
generalised correlations are given below.
1. Onda, Takeuchi and Okumoto Correlation3: Height of overall gas phase
transfer unit HOG and height of overall liquid phase transfer unit HOL are related
with HG and HL by following equation:

HOG = HG +   (9.19)

HOL = HG +   (9.20)

where, m is the slope of equilibrium line, Gm/Lm is the slope of operating line
and Gm and Lm are the molar flow rates of gas and liquid, respectively per unit
cross-sectional area of tower in kmol/(m2 · s). Height of gas phase transfer unit
(HG) is determined by the following equation:

HG =   (9.21)

and height of liquid phase transfer unit (HL) is determined by the following
equation:
HL =   (9.22)

where, Gw = Superficial mass velocity of gas, kg/(s · m2)

Lw = Superficial mass velocity of liquid, kg/(s · m2)
p = Operating pressure, Pa
Mg = Molar mass of gas, kg/mol
ρL = Liquid density, kg/m3
ae = Effective area for mass transfer, m2/m3
KG = Gas phase mass transfer coefficient, kmol/(m2 · s · Pa)
KLC = Liquid phase mass transfer coefficient, m/s
Effective area for mass transfer is obtained by the following equation:
=   (9.23)

where, ae = Effective area for mass transfer, m2/m3

at = Total area of packing, m2/m3 (Ref. Table 9.1)
σL = Liquid surface tension, m N/m (= dyn/cm)
Z = Packed height, m
CaL = Liquid capillary number = μL Lw /(ρL σL)  (9.24) (dimensionless group)
ρL = Density of liquid, kg/m3
μL = Viscosity of liquid, mPa · s (= cP)

ReG = Gas Reynolds number =   (9.25)

μG = Gas viscosity Pa · s
Gas phase mass transfer coefficient is determined by the following equation:

=   (9.26)

where, KG = Gas phase mass transfer coefficient, kmol/(m2 · s · Pa)

R = 8314 J/(kmol · K)
DG = Gas phase diffusion coefficient, m2/s
T = Operating temperature, K
ρG = Density of gas, kg/m3
dp = Packing size, m (derived from δp)
C1 = 2 for random packings of size less than 12 mm
= 5.23 for random packings of size more than 12 mm
Liquid phase mass transfer coefficient is determined by the following equation:
=   (9.27)

where, μL = Liquid viscosity, kg/(m · s )

g = 9.81 m/s2
LW = Mass velocity of liquid, kg/(m2 · s)
DL = Liquid phase diffusion coefficient, m2/s
dp = Equivalent diameter of packing, m (derived from δp)
aw = Wetted surface of packing, m2/m3
aw is determined by following equation

=   (9.28)

where, σL = Surface tension of liquid, mN/m

ReL = Liquid Reynolds number = (dimensionless)

FrL = Froude number for liquid = Lw 2 (dimensionless)

WeL = Weber number for liquid = Lw2 /(ρL σL at) (dimensionless)
σC = Critical surface tension of the packing material, mN/m
(Ref. Table 9.3 )
Table 9.3 Critical Surface Tension of Packing Materials3
Material of Packing σC, mN/m or dyn/cm

Carbon 56
Ceramic 61
Glass 73
Polyethylene 33
Polypropylene 33
Polyvinyl chloride 40
Steel 75

2. Cornell’s method: In this method height of overall gas phase transfer unit
HOG or height of overall liquid phase transfer unit HOL is calculated by the same
equations Eqs.   (9.19) and   (9.20) that are used for Onda’s method but
HG and HL are determined by the following equations:
Height of gas phase transfer unit HG is calculated by following equations.
For Raschig rings and Pall rings,
HG =   (9.29)

For saddle-type packings,

HG =   (9.30)

where HG = Height of gas phase transfer unit, m

ScG = Gas phase Schmidt number = (dimensionless)  (9.31)

DG = Gas phase diffusion coefficient, m2/s

ρG = Density of gas, kg/m3
μG = Viscosity of gas, kg/(m · s) or Pa · s
D = Column diameter, m
Z = Packed height, m
Lw = Mass velocity of liquid, kg/(m2 · s)

f1 = with μW = 1.0 mPa · s

f2 = with ρW = 1000 kg/m3

f3 = with σW = 72.8 mN/m

ψc = Correlation parameter which can be obtained from Fig. 9.6 for various types
of random packings.
Height of liquid phase transfer unit is calculated by the following equation:

HL =   (9.32)

where, HL = Height of liquid phase transfer unit, m

ϕCO = Correlation parameter for a given packing, m (can be obtained from Fig.
9.7)
C = Correlation factor for high gas rates (Refer to Fig. 9.8)
μL = Liquid viscosity, Pa · s
ρL = Liquid density, kg/m3
DL = Liquid diffusion coefficient, m2/s
Z = Height of packing, m
In this method, if the column diameter (D) is greater than 0.6 m then take D =
0.6 m for the calculation of HG.

Fig. 9.6 Correlation Parameter ψC for Cornell’s Method3

(Reproduced with the Permission of McGraw-Hill Education, USA)

(a) Ceramic Raschig Rings

(b) Ceramic Berl Saddles

 (c) Metallic Raschig Rings

(d) Metal Pall Rings

Lw = Liquid Rate, kg/(m2 · s)

Fig. 9.7 Correlation Parameter for Various Packings3

(Reproduced with the Permission of McGraw-Hill Education, USA)
 Fig. 9.8 Flooding Correction Factor, C3
(Reproduced with the Permission of McGraw-Hill Education, USA)

Design or Selection of Internals of Packed Tower Internals of packed tower

are packings, packing support, liquid distributor, liquid redistributor, mist
eliminator, hold-down plate, etc. Refer to Fig. 7.16 for packed tower details.
1. Packings: Suitable packing must be selected before the determination of
tower diameter. Packings can be divided in two broad classes:
(a) Regular packings: Grids, structured packings, stacked rings, etc., are of this
type. Grids are used for high gas rates as they offer very low pressure drop. They
are used in cooling towers.
Structured packings (Refer Section 9.3.1) are widely used in chemical process
industries. They are made from corrugated sheets with some perforations or wire
mesh. They provide high surface area with high void fraction. HETP (Height
Equivalent to Theoretical Plate) of structured packing is generally less than 0.5
m which could be less than half as compared to dumped Pall rings. They are
produced by many manufacturers. They are preferred for difficult distillation
operations, for high vacuum distillation, for high efficiency absorption with low
allowable pressure drop and also for increasing the capacity of existing
distillation column. Necessary data, correlations or graphs for calculating tower
diameter can be obtained from the manufacturer.
(b) Random packings: Various types of random packings are covered in Section
9.3.1.
2. Liquid distributors: Various types of liquid distributors, used with packed
tower, are shown in Fig. 9.9. Uniform initial distribution of liquid at the top of
packed bed is essential for the efficient mass transfer operation (distillation or
absorption).

Fig. 9.9 Different Types of Liquid Distributors

For small diameter packed columns (having diameter less than 0.3 m) single
point distributor like one spray nozzle is adequate. For large diameter packed
columns, multipoint distributors like perforated pipe distributor, trough-type
distributor, orifice distributor, etc., are used. As per the Perry’s Handbook3, for
each 194 cm2 cross-sectional area one distribution point is required. As per the
Eckert’s recommendations, number of streams or number of distribution points
required in packed tower are given in Table 9.4.
Table 9.4 Number of Distribution Points in Packed Tower3
Column diameter, m No. of points/m2 of cross-section

0.4 340
0.75 170
≥ 1.2 40

Perforated pipe distributors are preferred for column diameter ranging from
0.3 m to 1 m. It is used with clean liquids and offers minimum restriction to gas
flow. They are preferred with reasonably constant flow rate.
Trough-type distributors are preferred for columns having diameter greater
than or equal to 1.2 m. It can be used with liquids having solid in suspension.
Trough-type distributors are fabricated from metal sheets, plastics or ceramics.
This distributor consists of a series of troughs containing side notches. It
provides good liquid distribution with a large free area for gas flow.
Orifice-type liquid distributors are also preferred for large diameter packed
column. This type of distributor consists of flat perforated tray equipped with a
number of risers or short stand pipes. Ideally, gas rises upward through risers or
short stand pipes while liquid maintains the certain level over the perforated
plate and flows down through perforations. The risers should be sized to give
sufficient area for gas flow without creating a significant pressure drop. The
holes (perforations) should be small enough to ensure that there is a level of
liquid on the plate at the lowest liquid rate but large enough to prevent the
distributor overflowing at the highest rate.
In weir-type distributor gas and liquid (both) are flowing through the same
stand pipes. Notched weirs are provided in upper portion of stand pipes. Certain
level of liquid is maintained over the tray and liquid flows over notched weirs
and falling down through the same pipes from which gas rises upward. Weir-
type distributor is preferred with fluctuating liquid flow rates.
3. Liquid redistributor: After travelling a certain distance in a packed tower,
considerable fraction of liquid is migrated to the column wall and flows down in
the vicinity of column wall while the gas rises upwards through the central
portion. This phenomenon is called channelling. Liquid redistributor collects the
liquid that has migrated to the column wall and redistributes it evenly over the
next bed of the packing.
For small diameter column (D ≤ 0.6 m) wall wiper-type liquid redistributor is
preferred. It collects liquid from the column wall and redistributes it into the
central portion. Sometimes packing support plate itself acts as liquid
redistributor. Different types of liquid redistributors are shown in Fig. 9.10.

Fig. 9.10 Different Types of Liquid Redistributors

4. Packing support: The function of packing support is to carry the load of wet
packing bed without providing excessive restriction to gas and liquid flows. It
also acts as distributor for both streams. Poor design of packing support provides
higher pressure drop and can cause premature local flooding. Two types of
packing supports are used (Fig. 9.11).
(a) Countercurrent type: Example; Mesh-type packing support
(b) Separate flow or gas injection type: Example; Cap-type packing support,
beam-type gas injection support, etc.
 Fig. 9.11 Packing Supports

In countercurrent-type packing support liquid and gas both are flowing through
the same openings in countercurrent direction. Major open area of
countercurrent-type packing support is occluded by packing pieces resting on it.
Hence, this type of packing support provides the low value of net free area (50%
or less). Column diameter is decided based on the characteristics of packing
material. Berl saddles, Raschig rings, etc., provide the lesser free area for flow of
gas, hence with such packings, counter current-type packing support can be
used. But other packings like Pall rings, structured packings, etc., provide higher
% free area (ε) or percentage void space for the flow of gas (85% or more).
Hence, with these type of packings, counter flow-type packing support cannot be
used. Gas injection-type packing support can be designed for free area up to 90%
and because of their geometry there will have very little occlusion by the
packing. In this type of packing support, separate flow passages or openings are
provided for gas and liquid streams. Gas inlets are provided above the liquid
level. With packing materials like Pall rings, structured packings, etc., gas
injection-type packing support is recommended.
5. Hold-down plate: Hold-down plate is required to prevent damage to the
packing which can result due to a breakdown in normal operating conditions. At
high gas or vapour flow rates, packing at the top can be fluidised. This may
result in breaking or deshaping of packing. Ceramic packings can be easily
broken which may settle in voids of the packed bed and plug the flow channels.
In case of metal packings, deshaping may take place and the deshaped pieces can
plug the flow channels. In case of plastic packings, they may fly away with gas
or vapour and seat at various locations in the column. Also plastic packing may
break and seat randomly in the tower.
A typical hold-down plate is shown in Fig. 7.17.
Hold down plate, generally in the grid form or the screen form, can be used to
prevent such fluidisation. While heavier hold down plates are used for ceramic
and metal packings, lighter hold down plates of similar construction are used for
packed towers with plastic packing. Hold down grids/screens have generally
poor open area (< 50% of internal cross section of the tower). However, size of
opening in the grid/plate are fixed in such a manner that fly off of the packing
material can be prevented.
Design a scrubber for absorbing acetone vapour from air–
acetone vapour mixture by using pure water as solvent. The temperature in the
scrubber is 25°C and scrubbing is isothermal. Operating pressure of scrubber is
near atmospheric. A mixture of air with acetone vapour containing 6% by
volume of acetone is passed through the scrubber. This mixture contains 1400
m3/h of air. The scrubber is required to absorb 98% of the acetone.
Solution
Refer to Fig. 9.3.
Let Gs = Molar flow rate of air, kmol/h.
G1 = Molar flow rate of incoming air-vapour mixture, kmol/h
y1 = Mole fraction of solute (vapour) in incoming air-vapour mixture
Y1 =

G2 = Molar flow rate of outgoing air-vapour mixture, kmol/h

y2 = Mole fraction of vapour in outgoing air-vapour mixture
Y2 =
x2 = Mole fraction of solute in incoming solvent = 0
X2 =

L2 = Ls = Molar flow rate of solvent at top, entering the scrubber, kmol/h

L1 = Molar flow rate of solution leaving the scrubber, kmol/h
x1 = Mole fraction of acetone in liquid solution leaving the absorber
X1 =

Flow rate of air, qs = 1400 m3/h at 1 atm and 25°C

Density of air, ρA =

ρA = 1.1859 kg/m3
Gs = = 57.25 kmol/h

y1 = 0.06 (mole % = volume %)

Y1 = = 0.063 83

Scrubber is required to absorb 98% of acetone.

Y2 = 0.02 × Y1 = 0.001 277
X2 = 0, Ls = ?
Ls: To find the value of Ls, first the minimum amount of solvent required Lsm for
the desired separation is to be determined.
Phase equilibrium equation
y = mx =

At 25°C, vapour pressure of acetone is given by

ln pv = 16.6513 – (Antoine equation)

where, pv = Vapour pressure of acetone at 25°C, Torr

T = Temperature, K
ln pv = 16.6513 –
pv = 228.416 Torr
m = = 0.3

y = 0.3 x and

Table 9.5 Equilibrium Data Calculations

y Y = x = X =

0 0 0 0
0.0012 0.001 2014 0.004 0.004 016
0.01 0.010 1 0.033 0.034 12
0.02 0.020 4 0.067 0.071 8
0.03 0.030 9 0.1 0.111
0.04 0.041 67 0.133 0.153 4
0.05 0.052 63 0.167 0.2
0.06 0.063 83 0.2 0.25

Solute balance at minimum solvent rate

Gs(Y1 – Y2) = Lsm(X1m – X2)
57.22 (0.063 83 – 0.001 277) = Lsm(X1m – 0)
where, Lsm = Minimum amount of solvent required, kmol/h
X1m = Value of X1 corresponding to the minimum amount of solvent
Value of X1m can be obtained from the graph. Here, the curve of Y vs X
(equilibrium curve; Fig. 9.12) is concave downward, hence, the tangent is drawn
from (X2, Y2) point and is allowed to intersect the horizontal line at Y1.
Intersection point is (X1m ,Y1)
X1m = 0.2304 (from Fig. 9.12)
57.22 (0.063 83 – 0.001 277) = Lsm(0.2304 – 0)
Lsm = 15.535 kmol/h
Actual amount of solvent should be the greatest of the following values:
1. Ls = 1.5 Lsm = 23.3 kmol/h L2 = 23.3 × 18 = 419.4 kg/h
2. Ls = = 55.55 kmol/h, i.e. about 1 m3/h (minimum flow rate required to
facilitate the use of a centrifugal pump)
3. Let rise in temperature of solvent = 10°C per pass
Heat of solution of acetone in water3
ΔHs = –10 467 kJ/kmol
Aceton absorbed = GS (Y1 – Y2)
= 3.58 kmol/h
Heat generation on absorption, ϕ = 10 467 × 3.58
= 37 472 kJ/h
≡ 10.4089 kW
ϕ = ṁs CL Δt

Fig. 9.12 Determination of Minimum Amount of Solvent

ṁs = 37 472/(4.1868 × 10)

= 895 kg/h
Lsm = 895/18 = 49.722 kmol/h
4. Lsm = Minimum wetting rate required to form a liquid film over all the
packing
But minimum wetting rate required can be calculated after finding the tower
diameter. Tower diameter can be determined after deciding the actual amount
of solvent. Hence, it requires trial and error calculations.
For the first trial let Ls = 55.55 kmol/h = 1000 kg/h; highest amongst first three.
Tower diameter required at top: Eq. (9.15):
Lw = Mass velocity of liquid kg/(m2 · s) =

= 0.2778/

Molar flow rate for gas-vapour mixture at the top

G2 = Gs (1 + Y2) = 57.25 (1 + 0.001 201 44 ) = 57.3189 kmol/h
Average molar mass of gas-vapour mixture at the top
Mav = Σ (Mi yi)
Mav = 0.012 × M of acetone + 0.9988 × M of air
Mav = 0.0012 × 58 + 0.9988 × 29
Mav = 29.0348 kg/kmol

Gw = kg/(m2 · s)

Density of vapour gas mixture at top

ρG = = 1.187 kg/m3

FLG = = 0.020 72

From Fig. 9.5, KF = 0.2

Let actual velocity of gas = 66% of folding velocity
× 100 = 66%

K = (0.66)2 × 0.2 = 0.087 12

For K = 0.087 12 and FLG = 0.020 72
Pressure drop = 60 mm H2O/m of packing – acceptable (From Fig. 9.5)
Mass velocity of gas through tower
ψ = = 1, ρL = 1000 kg/m3

g = 9.81 m/s2
Viscosity of water at 20°C, μL = 1 cP = 10–3 kg/(m · s)
Select 25 mm polypropylene Pall rings as packing,
Packing factor FP = 170 m–1 (Ref. Table 9.1)

Gw =

= 2.4428 kg/(m2 · s)
Tower area required at top =

= = 0.1892 m2

Tower diameter required at top = 0.49 m ≈ 0.5 m ≡ 500 mm

= 20 > 10 (satisfactory)

Liquid rate = = 5.09 m3/(m2 · h)

Hence, it is greater than 4 m3/(m2 · h) (minimum wetting rate required for

polypropylene surface). Thus condition No. 4 (p. 680) is satisfied.

Liquid rate = = 5093 kg/(h · m2)

It is less than 7340 kg/(h · m2), minimum wetting rate specified in literature
(Ref. 2).
Based on Morris and Jackson’s equation, minimum solvent required
MWR = R · at  (9.5c)
For 25 mm polypropylene Pall rings at = 205 m2/m3 (From Table 9.1)
For 25 mm polypropylene Pall rings R = 0.08 m2/h
Lsm = 0.08 × 205 = 16.4 m3/(m2 · h)
Considering the last method for predicting minimum wetting rate,
let Ls = 3500 kg/h = 194.4 kmol/h

At top FLG = = 0.0725

From Fig. 9.5, KF = 0.17

Let actual velocity = 66% of flooding velocity
K = (0.66)2 × 0.17 = 0.074
Pressure drop = 50 mm H2O/m of packing
Mass velocity of gas through tower

Gw =

= 2.2514 kg/(m2 · s)
Tower area required at top, A1 =

= 0.2052 m2 =

D = 0.511 m ≈ 0.52 m

Liquid rate = = 16 480.54 kg/(m2 · h) ≡ 16.48 m3/(m2 · h)

Hence, it satisfies the conditions.

Tower diameter required at the bottom:
Average molar mass of gas-vapour mixture
Mav = ΣMi yi = y1 × 58 + (l – y1) × 29 = 0.06 × 58 + (1 – 0.06) × 29
= 30.74 kg/kmol
Density of gas-vapour mixture at bottom
ρG1 = = 1.257 kg/m3
For Ls = 3500 kg/h = 194.44 kmol/h
Acetone balance
Gs(Y1 – Y2) = Ls(X1 – X2)
57.25(0.063 83 – 0.001 277) = 194.44(X1 – 0)
X1 = 0.018 42

x1 = = 0.0181

Density of water = 1000 kg/m3

Density of acetone = 790 kg/m3
Average density of liquid solution leaving the scrubber

ρL1 =

w1 = x1 × 58/[x1 × 58 + (1 – x1) × 18]

= = 0.056 (acetone)

ρL1 = = 985.33 kg/m3

L1 = Ls(1 + X1) = 194.44(1 + 0.018 42) = 198.025 kmol/h

Average molar mass of liquid solution
Mav = 0.0181 × 58 + 0.9819 × 18 = 18.724

L1W = kg/(m2 · s)

G1 = Gs(1 + Y1) = 57.25(1 + 0.063 83) = 60.904 kmol/h

G1w = kg/(m2 · s)

At bottom, FLG =   (9.15)

= 0.0707
KF = 0.175 (From Fig. 9.5)
Let actual velocity of gas = 66% of flooding velocity
K = (0.66)2 × 0.175 = 0.076 23
Δp of packing ≅ 50 mm of H2O/m of packing
Mass velocity of gas through tower

G1w =

= 2.316 kg/(m2 · s)
Tower area required at bottom A = = 0.2245 m2

D2 = 0.2245 m2

D = 0.5346 m
Thus the diameter of the entire tower, D ≈ 0.54 m
NOG: (Number of overall gas phase transfer units)
To calculate the NOG, draw the operating line on the graph which starts from (X2,
Y2) and terminates at (X1, Y1), i.e., starts from (0, 0.001 277) and terminates at
(0.018 43, 0.06 383). (Ref. Fig. 9.12)
Given system is a dilute system (x1 < 0.1). Here the portion of equilibrium curve
under the operating line can be approximated as straight line. Hence, in the given
case NOG can be calculated by the following equation:

NOG =   (9.12)

= 0.090 38
NOG =

= 4.1988
Here, acetone is highly soluble in water. Hence, liquid film resistance is
negligible as compared to the gas film resistance. Hence, height of packing in
the given case should be determined by equation,
Z = NOG · HOG  (9.7)
HOG: Height of Overall Gas Phase Transfer Unit
HOG = HG +   (9.19)

Based on Onda, Takeuchi and Okumoto’s method,

Height of gas phase transfer unit
HG =   (9.21)

Gw = Superficial mass velocity of gas, kg/(m2 · s)

= = 2.1439 kg/(m2 · s)

Effective area of mass transfer ae is determined by following equation:

=   (9.23)

at = 205 m2/m3
Surface tension data for acetone at the scrubbing temperature can be calculated
by a group-contribution method, given in Ref. 3.
σL = Liquid surface tension in or

xi = xav = = 9.05 × 10–3 = 0.0095

O
[ Pi ] for Acetone: | |
(CH3—C—CH3)
[Pi ] = 55.5 × 2 + (9 + 22.3) = 141.7 (From Table 3.343 of Ref. 3)
Surface tension of Acetone
σ 1/4 = [Pi ]( ρL – ρG) (Ref. Eq. 3.151 of Ref. 3)

Density of pure acetone vapour, ρG = = 2.372 kg/m3

where, ρL, ρG are in mol/cm3

σ 1/4 = 141.7

= 1.9242
σ = 13.71 dyn/cm
Actual surface tension data are reported by Timmermans (Ref. 8).
σ = 24.5 dyn/cm at 25oC
Thus it can be seen that group-contribution method may give erroneous results.
It is advised to use the actual value as far as possible.
For pure water, σ = 70 dyn/cm
For liquid mixture, using Eq. (7.37),
σL1/4 = Σ (σi1/4 xi) = 24.51/4 × 0.009 05 + (1 – 0.009 05)701/4
σL = 69.58 dyn/cm or mN/m
For dilute aqueous solution, surface tension of water may be taken for
calculations.
CaL = Liquid capillary number =   (9.24)

xav = 0.009 05
Wav = (0.009 05 × 58)/(0.009 05 × 58 + (1 – 0.009 05) × 18)
Wav = 0.028 58 (average mass fraction)

ρL = = 992.46 kg/m3

σL = 69.58 × 10–3 N/m

Lw =
= = 4.3712 kg/(m2 · s)

μL = Viscosity of liquid mixture in kg/(m · s)

ln μmix = Σ (wi ln μi) (Eq. 3.107 of Ref. 3)
Viscosity of acetone at 25°C, μac = 0.33 × 10–3 kg/(m · s)
Viscosity of water at 25°C, μw = 0.95 cP = 0.95 × 10–3 kg/(m · s)
ln μL = 0.028 58 ln (0.33 × 10–3) + (1 – 0.028 58) ln (0.95 × 10–3)
μL = 9.217 × 10–4 kg/(m · s)

CaL =

= 5.834 × 10–5
Reynolds number:
GW = 2.1439 kg/(m2 · s), at = 205 m2/m3
Viscosity of gas mixture:

μG = (6.77)

i represents acetone while j represents air.

yav = = 0.0306

Θij =

and Θji =

Viscosity of acetone vapour, μac = 750 × 10–7 P = 750 × 10–5 cP

Viscosity of air μair = 1850 × 10–7 P = 1850 × 10–5 cP

Θij = = 0.4812
Θji = × 0.4812 = 2.374

μG =

= 0.033 69 cP ≡ 0.033 69 × 10–3 kg/(m · s)

ReG = = 1862

Z = NOG · HOG = 4.1988 HOG

=   (9.23)

For the first trial calculations, let HOG = 0.6 m

Approximate values of HOG for random packings of size 25 mm is 0.3 to 0.6 m,
for 38 mm size it is 0.5 to 0.75 m and for 50 mm size it is 0.6 to 1.0 m (Ref. 2).
Z = 4.1988 × 0.6 = 2.52 m
= = 0.748 42

ae = 153.43 m2/m3 of packing

Gas phase transfer coefficient KG is determined by Eq. (9.26).
where, R = 8314 J/(kmol · K)
T = operating temperature = 273 + 25 = 298 K
Mav = ΣyiMi = 0.0306 × 58 + (1 – 0.0306) × 29 = 29.8874
ρG = ,

= = 1.2222 kg/m3

DG = Gas phase diffusion coefficient, m2/s

From Table 3.318, of Ref. 3,
At 25°C diffusivity of acetone in air
DG = 0.109 cm2/s = 0.109 × 10–4 m2/s
DG ∝ T1.75
DG = 0.109 × 10–4 × = 1.2706 × 10–5 m2/s

C1 = 5.23 for dp > 12 mm

KG × = 5.23

× × (205 × 0.025)–2

KG × 9.511 82 × 108 = 5.23 × 55.509 × 1.2945 × 0.038 07.

KG = 1.504 × 10–8 kmol/(m2 · s · Pa)
Height of gas phase transfer unit, Eq. (9.21):
GW = Average superficial mass velocity of gas, kg/(m2 · s)
GW = 2.1435 kg/(m2 · s)
KG = 1.504 × 10–8 kmol/(m2 · s · Pa)
ae = 153.43 m2/m3
p = operating pressure = 101.325 × 103 Pa
Mg = average molar mass of gas = 29.8874 kg/mol
HG =

= 0.3068 m
Height of liquid phase transfer unit, Eq. (9.22)
LW = Average mass velocity of liquid through tower

= = 4.3712 kg/(m2 · s)

ae = 153.41 m2/m3, ρL = 992.46 kg/m3

HL =
Liquid phase mass transfer coefficient KLC can be determined by Eq. (9.27).
ρL = 992.46 kg/m3, μL = 9.217 × 10–4 kg/(m · s)
LW = 4.3712 kg/(m2 · s), g = 9.81 m/s2, at = 205 m2/m3
dp = 0.025 m
aw = wetted surface of packing, m2/m3

= 1 – exp   (9.28)

For polypropylene packing, σc = 33 × 10–3 N/m (Table 9.3)

σL = 69.58 × 10–3 N/m

Liquid Reynolds number, ReL =

= 23.134
Froude number of liquid, FrL = LW2at /( ρL2g)

= = 4.0538 × 10–4

Weber number for liquid,

WeL = /( ρL σL at ) = = 1.35 × 10–3

= 1 – exp

= 1 – exp (– 0.4472)

= 0.3606

aw = 0.3606 × 205 = 73.923 m2/m3

DL = Diffusivity of acetone in water, m2/s
DL can be determined by Wilke and Chang equation.

DL = (Equation 2.44, of Ref. 1)  (9.33)

where, VA = Molar volume of solute at normal boiling point, m3/kmol

ϕas = Association factor for solvent = 2.26 for water (a parameter)
Ms = Molar mass of solvent = 18 kg/kmol
μ = Solution viscosity = 9.217 × 10–4 kg /(m · s)
T = Temperature, K = 273 + 25 = 298 K
For calculating VA, a group contribution method, given in Treybal is
recommended.
From Table 2.3 of Ref. 1
Table 9.6 Group Contributions for Diffnsivity of Acetone1
Atom Volume, m3/kmol Total No. of atoms

C 0.0148 3
H 0.0037 6
O 0.0074 1

VA = 0.0148 × 3 + 0.0037 × 6 + 0.0074 × 1 = 0.074 m3/kmol

DL = = 1.154 × 10–9 m2/s

KLC = 0.0051

× × (205 × 0.025)0.4

KLC × 47.88 = 0.0051 × 16.02 × 0.03525 × 1.9225

KLC = 1.1564 × 10–4 m/s

HL =

= 0.248 m
HOG = HG +   (9.19)
= 0.3068 + 0.09038 × 0.248 = 0.329 m
Z = NOG · HOG = 4.1988 × 0.329 = 1.38 m
For second trial calculations, let Z = 1.4 m
ae = = 194 m2/m3

HG = 0.3068 × = 0.2426 m

and HL = 0.248 × = 0.196 m

HOG = 0.2426 + 0.090 38 × 0.196 = 0.2603 m

Z = 4.1988 × 0.2603 = 1.093 m < 1.382 m
HOG for 25 mm size random packing ranges from 0.3 to 0.6 m.
Let HOG = 0.3 m
Z = 0.3 × 4.1986 = 1.26 m ≈ 1.3 m
This is the value of packed bed height based on Onda, Takeuchi and Okumoto’s
method.
Cornell’s method
For Pall rings, height of gas phase transfer unit [by Eq. (9.29)]
ψc = Correlation parameter = 60 m for metal Pall rings (Fig. 9.6)
For polypropylene Pall rings, data for ψc are not available. Hence, assume that
the values of ψc for polypropylene Pall rings and ψc for metal Pall rings are
same.
For PP Pall rings ψc = 60 m
D = 0.54 m, Z = 1.3 m (Assume for the first trial calculations)
ScG =

= 2.1695

f1 = with μ w = 10–3 Pa · s

= = 0.987
f2 = with ρw = 1000 kg/m3

= = 1.01

f3 = where σw = 72.8 mN/m

= = 1.038

Lw = Liquid phase mass velocity = 4.3712 kg/(m2 · s)

HG = = 0.3085 m

Height of liquid phase transfer unit, can be determined by Eq. (9.31)

HL =

From Fig. 9.7, for L = 4.3712 kg/(m2 · s),

Correlation parameter ϕCO for metal Pall rings = 0.05 m
Assume that ϕCO for PP Pall rings is ϕCO for SS Pall ring.
ϕCO = 0.05 m
C = 0.75 for 66% Flooding (From Fig. 9.8)

HL =

= 0.2854 m
HOG = HG + = 0.3085 + 0.09038 × 0.2854

= 0.334 m
Height of packed bed, Z = NOG · HOG
= 4.1988 × 0.334 m = 1.4 m
For second trial calculations, let Z = 1.4 m
HG = 0.3058 × = 0.3134 m

HL = 0.2854 × = 0.2856 m

HOG = 0.3134 + 0.090 38 × 0.2856 = 0.3392 m

Z = 4.1988 × 0.3392 = 1.424 m
Let height of packed bed Z = 1.5 m (higher than the values of Z obtained based
on Cornell’s method as well as based on Onda, Takeuchi and Okumoto’s
method.)
No. of packing sections required = one
Liquid distributor
For clean solvent and for small diameter tower perforated pipe-type distributor is
satisfactory. It is recommended that one liquid stream must be provided for each
194 cm2 area (Ref. 3).
No. of holes (distribution points) required in pipe type distributor,
nh =

nh = = 11.8, say 12

Let the velocity of water through the pipe = 2 m/s

Flow rate of water (solvent) = 3500 kg/h
Volumetric flow rate of solvent, qv = = 9.7222 × 10–4 m3/s

d = = 0.025 m

= 25 mm (size of pipe)
Let the velocity of solvent through holes = 3 m/s
Let dh = diameter of hole, m

12 × =

dh = 0.006 m ≡ 6 mm
Resulting pipe-type liquid distributor is shown in Fig. 9.12.

Fig. 9.13 Pipe-Type Liquid Distributor

Packing support
Packing support should be selected such that flow area provided by packing
support for the flow of gas should be greater than flow area (i.e., void) provided
by packing material.
% Void space for 25 mm size polypropylene Pall rings = 90% (Ref. Table 9.1)
With Pall rings any gas-injection-type packing support can be used.
Let type of packing support: Cap-type packing support
Actual outer diameter of packing support is greater than 540 mm as some
portion of packing support is sandwiched between two flanges.
Let h′ = height of slot in riser
x = width of slot in riser
d = diameter of riser
n = number of risers
DC = Inside diameter of column = 540 mm
If d = DC/6, then nine number of risers can be provided on packing support.
= = 90 mm

Let, ns = number of slots per riser

Total area of slots of risers = ns · n h′ x
The distance between two adjacent slots is x mm
Let x = 10 mm
ns = ≅ 14

Total area of slots of risers = 14 × 9 × h′ × 10 ≥ 0.9 × (540)2

h′ ≥ 163.58 mm
Consider h′ = 170 mm
Let velocity of liquid through weep hole = 0.5 m/s
Let number of weeping holes = 34
34 × (dw)2 = 9.733 × 10–4/0.5

dw = 8.5 mm ≈ 9 mm
Refer to Fig. 9.14 for packing support plate details.
 Fig. 9.14 Cap-Type Packing Support

9.4 ABSORPTION WITH CHEMICAL REACTION

Many industrial absorption systems involve chemical reaction. Majority of
pollution control towers are accompanied by chemical reaction. In absorption
with chemical reaction, a chemical reagent is added in solvent which reacts with
solute gas of feed gas mixture. If recovery of solute gas or solvent is essential
then reaction should be reversible. Otherwise in most of the cases it is
irreversible. For example, if CO2 from the feed gas mixture is required to be
separated in pure form then mono ethanol amine is selected as reagent as it
reacts with CO2 by reversible reaction. Otherwise NaOH can be selected as
reagent if CO2 recovery is not required.
Absorption with chemical reaction enhances the rate of absorption considerably.
It converts the absorbed solute into other compound and hence, increases driving
force for mass transfer. It also increases liquid phase mass transfer coefficient by
a magnitude which contributes in increasing absorption rates.
Process design of packed tower type absorber involving chemical reaction
includes following steps of calculations:
1. From the given feed gas flow rate and composition, Find G1, Gs, G1w and y1.
where,
G1 = Molar flow rate of feed gas mixture, kmol/h
Gs = Molar flow rate of non-solute gas mixture, kmol/h
y1 = Mole fraction of solute in feed gas
G1w = Mass flow rate of feed gas mixture, kg/h
2. Fix the value of y2, mole fraction of solute in exit gas. Decrease in the value of
y2 increases fixed cost and operating cost of absorber. In case of pollution
control tower y2 should be fixed based on the maximum permissible
concentration in the vent gas, limited by pollution control norms.
3. Find the minimum amount of solvent Lsm, required for the desired separation.
In this case Lsm is determined based on the stoichiometry of the chemical
reaction.
4. Find the actual amount of solvent Ls. Actual amount of solvent Ls should be
the greatest among the following values:
(a) Ls = 1.5 Lsm
(b) Ls ≥ 1 m3/h to facilitate the use of centrifugal pump. Metering pump gives
pulsating flow.
(c) Actual wetting rate should be greater than minimum wetting rate required for
packing (MWR). Value of MWR depends on type and MOC of packing.
For example, MWR for polypropylene Pall ring is 4 m3/(h · m2).

Actual wetting rate   (9.34)

where, Lsw = Mass flow rate of solvent, kg/s

ρL = Density of liquid solvent, kg/s
Di = Inside diameter of tower, m
(d) Ls should be such that per pass rise in temperature of solvent is 5 to 10°C.
Accompanying chemical reaction could be exhothermic in nature. Hence, if
solvent is to be recirculated, a cooler is required in recirculation line of
solvent to remove exhothermicity of reaction and to control the temperature
of solvent.
5. Find the inside diameter of absorption tower. Method for finding tower
diameter is same as that for physical absorption. Actual wetting rate can be
calculated after finding Di. Hence, for the first trial calculation Ls should be
greatest of three values given by (a), (b) and (d). Value of (c) can be verified
at the end of trial calculation. Thus calculations for finding Ls and Di require
trial-and-error calculations.
6. Find the number of overall gas phase transfer units.

NOG =   (9.35a)

With irreversible reaction and excess of chemical reagent , ye = 0.

NOG =   (9.35b)

7. Find the height of overall gas phase transfer unit, HOG.

HOG =   (9.36)
where, HOG = Height of overall gasphase transfer unit, m
G1 = Molar flow rate of feed gas, kmol/s
Di = Inside diameter of tower, m
KGa = Gas phase mass transfer coefficient, kmol/(h · m3)
Table 9.7 Recommended Values of KGa3
Gas phase reactant Liquid phase reactant KG a, kmol/(h · m3)

Cl2 8% NaOH solution 272

SO2 11% Na2CO3 solution 229

Br2 5% NaOH solution 131

H2S 4% NaOH solution 96

CO2 4% NaOH solution 38

HCl* H2O 353

NH3* H2O 337

HF* H2O 152

HCN* H2O 114

HCHO* H2O 114

HBr* H2O 59

*Physical absorption

These values of KGa are derived for operating conditions: 1 atm, 16–24°C, 38
mm Intelox saddles, Gas rate = 1.3 kg/(m2 · s), Packed height = 3.05 m, Tower
diameter = 0.76 m, Solute gas concentration = 0.5 to 1% (by mole)
Ideally these values of KGa can be used in the given case if
1. Solute concentration ≥ 0.5 %
2. Operating pressure ≥ 1 atm
3. Operating temperature ≤ 24°C
4. Tower diameter ≤ 0.76 m
5. Packing size ≤ 38 mm
6. Type of packing: Intelox saddles or better packings like Pall-rings, structured
packing, etc.
7. Superficial mass flow rate of gas ≥ 1.3 kg/(s · m2)
If one (or more) conditions are not fulfilled in the given case, then use of this
value of KGa may result in underdesign of absorption unit.
KGa can also be derived from the performance data of commercial gas
absorption tower working on the same system. Derived value of KGa from the
performance of commercial tower can be used for the new design, if in the new
case all conditions related to absorption are similar or better. Even today, most
commercial gas absorption units are designed primarily on the basis of prior
commercial experience.
8. Height of packing of packed tower
Z = NOG · HOG  (9.38)
where Z = Height of packing, m
9. Decide number of packing sections. Height of each packing section should be
less than 3 m to avoid channelling.
10. Design the suitable type of cooler to remove the exothermic heat of
absorption.
If heat duty requirement is low, then a coil-type cooler can be inserted in the
liquid pool of circulating solvent hold-up tank. For the higher heat duty a
separate shell and tube-type heat exchanger can be selected and installed in
recirculation line of solvent.
11. Design the suitable type of liquid distributor, liquid redistributors, packing
support, etc.
Design a scrubber for bromine for following service:
Feed gas to bromine scrubber = 25 kg/h of Br2 and 250 kg/h of air (Here air is
added for dilution.)
Design data of an existing bromine scrubber are available in literature as under.
Feed gas = 10 kg/h of Br2 + 50 ft3/min air measured at 1 atm and 25°C
Molar mass of Br2 = 160
Concentration of Br2 in outlet gas = 0.1 ppm
Gas temperature = Ambient
Type of scrubber = Counter flow packed bed
Column ID = 150 mm
Packed bed height = 2 m
No. of packing sections = 2
Type of packing = 16 mm (5/8 in) polypropylene Pall rings
Circulating liquid = 10% caustic soda solution
Pressure drop across packing = 45 mm WC
Capacity of circulation tank = 1200 L
Tank MOC = FRP
Solvent rate = 1 m3/h
Gas pressure = Atmospheric
Type of fan: Induced draft fan 100 mm (4 in) WC pressure, 50 ft3/min capacity
Solution
Feed gas flow rate = G1w = 250 + 25 = 275 kg/h

G1 =

y1 = = 0.0178, y2 = 0.1 × 10–6 = 10–7

NOG =

Reaction between Br2 and NaOH is instantaneous and irreversible. Hence ye = 0

Minimum amount of solvent:
Reaction
2 NaOH(aq) + Br2(g) → NaOBr(aq) + NaBr(aq) + H2O(l)
NaOH required to absorb 25 kg/h Br2
= 25 × = 12.5 kg/h

Lsm = = 125 kg/h of 10% NaOH solution

Actual amount of solvent should be the greatest of the following:

1. Ls = 1.5 Lsm = 125 × 1.5 = 187.5 kg/h
2. To facilitate the use of a centrifugal pump for solvent.
Ls ≥ 1 m3/h
ṁs = 1 m3/h × Density of 10% NaOH solution
= 1 × 1100 = 1100 kg/h
3. Rise in temperature of circulating solvent should be in the range of 5 to 10°C.
Heat of reaction.
ΔHor =
ΔHor = –285.83 – 334.3 – 361.67 – 2(–470.11) – 30.91
ΔHor = –72.49 kJ/mol = ΔHr
Heat of absorption, ϕ1 = 72.49 × 1000 × = 11 326.6 kJ/h
≡ 3.13 kW
Let rise in temperature of solvent = 5°C
LSW = = 580.85 kg/h

4. Actual wetting rate should be greater than minimum wetting rate required.
For the first trial calculation,
Let Lsw = 1100 kg/h (= 1 m3/h)
ρL = 1100 kg/m3
G1w = 250 + 25 = 275 kg/h
Average molar mass of gas mixture at inlet
Mav =

ρG1 =

Based on Eq. (9.15):

FLV =

For 45 mm WC/m of packing

K = 0.045 and KF = 0.15 (From Fig. 9.5)

% Flooding =

Hence, it is in the range of 50 to 70%.

For 16 mm (5/8 in) polypropylene Pall rings, packing factor Fp = 310 m–1 (Ref.
Table 9.1)
Mass velocity of gas through tower can be calculated using Eq. (9.18):
ψ =

μL = 5 cP

Gw =
= 1.2654 kg/(m2·s)
=

Di = 0.277 m ≡ 0.28 m

Actual wetting rate =

MWR for polypropylene Pall ring = 4 m3/(h · m2)

Hence, AWR > MWR (satisfactory)

Eq. (9.36) HOG =

For existing scrubber:

G1w = 10 kg/h + 50 ft3/min (≈ 85 m3/h) at 1 atm and 25°C
Density of air
ρa =

= 1.185 kg/m3
Mass flow rate of air = 85 × 1.185 = 100.7 kg/h
G1W = 10 + 100.7 = 110.7 kg/h
G1 =

y2 =

y2 = 0.1 ppm = 0.1 × 10–6

NOG = = 12.08
GM = = 200 kmol/(m2 · h)

HOG =

KGa =

For the new tower design

HOG =

Height of packing Z = NOG × HOG

= 12.0895 × 0.118 = 1.4265 m
Actually increase in diameter decreases the value of KGa. Hence, take actual as
height of packing for new tower as 2 m.
Let number of packing section = 2
Consider 20% extra heat duty of cooler.
Heat duty of cooler
ϕ2 = 11 326.6 × 1.2 = 13 600 kJ/h ≡ 3.778 kW
Rise in temperature of solvent
Δt =

Coil can be immersed in liquid pool of circulation tank.

Let U = 350 W/(m2 · C)
Cooling water temperature:
Inlet temperature of cooling water = 32°C
Outlet temperature of cooling water = 34°C
Let ΔTm = 7°C
Hence, average temperature of solvent in absorber is 40°C (= 33 + 7).
Cooling water required
ṁa = ≡ 1.624 m3/h

Let velocity of water through coil = 2 m/s

 d2ci =

dci = 0.017 m = 17 mm
dco = 22 mm = outside diameter of pipe
Let helix diameter = 0.5 m
Heat transfer area per turn.
HTA = π × 0.5 × (π × 0.022)
= 0.1086 m2
Area of coil required
Ar =

No. of turns of coil required =

Let number of turns = 15

Design of liquid distributor:
Pipe-type distributor may be selected.
Rate of solvent = 1 m3/h
It is recommended that one distribution point must be provided for each 194 cm2
area.
No. of holes (distribution points) required in pipe-type distributor

nh =

Let nh = 4
Let velocity of solvent through pipe = 2 m/s
If di = ID of pipe, then
di2 =

di = 0.0133 m ≡ 13.3 mm
Let velocity of solvent through hole = 3 m/s
If dh = diameter of each hole, then

=
dh = 5.42 mm
Let dh = 5 mm
Design of liquid redistributor:
Let 45° wiper-type liquid redistributors be employed.

Fig. 9.15 Suggested Redistributor

Design of packing support

Design criteria: Free area for the flow of gas provided by packing support should
be greater than free area for the flow of gas provided by packing material.
% free space provided by 16 mm PP Pall ring = 87%
Let type of packing support = Cap type
Let diameter of rises or cap =

Total number of caps = 9

Let width of slot x = 5 mm
No. of slots per cap = = 14

Let h′ = height of slot

h′ × 9 × 14 × 5 × ≥ 0.87

h′ ≥ 85 mm
Let velocity of liquid through weeping hole = 0.5 m/s
Let number of holes = 12
=
dW = 7.68 × 10–3 m say 8 mm

Fig. 9.16 Suggested Cap-Type Packing Support Plate

9.5 PROCESS DESIGN OF SPRAY CHAMBER OR

SPRAY TOWER-TYPE ABSORBER3, 9
Spray absorbers are used for removing SO2 (sulphur dioxide gas) form boiler
flue gases. When very low pressure drop is essential and when incoming gas
stream is contaminated by particulate matter, spray absorbers are preferred over
the packed tower type absorber. In the large scale plant, boiler flue gases
entering the absorber for the removal of SO2 are available at very low pressure
(100 to 200 mm WC) and with very high volumetric flow rate. These gases are
also contaminated by particulate matters. If coal is used as a fuel then
concentration of SO2 and suspended particulate matter (SPM) in boiler flue
gases are relatively higher and it favours selection of spray tower.
Spray tower facilitates the use of lime slurry as a solvent against caustic soda
solution. Former is cheaper than the latter. Lime slurry as such cannot be
considered in packed tower.
Spray chambers (absorbers) are having either circular or rectangular cross
section. Normally there is no packing in the spray absorber.
They are of two types:
1. Vertical spray tower
2. Horizontal spray chamber
Liquid phase residence time in spray absorbers is in the range of 1 to 10 s which
is very low. To compensate this, liquid phase resistance must be negligible.
Hence, a chemical reagent is added in the liquid phase to get the higher rate of
absorption.
9.5.1 Process Design of Vertical Spray Tower3
In vertical spray tower, gas stream is flowing vertically upward and liquid is
sprayed downward in various sections within the tower as well as from the sides
by spray nozzles.
Process design steps for vertical spray tower are as follows:
1. Determination of tower diameter: Maximum permissible gas velocity through
vertical spray tower is 2.3 m/s. Actual velocity of gas through tower should be
less than 2.3 m/s to avoid the excessive entertainment of liquid droplets in the
exist gas. On fixing the actual velocity, tower diameter can be determined by
following equation:

D =   (9.37)

where, D = Tower diameter, m

qv = Volumetric flow rate of gas, m3/s
ug = Velocity of gas through tower, m/s < 2.3 m/s
2. Determine the number of gas phase transfer units required for the desired
separation by the following equation:
NG =   (9.38)

where, NG = Number of gas phase transfer units

y1 = Gas phase solute mole fraction at inlet (at bottom)
y2 = Gas phase solute mole fraction at outlet (at top)
Actually the correlation for NG is

NG =   (9.39)

Considering irreversible chemical reaction and use of excess reactant, ye = 0.

Also in vertical spray tower, gas tends to become completely backmixed by the
action of liquid spray. Hence in spray section, concentration of solute is uniform
and y = y2 in Eq. (9.39).
3. Calculate the total volume of spray section by the following equation:
V =   (9.40)

where, V = Volume of spray section, m3

KGa = Overall volumetric gas phase mass transfer coefficient kmol/ [(s · m3)
(mole fraction of solute in gas)]
G = Flow rate of gas phase, kmol/s
The best method for finding the KGa is from the similar type of commercial gas
absorption unit noting that same value of KGa can be used only with same value
of liquid flow rate to gas flow rate ratio.
9.5.2 Process Design of Horizontal Cross Flow Spray
Chamber9
For the given absorption duty total volume required by horizontal cross-flow
chamber is considerably less than that of a vertical spray tower. Horizontal
chamber is having either rectangular or circular cross section. Gas passes
horizontally through it and liquid is sprayed vertically downward perpendicular
to the direction of gas flow. Process design steps for horizontal cross flow
chamber are as follows.
1. Determination of duct area or diameter: Maximum permissible gas velocity
through horizontal spray chamber is 7 m/s. On fixing the actual gas velocity
through horizontal chamber, duct area or diameter can be determined.
2. Determine the number of gas phase transfer units, required for the desired
separation by the following equation:
NG = ln   (9.41)

where, NG = Number of gas phase transfer unit

y1 = Gas phase solute mole fraction at inlet
y2 = Gas phase solute mole fraction at outlet
3. Total volume of spray section is determined by the same equation that is used
for vertical spray tower. (Eq. 9.40).
In a horizontal spray chamber, gas is well mixed radially by the action of liquid
spray in a cross-current contact but not axially or in the direction of gas flow.
Hence, concentration of solute in gas mixture is not uniform in the spray section.
Design the spray chamber for the following duty:
Volumetric flow rate of boiler flue gas = 24 000 Sm3/h
Pressure of gas = 150 mm WC (gauge)
Temperature = 80 to 90 °C
SO2 concentration in boiler flue gas = 4000 ppm = 4000 × 10-6 kmol/kmol gas
Desired concentration of SO2 in the outgoing gas
= 200 ppm = 200 × 10-6 kmol/kmol gas
Solvent = 1% lime solution
Solution
Type of scrubber: Spray chamber
Position: Horizontal or vertical
(a) Tower diameter
(i) For vertical spray chamber:
Volumetric flow rate of gas = 24 000 Sm3/h
Volumetric flow rate of gas at actual operating condition
qv =

= 31 455.5 m3/h ≡ 8.7376 m3/s

Maximum permissible gas velocity through vertical spray chamber, ug max = 2.3
m/s
Let actual velocity of gas through spray tower.
ug = 66% of ug max
ug = 0.66 × 2.3 = 1.518 m/s
 =

where, Di = Inside diameter of spray tower

Di = = 2.707 m

Let Di = 2.71 m ≡ 2710 mm

(ii) Horizontal cross-flow spray chamber:
Maximum permissible gas velocity through horizontal spray chamber ug max = 7
m/s
Let actual velocity of gas through horizontal spray chamber, ug = 66% of ug max
Let cross-sectional area of horizontal spray chamber, A = W2 for square duct
where, W = Width of chamber = Height of chamber
W 2 = = 1.8913

W = 1.375 m
(b) Number of gas phase transfer units, NG
(i) Number of gas phase transfer units required for desired separation, for
vertical spray tower:
NG =   (9.38)

= = 19

(ii) For horizontal spray chamber:

NG = ln = ln = 3  (9.41)

(c) Circulation rate of 1% lime solution

(i) For vertical spray chamber:
To decide the circulation rate and height of gas phase transfer unit duplication of
similar/same type of commercial gas absorption unit is the best method.
From the existing vertical spray tower of one of the large scale industry,
following data are available:
Flow rate of boiler flue gas = 12 000 Sm3/h
Temperature of incoming gas = 80 to 90°C
SO2 concentration in inlet stream = 2500 to 4000 ppm
SO2 concentration in outlet stream = Less than 400 ppm
Circulation rate of 1% lime solution = 30 m3/h
Tower diameter = 1.7 m
Height of spray zone = 3.5 m
Pressure drop = 100 mm WC (maximum)
Based on these data, circulation rate required for new tower = 30 × = 60
m3/h
(ii) For horizontal spray chamber:
For the trial run, let the circulation rate of 1% lime solution for horizontal spray
chamber = 60 m3/h
(d) Height of tower
(i) Vertical spray tower: Eq. (9.40)
For existing tower
NG = = 9

V = (1.7)2 × 3.5 = 7.944 m3

Gm = = 535.38 kmol/h ≡ 0.1487 kmol/s

For existing tower, KGa = = 0.1685 kmol/(m3 · s)

Volume of spray section required for new tower

V =

= 33.54 m3 = (π /4)D2iH
where, H = Height of spray section required
33.54 = (π /4) × (2.7)2 × H
H = 5.86 m
Let H = 6 m
(ii) For horizontal spray chamber:
Mass transfer data are available for vertical spray tower only. Assume the overall
volumetric mass transfer coefficient KGa for cross flow arrangement in
horizontal spray chamber (less than the same for vertical position), to be 0.1
kmol/(m3 · s).
V = = 8.922

V = 8.922 m3 = W 2L = (1.375)2 L
Length of spray section required with horizontal position, L = 4.72 m
Let L = 5 m
(e) Approximate pressure drop calculations
(i) For vertical spray chamber:
Pressure drop Δp ∝ m.1.84 d–4.84L [Ref. Eq. (5.5a)]
Pressure drop in actual vertical spray tower is 100 mm WC (maximum).
Pressure drop in proposed new vertical spray tower,
Δp = 100 ×

Δp = 64.24 mm WC (within permissible limit)

(ii) For horizontal spray chamber:
Δp = 100 ×

Δp = 1428 mm WC (very high)

To decrease the pressure drop, duct area must be increased.
Let W = 2 m and L = 2.5 m
Δp = 100 ×

= 116.45 mm WC (within permissible limit)

Note: For this approximate calculation, width or height of chamber W is
considered as diameter.
Actually duct area (W2) is higher than circular cross-sectional area and
hence it will offer less pressure drop than the circular cross section.
Volume of spray section, V = (2)2 × 2.5 = 10 m3 (revised)
(f) Make-up of lime solution
Lime required = Lime consumed in reaction + Drift loss
Liquid particles entrained in the outgoing gas.
Let drift loss = 0.2% of circulation rate (based on experience in cooling tower).
Drift loss = × 60 = 0.12 m3/h ≅ 120 kg/h

Loss of lime as drift = 1.2 kg

Lime is consumed in reaction.
Ca(OH)2 + SO2 → CaSO3 + H2O

SO2 consumed in reaction = × (4000 – 200) × 10–6 = 4.07 kmol/h

Ca(OH)2 consumed in reaction = 4.07 kmol/h = 4.07 × 74 = 301.18 kg/h

Lime required as CaCO3 = (301.18 + 1.2) × = 408.62 kg/h

Loss of water = Drift loss + Evapouration rate + Blow down ≅ 1% of circulation

rate
CaCO3 addition required = 408 kg/h
Water addition required = 6 m3/h = 6000 kg/h
Summary of results
Vertical spray tower Horizontal spray chamber
1. Tower diameter = 2.71 m 1. Duct width and height = 2 m
2. Height of spray section = 6 m 2. Length of spray section = 2.5 m
3. Volume of spray section = 33.54 m3 3. Volume of spray section = 10 m3
4. Pressure drop = 64 mm WC 4. Pressure drop = 116.45 mm WC
5. Circulation rate = 60 m3/h 5. Circulation rate = 60 m3/h
6. Lime as CaCO3 required = 408 kg/h 6. Lime as CaCO3 required = 408 kg/h
7. Makeup water required = 6 m3/h 7. Makeup water required = 6 m3/h
Comment: From the resulting data, it is clear that horizontal spray chamber is
a better option. But it is less common in operation as compared to vertical
spray tower. Here mass transfer data used for calculation are available for
vertical spray tower. Hence, design of the vertical spray tower is more reliable
as compared to the horizontal spray chamber.

9.6 VENTURI SCRUBBER3, 10

Venturi scrubber is normally preferred for removing particulate matter from gas
stream as opposed to absorbing soluble vapour. Efficient contact between liquid
and gas is obtained in venturi scrubber. Gas is drawn into the throat of the
venturi by a stream of absorbing liquid solvent sprayed into the convergent duct
section. Compared to spray absorber, more efficient contact between gas and
liquid is provided by the venturi scrubber but at the expense of relatively large
pressure drop and a consequent higher power consumption. Spray absorber
cannot remove very fine particles (below 10 μm size) while the venturi scrubber
can remove the fine particles up to the size 2 μm effectively.
Venturi scrubber provides cocurrent contact (over all) between liquid and gas.
The fractional solute removal for a dilute system in venturi scrubber is given by
the following equation:

Fractional solute removal =   (9.42)

where, y1 = Mole fraction of solute in gas phase at outlet

y2 = Mole fraction of solute in gas phase at inlet
m = Slope of equilibrium curve
x2 = Mole fraction of solute in incoming solvent
GM = Molar gas phase velocity, kmol/(m2 · s)
LM = Molar liquid phase velocity, kmol/(m2 · s)
η = Mass transfer efficiency
η = 1 – e–NG  (9.43)
where, NG = Number of overall gas-phase mass transfer units
= (KGaRT/pt)θc   (9.44)
where, KGa = Overall volumetric gas-phase mass-transfer coefficient for dilute
systems, kmol/(s · m3)
R = Ideal gas constant = 8314 J/(kmol · K)
T = Operating temperature of gas, K
θc = Effective gas-liquid contact time, s
pt = Operating pressure, Pa
The number of transfer units, NG that can be achieved in venturi scrubber is in
the range of 1 to 2 for physical absorption.
The liquid to gas ratio for venturi scrubber is in between 0.7 to 2.7 L/m3. Throat
velocity is kept in between 60 to 150 m/s. The estimated pressure drop or
vacuum created by the venturi scrubber is given by following equation (Hesketh
equation9).
Δp = 2.584 × 10–3 v2G ρG Ath0.133(L′/G′ )0.78  (9.45)
where, Δp = Venturi pressure drop from inlet duct to the outlet duct, cm WC
vG = Gas velocity in the throat, m/s
ρG = Gas density, kg/m3
Ath = Venturi throat area, cm2
L′/G′ = Liquid to gas ratio, L/m3
Venturi scrubber can be selected as an absorber and placed before a packed
tower type absorber in the following two cases:
1. If the feed gas to absorber is contaminated by fine solid particles.
2. If vent gases from the process are available at very low pressure. In this case,
venturi scrubber is designed such that suction or negative pressure to be
developed by the venturi scrubber is sufficient to overcome the pressure drop
across the packed tower.
In both cases, venturi scrubber also decreases load of absorption of the packed
tower as some absorption takes place in the venturi scrubber. Absorption
equivalent to 1 to 2 theoretical stages in case of physical absorption and could be
higher than two stages in case of absorption with chemical reaction.
Venturi scrubber itself acts as liquid jet ejector and creates negative suction or
vacuum. Typical entrance cone angle is 21 ± 2° and exit cone angle is 5 to 15°.
Throat length is equal to diameter of throat (d). Diameters of both cones at large
ends (other ends) are equal to 3d.
Predict the fractional solute removal and pressure drop in a
venturi scrubber based on the following data:
Volumetric flow rate of boiler flue gas = 24 000 Nm3/h
Discharge pressure of gas from venturi = Atmospheric
Temperature of gas = 80 to 90°C
SO2 concentration in boiler flue gas = 4000 ppm (or mg/kg)
Solvent = 1% lime solution
Solvent to gas ratio = 1.4 L/m3
Recommended throat velocity of gas phase = 100 m/s
Average molar mass of flue gas = 29.48 kg/kmol
Solution
Volumetric flow rate of gas at actual operating condition.
qV = × (273 + 90) = 31 912.1 m3/h

≡ 8.8645 m3/s
Gas velocity in throat = 100 m/s
Inside diameter of throat,
di = = 0.336 m

Let angle of upstream cone = 21°

angle of downstream cone= 10°
Diameters of cones at large ends = 3 di = 1.008 m
Height of venturi = Height of upstream cone
+ Height of throat + Height of downstream cone
= [0.336/tan(21/2)] + 0.336 + [0.336/tan(10/2)]
= 5.99 m, say 6 m
Flow rate of solvent (1% lime solution) = 1.4 × 8.8645 = 12.41 L/s ≡ 44.676
m3/h
Density of 1% lime solution = 1012.5 kg/m3
Mass flow rate of solution = 45 234.5 kg/h
ṁL = 12.565 kg/s

Molar flow rate of liquid, LM = = 0.693 kmol/s

Molar flow rate of gas, GM =

= 0.2978 kmol/s
Assuming NG = 1 (safe value)
Mass transfer efficiency,
η = 1 – e–NG
η = 1 – e–NE = 1 – e–1 = 0.6321  (9.43)
Fractional solute removal,
1 – =

= = 0.43

x2 = 0, y2 = ?

1 – =

y1 = 4000 × 10–6 = 0.004

Partial pressure of SO2 in gas phase ≅ 0.004 × 760 = 3.04 Torr
Assume the average temperature of scrubber = 50°C
Equilibrium mass of SO2 per 100 mass of H2O = 0.035 (Table 3–144 of Ref. 3)
Equilibrium mass fraction = ≅ 3.5 × 10–4

Equilibrium mole fraction =

xe = 9.846 × 10–5
ye = 0.004
m = = 40.63

Equilibrium constant for physical absorption at 50°C and 1 atm pressure, m =

40.63
For low concentration range m can be assumed constant.
1 – = = 0.034 22

= 0.9658

y2 = 0.003 86
Here, presence of 1% lime in solution or chemical reaction between lime and
SO2 is not considered. Chemical reaction enhances the rate and extent of
absorption considerably. To find this, lab-scale or pilot plant scale data are
required on the same system and also on the same type of absorber. Reliable
theoretical equations are not available for the same.
Pressure drop by Eq. (9.45).
vG = 100 m/s
Ath = Throat area = (33.6)2 = 886.7 cm2

Density of gas ρG = = = 0.99 kg/m3

Δp = 82 cm WC ≡ 820 mm WC
Pressure drop provided by venturi scrubber (820 mm WC) is negative pressure
or vacuum generated by the same.

9.7 PROCESS DESIGN OF FALLING FILM

ABSORBER11, 12
Falling film absorber is a shell and tube heat exchanger [Fig. 9.17(a) and Fig.
10.12] in which absorption of gas in liquid is carried out inside the tubes while
cooling medium like cooling water is circulated on the shell side. It is used for
highly exothermic absorption like absorption of hydrogen chloride gas in water,
absorption of ammonia gas in water, absorption of sulphur trioxide in alpha
olefins, absorption of sulphur trioxide in methyl esters of saturated fatty acids,
etc.
Use of graphite block-type heat exchanger is increasing for similar applications
in which cocurrent absorption is carried out in vertical holes and cooling
medium is circulated in horizontal holes. Schematic representation of block-type
heat exchanger is given in Fig. 9.17(b).
 Fig. 9.17 Graphite Heat Absorbers

9.7.1 Advantages of Falling Film Absorber over

Packed Tower-Type Adiabatic Absorber
1. Falling film absorber gives better quality of product solution. Concentration of
HCl in product solution from falling film absorber can be 35% to 37% (by
mass), while the same from adiabatic packed tower ranges from 25% to 30%.
2. Absorption is more efficient in falling film absorber for the given absorption
duty. Falling film absorber requires less space compared to adiabatic packed
tower-type absorber.
3. Other advantages of falling film absorber are lower temperatures, lower
pressure drops and higher turn down.
9.7.2 Disadvantages of Falling Film Absorber
1. More difficult to control and is sensitive to liquid and gas distribution
problems.
2. It has greater tendency to absorb undesirable water soluble low boiling
impurities.
9.7.3 Exothermic Absorption of Gases Like HCl and
NH3
In case of exothermic absorption of gas like HCl and NH3 in water, there are two
requirements:
1. Large liquid surface area to get the higher rate of mass transfer.
2. Rapid removal of the heat of absorption.
Both can be done most effectively by causing the liquid to flow down as a film
inside of a vertical tube which is cooled on the outside.
Falling film absorber with countercurrent flow of the gas and liquid suffer from
the limitation that flooding occurs when the gas velocity exceeds a limiting value
of the order of 4.5 m/s for 25 mm OD tube. A sample calculation shows that at
the usual ratio of liquid and gas flow rates, the liquid rate is barely sufficient to
maintain a liquid film on the tube wall. Hence, countercurrent contact is not
preferred. Cocurrent contact of gas and liquid allows the gas velocity up to 18
m/s.
For the absorption of hydrogen chloride gas falling film absorber (with cocurrent
contact of gas and liquid) followed by tail gas scrubber is preferred if the
concentration of hydrogen chloride gas in feed gas mixture is more than 40% (by
mole). Nearly, 66 to 85% of absorption is carried out in falling film absorber
while the balance is carried out in the tail gas scrubber. If the concentration of
hydrogen chloride in feed gas mixture is less than 40% (by mole), then it is
difficult to produce the product solution of commercial strength. With cocurrent
contact, partial pressure of hydrogen chloride in exist gas is greater than or equal
to exit gas equilibrium partial pressure of hydrogen chloride in product solution.
While most HCl gas will be scrubbed in these two scrubbers, local
environmental standards could mandate an additional vent scrubber in which
final wash of the outgoing gas mixture is carried out by spraying small quality of
water. However, the bottom solution from the vent scrubber will contain some
HCl and its disposal may require neutralisation with alkali.
Low concentration of HCl in exist gas gives low concentration of HCl in product
solution. The arrangement of falling film absorber with tail gas scrubber is
shown in Fig. 9.18.

Fig. 9.18 Typical Falling Film Absorption System

Feed gas and weak acid from the tail gas scrubber (5 to 10% HCl solution) is
supplied at the top head of falling film absorber. Tubes are extended above the
top tube sheet to facilitate the uniform distribution of liquid in each tube.
Notches may be cut in the top ends of the tubes. Tubes are made from graphite,
glass or tantalum. As per the recommendations tubes must be extended at least
by 125 mm above the top tube sheet to minimise the effect of hydraulic gradient
in the liquid. With cocurrent contact in upper portion of tubes, rate of absorption
is very high. It may lead to the overheating of tubes in upper portion.
Overheating could break the glass tubes. To avoid this, inverted U in the
discharge line of water is essential. Inside the tubes there is a two-phase flow of
liquid and gas in which liquid forms the film over the inside periphery of tubes
while gas flows in the central portion. Flow rates of liquid and gas should be so
adjusted to get annular flow inside the tubes. Absorption of gas in liquid film
generates the heat. Temperature of liquid film is controlled by cooling water
circulated on shell side. In the bottom head, gas and product solution are
separated. Exit gas is in the equilibrium of product solution at outlet condition.
Hence, complete absorption of HCl with this arrangement is impossible.
Remaining HCl is absorbed by contact with pure water in tail gas scrubber
which is commonly used and is normally a packed tower-type absorber.
9.7.4 Process Design of Falling Film Absorber
Consider absorption of HCl in water.
1. Fix the inlet and outlet concentrations of aqueous hydrochloric acid. Also, fix
the concentration of hydrogen chloride in the exist gas stream. Partial
pressure of hydrogen chloride in the exist gas stream at bottom should be
greater than vapour pressure of hydrogen chloride in product solution at outlet
temperature. Based on this fact, outlet temperature and concentrations of
liquid and gas streams can be fixed.
2. Determine the overall rate of absorption and based on the same, determine the
total heat duty. Actually rate of absorption and hence, rate of heat transfer
required are changing from point to point of the heat exchanger (i.e., falling
film absorber)
3. Based on the heat balance, find the flow rate of cooling medium.
4. In the end, heat transfer area of falling film absorber must be determined:
A = Nt π doL = (6.17)

To facilitate two-phase flow, do (tube outside diameter) can be selected as 25.4

mm. Number of tubes must be decided based on following criteria:
1. For cocurrent contact velocity of feed gas mixture at inlet should be less than
or equal to 18.3 m/s. To get the higher rate of mass transfer gas velocity at
tube, inlet should be around 15 m/s as per the recommendations. (p. 409 of
Ref. 11)
2. For efficient operation of falling film absorber certain minimum liquid flow
rate is required to maintain a continuous film over the tube wall. Minimum
required flow rate of liquid is 150 kg/(h · m of inside periphery of tubes).
Actually once velocity of gas is fixed, flow rate of liquid per unit periphery of
tube is fixed by material balance and reverse is also true.
Tube length can be easily determined based on average value of overall heat
transfer coefficient U and based on the overall values of heat duty (ϕ) and overall
mean temperature difference (ΔTm) (difference between temperature of liquid
film and temperature of water). Actually rate of absorption, rate of heat transfer,
overall heat transfer coefficient and temperature difference across the heat
transfer wall are changing from point to point. Hence, for the precise
calculations, heat exchanger can be divided in different zones and for each zone,
tube length is determined. Also for each zone, rate of absorption, heat duty and
mean temperature difference are calculated. Overall coefficient U is assumed as
constant. In trial-and-error calculations, film temperature is assumed. Based on
that rate of absorption per unit area, heat duty per unit area and ΔTm are
calculated. Check ϕ/A = UΔTm. Rate of absorption of gas is given by equation:
Rate of absorption, NA = KG M   (9.46)
where, KG = Overall mass transfer coefficient based on the arithmetic mean
partial pressure difference in kmol/(m · atm · s)
M = Molar mass of gas absorbed kg/kmol = 36.5 for HCl
or 17 for NH3
= Partial pressure of solute gas in gas mixture, atm
= Equilibrium partial pressure of solute A at gas liquid interface, atm
Mass transfer coefficient for hydrogen chloride absorption in the tubular
absorber is calculated by following equation11:
KG =   (9.47)

where, di = Inside diameter of tube, m

Gt = Mass velocity of gas mixture at inlet, kg/(m2 · s)
μ = Viscosity of gas mixture kg/(m · s)
Mav = Average molar mass of gas mixture at inlet, kg/kmol
Heat transfer coefficient of liquid film can be calculated by McAdams, Drew and
Bay’s equation (Ref. 11).
h = 9136 × τ 1/3  (9.48)
where, h = Tube side heat transfer coefficient, W/(m2 · °C)
τ = Liquid flow rate per unit perimeter, kg/(m · s)
This correlation is valid for values of τ, ranging from 0.25 to 6.2 kg/(m · s).
Gas mixture, obtained after the chlorination reaction and
separation of organic vapours, contains 20% Cl2 and 80% HCl by mole. From
the mixture, hydrogen chloride gas is to be separated by absorption in water. For
the absorption of HCl, falling film absorber (Fig. 9.19), followed by tail gas
scrubber are used. Design the falling film absorber.
Flow rate of gas mixture = 2500 kg/h
Temperature of gas mixture = 10°C
% of HCl in the liquid stream leaving the tail gas scrubber and entering at the top
of falling film absorber = 5% (by mass)
Desired concentration of HCl in product stream = 32% (by mass)
Operating pressure = 30 kPa g (at inlet)
 Fig. 9.19 Falling Film Absorption System

Solution
Amount of product solution = qm kg/h
HCl balance
2500 × = qm × 0.32

2500 × 0.6728 = qm × 0.32

qm = 5256.25 kg/h
Tail gas scrubber is operated in adiabatic manner. Hence, the total heat of
solution of hydrogen chloride in water (which is highly exothermic) is removed
in falling film absorber to maintain nearly isothermal condition. Heat duty of
falling film absorber can be calculated based on heat of formation data of
aqueous hydrochloric acid solution (Table 9.8).
Table 9.8 Heat of Formation of Aqueous Hydrochloric Acid Solution13
Formula State Heat of formation (ΔHfo) at K, 298 kJ/mol HCl mass % HCl in aqueous solution

HCl 75 H2O aq –165.72 2.63

HCl 50 H2O aq –165.36 3.90

HCl 40 H2O aq –165.10 4.82

HCl 30 H2O aq –164.67 6.33

HCl 25 H2O aq –164.34 7.50

HCl 20 H2O aq –163.85 9.21

HCl 15 H2O aq –163.03 11.9

HCl 10 H2O aq –161.32 16.9

HCl 8 H2O aq –160.00 20.2

HCl 6 H2O aq –157.68 25.3

HCl 5 H2O aq –155.77 28.9

HCl 4 H2O aq –152.92 33.6

HCl 2 H2O aq –140.96 50.3

HCl H2O aq –121.55 67.0

HCl gas –92.31 100.0

Note: The number in the first column indicates the number of moles of water
mixed with one mole of hydrogen chloride.
Heat of solution = (Δ Hfo ) of 32% solution – (Δ Hfo ) of HCl gas
Heat of formation of 32% HCl solution = –153.9 kJ/mol HCl (by interpolation)
Heat of solution = (–153.89) – (–92.31) = –61.58 kJ/mol HCl
Moles of HCl absorbed = = 46.082 kmol/h
Heat duty of falling film absorber (including that of tail gas scrubber),
ϕt = 61.58 × 1000 × 46.082 = 2837 729.6 kJ/h ≡ 788.26 KW
(Actually in falling film absorber, concentration of HCl in aqueous solution is
increased from 5% to 32% but the tail gas scrubber is operated in adiabatic
manner hence heat of solution generated on increasing the concentration from 0
to 5% of HCl is to be added in heat duty of the falling film absorber.)
Temperature of 5% weak acid:
Heat of formation for 5% HCl solution = –165.05 kJ/mol (by interpolation)
Heat of solution = (–165.05) – (–92.31) = –72.74 kJ/mol
Moles of HCl absorbed in adiabatic absorber
=

= 5.154 kmol/h
≡ 188.121 kg/h
Heat generated in adiabatic absorber
ϕ1 = 72.74 × 1000 × 5.154 = 374 902 kJ/h ≡ 104.14 kW
Mass flow rate of pure water entering to adiabatic absorber = Mass flow rate of
32%
solution × (1 – 0.32)
m.w = 5256.35 × (1 – 0.32) = 3574.25 kg/h
Mass flow rate of 5% HCl solution
m.w1 = = 3762.37 kg/h

Enthalpy of 5% solution = Enthalpy of pure water + Enthalpy of incoming gas

mixture – Enthalpy of exist gas from adiabatic absorber + Heat of solution
Let reference temperature = 30°C
Here pure water enters at room temperature or at a temperature close to 30°C.
Temperatures of incoming and exist gas mixture are also close to 30°C. Heat of
solution must be determined at average temperature. Neglect the change in the
value of heat of solution with temperature.
Heat capacity of 5% HCl solution = 4.19 kJ/(kg · °C)
Enthalpy of 5% HCl solution = m.w1 · CL (t – 30) = ϕ1
3762.37 × 4.19(t – 30) = 374 902
t = 53.78°C
Mass flow rate of cooling water:
Let inlet temperature of cooling water to absorber = 32°C and outlet temperature
of cooling water from absorber = 38°C
Mass flow rate of cooling water,

m.cw = = 31.379 kg/s

= 112 964.4 kg/h ≈ 113 m3/h

Material of tube = Graphite
Outside diameter of tube do = 25 mm
Inside diameter of tube di = 20 mm
It is reported in literature that if the velocity of gas through the tubes is kept
more than 15.24 m/s then higher tube side or film side heat transfer coefficient is
obtained. There is a sharp increase in the tube side heat transfer coefficient when
the velocity of gas at the tube inlet exceeds 15.24 m/s.
Let the velocity of gas at tube inlet = 16 m/s
Average molar mass of feed gas mixture at inlet
Mav = Σ(Mi · yi) = 0.2 × 71 + 0.8 × 36.5 = 43.4 kg/kmol
Density of feed gas at inlet
ρ =

ρ = 2.42 kg/m3
Volumetric flow rate of gas mixture at tube inlet, qv =

= 1033 m3/h

Velocity of gas at tube inlet, v = = 16

Nt =

Tube side mass velocity, Gt = ut ρ = 16 × 2.42 = 38.72 kg/(m2 · s)

Tube side heat transfer coefficient
Tube side or falling film side heat transfer coefficient
hi = 9136 × τ1/3  (9.48)
where, hi is in W/(m2 · °C)
Liquid flow rate at inlet = 3762.37 kg/s = qw1
τ =   (9.49)

τ = 0.2918 kg/(m · s) [= 1050.5 kg/(m · h) > 150 kg/(m · h)]

hi = 9136 × (0.2918)1/3 = 6059.7 W/(m2 · °C)
Shell side heat transfer coefficient
Shell side flow area, As = (6.29)

Pt = Tube pitch = 1.25 do = 1.25 × 25 = 31.25 mm, do = 25 mm

Tube bundle diameter
Db = (6.1)

For Pt/do = 1.25, triangular pitch, 1-1 shell and tube fixed tube sheet heat
exchanger, k1 = 0.319, n1 = 2.142 (From Table 6.2)

Db = = 281.4 mm

Ds = Db + 18.6 mm = 300 mm, Let Bs = 100 mm

As = × 0.1 × 0.3 = 0.006 m2

Shell side equivalent diameter

de = (6.33)

= (31.252 – 0.907 × 252) = 18.026 mm

Shell side mass velocity, Gs = = 5229.8 kg/(m2 · s)

Shell side velocity, uS = = 5.26 m/s > 3 m/s (too high)

For the liquid flow on shell side, it is better to restrict the shell side velocity
below 3 m/s to avoid the liquid induced vibrations.
Change the baffle spacing from 100 to 300 mm.
Bs = 300 mm
As = × 0.3 × 0.3 = 0.018 m2

Shell side mass velocity, Gs = = 1743.3 kg/(m2 · s)

us = = 1.7538 m/s

Shell side Reynolds number, Re = = 39 280.9

(Viscosity of water at 35°C, μ = 0.8 cP)

Prandtl number for water at 35°C
Pr = = 5.333

Evaluation of ho (shell side heat transfer coefficient):

ho = 0.36 × × (39 280.9)0.55 × (5.333)0.33 ×

where, ≅ 1

ho = 7329.3 W/(m2 · °C)

Overall heat transfer coefficient:
Fouling coefficients, hod = hid = 5000 W/m2 · °C)
Thermal conductivity of graphite = 150 W/(m · K) from Appendix of Ref. 14.

Uo = 1232.6 W/(m2 · °C)

The temperature of the acid film on inside surface of tubes attains an equilibrium
value at which the rate of heat transfer to the cooling water balances the rate of
heat release due to absorption. This temperature is calculated by trial-and-error
method for a series of points in the absorber.
Point 1: Bottommost point of tubes: At this point aqueous acid film
concentration of liquid film is 32% (by mass). Liquid composition at point 1:
Table 9.9 Composition of Bottom Solution (Point 1)
Component kg/h mass % kmol/h

HCl 1682.00 32 46.082

H2O 3574.25 68 198.570

Total 5256.25 100 244.650

HCl amount in 5% acid stream = 188.12 kg/h

HCl balance around falling film absorber:
HCl in exist gas stream = HCl in inlet gas stream + HCl in 5% acid stream
– HCl in product stream
HCl in exist gas stream = 2500 × 0.6728 + 188.12 – 1682 = 188.12 kg/h
Cl2 in exist gas stream = 2500 (1 – 0.6728) = 818 kg/h
Table 9.10 Gas Composition at Point 1
Component kg/h mass % kmol/h mole %

HCl 188.12 18.7 5.154 30.91

Cl2 818.00 81.3 11.520 69.09

Total 1006.12 100.0 16.674 100.00

Partial pressure of HCl gas in gas mixture at point 1

= pt × yHCl
= 131.325 × 0.3091
= 40.6 kPa ≡ 0.4007 atm
Equilibrium Partial Pressure of HCl (p–*AL)
Equilibrium Partial Pressure of HCl gas over aqueous solution of hydrochloric
acid can be calculated by the following equation:
log =

where, is in Torr and T is in K (Table 3.11 of Ref. 3)

Table 9.11 Antoine Equation Constants3
% HCl by mass A B

2 11.8037 4736
4 11.6400 4471
6 11.2144 4202
 8 11.0406 4042
10 10.9311 3908
12 10.79 3765
14 10.6954 3636
16 10.6261 3516
18 10.4957 3376
20 10.3833 3245
22 10.3172 3125
24 10.2185 2995
26 10.1303 2870
28 10.0115 2732
30 9.8763 2593
32 9.7523 2457
34 9.6061 2316
36 9.5262 2279

Let the temperature of acid film at point 1 = 35.5oC. For 32% HCl solution,
equilibrium partial pressure of HCl over solution,
log = 9.7523 –

= 61.37 Torr ≡ 0.08 atm

Rate of Absorption, NA = KGM kg/(m2 · atm · s)  (9.46)
where, M = Molar mass of HCl = 36.5 kg/mol
Mass transfer coefficient

KG =   (9.47)

Average molar mass Mav = Σ( yiMi)

Mav = 0.3091 × 36.5 + 0.6909 × 71 = 60.336 kg/mol

di = 0.02 m, Gt = = 15.607 kg/(m2 · s)

Viscosity of gas mixture at low pressure can be calculated by the following

equations:

μ mix = (6.77)
where, Θij =

and Θji =

At 35.5°C, μHCl = μi = 0.014 cP

μCl2 = μi = 0.014 cP
μmix = 0.014 cP = 0.014 × 10–3 kg/(m · s)

KG = = 2.823 × 10–4 kmol/(m2 · atm · s)

NA = 2.823 × 10– 4 × 36.5 × (0.4007 – 0.08) = 3.3 × 10–3 kg/(m2 · s)

Heat of solution of 32% HCl
Δ Hs = – 61.58 kJ/mol ≡ – 61 580 kJ//kmol ≡ –1687.12 kJ/kg
Heat release due to absorption = NA × (–ΔHs) = 3.3 × 10–3 × 1687.12
= 5.567 kW/m2
Rate of heat transfer at point 1 between acid film and cooling water
= UΔt = 1232.6 × (35.5 – 32)

= 4314.1 W/m2 ≡ 4.314 kW/m2

Heat release due to absorption ≠ UΔt
Second trial
Let acid film temperature at point 1 = 36.5°C
Equilibrium partial pressure of HCl for 32% solution at 36.5°C:
= 65.116 Torr = 0.0857 atm
Heat release due to absorption = 2.8233 × 10–4 × 36.5 (0.4007 – 0.0857)
= 3.25 × 10–3 kg/(m2 · s)
Heat release due absorption = 3.25 × 10–3 × 1687.12 = 5.48 kW/m2
= UΔt = 1232.6 × (36.5 – 32)

= 5546.7 W/m2
= 5.546 kW/m2
Heat release due to absorption ≈ UΔt = q/A
Hence, temperature of acid film at point 1 = 36.5°C
Point 2: At this point concentration of aqueous acid film is 26% (by mass).
Mass flow rate of 26% HCl solution = = 4830.07 kg/h

Amount of HCl absorbed in between points 1 and 2 = 5256.25 – 4830.07

= 426.18 kg/h
Composition of liquid mixture at point 2:
Table 9.12 Liquid Composition at Point 2
Component kg/h mass %

HCl 1255.82 26
Water 3574.25 74
Total 4830.07 100

Table 9.13 Composition of Gas Mixture at Point 2

Component kg/h kmol/h mole %

HCl 614.3 16.83 59.365

Cl2 818.0 11.52 40.635

Total 1432.3 28.35 100.000

For the first trial calculations let the temperature of 26% acid film at point 2 =
40°C
log = 10.1303 –

= 9.14 Torr = 0.012 atm

= yHCl × pt = 0.593 65 × 131.325 = 77.96 kPa ≡ 0.7694 atm
Average molar mass of gas mixture
Mav = Σ( yiMi) = 0.593 65 × 36.5 + 0.403 35 × 71 = 50.306 kg/kmol
Mass velocity of gas mixture

Gt = = 22.218 kg/(m2 · s)

At point 2
At 40°C, μCl2 = 0.0145 cP, μHCl = 0.0145 cP
KG =

= 5.334 × 10–4 kmol/(atm · m2 · s)

Rate of absorption, NA = KG M
NA = 5.334 × 10–4 × 36.5 × (0.7694 – 0.012) = 0.014 746 kg/(m2 · s)
Heat of solution for 26% HCl solution
ΔHs = (ΔHfo )26% HCl – (ΔHfo )HCl gas
= (–157.3) – (92 .31) = –65 kJ/mol
≡ –65 000 kJ/kmol ≡ –1780.82 kJ/kg
Heat release due to absorption (per unit area)
ϕ = NA × (–ΔHs ) = 0.014 746 × 1780.82 = 26.26 kW/m2
Heat duty required in between 1 and 2
ϕ12 = Total heat duty of falling film absorber – Heat duty required to form 26%
HCl solution
ϕ12 = 788.26 kW – (–ΔHs) × kmol/h of HCl absorbed to form 26% solution

ϕ12 = 788.26 – 65 000 ×

= 167.0435 kW = qcw CL(t – 32)

167.0435 = 31.379 × 4.1868 × (t – 32)
Temperature of cooling water at point 2, t = 33.27°C
= U Δt = 1232.6 × (40 – 33.27) = 8295.4 W/m2 ≡ 8.295 kW/m2

= 8.295 kW/m2 ≠ Heat release due to absorption.

Second trial
Let temperature of 26% acid film at point 2 = 54°C
Equilibrium partial pressure of HCl over solution
log = 10.1303 –

= 22.57 Torr ≡ 0.0297 atm

At 54°C, μHCl = μCl2 = 0.015 cP (Fig. 3.14 of Ref. 3)
KG = 5.334 × 10–4 ×
= 4.9899 × 10–4 kmol/(m2 · atm · s)
NA = 4.9899 × 10–4 × 36.5 × (0.7694 – 0.0297)
= 0.013 47 kg/(m2 · s)
Heat release due to absorption = 0.013 47 × 1780.82 = 23.9876 kW/m2
= U Δt = 1232.6 × (54 – 33.27)

= 25 551.8 W/m2 = 25.55 kW/m2 ≠ Heat release

Third trial
Let temperature of 26% Acid film = 52.5°C
= 20.563 Torr = 0.027 atm
NA = 4.9899 × 10–4 × 36.5 (0.7694 – 0.027) = 0.013 52 kg/(m2 · s)
Heat release due to absorption = 0.013 52 × 1780.82 = 24.076 kW/m2
= U Δt = 1232.6 × (52.5 – 33.27) = 23 703 W/m2 ≡ 23.7 kW/m2

= U Δt ≈ Heat release due to absorption (close enough)

Temperature of acid film = 52.5°C

Heat transfer between points 1 and 2:
ϕ12 = 167.0435 kW, Uo = 1232.6 W/(m2 · °C)

Δ T ln = = 10.142°C

Area required A12 = = 13.362 m2

Nt π do L12 = A12 = 13.362 m2 or L12 = = 2.984 m

Point 3: At this point aqueous acid film contains 20% by mass HCl.
Let liquid film temperature at point 3 = 68°C
Table 9.14 Liquid Composition at Point 3
Component kg/h mass % kmol/h

HCl 893.56 20 24.48

H2O 3574.25 80 198.57
 Total 4467.81 100 223.05

Table 9.15 Gas Composition at Point 3

Component kg/h kmol/h mole %

HCl 976.56 26.755 69.9

Cl2 818.00 11.52 30.1

Total 1794.56 38.275 100.0

Equilibrium partial pressure of HCl over 20% HCl solution:

log = 10.3883 –

= 10.3833 –

= 7.365 Torr ≡ 0.0097 atm

= pt × yHCl = 131.325 × 0.699 = 91.8 kPa ≡ 0.906 atm
Average molar mass of gas mixture,
Mav = Σ(Mi yi) = 0.699 × 36.5 + 0.301 × 71 = 46.885

Mass velocity, Gt = = 27.8375 kg/(m2 · s)

At 68°C, μ = μHCl ≅ = 0.0162 cP

KG =

KG = 6.766 × 10–4 kmol/(atm · m2 · s)

Rate of absorption,
NA = KG M
= 6.766 × 10–4 × 36.5 × (0.906 – 0.0097)
= 0.022 14 kg/(m2 · s)
Heat of solution of 20% HCl solution ΔHs = (–160.08) – (–92.31)
ΔHs = –67.77 kJ/mol ≡ –67 770 kJ/mol ≡ –1856.7 kJ/kg
Rate of heat generated due to absorption = NA × (–ΔHs)
= 0.022 14 × 1856.7 = 41.1073 kW/m2
Amount of HCl absorbed in between point 3 and 2 = 362.26 kg/h
Heat duty required in between points 2 and 3
ϕ23 = ϕt – ϕ12 – Heat required to form 20% HCl solution
Heat required to form 20% HCl solution
= –ΔHs × 893.56 kg/h
= 1856.7 × 893.56 = 1659 072.9 kJ/h ≡ 460.85 kW
ϕ23 = 788.26 – 167.044 – 460.85 = 160.366 kW
= 160.336 kW = m.cw · CL (t – 33.27)
160.336 = 31.379 × 4.1868 × (t – 33.27)
t = 34.49°C = Temperature of cooling water at point 3
= UΔt = 1232.6 × (68 – 34.49) = 41 304 W/m2 ≡ 41.304 kW/m2

= Rate of heat generated due to absorption

Temperature of acid film at point 3 = 68°C
ϕ23 = 160.336 kW, U = 1232.6 W/(m2 · °C)

ΔTlm = = 25.71°C

A23 = = 5.0595 m2 = Nt π do L23

57 × π × 0.025 × L23 = 5.0595 m2

L23 = 1.13 m
Point 4: Assume that at this point aqueous acid contains 12% HCl.
Table 9.16 Liquid Composition at Point 4
Component kg/h mass %

HCl 487.4 12
H2O 3574.25 88

Total 4061.65 100

Table 9.17 Gas Mixture Composition at Point 4

Component kg/h mass % kmol/h mole %

Cl2 818 37.17 11.52 23.32

HCl 1382.72 62.83 37.883 76.68

Total 2200.72 100.00 49.403 100.00

Let temperature of acid film at point 4 = 72°C

Equilibrium partial pressure of HCl over 12% HCl solution
log = 10.79 – = 10.79 –

= 0.753 Torr ≡ 9.9 × 10–4 atm

p–AG = yHCl × pt = 0.7668 × 131.325
= 100.7 kPa ≡ 0.9938 atm
Mav = Σ(yiMi) = 0.7668 × 36.5 + 0.2332 × 71 = 44.55 kg/mol

Gt = = 34.138 kg/(m2 · s)

Mass transfer coefficient

At 72°C, μ = μHC l = μC l2 = 0.0163 cP

KG =

= 9.0185 × 10–4 kmol/(m2 · atm · s)

Rate of absorption
NA = KGM
= 9.0185 × 10–4 × 36.5 (0.9938 – 9.9 × 10–4)
= 0.032 68 kg/(m2 · s)
Heat of formation of 12% HCl solution = –163 kJ/mol
Heat of solution of 12% HCl solution
ΔHs =
= –163 – (–92.31) = –70.69 kJ/mol
≡ –70 690 kJ/mol ≡ – 1936.7 kJ/kg
Heat release due to absorption
= 0.032 68 × 1936.7 = 63.29 kW/m2
Heat duty required in between points 3 and 4
ϕ 34 = ϕ t – ϕ 12 – ϕ 23 – Heat duty required to form 12% HCl solution
ϕ 34 = 788.26 – 167.044 – 160.366 – (1936.7 × 487.4) ×

= 198.64 kW = qcw Cl Δt34 = 31.379 × 4.1868 × (t – 34.49)

t = 36°C = Temperature of cooling water at point 4
 = U Δt = 1232.6 × (72 – 36) = 44 374 W/m2 ≡ 44.374 kW/m2

Hence UΔt ≠ heat release due to absorption

Second trial
Let film temperature of 12% HCl solution = 85°C
log = 10.79 –

= 1.876 Torr ≡ 0.0025 atm

Viscosity of gas mixture μ = 0.017 cP
KG = 9.0185 × 10– 4 ×

= 8.647 × 10–4 kmol/(m2 · atm · s)

NA = = 8.64ss7 × 10– 4 × 36.5 × (0.9938 – 0.0025)
= 0.031 29 kg/(m2 · s)
Heat release due to absorption = NA × (–ΔHs ) = 0.031 29 × 1936.7 = 60.6
kW/m2
= UΔt = 1232.6 × (85 – 36) = 60 397 W/m2 = 60.397 kW/m2

≠ Heat release due to absorption

Third trial
Let film temperature of 12% HCl solution = 85.1°C
= 1.889 Torr = 0.0025 atm
NA = 8.647 × 10–4 × 36.5 × (0.9938 – 0.0025) = 0.031 287 kg/(m2 · s)
Heat release due to absorption = 0.031 287 × 1936.7 = 60.59 kW/m2
= U Δt = 1232.6 × (85.1 – 36) = 60 521 W/m2 ≡ 60.52 kW/m2

= U Δt = Heat release due to absorption

ϕ 34 = 198.64 kW, Uo = 1232.6 W/m2

ΔTln = = 40.81°C
A34 = = 3.949 m2 = Nt π do L34

L34 = = 0.88 m

Point 5: At this point concentration of aqueous acid film is 5% HCl (by mass).
Table 9.18 Liquid Composition at Point 5
Component kg/h mass %

HCl 188.12 5
H2O 3574.25 95

Total 3762.37 100

Let film temperature of 5% HCl solution = 97.5°C

For 5% HCl solution
log = 11.4272 –

= 0.5281 Torr = 6.95 × 10–4 atm

= 0.8 × 131.325 = 105.06 kPa ≡ 1.0369 atm
Mav = 0.8 × 36.5 + 0.2 × 71 = 43.4 kg/kmol

Gt = = 38.78 kg/(m2 · s)

Viscosity of gas mixture at 97.5°C, μ = 0.018 cP

KG = = 9.7116 × 10–4 kmol/(atm · m2 · s)

NA = 9.7116 × 10–4 × 36.5 × (1.0369 – 6.95 × 10–4) = 0.036 73 kg/(m2 · s)

Rate of heat generated due to absorption = NA × (–ΔHs)

= 0.036 73 ×

= 73.2 kW/m2
At point 5
= UΔt = 1232.6 (97.5 – 38) = 73 340 W/m2 ≡ 73.34 kW/m2
≅ Heat release due to absorption
Table 9.19 Gas Composition at Point 5
Component kg/h kmol/h mole %

HCl 1682 46.082 80

Cl2 818 11.521 20

Total 2500 57.603 100

Equilibrium acid film temperature at point 5 = 97.5°C

ϕ 45 = 788.26 – 167.044 – 160.366 – 198.64
ϕ 45 = 262.2 kW, U = 1232.6 W/(m2 · °C)

ΔTlm = = 54.134°C

A45 = = 3.9295 m2

L45 = = 0.878 m

Aqueous 5% HCl solution enters the falling film absorber at 53.78°C. Hence, it
will absorb the heat to reach equilibrium temperature (97.5oC). Actually this
direct heating of liquid film will reduce some heat duty of heat exchanger. So,
further addition in heat transfer area is not required.
Table 9.20 Summary of Zonewise Calculations for Falling Film Absorber
% HCl Water, Temperature of acid film, Temperature of water Heat Transfer Area, Length from, bottom,
Point
(mass) kg/s °C °C m2 m

1 32 31.379 36.5 32
(bottom)
2 26 31.379 52.5 33.27 13.362 2.984
3 20 31.379 68 34.49 5.059 1.13
4 12 31.379 85.1 36 3.949 0.95
5 (top) 5 31.379 97.5 38 3.9295 0.88
Total 26.300 5.944

Normally 10% excess area is provided.

Desired HTA = 26.3 × 1.1 = 28.93 m2
Select two heat exchangers with equal HTA in series.
HTA per heat exchanger = 28.93/2 = 14.465 m2
Let tube length be L.
57 × L × π × 0.025 = 14.465
L = 3.23 m
Carbate tube length equal to 3 m is standard. To provide sufficient area, provide
58 tubes instead of 57. This will mean 14.137 m2 HTA. Excess area with two
heat exchangers will be 4.0%.
Comment: For the given case, falling film absorber with glass tubes can be
considered as another option. But glass falling film absorber with once through
operation for the given case cannot be used. Conductive coefficient, offered by
glass tubes, is very low hence, it will control the value of overall coefficient.
For absorption of HCl in water with once through operation in glass tube heat
exchanger, temperature of acid film in upper portion of absorber reaches
boiling temperature. However, glass falling film absorber can be used for the
given case with only recirculation of hydrochloric acid solution. Further to
reduce the acid film temperature, gas feed can be diluted by air. Recirculation
of hydrochloric acid solution or addition of air in gas feed decreases acid film
temperatures but increases heat transfer area required for the falling film
absorber.

9.8 PROCESS DESIGN OF SIEVE PLATE

ABSORBER
Various process design aspects of a sieve-plate column were covered in Sec.
8.4.7.1. An example of the sieve-plate column for absorption is given below:
Design a crossflow-type sieve plate absorber for the absorption
of formaldehyde from the Reactor Exit Gas Mixture by water to make 37% (by
mass) of formaldehyde solution in water. Material balance data around absorber
of formaldehyde plant (Via Formox process) are given as follows.
Reactor exit gas is at 110°C (from recuperative heat exchanger) and at 135 kPa a
pressure.
Enthalpies of reactor exit gas mixture, absorber outlet gas mixture, bottom
product solution and fresh water with reference to 298.15 K (25°C) are 1141.77
kW, 295.38 kW, 216.35 and 32.17 kW, respectively. Exothermic heat of solution
of formaldehyde gas in water is substantially independent of concentration and
has a value of 62.75 kJ/mol HCHO at 298.15 K up to a concentration of about
40% (by mass) of HCHO. Fresh water enters at 30°C and product solution leaves
the absorber at 50°C. Temperature of exit gas from absorber is 50°C (Ref. 13).
 Table 9.21 Stream Compositions across Absorber13
Reactor Exit Gas Absorber Outlet Gas Bottom Product
Component kg/h mass % kg/h mass % kg/h mass %

CH3OH 40 0.09 nil — 40.0 0.44

O2 6 806.8 15.81 6 806.8 17.21
N2 29 626.8 68.81 29 626.8 74.9
H2O 2 766.0 6.42 2 649.0 6.7 5649.1 62.56
HCHO 3 341.3 7.76 nil — 3341.3 37.00
CO2 386.6 0.9 386.6 0.98
CO 27.7 0.07 27.7 0.07
H2 4.0 0.01 4.0 0.01
CH4 9.9 0.02 9.9 0.02
(CH3)2O 45.5 0.11 45.5 0.11

Total 43 054.6 100.00 39 556.3 100.00 9030.4 100.00

Solution
Total moles of formaldehyde absorbed = = 111.376 kmol/h

Heat of solution = 62.75 kJ/mol

Total heat evolved by absorption
= 111.376 × 62.75 × 1000 = 6988 844 kJ/h ≡ 1941.35 kW
This exothermic physical absorption is carried out either in tray tower or packed
tower in actual practice. In packed tower or tray tower, recirculation streams are
provided and these recirculation streams are passing through external coolers.
Recirculation streams are provided to control the temperature of absorption. To
achieve the complete absorption, final contact of exit gas is with fresh water,
which is fed from top.
In tray tower, intercoolers are provided. These intercoolers may be shell and tube
type or plate type or spiral heat exchangers. Here tray tower with sieve trays is
designed.
Total heat to be removed by coolers
= Total enthalpy in + Heat generated due to absorption – Total enthalpy out
= (1141.77 + 32.17) + 1941.35 – (295.38 + 216.35) = 2603.56 kW
Flow rate of fresh water fed at the top of tower = 5649.1 + 2649 – 2766 = 5532.1
kg/h
Absorption of formaldehyde in water is highly exothermic but it can be carried
out approximately in isothermal condition by using intercoolers and providing
recirculation streams.
Let the average temperature of absorption in the entire tower is 45°C.
Consider the topmost section of tower [Refer to Figs. (4.2) and (4.8)]. To achieve
the complete absorption of formaldehyde, gas mixture must be finally contacted
with pure water. Recirculation cannot be provided in the topmost section. If
dilute solution of formaldehyde is fed from the top with fresh water, then it
limits the concentration of formaldehyde in the exist gas.
Assuming that the heat of solution is utilised only in increasing the temperature
of formaldehyde solution. Fix the temperature of formaldehyde solutions leaving
the topmost section as 60°C.
Enthalpy of solution leaving the top most (first) section – Enthalpy of fresh
water
= Heat of solution × Formaldehyde absorbed in top section

= (a)

where, w1 = Mass % of formaldehyde in solution leaving the first section

Specific heat capacity (CL) of formaldehyde solution is given by following
equation15:
CL = 4.1868 [1 – 0.005 34 w + (0.004 + 0.000 01w) t](b)
where, CL is in kJ/(kg · °C), w is mass %, t is temperature in °C.
Here, t = = 42.5°C and w = w1

CL = 4.9 – 0.0206 w1(c)

From (a) and (c),
(4.9 – 0.0206 w1) (60 – 25) – 32.17

= 0.581 × 5532.1

(4.9 – 0.0206 w1) – 32.17

= 3214.15

By trial-and-error calculations
For w1 = 7.0, LHS ≅ RHS = 242 kW
Hence, mass % of formaldehyde in solution leaving the topmost (first) section =
7.0%
This solution is leaving at 60°C.
Provide the first intercooler to cool this solution from 60°C to 40°C by cooling
water.
Heat duty of first intercooler.
ϕI = ṁ CL(60 – 40)

ṁ = = 5948.5 kg/h ≡ 1.6524 kg/s

CL = 4.1868

CL = 4.88 kJ/(kg · °C)

ϕ 1 = 1.6524 × 4.88 (60 – 40) = 161.27 kW
Stream leaves the first intercooler and enters at the top of second section which
is 7% (by mass) formaldehyde solution at 40°C.
This stream is mixed with recirculated stream as shown in Figs. (4.2) and (4.8).
Let recirulation ratio = R
Mass flow rate of recirculated stream = R × 1.6524
Total mass flow rate of formaldehyde solution feed to the top of the second
section
= (R + 1) × 1.6524.
% by mass of formaldehyde in liquid solution leaving the second section = 37%
Mass flow rate of formaldehyde solution leaving the second section

= R × 1.6524 + = (R × 1.6524 + 2.4392) kg/s

This stream is leaving at 50°C.

Enthalpy balance around second section:
Enthalpy of solution leaving the second section – Enthalpy of recirculated
stream
– Enthalpy of solution coming from 1st section = Heat of solution ×
Formaldehyde absorbed in second section.
(1.6524 × R + 2.4392) CL (50 – 25) – 1.6524 R C′L × (40 – 25) – 1.6524 C″L ×
(40 – 25)
= (2.4392 – 1.6524)(I)

where, CL = 4.1868

CL = 4.0457 kJ/(kg · °C)(II)

C′L = 4.1868

C′L = 3.9542 kJ/(kg · °C)(III)

C″L = 4.1868

C″L = 4.5841 kJ/(kg · °C)(IV)

From (I), (II), (III) and (IV)
(1.6524 × R + 2.4392) × 4.0457 (50 – 25)
= 1.6524 R × 3.9542 (40 – 25)
– 1.6524 × 4.5841 (40 – 25)
= (2.4392 – 1.6524)

69.119 R + 133.085 = 1645.723

R = 21.8845
Recirculated stream back to the top of second section = 21.8845 × 1.6524
= 36.16 kg/s
Second intercooler will cool this stream from 50°C to 40°C.
Heat duty of second intercooler,
ϕ 2 = ṁCL(50 – 40) = 36.16 × CL (50 – 40)

CL = 4.1868
CL = 4.1829 kJ/(kg · °C)
ϕ 2 = 1512.5 kW
Let cooling water enters to this cooler at 32°C and leave at 37°C. Mass flow rate
of cooling water required.
ṁw = = 72.25 kg/s

Flow rate of cooling water required = 72.25 × 3600 = 260 100 kg/h ≡ 260 m3/h.
It is planned to use the same stream of cooling water at 37°C as a cooling
medium for both, top section intercooler and inlet gas cooler on parallel basis.
Let the outlet temperature of cooling water from top section intercooler is equal
to 39°C.
ϕ 1 = 161.27 kW = ṁ1 × 4.1868 × (39 – 37)
ṁ1 = 19.26 kg/s
Remaining, ṁ2 = 72.25 – 19.26 = 52.99 kg/s will be used for the gas cooler.
Calculations for heat duty of gas cooler
Reactor exist gas is at 110°C and at 135 kPa pressure. This gas mixture is cooled
from 110°C to 55°C in cooler.
Dew point temperature calculation
Mass % of water in Reactor exist gas = 6.42%
Molar flow rate of reactor exit gas
=

= 1550.49 kmol/h
Moisture (H2O) content of gas mixture

= =

At dew point
= 0.110 01

= 0.110 01
1.110 01 pv = 135 × 0.110 01 = 14.851 35
pv = 13.379 kPa
At 51.6°C, vapour pressure of water is 13.379 kPa (Ref. 13). Hence, dew point
temperature of gas mixture is 51.6°C. Therefore on cooling the gas mixture from
110°C to 55°C, no condensation is expected.
Heat duty of cooler,

ϕ 3 =

ϕ 3 = Σ ai (383.15 – 328.15) +

Heat capacity equation constants; Σai, Σbi, Σci and Σdi are taken from Ref. 13.
ϕ 3 = 47 588.5 (383.15 – 328.15) – (383.152 – 328.152)

+ (383.153 – 328.153) – (383.154 – 328.154)

= 2617 367.5 – 44 346.20 + 106 035.4 – 14 476.6

= 2664 580.1 kJ/h ≡ 740.16 kW
Total heat removed by two intercoolers and one gas cooler
= ϕ1 + ϕ2 + ϕ3
= 161.27 + 1512.5 + 740.16 = 2413.93 kW
ϕ3 = ṁ × 4.1868 (to – 37) = 740.16
= 52.99 × 4.1868 (to – 37) = 740.16
to = 40.336°C
Calculations for number of trays:
(b) Top section
Flow rate of fresh water at top, L2w = 5532.1 kg/h
L2 = Ls = = 307.34 kmol/h

x2 = 0, X2 = = 0
Mass flow rate of exist gas at top = 39 556.3 kg/h
G2w = 39 556.3 kg/h
Molar flow rate of gas mixture at the outlet of absorber,
ṁ2 =

= 1431.36 kmol/h
Let concentration of formaldehyde in outgoing gas mixture = 120 ppm
y2 = 120 × 10–6 = 0.000 12

Y2 = = 0.000 12

Let x′2 = Mole fraction of formaldehyde in solution entering to the second

section.
y′2 = Mole fraction of formaldehyde in gas mixture leaving the second
section.
ṅF = Moles of formaldehyde in liquid solution entering to second section
=

= 0.528 33 kmol/s
Moles of water in the same solution
ṅw =

= 1.5737 kmol/s
x′2 = = 0.251 34

X′2 = = 0.335 72

Formaldehyde absorbed in top section

GS (Y′2 – Y2) =

GS (Y′2 – Y2) = 13.88 kmol/h

GS = G2(1 – y2) = 1431.19 kmol/h
1431.19 (Y′2 – 0.000 12) = 13.88
Y′2 = 0.009 818
y′2 = = 0.009 72

At the base of tower

G1 = 1550.49 kmol/h
Mole fraction of formaldehyde in gas mixture

y1 = = 0.071 83

Y1 = = 0.0774

Flow rate of liquid solution leaving the tower

L1w = R × 1.6524 × 3600 + 9030.4 (R = 21.8845)
L1w = 139 213 kg/h
Note: From this stream 9030.4 kg/h is separated as the product stream and the
remaining is recirculated back.
Average molar mass of product solution

Mav = = 21.175 kg/mol

L1 = = 6574.4 kmol/h

Mole fraction of formaldehyde in this solution.

x1 = = 0.261 16

X1 = = 0.353 47

Mass % of formaldehyde at the top of second section, w2

= × 100
=

w2 = 35.878%
Average mass % of formaldehyde in second section = = 36.44%

Average temperature of absorption = 45°C

Partial pressure of formaldehyde in atmospheric air over formaldehyde solution
is given by Lacy’s empirical equation (Ref. 15).
Log p–v = 9.942 – 0.953 (0.488)w/10 – 2905/T
At T = 45 + 273 = 318 K and w = 36.44%
p–v = 5.46 Torr
At atmospheric pressure,
equilibrium constant, m =

= 0.028 at atmospheric pressure

Operating pressure in the given absorber is close to atmospheric, hence, the
average value of m for second section is 0.028.
= 0.028

Y 0 0.005 0.01 0.015 0.02 0.025 0.028

X 0 0.216 0.547 1.1177 2.336 6.76 ∞
Average mass % of formaldehyde for top section, w = = 3.5%

Average temperature of absorption = = 45°C

log p–v = 9.942 – 0.953 – 2905/T

p–v = 1.1625 Torr [for T = (273 + 45) K]

m = = 0.071 82

= 0.071 82

Equilibrium data for top section:

Y 0 0.001 0.002 0.003 0.004 0.006 0.008 0.01
X 0 0.014 0.0286 0.0434 0.0587 0.09 0.1242 0.16
For first section operating line starts from (0, 0.000 12) to (X, 0.009 817).

where X = and x = = 0.0432

X = 0.045 16
 Fig. 9.20 Theoretical Stage Calculations for Top Section of Formaldehyde Absorber

Number of theoretical stages required in top section is n1 = 3.73

For second section, operating line starts from (0.335 72, 0.009 818) and
terminates at (0.353 47, 0.0774).
Number of theoretical stages required in bottom section is n2 < 1 but assumed as
1.
Total number of theoretical stages required for the desired separation, nT =
4.7304

Fig. 9.21 Theoretical Stage Calculations for Bottom Section of Formaldehyde Absorber

Refer to Fig. 4.8 for two sections of the tray tower.

Tower diameter calculations
At the base of second section
G1w = 43 054.6 kg/h, L1w = 139 213 kg/h
Average molar mass of gas mixture that enters the tower,
Mav = = 27.768 kg/kmol

Liquid vapour flow factor FLV for sieve tray tower

FLV = (8.66)

Lw = L1w = 139 213 kg/h

Gw = G1w = 43 054.6 kg/h

ρV = = 1.375 kg/m3

(Density of water at 50°C, ρw = 988 kg/m3, ρFo = 815 kg/m3, ρme = 780 kg/m3)

ρL =

= 919.37 kg/m3
Volumetric flow rate of liquid =

= 0.042 m3/s
Select two pass sieve tray (Ref. Table 8.33) as liquid flow rate is high.
FLV = = 0.0625

For tray spacing, S = 0.45 m

From Fig. 8.20, Cf = 0.079
Flooding velocity
vF = (8.65)

σ L = Surface tension of liquid, N/m

= [Pi] ; ρ ′ is in kmol/m3
[Pi] for HCHO = 15.5 + 66 = 81.5
(Table 3–343 of Ref. 3)
ρV = = 1.5 kg/m3

= = 2.21

σ Fo = 23.85 = 23.85 × 10–3 N/m

σ L1/4 = Σ σ i1/4 xi
σ L1/4 = 0.26 × (23.85 × 10–3)1/4 + 0.74 × (70 × 10–3)1/4 = 54.3 × 10–3 N/m

vF =

= 2.5257 m/s
Let actual velocity of gas-vapour mixture through tower
v1 = 0.66 vF = 1.667 m/s
Volumetric flow rate at bottom (of gas mixture)
qV1 = = 8.7 m3/s

Net area of tray required at bottom

An = = 5.219 m2

Let downcomer area, Ad = 0.1 Ac per downcomer

where, Ac = Inside cross-sectional area of tower
An = Aa – Ad = Ac – 0.2 Ac = 0.8 Aa
0.8 Ac = 5.219 m2
Ac = 6.524 m2 =

Di = 2.88 m
Tower diameter required at the base of top section:
At the base of top section.
Lw = 1.6524 kg/s
Gas flow rate at the base of first section
Gw = Gas flow rate at top of first section + Lw – Solvent flow rate at top
Gw = 39 556.3 + 1.6524 × 3600 – 5532.1
Gw = 39 972.84 kg/h ≡ 11.1036 kg/s
Molar flow rate of gas mixture at the entrance of first section,
G = Gs (1 + Y′2) = 1431.36 (1 + 0.009 818)
G = 1445.41 kmol/h
Mav = = 27.655 kg/kmol

ρV = = 1.35 kg/m3

ρL = = 969.2 kg/m3

(Density of water at 60°C = 983.2 kg/m3)

FLV = = 5.554 × 10–3

for tray spacing S = 0.45 m

from Fig. 8.20, Cf = 0.08
Surface tension of 7% formaldehyde solution.
σ L1/4 = Σσ i1/4xi

x = = 0.0432

σ L1/4 = 0.0432 × (23.843 × 10–3)1/4 + (1 – 0.0432) × (70 × 10–3)1/4

σL = 0.068 N/m
Flooding Velocity

vF =

= = 2.736 m/s

For 66% flooding,

v = 0.66 × 2.736 = 1.8 m/s
Volumetric flow rate of gas mixture at the base of first section.
qV = = 8.225 m3/s

Net area of tray required,

An = = 4.57 m2

Let downcomer area = 12% of column area (Ac)

Ac = = 5.19 m2 =

Di = = 2.57 m

Checking of weeping for bottom section:

The minimum design vapour velocity through holes to avoid weeping is given
by following equation:
uh =   (9.70)

uh = Minimum gas velocity through holes, m/s

K = Constant can be obtained from Fig. 9.19
dh = hole diameter, mm
Let dh = 5 mm, ρG = 1.375 kg/m3
Height of weir, hw = 50 mm
Height of liquid crest over the weir

how =   (9.69)

Lw = Liquid flow rate, kg/s

Lw = 139 213 kg/h = 38.67 kg/s
ρL = 919.37 kg/m3
lw = Length of weir, m

For = 0.1 = 0.727 (Ref. Table 8.34)

lw = 0.727 × 2.88 m = 2.0938 m

how = = 34.9 mm

hL = hw + how = 84.9 mm, K = 30.8

uh =

= 10.61 m/s
Maximum hole area required
Ahm = = 0.82 m2

= = 0.1257

Let Ah = 0.1 Ac1 = 0.391 44 m2

Actual velocity, of gas through holes, vh = 22.22 m/s > 10.61 m/s
Hence, weeping will not take place.
A check by Eq. (8.72) would confirm no weeping.

FrL =

= 0.9416 > 0.5

Checking of weeping for top section:
For the trays of top section, let hole diameter, dh = 5 mm
Height of weir, hw = 50 mm
Height of liquid crest over the weir

how = (8.69)

Lw = Liquid flow rate = 1.6524 kg/s

ρL = 969.2 kg/m3

For = 0.12 ⇒ = 0.77

Length of weir,
lw = 0.77 × 2.57 = 1.979 m
how =

how = 6.79 mm
hL = hw + how = 56.79 mm, K = 30.2
Minimum vapour velocity required through holes to avoid weeping
uh = (8.71)

= 10.19 m/s
Maximum hole area which can be provided,
Ahm = = 0.8072 m2

= = 0.1556

Let = 0.1

Ah = 0.1Ac = 0.518 75 m2
Actual velocity of gas through holes,
vh = = 15.855 m/s > 10.10 m/s

Hence, weeping will not take place.

Based on Eq. (8.72),

FrL =

= 0.8001 > 0.5

In this tray tower both sections are equally sensitive to weeping.
Checking of downcomer flooding for the bottom section:
Clear liquid back-up in downcomer,
hb = hw + how + ht + hdc(8.83)
where, ht = Total tray pressure drop, mm LC
hdc = head loss in downcomer, mm LC
ht = hd + (hw + how) + hr(8.75)
where, hd = Dry tray pressure drop, mm LC
hr = Residual pressure drop, mm LC

hd = (8.76)

where, Co = Orifice coefficient can be determined by using Fig. 8.24

Let plate thickness = 5 mm,
= = 1

= 0.1

Co = 0.84 (From Fig. 8.24)

hd = 51 × = 53.37 mm LC

Residual pressure drop,

hr = = 13.6 mm LC

ht = 53.37 + (50 + 34.9) + 13.6 = 151.87 mm LC

= 50 + 34.9 + 151.87 + hdc
head loss in downcomer,

hdc = (8.80)

L wd = Liquid flow rate in downcomer, kg/s

A m = Ad or Aap whichever is smaller
A ap = hap l w
hap = Clearance between bottom edge of downcomer apron and tray
hap = h w – (5 to 10 mm)

hdc =

Am = Ad (= 0.6524 m2) or Aap = 0.04 × 2.0938 = 0.083 75 m2

Am = 0.083 75 m2

hdc =

hdc = 10.46 mm
hb = 50 + 34.9 + 151.87 + 10.46
= 247.23 mm
S + hw = 450 mm + 50 mm = 500 mm
hb < 0.5 (S + hw)
Hence, downcomer flooding will not take place in bottom section.
Downcomer residence time,
θr =

= 7.66 s > 3 s (satisfactory)

Checking of downcomer flooding in top section:
Dry plate pressure drop,

hd = (8.76)

= = 1

= = 0.1315

Co = 0.87 (From Fig. 8.24)

hd = = 23.6 mm

Residual pressure drop,

hr = = 12.9 mm

ht = 23.6 + 50 + 6.79 + 12.9 = 93.3 mm

lw = 1.979 m, Aap = 0.045 × 1.979 = 0.089 m2
Head loss in downcomer,
hdc =

hdc = 0.06 mm
hb = 50 + 6.79 + 93.35 + 0.06 mm = 150.2 mm
S + hw = 450 + 50 = 500 mm
hb < 0.5 (S + hw)
Hence, downcomer flooding will not take place.
Downcomer residence time

θr =

= 54.77 s (satisfactory)
Tray efficiency
Overall tray efficiency based on O’Connell correlation for the absorption in tray
tower is given in Fig. 14.7, Perry’s Chemical Engineers’ Handbook, 7th Ed.,
1998.
The O’Connell parameter for top section
=

where, K = Equilibrium constant

ML = Average molar mass of liquid solution
μL = Viscosity of liquid solution
For top section
K = 0.071 82
ML = × 30 + × 18 = 18.26

μ = Viscosity of liquid solution = 10–3 Pa · s

= = 739 038.8 kmol/(m3 · Pa · s)

= = 46.2 lb mole /(ft3 · cP)

Plate efficiency, η is more than 60% (from Fig. 14.7, Perry’s Chemical
Engineers Handbook, 7th Ed.)
For bottom section
ML = 0.2563 × 30 + (1 – 0.2563) × 18 = 21.0756

= = 1298 288.2 kmol/(m3 · Pa · s)

= 81.143 lb mole/(ft3 · cP)

Plate efficiency, η is more than 60%.
(From Fig. 14.7, Perry’s Chemical Engineers Handbook, 7th Ed.)
However, plate efficiencies of commercial hydrocarbon absorbers are ranging
from 15 to 55%.
Let tray efficiencies be 0.34 for top section and 0.33 for bottom section.
Hence, actual number of sieve trays in top section will be 11 and in bottom
section will be 3. Total number of sieve trays required is 14.
Resulting data
Inside diameter of tower = 2570 mm
No. of trays in top section = 11
No. of trays in bottom section = 3
Flow rate of liquid (37% formaldehyde) solution recirculated and fed to the top
of bottom section = 153 239 kg/h
Details of the sieve trays in both sections are shown in Fig. 9.22 and Fig. 9.23.
Fig. 9.22 Single Pass Cross-flow Sieve Tray for Top Section for Example 9.6
 Fig. 9.23 Double Pass Cross-flow Sieve Tray for Bottom Section for Example 9.6

Figure 4.9 is the P&I diagram of the entire formaldehyde manufacturing. The
diagram shows four coolers in which cooling water is required. There are two
alternative cooling water circuits considered for design in Fig. 9.24.
 Fig. 9.24 Alternative Cooling Water Circuits for Example 9.6

Alternative I: Solution Cooler (HE-3) has the highest cooling load of 1512.5 kW.
Cooling water at the rate of 260.0 m3/h is supplied at 32°C to this cooler.
Cooling water leaves the HE-3 at 37°C. Cooling water stream at 37°C is now
split in three streams. One stream will go to Gas Cooler (HE-2) to cater cooling
load of 161.27 kW. Its flow rate can be calculated to be 69.33 m3/h for a
temperature rise of 2°C. Balance stream (190.76 m3/h) will flow through
Solution Cooler (HE-4) to cater cooling load of 740.16 kW. A small stream is
connected to Sample Cooler (SC-1) which is used only when sample is drawn.
Thus overall rise of 7.98°C in cooling water temperature will be realised.
Alternative II: In this alternative, it is considered that all coolers will be supplied
cooling water at 32°C. Increase of 5°C is considered in each cooler. This will
mean total cooling water flow of 415.12 m3/h.
It can be seen that total cooling water flow defers in both the alternatives
although total cooling load is same (2413.93 kW). While procuring cooling
tower for the plant, specifications will have different cooling water flow rates
with different temperature ranges. Although in both alternatives, it will be called
the tower to cater 2413.93 kW (686.85 TR).
In cooling tower design wet bulb temperature of ambient air is an important
consideration. Cold water temperature is limited by approach to the wet bulb
temperature. Cooling tower with a range of 7.98°C and flow rate 260.0 m3/h will
require lesser floor space than the cooling tower with a range of 5°C and flow
rate of 415.12 m3/h. Hence, cost of cooling tower with a cooling range of 7.98°C
will be lower. Also, cost of pumping cooling water will be lower.
Above discussion should not mislead to assume that coolers should always be
designed for series cooling as heat transfer area requirement of the coolers also
will have to be checked vis-à-vis cooling tower design. In general, a cooling
range of 8 to 10°C and approach of 5°C to design wet bulb temperature of
ambient air can be considered optimal for the design of cooling water system.

EXERCISES
9.1 Design a scrubber for absorbing dimethyl ether (DME) vapour from
nitrogen-dimethyl ether vapour mixture. The temperature in the scrubber is
20°C and scrubbing is approximated as isothermal. Operating pressure of
scrubber is atmospheric. A mixture of nitrogen—DME vapour contains 10%
DME (by volume) and its flow rate is 60 kmol/h. The scrubber is required to
absorb 98% of the DME. Molar mass of DME is 46 kg/kmol. Cold water is
used as solvent.
9.2 Gas mixture at the rate of 100 kmol/h, obtained from chlorination process,
contains 20% hydrogen chloride gas and 80% chlorine (by mole). From the
mixture, hydrogen chloride gas is to be separated by absorption in water. For
 the absorption of HCl, FRP (fibre glass reinforced plastic) packed tower is
used. Product hydrochloric acid solution, leaving the packed tower, must
contain 30% HCl (by mass). Design the FRP packed tower.
9.3 Falling film absorber is used in ammonia-water absorption refrigeration
plant12. Refrigeration duty of this plant is 200 kW (56.87 TR). Flow rate of
ammonia vapour, entering to falling film absorber is 0.1935 kg/s. It is
absorbed by weak aqueous solution of ammonia having 30% strength (by
mass). Product strong solution of ammonia leaving from the bottom of
absorber is having 36.5% strength. Average value of heat of absorption of
ammonia for the required concentration change is 2046.6 kJ/kg. Design the
falling film absorber. Temperature of ammonia gas at inlet to absorber is –
10°C.
Operating pressure = 280 kPa a
Assume that Eq. (9.46), Eq. (9.47) and Eq. (9.48) are also valid for the
absorption of ammonia in water (in falling film absorption) to calculate mass
transfer coefficient and heat transfer coefficient, respectively.
9.4 Venturi scrubber is planned for absorbing NOx gases from the exist gas
stream of nitric acid plant by using 10% NaOH solution (by mass). Exist gas
mixture, leaving the absorber of nitric acid plant, is having the following
composition (Ref. Table 2.9 of Chapter 2).
Component kmol/h
N2 1275.097
O2 61.537
Ar 16.330
NO 2.720
N2O 2.190
H2O 169.191
Total 1527.065
Discharge pressure of gas from venturi = atmospheric
Solvent to gas ratio = 1.5 L/m3
Temperature of gas mixture, entering the venturi scrubber = 50°C
Determine: (a) throat diameter of venturi scrubber,
(b) % removal of NOx gases,
(c) consumption of NaOH, and
(d) pressure drop in venturi scrubber.
9.5 Design a sour-water stripper based on following information16:
Feed liquid mixture flow rate = 50 000 kg/h
% of H2S in feed = 0.73% (by mass)
% of NH3 in feed = 0.556% (by mass)
% of H2O in feed = 98.714% (by mass)
Feed condition = Saturated liquid at operating condition
Stripping steam rate = 5100 kg/h
Operating pressure at bottom = 2.4 atm a
Concentration of H2S and NH3 in bottom stream = 5 ppm and 20 ppm,
respectively
Operating temperature at bottom = 126°C
Steam is generated by using reboiler at bottom. Stripped overheads are to be
condensed in overhead condenser. Condensate and tail gas from this
condenser are separated in a reflux drum. Liquid phase from drum is recycled
back to top tray. Tail gas will be sent to sulphur recovery plant.
9.6 From a 10 mole % NH3 in NH3-air mixture, nearly 100% of the ammonia is
to be removed by countercurrent scrubbing at 1 atm and 20°C. Consider
concentration of ammonia from scrubber to be 20 ppm. Entering gas rate is
3600 kg/(h · m2). Calculate the following for both solvents, water and 0.1 N
H2SO4 solution.
For NH3-H2O system, Henery’s law constant at 30°C, m = 0.85 (Ref. 1).
(a) Minimum amount of solvent
(b) Actual amount of solvent
(c) The number of overall gas transfer units, NOG
(d) HOG based on KGa17.
KGa = 78.1 kmol/(h · m3 · atm) for 0.1 N H2SO4 solution
(e) Column height
9.7 Enrichment of a lean natural gas stream is carried out by absorbing CO2 and
H2S from a gas stream containing 45% CO2, 3% H2S and 52% CH4 (by mole).
A packed tower with 25 mm polypropylene Pall rings is considered for
absorption of acidic gases in 10% (by mass) aqueous NaOH solution. Feed
gas is introduced at a rate of 900 Nm3/h at 1.2 bar a and 40°C. Fresh aqueous
solution is introduced at 35°C. Desired purity of gas stream is 100 mg/L acid.
Design packed tower system with liquid circulation requirement. Calculate heat
duty of heat exchanger(s), installed in the recirculation system.
Stoichiometry:
2 NaOH(aq) + CO2(g) = Na2CO3(ag) + H2O(l) ΔH°r1 = –104.67 kJ/mol
2 NaOH(aq) + H2S(g) = Na2S(aq) + 2 H2O(l) ΔH°r2 = –52.73 kJ/mol

REFERENCES
1. Treybal, R. E., Mass Transfer Operations, 3rd Ed., McGraw-Hill, 1980.
2. Ludwig, E. E., Applied Process Design for Chemical and Petrochemical
Plants, 3rd Ed. Vol. 2, Gulf Publishing Co., USA, 1997.
3. Perry, R. H. and Green D., Perry’s Chemical Engineers’ Handbook, McGraw-
Hill, 6th edition, 1984, Ch – 14 and Ch – 18.
4. Kister, H. Z. and D. R. Gill, Chem. Engg. Progress, 87(2), 1991, p. 32.
5. Bennett, D., Chem. Engg. Progress, 96(5), 2000, p. 27.
6. Billet, R. and M. Schultes, Trans. Inst. Chem. Engrs., UK, Vol. 77, Part A,
Sept. 1999, p. 498.
7. Sinnott, R. K., Coulson and Richardson’s Chemical Engineering, Vol. 6,
Revised 2nd Ed., Asian Books Private Limited, New Delhi, 1988.
8. Timmermans, J., Physico-Chemical Constants of Pure Organic Compounds,
Elsevier Publishing Co., USA, 1950.
9. Edwards, W. M. and P. Huang, Chem. Engg. Progr., 73(8), (1977) p. 64.
10. Niessen, W. R., Combusion and Incineration Processes, 3rd Ed., Marcel
Dekker, Inc., USA, 2002.
11. Norman, W. S., Absorption, Distillation and Cooling Towers, Longmans,
USA, 1961
12. Kister, H. Z., G. Balekjian, J. F. Litchfield, J. P. Damm and D. R. Merchant,
Chem. Engg. Progr., 88(6), (1992) p. 41.
13. Bhatt, B. I. and S. M. Vora, Stoichiometry, 4th Ed., McGraw-Hill Education
(India), 2004.
14. Kern, D. Q., Process Heat Transfer, McGraw-Hill, USA, 1950.
15. Mcketta, J. J. and W. A. Cunningham (Ed.). Encyclopedia of Chemical
Processing and Design, Marcel Dekker INC., Vol. 23, 1985, p. 353.
16. Wild, N. H., Chem. Engg., 86(4), Feb. 12, 1979, p. 103 to 113.
17. Henley, E. J., and J. D. Seader, Equilibrium-stage Separation Operations in
Chemical Engineering, John Wiley & Sons, USA, 1981.
18. Sherwood, T. K., R. L. Pigford, and C. R. Wilke, Mass Transfer, McGraw-
Hill, USA, 1975.
 10
Process Design of Reactors
10.1 INTRODUCTION
Reactor is the most important equipment of any chemical plant. Minor change or
improvement in the reactor design or in its performance greatly affects the
process and mechanical design of other equipments. Sometimes it may even
change the entire flow sheet. For example, in the process of manufacturing of
monochloroacetic acid, reaction is carried out between acetic acid and chlorine.
Conversion of acetic acid is about 90%. Product mass contains a mixture of
monochloroacetic acid, dichloroacetic acid and unconverted acetic acid from
which monochloroacetic acid is separated by crystallization. But, if in the same
reactor, conversion of acetic acid is restricted to 50%, then dichloroacetic acid is
not formed in noticeable quantity and the product mixture contains mainly acetic
acid and monochloroacetic acid from which monochloroacetic acid can be
separated by vacuum distillation. Thus, change in conversion of reactant in the
reactor changes the entire flow sheet. Hence, for any chemical process,
development in the reaction kinetics or reactor design is important in research.
Major area of the research is the development of a catalyst for the reaction. By
developing a new catalyst, entirely a new process route can be developed. Final
goal of the research for all the reactions and reactors is to carry out the reaction
at atmospheric pressure and at ambient temperature with sufficiently high rate of
reaction, 100% conversion and 100% yield. At the same time, heating or cooling
medium requirement to run the reaction should be minimum. To achieve this
goal one can take inspiration from natural reactor like human body. In the human
body, thousands of reactions are running at about 37°C temperature and at a
pressure close to one atmosphere. For the process design of the reactor,
knowledge of various subjects of chemical engineering is required like of
chemical reaction engineering, thermodynamics, stoichiometry, heat transfer
operation, mass transfer operation, fluid flow operation, chemistry, etc. Process
design of various types of reactors are outlined in the subsequent sections.

10.2 DIFFERENT TYPES OF REACTORS

Several types of reactors are in use in industry. Basically they fall into two
categories; batch and continuous. In each category, different designs are
available. In fact, development of a reactor can be an innovative idea and a
process engineer can use his skills for such a development.
In Fig. 10.1, different types of reactors are represented schematically. Designs
date back to a simple batch reactor to modern continuous catalytic reactors such
as ammonia converter. Designs also vary with the type of reactions; such as
homogenous, heterogeneous, run-away reactions, etc.
In many cases, a number of reactors are used in series. Continuous stirred tank
reactors [Fig. 10.1(c)] can be cited as an example. In some cases, number of
parallel reactors are used. For example, in reforming of hydrocarbons (such as
naphtha or natural gas), several hundred tubular reactors are used in parallel in
which catalyst is packed in different shapes and sizes. Selection is made on the
basis of degree of conversion, ease of heat removal/addition, safety and other
considerations.

10.3 BATCH REACTORS1, 2, 3

In a batch reactor, all the reactants are initially charged into the reactor. In most
cases, then they are allowed to react for certain period of time. The resultant
mixture is then discharged. In any batch reactor, composition of reaction mass
changes with time, hence, it is called an unsteady state reactor.
In an ideal batch reactor at any instant, composition throughout the reactor is
uniform while the same is not true for a non-ideal batch reactor. Size of batch
reactor does not directly depend on the reaction time but it depends on volume of
material charged to the reactor. For small capacity plant having production
capacity less than 5 tonnes per day, batch process with a batch reactor is more
economical than continuous process (continuous reactor).
In a batch reactor, variety of agitators are used. Each one has its specific
advantages and limitations. Figure 10.2 gives different types of mixing
equipment, used in chemical industry. Selection of an agitator plays an important
role in achieving desired conversion in the reactors. Selection criteria for the
different types of agitators are discussed in Sec. 10.6. In majority of reactors,
turbine-type agitators are used. While for simple mixing, flat blade turbine
agitator is used, pitched blade turbine agitator is preferred for a slurry. For
viscous liquids (such as manufacture of amino/alkyl resins), anchor/paddle type
agitators of different designs are used. For homogenising applications, such as
milk powder dissolution in water, degumming of edible oils (with phosphoric
acid), etc., shear mixer is an ideal choice but is a costly equipment. For gas-
liquid reactions gas induction-type agitator [Fig. 10.1(k)] and jet loop reactor
[Fig. 10.1(i) or (j)] are preferred. Design of agitation equipment is discussed at
length in literature. Reference (3) gives a good treatise on the subject.
10.3.1 Industrial Examples of Batch Reactor
1. Synthesis of dyes, e.g., sulphonation of naphthalene gives β-sulphonate
2. Benzene sulphonic acid from benzene with oleum
3. Nitrobenzene from benzene with mixed acid
4. Synthesis of polyurethane from toluene diisocyanate and polyglycol
5. Synthesis of cypermathrin
6. Batch fermenter, e.g., fermentation of molasses to ethanol
7. Manufacture of active pharmaceutical ingredients (API)
8. Manufacture of amino and alkyd resins, etc.
Fig. 10.1 Different Types of Reactors
 Fig. 10.2 Different Types of Agitators

10.3.2 Advantages of Batch Reactors over Continuous

Flow Reactors
1. For small scale production, batch process with a batch reactor is more
 economical than continuous process with a continuous reactor.
2. Batch reactors provide more flexibility. For example, it is readily possible to
increase the reaction time in the batch reactor while the same is difficult in
continuous flow reactor without sacrificing the production rate.
3. Some process steps (upstream or downstream processing) of the reaction can
also be performed in the reactor.
4. Batch reactors require less and relatively simple or cheaper instruments while
continuous flow reactors require more sophisticated and costlier instruments.
10.3.3 Disadvantages of Batch Reactors over
Continuous Flow Reactors
1. High manpower requirement.
2. Time is wasted in between two batches.
3. Temperature control is poor, especially for highly endothermic or exothermic
reaction. While starting reaction in a batch reactor, rate of reaction is high and
hence, heat duty is substantially high.
Some of these aspects were discussed in Chapter 3.
4. Quality of the product may vary from a batch to batch.

Performance equation of ideal batch reactor is θ = NAo   (10.1)

where, NAo = Initial amount of limiting reactant, kmol

θ = Time required to achieve the conversion XA, s
XA = Fractional conversion of limiting reactant A
–rA = Rate of reaction kmol/(m3 · s)
V = Working volume of reactor, m3
For the constant volume reaction system,

θ =   (10.2)

where, CAo = Initial concentration of limiting reactant A, kmol/m3

Butyl acetate formation is carried out in a batch reactor at 90°C
with sulphuric acid as a homogeneous catalyst. The feed contained 4.97 moles of
n-butanol per mole of acetic acid and the catalyst concentration is 0.032% by
mass as H2SO4.
Rate equation for this reaction is – rA = kC2A.
where, CA = Concentration of acetic acid in mol/cm3
k = 17.4 cm3/(mol · min) (Ref. 4)
Density of the reaction mixture at 90°C can be assumed constant and equal to
0.75 g/cm3.
(a) Calculate the time required to obtain conversion of 50%.
(b) Determine the size of reactor and the mass of reactants that must be charged
to the reactor in order to produce ester at the average rate of 100 kg/h.
Conversion is 50%. Reactor will be shut down for 30 min between batches
for removal of product, cleaning and start-up. Assume that batch reactor is
ideal. Also decide height and diameter of the reactor.
(c) Determine the heat duty required and heat transfer area required for starting
conditions of the reaction.
Solution
(a) For ideal batch reactor, time required to obtain the conversion XA can be
obtained using Eq. (10.1).
Given reaction is a liquid phase reaction in homogenous phase hence working
volume can be assumed constant.

θ =

θ =

θ =

CAo =

CAo =
= × 0.75 = 0.0018 mol/cm3

Mole ratio in feed =

x°A = 1/5.97
Mav = = 71.655

where M of acetic acid = 60 and M of butyl alcohol = 74

k = 17.4 cm3/(mol · min)
θ = = 32 min or 0.53 h

Time required to obtain 50% conversion is 32 min.

(b) Working volume of reactor =

Assume that volume of liquid mass inside the reactor will remain same during
the reaction (or change in liquid volume during the reaction is negligible).
For 100 kg/h ester production with 50% conversion,
Molar mass of butyl acetate (CH3COOC4H9) = 6 × 12 + 2 × 16 + 1 × 12 = 116
Mass of acetic acid required per hour
=

= = 103.45 kg/h

Mass of acetic acid required per batch

mA = 103.45 × (Time between batches = 30 min)

= 106.9 kg/batch
Reaction
CH3COOH(l) + C4H9OH(l) ⇌ CH3COOC4H9(l) + H2O(l)
Reaction is carried out at 90°C and with 50% conversion.
Mass of butanol required per batch,
mB = × 74 × 4.97 = 655.26 kg/batch
Mass of H2SO4 required per batch

ms = × 0.000 32 = 0.244 kg/batch

Total mass of raw materials = 106.9 + 655.26 + 0.244 = 762.404 kg

Vworking = = 1.0165 m3

Let = 1,

where, h = Depth of liquid in reactor shell, m

Di = Diameter of vessel (inside diameter), m
Let the type of bottom head = Torispherical
Inside volume of Torispherical head,
V = 0.084 672 SF (Ref. 2)  (10.3)

where, Di = Inside diameter of reactor, m

V = Inside volume, m3
SF = Straight flange, m
Vworking = h + 0.084 672 SF = 1.0165 m3

Let SF = 1.5 in = 0.0381 m and = 1

1.0165 = + 0.084 672 D3i + 0.029 92 × D2i

1.0165 = 0.870 007 D3i + 0.029 92 D2i

Di = 1.042 m = h
Conside provision of 20% extra space for vapour-liquid disengagement, then
actual height of shell of reactor H = 1.2 m.
(c) To determine heat duty required at starting condition of reactor, heat of
reaction ΔHr is required.
Table 10.1 Standard Heat of Formation and Heat Capacity Data
Compound ΔHfo, kJ/mol at 25°C CLi, kJ/(kmol · °C) at 57.5°C

Acetic acid –484.2 129.67

Butanol –327.3 215.41
 Water –285.83 75.36
Butyl acetate –522.766 267.12

ΔHor = (–522.766 – 285.83) – (–327.3 – 484.2)

= 2.904 kJ/mol (at 25°C) (endothermic)
In a reactor, when reaction starts, both reactants and products are at the reaction
temperature.
Heat of reaction at reaction temperature (90°C)
ΔHr = ΔHor + (Σvi CLi Δt)/1000
where, vi = Stoichiometric coefficient = 1 for all reactants
Assuming heat capacities to be nearly constant up to 90°C,
ΔHr = 2.904 + (267.12 + 75.26 – 215.41 – 129.67) ×

= 2.904 – 0.1755 = 2.7285 kJ/mol (endothermic)

Heat required at the starting of reaction
ϕ = ΔHr × 1000 × kmol/h of acetic acid consumed
+ Heat required to remove the water continuously from the reaction mass
of ternary azeotrope
ϕ = × 1.05 (Considering 5% heat loss)

Water will leave the reaction mass either as binary azeotrope with butanol
(normal boiling point = 92.2°C) or as ternary azeotrope with butyl acetate and
butanol (normal boiling point = 89.4°C). Boiling point of ternary azeotrope is
less than binary azeotrope.
Composition of ternary azeotrope:
60.86% butyl acetate, 10.47% butanol, 28.92% water (by mass).
Ratio of = = 6.44 > (= 2.1)

Hence, it will be logical to assume that water will leave as ternary azeotrope.
Rate of reaction at the starting of reaction
–rA = kC2Ao (CA = CAo when θ = 0)

or – =

Moles of acetic acid consumed per unit volume of reactor

= = 5.6376 × 10–5

Moles of acetic acid consumed = = 5.6376 × 10–5 × V

= 5.6376 × 10–5 × 1.0165 × 106 × 60 ×

= 3.438 37 kmol/h (when θ = 0 or at starting of reaction)

Water formation rate = 3.438 37 kmol/h = 61.89 kg/h
Total flow rate of ternary azeotrope in starting of reaction
= = 214 kg/h

It will contain 130.24 kg/h of butyl acetate, 22.4 kg/h of butanol and 61.89 kg/h
of water.
ϕv at starting of reaction = Σmi λi
Table 10.2 Latent Heat of Vaporization Data2
Component Latent heat of vaporization at 89.4°C (λ), kJ/kg

Butanol 625.39
Butyl acetate 330.2
Water 2284

ϕv = 130.24 × 330.2 + 22.4 × 625.39 + 61.89 × 2284

= 198 370.7 kJ/h ≡ 55.103 kW
ϕ = (2.7285 × 1000 × 3.43837 + ϕv) × 1.05 = 218 139.9 kJ/h
≡ 60.5944 kW
As a heating medium saturated steam is used.
Let saturated steam pressure = 0.2 MPa a
Saturation temperature = 120.23°C
Latent heat of vaporization of steam, λs = 2201.6 kJ/kg

Steam required ṁs = = 0.0275 kg/s ≡ 99.08 kg/h

Saturated steam side heat transfer coefficient h′o = 6000 W/(m2 · °C) (assumed)
Inside heat transfer coefficient, hi:
It is difficult to find the suitable correlation for the reacting mass side heat
transfer coefficient as the reaction is taking place. hi can be governed by boiling
coefficient or by convective film coefficient. If flat blade disc turbine agitator is
used to improve the reaction rate and to improve the convective film heat
transfer coefficient, then
hi = hnB or hiC whichever is less
where, hiC = Convective film coefficient, W/(m2 · °C)
hnB = Nucleate boiling coefficient, W/(m2 · °C)
hic
Normally, tip velocity of turbine agitator is in the range of 150 to 250 m/min.
To calculate the convective film coefficient, let the tip velocity of turbine
agitator
v = 200 m/min = πDan
where, Da = Diameter of agitator
n = Rotational speed of agitator in revolutions per min
For flat blade disc turbine
Da = = 0.347 m

n = = 183.5 ≅ 184 rpm

For flat blade disc turbine, baffled vessel for Re > 400 (Ref. 5)

= 0.74 Re0.67   (10.4)

where, Re =   (10.5)

Table 10.3 Composition of Reaction Mass at the Starting of Reaction

Component Mass fractions at θ = 0

Acetic acid 0.1405

Butanol 0.8595
Total 1.0000

ρmix = 750 kg/m3

Viscosity of organic liquid mixture
= (7.31)
μBu = 0.65 mPa · s at 90°C
μAA = 0.53 mPa · s at 90°C
μmix = 0.63 mPa · s
Reynolds number, Re = = 439 588

Pr =   (10.6)

Table 10.4 Thermal Properties of Acetic Acid and Butanol

Component Cpi, kJ/(kg · °C) k, W/(m · °C)

Acetic acid 2.1855 0.1557

Butanol 2.8763 0.164

Cp mix = ΣCpi wi = 2.1855 × 0.1405 + 2.8763 × 0.8595 = 2.78 kJ/(kg · °C)

Thermal conductivity of liquid mixture can be calculated by Li method.
(Equations 3.127 to 3.130 of Ref. 2)
kmix = ϕ12 kL1 + 2ϕ1ϕ2 k12 + ϕ22 kL2   (10.7)
In Eq. (10.7), ϕ1 and ϕ2 are parameters and do not represent heat loads.
kL1 = 0.1557 W/(m · °C) kL2 = 0.164 W/(m · °C)
k12 = 2(kL1–1 + kL2–1)–1   (10.8)
= 0.159 74 W/(m · °C)
ϕ1 = and ϕ2 =   (10.9)

x1 = Mole fraction of acetic acid = = 0.1678

x2 = 1 – 0.1678 = 0.8322 = Mole fraction of butanol

V1 = Molar volume of pure acetic acid

= = 0.057 25 m3/kmol

V2 = Molar volume of pure butanol

= = 0.086 m3/kmol

ϕ1 = = 0.118 34

ϕ2 = 0.881 66
kmix = 0.118 342 × 0.1557 + 2 × 0.118 34 × 0.881 66 × 0.159 74
+ 0.881 662 × 0.164
= 0.163 W/(m · °C)
Pr = = 10.744

Using Eq. (10.4): hi = 0.74 × × (439 588)0.67 ×   (10.744)0.33 × 1

= 4594.25 W/(m2 · °C) = hnC

Nucleate boiling coefficient for the boiling of azeotropic mixture can be
determined by Mostinski’s equation:
hnB = 0.104 (6.82)

It is valid for single component boiling. But it can be used for close boiling
range and azeotropic mixtures.
ϕ = 60.5944 × 103 W
For nucleate boiling ΔTm = 30.83°C (as decided earlier)
Heat transfer area provided with full height covered jacket,
Aopr ≅ πDoh ≅ π × 1.0547 × 1.042 = 3.4526 m2
(Assuming shell thickness, ts = 6.35 mm)
p = operating pressure = 0.9869 bar
pc = critical pressure, bar
Pseudo-Critical pressure of mixture pcm = Σpci yi (Kay’s rule)
Table 10.5 Critical Pressure of Azeotropic Mixture
Component of Azeotropic mixture Mass fraction Mole fraction Molar mass Critical pressure, pc, bar

Butyl acetate 0.6086 0.23 116 31.4

Butanol 0.1047 0.06 74 44.2
Water 0.2892 0.71 18 220.5

Total 1.0000 1.00

pcm = 0.23 × 31.4 + 0.06 × 44.2 + 0.71 × 220.5 = 166.43 bar

pr = = 5.9298 × 10–3

hnB = 0.104 (166.43)0.69 (1.8pr0.17 + 4pr1.2 + 10pr10)

hnB = 2524.95 W/(m2 · °C)

Convective film coefficient, hi = 4594.25 W/(m2 · °C) > hnB
Since nucleate boiling coefficient is less than the convective coefficient, it is
taken for heat transfer area calculations.
hi = 2524.95 W/(m2 · °C)
Jacket side or saturated steam side heat transfer coefficient,
h′o = 6000 W/(m2 · °C)
Thermal conductivity of reactor shell material (SS-304) = 16 W/(m · °C)
Let thickness of reactor shell, ts = 6.35 mm (Assumed but it should be decided
based on mechanical design).
Overall heat transfer coefficient: Eq. (6.46)

Di = 1.042 m, Do = Di + 2ts = 1.042 + 2 × 0.006 35 = 1.0547 m

Uo = 737.66 W/(m2 · °C)

Heat duty, ϕt = 60.5944 × 103 W, ΔTm = 30.23°C
Heat transfer area required
A = = 2.7173 m2

Outside heat transfer area provided with full height covered jacket
Aoav ≅ 3.4526 m2
Excess heat transfer area = 27.06%
Thus available heat transfer area is sufficient.

10.4 CONTINUOUS FLOW REACTORS

All continuous flow reactors are steady state flow reactors. In a steady state flow
reactor, composition at any point is unchanged with time.
For continuous flow reactors, in-line mixers are used. In-line shear mixers are
developed for continuous degumming of edible oils. In-line static mixers are
effective in achieving desired mixing in continuous systems. In continuous
nitration, mixing of a reactant with mixed acid is very effective through a static
mixer. Similarly, mixing of natural gas/naphtha vapour with steam prior to
reforming can be carried out in an in-line static mixer. For removal of free fatty
acids from edible/non-edible oils with caustic soda solution, a static mixer is
used in a continuous plant. In Fig. 10.3, different designs, of static mixers are
shown.
Continuous flow reactors are used for achieving high throughputs. Also quality
of product can be maintained at desired level once the operating conditions are
stabilised.
There are two ideal steady state flow reactors:
(a) Ideal plug flow reactor (b) Ideal mixed flow reactor
10.4.1 Ideal Plug Flow Reactor1, 2, 3
It is also known as ideal tubular, unmixed flow, slug flow or piston flow reactor.
Ideal plug flow reactor has flat velocity profile. In this type of reactor, lateral or
radial mixing is permitted, but axial mixing is not permitted. Hence, in ideal
PFR, the residence time in the reactor is same for all elements or particles of
fluid.
 Fig. 10.3 Different Types of Static Mixers

Ideal PFR is defined by two criteria:

1. Over any cross section of ideal PFR, the mass flow rate, velocity and the fluid
properties like pressure, temperature and composition are uniform and the
same do not change with time.
2. There is negligible diffusion relative to bulk flow or no element of fluid
overtaking or mixing with any other element ahead or behind.
In ideal PFR, composition at any point is not changing with time but it changes
with length or axial distance in reactor.
 Fig. 10.4 Plug Flow Reactors

Industrial plug flow reactors can be fixed bed catalytic reactor, packed tower-
type reactor, tubular reactor with very high length to diameter ratio and zigzag
path, shell and tube heat exchanger, double pipe heat exchanger, sieve tray
tower, etc. Working volume of ideal PFR can be determined by using the
following equation:

=   (10.10)

where, V = Working volume of ideal PFR, m3

FAo = Molar flow rate of limiting reactant A at inlet, kmol/s
XA = Fractional conversion of liming reactant required
–rA = = Rate of chemical reaction, kmol/(m3 · s)

Residence time required in ideal PFR for the desired extent of given reaction can
be determined by the following equation:

θ = CAo   (10.11)

where, θ = Residence time required, s

Vo = Volumetric flow rate of reacting fluid at inlet, m3/s
Advantages and disadvantages of plug flow reactors over mixed flow reactors
Advantages
1. For any positive order reaction, plug flow reactor require minimum volume
among all types of continuous flow reactors. Volume required by ideal PFR is
very much less than the same by ideal mixed flow reactor, particularly for a
higher order reaction with higher desired conversion.
2. Plug flow reactor does not use moving mechanical parts and hence it is
suitable for high pressure reaction system, corrosive reaction system and for
the reaction system which involves the use of a toxic gas.
Disadvantages
1. Cost per unit volume of plug flow reactor is higher than the same of mixed
flow reactor.
2. It requires high investment cost for lower conversion and lower order reaction.
Industrial examples of plug flow reactors
1. Conversion of ethylene to low density polythylene (LDPE).
2. Hydrolysis of ethylene oxide to ethylene glycol.
3. Fixed bed catalytic reaction.
Some examples in this category are as follows:
(a) Conversion of NH3 from a mixture of N2 and H2.
(b) Methanol synthesis reaction from a mixture of CO, CO2 and H2.
(c) Steam reforming reaction.
(d) Air oxidation of methanol to formaldehyde.
(e) Oxidation of SO2 to SO3 in sulphuric acid process.
(f) Oxidation of ethylene to ethylene oxide.
4. Olefins production by steam cracking, e.g., ethylene.
5. Conversion of ammonium carbamate to urea in an autoclave.
6. Splitting of edible oil to fatty acids.
7. Trickel bed reactors; Examples in this category are:
(a) Hydrocracking of heavy gas oil.
(b) Hydrodenitrogenation of a lube oil distillate.
10.4.2 Ideal Mixed Flow Reactor
It is also known as the back mix reactor (BMR), ideal continuous stirred tank
reactor or as constant flow stirred tank reactor. In ideal mixed flow reactor
(MFR), composition within the reactor is uniform throughout and does not
change with time. Hence, in ideal MFR (or in ideal BMR or CSTR), exit stream
from the reactor has the same composition as the fluid within the reactor.
Among all types of continuous flow reactors, ideal mixed flow reactor requires
maximum volume for the given positive order reaction. So, volume required by
ideal mixed flow reactor (Vm) is greater than volume required by ideal plug flow
reactor (Vp). Ratio Vm/Vp increases with increase in order of reaction and also
with increase in required conversion.
In an ideal CSTR, once the reactants enter into the reaction mass, they
immediately (practically within no time) achieve final required composition.
This is not due to very fast reaction but this is merely due to the dilution of
reactants by the products. Actually, the rate of reaction is minimum in ideal
CSTR, compared to all types of continuous flow reactors as it provides
minimum concentration of reactants during reaction.
In an ideal CSTR, not only composition but temperature is also uniform
throughout. Temperature of product stream is same as that within the reactor.
Industrial mixed flow reactors can be
(a) Agitated vessel-type reactor.
(b) Sparged vessel (including gas induced agitator type) or bubble column-type
reactor.
(c) Loop reactor (with recirculation, recirculating pump and external heat
exchanger).
(d) Jet reactor with recirculation through an ejector.
Volume required by ideal mixed flow reactor is determined by the following
equation:
=   (10.12)

Residence time required in ideal mixed flow reactor for the desired extent of
given reaction is determined by the following equation [modified form of Eq.
(10.2)].
θ = CAo   (10.13)

where, θ = Residence time required, s

CAo = Initial concentration of limiting reactant A, kmol/m3
V = Working volume of reactor, m3
Vo = Volumetric flow rate of reacting fluid at inlet, m3/s
XA = Fractional conversion of limiting reactant A
Advantages of ideal mixed flow reactor
1. Good mixing and lower rate of reaction (lower rate of heat generation in case
of exothermic reaction) makes the temperature control easier. In the entrance
portion of plug flow reactor and in starting period of batch reaction, rate of
reaction is very high and hence, temperature control is difficult. Agitated
vessel-type ideal mixed flow reactor also provides higher heat transfer
coefficient compared to packed tower-type or sieve tray-type plug flow
reactor which are normally operated in adiabatic manner.
2. Ideal mixed flow reactor may require less fixed cost compared to ideal plug
flow reactor for lower order reaction with lower equilibrium conversion
because cost per unit volume of mixed flow reactor is less than the same of
plug flow reactor.
3. Polymerisation reactions can be better controlled in ideal CSTR compared to
 the same in ideal PFR (e.g., desired molar mass of polymer can be easily
achieved).
4. Fine catalyst particles can be effectively suspended throughout the liquid
reaction system with agitated vessel-type mixed flow reactor.
5. For highly viscous fluid or slurry solution, agitated vessel-type mixed flow
reactor provides higher heat transfer coefficient and mass transfer coefficient,
compared to plug flow reactor.
Disadvantages of ideal mixed flow reactor
1. Ideal mixed flow reactor provides the minimum rate of reaction among all
types of continuous flow reactors. Hence, it requires maximum volume
among all types of continuous reactors for any given positive order reaction
and required conversion. For the higher order reaction with high conversion;
e.g., second order reaction with 90% conversion, volume required by ideal
mixed flow reactor is many times higher than that required for ideal PFR.
2. Changeover to other product is difficult compared to stirred tank batch reactor.
Industrial examples of mixed flow reactors
1. Carbonylation of methanol gives acetic acid in presence of liquid phase
catalyst.
2. Oxidation of acetaldehyde to acetic acid in presence of fine KMnO4 particles
as catalyst.
3. Syrene to polystyrene by solution polymerisation.
4. Slurry polymerisation of propylene to polypropylene.
5. Chlorination of benzene to chlorobenzene by continuous process.
6. Oxidation of toluene to benzoic acid.
7. Vinyl chloride to polyvinyl chloride by emulsion and suspension
polymerisation.
A homogeneous liquid phase reaction
A → R
takes place with 50% conversion in a mixed flow reactor. Its rate equation is
–rA =

(a) What will be the conversion if this reactor is replaced by the one, six times
larger reactor and all else remaining unchanged?
(b) What will be the conversion if the original reactor is replaced by a plug flow
reactor of equal size and all else remaining unchanged?
Solution
Performance equation of ideal mixed flow reactor is
=   (10.11)

(a) For V ′ = 6V (where V ′ = volume of larger reactor)

= (1)

Also = (2)

From Eqs. (1) and (2),

12(1 – XA)2 = XA
XA2 – 2.083 33 XA + 1 = 0
XA = 1.333 or XA = 0.75
XA cannot be greater than 1; hence XA = 0.75
Hence, six times larger ideal mixed flow reactor will give 75% conversion.
(b) Vmixed = Vplug = V
Performance equation of ideal plug flow reactor

=   (10.9)

where, XAP = Fractional conversion of reactant A in ideal plug flow reactor

Let p = 1 – XAP, hence, = –1

=
For ideal mixed flow reactor =

V = = Vmixed = Vplug

2 – 2XAP = XAP
XAP = 2/3 = 0.6667
Hence, if the original ideal mixed flow reactor (giving 50% conversion) is
replaced by an ideal plug flow reactor of equal size, then conversion will be
66.67%.
The following data have been obtained on the decomposition
of gaseous reactant A in a constant volume batch reactor at 100°C.
Table 10.6 Time vs Pressure Data of Decomposition of A
θ, s pA, atm θ, s pA, atm

0 1 140 0.25
20 0.8 200 0.14
40 0.68 260 0.08
60 0.56 330 0.04
80 0.45 420 0.02
100 0.37

The stoichiometry of the reaction is 2 A = R + S.

(a) What size of plug flow reactor (in litres) operating at 100°C and 1 atm can
treat 100 moles of A/h in a feed consisting of 20% inerts to obtain 95%
conversion of A?
(b) What conversion of A can be expected in mixed flow reactor of volume V =
208 L, for the same identical feed and identical operating conditions?
Solution
First rate equation, based on constant volume batch reactor data, must be
determined.
From the data, at θ = 420 s, pA is 0.02 atm means it is reduced to a value very
close to zero. Hence, given reaction is irreversible reaction,
Assuming first order irreversible reaction,
–rA = = kCA

pA = CART
= k

= kpA

–ln = kθ

From the given data, following table can be prepared:

Table 10.7 Tabulation for Calculation of Integral

θ, s pA, atm –ln

0 1.00 0.0000

20 0.80 0.2230
40 0.68 0.3856

60 0.56 0.5800
80 0.45 0.7980
100 0.37 0.9940

140 0.25 1.3860

200 0.14 1.9660

260 0.08 2.5260

330 0.04 3.2200

420 0.02 3.9100

The graph of –ln vs θ (Fig. 10.5) is a straight line passing through origin.

Hence, given reaction is irreversible reaction and is of first order.

k = 0.009 76 s–1
–rA = 0.009 76 CA mol/(L · s) ≡ 35.136 CA mol/(L · h)
(a) Performance equation of ideal plug flow reactor,
θ = CAo

Reaction 2 A → R + S, with 20% inert:

εA =

εA = 0, as total number of moles of reaction system is not changing

θ = CAo
 Fig. 10.5 Determination of Order of Reaction

–ln (1 – XA) = kθ
θ = – ln(1 – XA) = – ln (1 – 0.95) = 307 s

θ = CAo = 307 s
Working volume of reactor, V = θ

CAo = , pAo = pt × 0.8 = 0.8 atm (20% inerts)

CAo = = 0.026 16 mol/L

FAo = 100 mol/h

V = = 326 L

(b) For ideal mixed flow reactor,

=

= 1.912

XA = 0.6565
65.65% conversion of reactant A can be expected in ideal mixed flow reactor.
It has been reported that the reaction
CH2OH + NaHCO3 → CH2OH + NaCl + CO2

CH2Cl CH2OH
A B
Ethylene Sodium Ethylene Salt Carbon
Chlorohydrin Bicarbonate Glycol Dioxide
is elementary with rate constant k = 5.2 L/(mol · h) at 82°C. On the basis of this
information it is intended to construct a pilot plant to determine the economical
feasibility of producing ethylene glycol from two available feeds, a 15 mass %
aqueous solution of sodium bicarbonate and a 30 mass % aqueous solution of
ethylene chlorohydrin.
(a) What volume of tubular (plug flow) reactor will produce 20 kg/h ethylene
glycol at 95% conversion of an equimolar feed produced by intimately
mixing appropriate quantities of the two feed streams?
(b) What size of mixed reactor is needed for the same feed, conversion and
production rate as in Part (a)?
Assume all operations at 82°C at which specific gravity of the mixed reacting
fluid is 1.02.
Solution
(a) Performance equation of ideal tubular reactor

=   (10.9)

Here given reaction is elementary reaction. Hence, rate equation is

corresponding to stoichiometric equation.
–rA = kCA CB
for equimolar feed, CAo = CBo and CA = CB
–rA = kC2A

5.2 × =

For equimolar feed and based on stoichiometric equation,

FAo × 0.95 = Moles of ethylene glycol produced (for 95% conversion)

FAo =

= = 0.3396 kmol/h

Total Mass of Feed:

ṁF = Mass flow rate of ethylene chlorohydrin solution + Mass flow rate
of sodium bicarbonate solution
= (0.3396 × MEC + Mass of associated water) + (0.3396 × MSB +
mass of associated water)
where, MEC = Molar mass of ethylene chlorohydrin = 80.5 kg/kmol
MSB = Molar mass of sodium bicarbonate = 84 kg/kmol

ṁF =

= (27.3378 + 63.7882) + (28.5264 + 161.6496)

= 281.302 kg/h
Density of reacting fluid = 1020 kg/m3
Vo = = 0.2758 m3/h

CAo = = 1.2313 kmol/m3

k = 5.2 L/(mol · h) = 5.2 m3/(kmol · h)

5.2 ×

5.2 ×

Vp = 0.818 45 m3 ≡ 818.45 L
(b) For ideal mixed flow reactor
=

Vm =

Vm =

Vm = 16.369 m3 ≡ 16 369 L
Note: (i) In general, for any positive order reaction, Vm is greater than Vp.
(ii) Ratio Vm/Vp increases with increase in the order of reaction and increase in
conversion.
10.5 DEGREE OF COMPLETION OF REACTION
Chemical engineers are familiar with degree of conversion of a reaction which is
well defined as the ratio of moles of limiting reactant consumed to that in the
feed. However, it is not possible to define the degree of conversion of reaction(s)
in all cases. Chief reason for not able to define the conversion clearly is (known
or unknown) composition of feed in which more than one reactant react
simultaneously (i.e., parallel reactions) and degree of completion of each
reaction is unknown. Under these circumstances, degree of completion of
reaction is defined in several ways.
Consider a classical reforming reaction of hydrocarbons with steam. There are a
number of hydrocarbons present in the feed. They react simultaneously with
steam and produce carbon monoxide, carbon dioxide and hydrogen.
C6H14 + 6 H2O ⇌ 6 CO + 13 H2
C5H12 + 5 H2O ⇌ 5 CO + 11 H2
C2H6 + 2 H2O ⇌ 2 CO + 5 H2
CH4 + H2O ⇌ CO + 3 H2
CO + H2O ⇌ CO2 + H2
Study of equilibrium conversions of competing reactions reveal that while most
hydrocarbons reform to hydrogen, methane is not always fully reformed. Thus,
degree of completion of the reforming reaction is judged by methane slip (i.e.,
mole % methane in outgoing gas mixture from the reformer).
Edible oils are hydrogenated to produce hydrogenated fat. Basically, edible oil is
a mixture of (mostly) triglycerides of fatty acids. Each edible oil lot has different
composition of variety of fatty acids. It is not possible to know the exact
composition in terms of fatty acids for each batch of oil. These fatty acids may
be saturated or unsaturated. The degree of unsaturation is measured by titration
with potassium iodide and is known as iodine number. Higher the iodine
number, higher is the unsaturation. When hydrogenated, double and triple bonds
open up and saturation takes place with hydrogen. For this hydrogenation
reaction, degree of completion is measured by reduction in the iodine value of
the oil.
Alkyd resins are produced by reacting an organic acid such as phthalic
anhydride, benzoic acid, etc., with glycerine or other polyols. During the course
of esterification reaction, acid is consumed. Degree of completion of this process
is measured by the final acid value (refer Exercise 2.9) of the product.
In demineralisation of water, cations are removed in the cation exchanger while
anions are removed in the anion exchanger. While bi- and tri-valents are
removed on a priority, monovalent ions are not easily removed. Hence, degree of
demineralisation (i.e., degree of completion of each ion exchanger) is measured
by sodium and chloride slips, respectively.
Several other examples can be cited for the degree of completion.
It can, therefore, be visualised that the degree of completion of reaction is
defined in terms of a measurable parameter.
These reactors are designed for desired completion of reaction. For such
reactions, rate of reaction need not be defined as for pure chemical reaction (see
Example 10.5). It is normally expressed in terms of measurable index with
respect to other parameters.
Rate of reaction of manufacture of alkyd resin can be expressed in terms of
disappearance of acid value. Typical kinetic expression could be
AV = Xθ + Y
where, AV = Acid value after time θ
θ = Time
X and Y are constants. These could be determined by laboratory and pilot plant
studies.
In case of hydrocarbon reforming, methane slip is dependent of pressure and
temperature. It may be defined in an empirical form in terms of pressure and
temperature.
CCH4 = X pa Tb + Y
where, CCH4 = Concentration of methane (mole %) at exit
p = Operating absolute pressure
T = Operating absolute temperature
X, a, b and Y are constants
Alternately, degree of approach is used to define completion of reforming
reaction. Based on actual methane slip and pressure, equilibrium temperature is
found from the standard chart. Difference between actual operating temperature
and equilibrium temperature (read from the graph) is defined as the approach.
While designing a chemical reaction system, degree of conversion is used for a
well-defined reaction, if known. In case of ammonia manufacture, conversion
per pass is commonly expressed as the reaction is well defined. For all other
systems, desired degree of completion is defined and the system is designed
accordingly.
10.6 MIXING FOR DIFFERENT TYPE OF
REACTION SYSTEMS6
For the better mixing or better mass transfer rate, an agitator, a circulating pump,
etc., are used as mixing tools. Some new techniques are also available for the
same. Selection of agitator for the given reaction system depends on the
viscosity of reaction mixture, homogeneity of the reaction system, desired duty,
etc.
Following are the guidelines for the selection of agitator (or other technique) for
the different types of reaction systems:
1. Homogeneous Liquid Phase Reaction System: If the viscosity of
homogeneous liquid phase system is less than 500 mPa · s (or cP) then flat
blade turbine stirrer (with four or six flat blades on disc as shown in Fig.
10.2(v)) or Pfaudler impeller (as shown in Fig. 10.2(ix)) can be selected.
They are used with side baffles. These agitators create the currents mainly in
radial and tangential directions. Use of side baffles are required to avoid the
vortex formation. Pitched blade turbine with 45° blades and with side baffles
can also be selected for this application. This agitator creates the currents in
axial, tangential and radial directions, it consumes less power but provides
lesser heat transfer coefficient for jacketed heat transfer as compared to flat
blade turbine agitator. The latter creates more currents towards the heat
transfer wall.
If the viscosity is in the range of 500 to 5000 mPa · s, cross beam, grid and blade
stirrers (Ref. Fig. 10.2(xiv), (xvii) and (xviii)) are suitable for homogeneous
liquids. For the lower viscosity (nearer to 500 mPa · s), side baffles are used
with these types of agitators. For the higher viscosity (nearer to 5000 mPa ·
s), side baffles are not required.
For highly viscous liquid (viscosity, μ > 5000 mPa · s), anchor stirrer (Ref. Fig.
10.2(xi)) is selected. Normally, the diameter of anchor agitator is 90% or
more of the inside diameter of tank. It rotates with keeping a close clearance
with inside surface of shell. It removes the sticky material from the heat
transfer surface and thereby improves heat transfer coefficient. But this
agitator provides poor mixing. Anchor agitator can be used in conjunction
with a higher speed paddle or other agitator to improve the mixing. Other
agitator is usually turning in the opposite direction. Miscible liquids are often
simply mixed in pipelines by using a static mixer. The deflection elements in
static mixer divide the stream into two streams and turn each through 180°, so
 after passing through N elements the stream has been blended 2N times.
2. Gas-Liquid Reaction:
(a) Gas itself creates the axial currents. Hence, for the gas-liquid reaction system
suitable agitator is that which creates the current in tangential and radial
directions. Hence, conventional flat blade turbine agitator (as shown in Fig.
10.2(ii) and 10.2(v)) is used for such applications in which separate sparger
is provided.
(b) Gas induction-type hollow agitator is new innovation for this application.
Special type of impeller (as shown in Fig. 10.1(k) and Fig. 10.17) is
attached with hollow shaft. In the upper part of the hollow shaft, windows
are provided for gas suction. Gas enters from these windows and discharges
through the lowest part of the impeller. The agitator operates on the
principle of water jet ejector. The suction so generated blows the stirrer
edges during the rotation and hence, gas enters through windows and
discharges from the bottom of impeller to liquid pool. A specially designed
impeller vigorously disperses the gas bubbles and creates a mixture akin to
a boiling liquid. Gas bubbles react with liquid as they rise. Unreacted gas is
reinduced into the liquid through the windows. Recirculation of gas is
important because bubbling of gas only once through the liquid does not
use it completely. It offers the following advantages:
(i) It provides vigorous gas-liquid mixing.
(ii) It substantially increases gas-liquid interfacial area of contact and enhances
gas-liquid mass transfer rate.
(iii) It reduces reaction time considerably for the gas-liquid reaction in which
overall reaction rate is governed by rate of mass transfer.
(iv) It provides very high vessel side (i.e., inside) coefficient which approaches a
boiling coefficient.
(v) It is also the best choice for the gas-liquid reaction with suspended solid
catalyst. For example, hydrogenation in presence of suspended Reiny Ni
catalyst.
It is used for hydrogenation, alkylation, ozonisation, oxidation, amination, etc.,
reactions.
(c) Jet reactor is a new design of reactor which can be used to achieve the
excellent gas-liquid mixing. As shown in Fig. 10.1(i), jet reactor consists of
a reaction autoclave, a circulation pump, an external heat exchanger and a
venturi-type ejector. Jet reactors are available in the capacities from 0.02
m3 to 100 m3, operating pressure up to 200 bar, operating temperature up to
 350°C and in a variety of materials of constructions like stainless steel,
Hastelloy, Monel, etc. This reactor can be used for the viscosity of reaction
mixture up to 500 mPa · s and for gas-liquid reaction with suspended solid
particles (solid catatyst load should be less than 10% by mass).
Following are the advantages of jet reactor over agitated vessel-type reactor:
(i) Length-to-diameter ratio of jet reactor is higher than the same of agitated
vessel. Hence, jet reactor requires less cost particularly for high pressure
reactions.
(ii) The external heat exchanger (instead of internal coil or jacket) can be built as
large as needed and is not limited by the reactor geometry. Sufficient heat
transfer area could be made available for accurate temperature control
even if the reactor is operated with reduced working volume.
(iii) The maximum power input per unit volume is often a limiting factor,
especially for large reactors with an agitator. Since there is no agitator in
the jet reactor, this limitation does not exist.
The circulation pump can provide very high power per m3 of working volume if
it is required to achieve the desired mass transfer rate.
(iv) The down flow jet ejector forms fine gas bubbles in the liquid and creates
high mass transfer rates.
Jet reactor is used for hydrogenation, alkylation, carbonylation oxidation,
halogenation, amination, phosgenation, etc., reactions. Jet reactor is a
type of loop reactor.
(d) Another loop reactor is a monolith catalytic reactor [Fig. 10.1(j)]. Monolithic
structure of the base carrier is impregnated with a nobel metal (such as
platinum, palladium, etc.). Liquid is circulated through a pump and passed
through the catalyst with induced gas. Originally developed for emission
control from auto vehicles (for catalytic oxidation of carbon monoxide),
monolithic structure is found equally effective for chemical reactors.
Monolithic structure [Fig. 10.14(g)] provides large surface area and hence,
low concentration of catalyst (0.5 to 1%) on the structure is sufficient for
accelerating the reaction. The design is claimed to be highly effective in
hydrogenating a nitro compound to an amine.
(e) Recent innovation for gas-liquid reaction is to convert the heterogeneous gas-
liquid reaction into single homogeneous phase; supercritical phase, by
changing the operating conditions. Foreign substance (such as carbon
dioxide, propane, etc.) is added to the reaction system to get the
homogeneous supercritical phase.
Examples
(i) Hydrogenation of oleochemicals at supercritical single phase conditions: In
this case, propane is added to reaction system. Supercritical propane
dissolves both hydrogen and oil and creates the single homogeneous
supercritical phase. This supercritical phase is contacted with catalyst and it
gives very high overall rate of reaction and higher selectivity. To create the
necessary single phase (supercritical phase) conditions, operating pressure
is kept near 150 bar, temperature of 280°C and propane is added to the
extent of 5 to 6 times product mass. Hydrogenated product will have less of
trans-isomer and less free fatty acids.
(ii) Oxidation of waste water at supercritical phase conditions7: It is carried out
at above 374°C and 22.1 MPa a pressure. It is carried out between
supercritical wastewater and oxygen. Enhanced solubility of oxygen in
supercritical water eliminates mass transfer resistance and provides very
high mass transfer rate. Residence time required in reactor is only 60
seconds for 99.98% COD reduction.
3. For gas-liquid reaction with suspended solid particles: Simple loop reactor
or jet reactor are also valid for gas-liquid reaction with suspended solid
particles, except with static mixer. Oxidation of acetaldehyde to acetic acid in
presence of potassium permanganate can be classified in this category.
4. For liquid-liquid reaction: Such type of reaction mixture is a mixture of two
immiscible liquids, e.g., nitration of benzene using mixed acid, reaction of
phenol with an alkali, sulphonation of benzene with oleum, etc.
For better mixing of two immiscible liquids, axial currents are more
important. If the viscosity of liquid-liquid mixture is less than 500 mPa · s,
then pitched blade turbine or propeller are more effective. For smaller vessel
propeller (Fig. 10.2(i)) and for larger vessel pitched blade turbine (Fig.
10.2(iv)) is suitable. For higher viscosity (μ > 400 mPa · s), agitator which
contains cross beam with inclined blades (Fig. 10.2(xiv)), can be selected for
such a liquid-liquid reaction system.
For nitration reaction, a loop reactor with a static mixer can be used. Mixed
acid can be fed into the static mixer (at a controlled rate). Heat exchanger,
located on downstream of the static mixer can remove exothermic heat of
reaction effectively. It is claimed that in such a system chances of side
reaction (such as dinitro formation) are minimal.
5. For liquid-solid reaction: In this case, entire surface of solid particles must
be accessible to the liquid. Hence, all solid particles must be suspended in the
 liquid. For the suspension of solids, axial currents are important. Hence, for
the lower viscosity of slurry, 45° pitched blade turbine or propeller can be
selected. For more viscous slurry (500 mPa · s < μ < 5000 mPa · s) cross
beam with inclined blades and for highly viscous slurry, helical ribbon (Fig.
10.2(xvi)) can be selected. If solids are sticky and heat transfer by jacket is
important, then anchor agitator in conjunction with off-centre propeller or
pitched blade turbine is preferred.
Following guidelines can be followed in scale up of agitated vessel type batch
reactor.
1. Large scale and lab scale batch reactors should be geometrically similar. Type
of agitators and Shape factors of both reactors should be same. h/D, Da/D,
Da/E, J/D, etc. of both reactors should be same. Where h, D, Da, E and J are
height of liquid in vessel, diameter of vessel, diameter of agitator, clearance
between agitator and bottom surface of vessel and width of jacket,
respectively.
2. For homogeneous liquid phase reaction tip velocities of both reactors should
be same or n1Da1 = n2Da2. Where n1 and n2 are speeds of smaller and larger
sized agitators in rpm and Da1 and are Da2 diameters of smaller and larger
sized agitators. For gas-liquid reactions and for liquid-liquid reactors, power
required per unit volume should be same in both reactors or n1Da10.66 =
n2Da20.66
For slurry reactions, for both reactors n1Da10.75 = n2Da20.75
Hydrogenation of edible oil is carried out to produce
‘Vanaspati’ (hydrogenated fat) in presence of nickel catalyst in a batch reactor. In
the standard age-old process, edible oil is hydrogenated at about 2 bar g and
140–165°C in 8 to 10 hours (excluding heating/cooling). During this period,
iodine value of the mass is reduced, say from 128 to 68. Final mass has a melting
(slip) point of 39°C. The batch reactor [Fig. 10.1(a)] has a jacket for heating the
initial charge with circulating hot oil. Cooling requirements are met by passing
cooling water in internal coils.
In a newly developed Jet Reactor [Fig. 10.1(i)], it is planned to complete the
reaction in five hours by improving mass transfer in the reactor and cooling the
mass in external heat exchanger, thereby maintaining near isothermal conditions.
Figure 10.6 shows the suggested scheme.
Soybean oil, having iodine value (IV) of 128 is to be hydrogenated in the jet
reactor at 5 bar g and 165°C. Initially the charge is heated from 30°C to 140°C
with the circulating hot oil in external heat exchanger. Hydrogen is introduced in
hot soybean oil and pressure is maintained in the reactor at 5 bar g. Reaction is
exothermic and the temperature of mass increases. Cold oil flow in the external
heat exchanger controls the temperature at 165°C as per the requirement. IV
reduction is desired up to 68 when the reaction is considered over. Thereafter,
hydrogenated mass is cooled to 60°C in about 1.5 h before it is discharged to
filter. 150 kg spent nickel catalyst is charged with soybean oil while fresh 5 to 10
kg nickel catalyst is charged at intervals in the reactor under pressure. A bleed is
maintained from the system to purge out water vapour and non-condensables.
Design the jet reactor for the following duty:
(i) Charge = 10 t soybean oil with 128 IV
(ii) Average molar mass of soybean oil = 278.0
(iii) Average chain length of fatty acids = 17.78
 Fig. 10.6 Jet Reactor System for Hydrogenation of Edible Oil

(iv) Product specifications: 68 IV, 39°C melting point (max.)

Assume linear drop of IV in five hours.
(v) Average exothermic heat of reaction = 7.1 kJ/(kg IV reduction)
(vi) Hydrogen feed rate = 110 to 125 Nm3/h
Bleed rate = 1 to 2 Nm3/h
(vii) Thermic fluid or oil is used as both, heating medium in starting of reaction
and cooling medium during the reaction.
(viii) Cooling water is available at 2 bar g and 32°C. A rise of 5°C is permitted.
Cooling water is used for cooling the oil from 80°C to 70°C in oil cooler
(HE-2) of oil cycle.
(ix) Assume the following properties of fluids for the design.
Table 10.8 Average Properties of Edible Oil and Circulating Oil
Property Soybean oil or Hardened fat Circulating oil (thermic fluid)

Density, kg/L 0.825 0.71

Specific heat, kJ/(kg · °C) 2.56 2.95
Viscosity, mPa · s 2.0 0.5
Thermal conductivity, W/(m · °C) 0.16 0.1

Solution
Volume of liquid inside the jet reactor, VL = = 12 121.2 L ≡ 12.12 m3

VL = Di2 hL + Inside volume of bottom head

where, Di = Inside diameter of jet reactor, m

hL = Height of liquid inside the shell of jet reactor, m
Let hL = 1.5 Di
Type of bottom head = Torispherical
Inside volume of torispherical head = 0.084 672 Di3 + Di2 SF

Let SF = 1.5 in = 0.0381 m

12.12 = Di2 (1.5 Di) + 0.084 672 Di3 + × 0.0381 Di2

12.12 = 1.26 28 Di3 + 0.029 92 Di2

Di = 2.115 m
hL = 1.5 × 2.115 = 3.1725 m
Let total height of shell of reactor, H = 2 × 2.115 = 4.23 m
Design of external heat exchanger:
Type of heat exchanger: BEM-type shell and tube heat exchanger.
Heat duty of HE-1 for cooling period:
ϕC = Average heat of reaction × IV reduction ×

ϕC = 7.1 × (128 – 68) × = 852 000 kJ/h ≡ 236.67 kW

Let circulation rate of soybean oil = 71 m3/h ≡ 58 575 kg/h

This assumption means entire volume gets recirculated in 8.45 min.
ṁ = 58 575/3600 = 16.27 kg/s, CL = 2.56 kJ/(kg · °C)
ΔT = 5.682°C
Hence, in heat exchanger, temperature of circulating stream of reacting mass is
to be reduced from 165°C to 159.318°C. It is cooled by thermic fluid entering at
70°C and leaving at 80°C.
Mass flow rate of cooling oil ṁo:
ṁo = = 8.023 kg/s ≡ 28 883 kg/h ≡ 40.68 m3/h

Mean temperature difference:

ΔTm = ΔTlm × Ft

ΔTlm =

= 87.141°C
For 1-1 heat exchanger, Ft = 1.
Hence, ΔTm = 87.141°C
Allocating reacting oil stream on tube side and thermic fluid (cooling medium)
on shell side.
Calculations of hi: Let tube OD, do = 25.4 mm
For 16 BWG tube, tube ID, di = 22.098 mm (Table 11–2 of Ref. 2)
Let tube side velocity, ut = 1.5 m/s
ut = Gt/ρ
Gt = ut ρ = 1.5 × 0.825 × 103 = 1237.5 kg/(m2 · s)

Gt =

at = = 0.013 148 m2
at = di2

where, Nt = Total number of tubes, NP = number of tube side passes = 1

Nt = ≈ 34

Ret = = 13 673

Pr = = 32

Dittus–Boelter’s equation

= 0.023 Re0.8 Pr0.33 (6.19)

hi = 0.023 × × (13 673)0.8 × (32)0.33

hi = 1063.87 W/(m2 · °C)

Calculations of ho:
For 25.4 mm (1 in) OD and 31.75 (1.25 in.) triangular pitch, from Table 6.1 (f),
for
Nt = 24, BEM-type 1.1 heat exchanger:
Shell ID, Di = 203 mm
Let baffle spacing, Bs = 150 mm
Type of baffle = 25% cut segmental
Shell side flow area

As = × Ds × Bs (6.33)

As = × 0.254 × 0.15 = 7.62 × 10–3 m2

Shell side mass velocity, Gs = (6.34)

Gs = = 1052.89 kg/(m2 · s)

us = = 1.483 m/s

Shell side equivalent diameter, de:

de = (Pt2 – 0.907 do2) (6.37)

= (31.752 – 0.907 × 25.42) = 18.3147 mm

Shell side Reynolds number:

Re = = 38 566.7

Prandtl number:
Pr = = 14.75

= 0.36 Re 0.55 Pr 0.33 (6.39)

ho = × (38 566.7)0.55 × (14.75)0.33

ho = 1590.82 W/(m2 · °C)

Calculations of overall heat transfer coefficient, Uo:
Take thermic fluid (oil) side fouling coefficient, hod = 5000 W/(m2 · °C) and
soybean oil side fouling coefficient, hid = 3000 W/(m2 · °C) (Ref.: Table 6.9)
Tube material = SS 316
Thermal conductivity of tube material, kw = 16.26 W/(m2 · °C) (Table 3–322, of
Ref. 2)

+
Uo = 416.5 W/(m2 · °C)
Heat transfer area:
Heat transfer area required, Ar = = = 6.52 m2

Length of tubes required, Lr = = = 2.403 m

Let tube length L = 3 m

Heat transfer area provided, A = Nt π do L
A = 34 × π × 0.0254 × 3 = 8.139 m2
% excess heat transfer area = × 100 = 24.83 % (satisfactory)

Tube side pressure drop, Δpt:

For Re = 13 673.14, Jf = 0.0045 (from Fig. (6.13))

Δpt = 1

= 6856.4 Pa ≡ 6.856 kPa (acceptable)

Shell side pressure drop, Δps:
For Re = 38 566.73 and 25% cut segmental baffles
Jf = 0.04 (From Fig. 6.15)

Δps = 8 × 0.04 × = 69 299 Pa = 69.3 kPa

Δps = 69.3 kPa which is (optimum) pressure drop < Δps max = 70 kPa
(Ref. Table 6.8)
Resulting data for heat exchanger (HE-1) are presented in TEMA format in Fig.
10.7.
 Fig. 10.7 Soybean Oil Cooler Specifications

Heating before starting of reaction:

Let time required for heating the soybean oil from 30°C to 140°C,
θ = 2 h = 3600 × 2 = 7200 s
The same thermic fluid (oil) will be used as heating medium. Let t1 is the inlet
temperature of hot oil (thermic fluid) to heat exchanger in heating period. Both
side flow rates (shell side and tube side) for heating period can be kept same as
that for the cooling period.
For heating in batch reactor by circulation through external exchanger.

ln = θ  (10.14)

(Eq. (10–147d) of Ref. 2)

t1 = temperature of heating medium at inlet, °C
t1′ = temperature of cold fluid at the beginning of heating period, °C
t2′ = temperature of cold fluid at the end of heating period, °C
m· = flow rate of hot fluid through external heat exchanger, kg/s
θ = time of heating period, s
m = mass of cold fluid in batch reactor, kg
K2 =   (10.15)
U = Overall heat transfer coefficient, W/(m2 · °C)
A = heat transfer area, m2
CLH, CLC = specific heats of hot and cold fluids, kJ/(kg · °C)
t1′ = 30°C, t2′ = 140°C, m· = 16.27 kg/s, CLH = 2950 J/(kg · °C)
A = 8.139 m2, m = 10 000 kg, θ = 7200 s
U = 400 W/(m2 · °C) (approximately same as that for the cooling period)
CLC = 2560 J/(kg · °C)
Here, flow rates of oils are same as that for the cooling period. However, small
change in the value of U can be expected due to the change in the values of
thermal properties of both oils with temperature.
= = 0.067 83

K2 = = 1.07

ln = × 7200 = 0.8831

= e0.8831 = 2.4184

t1 = 217.55°C (minimum)
Let inlet temperature of hot oil, t1 = 230°C
Outlet temperature of hot oil, t2 = 220°C
Design of cooler (HE-2) of oil cycle:
Type of heat exchanger: BEM type, Fixed tube sheet
Tube side fluid: Cooling water
Shell side fluid: Oil (thermic fluid)
Cooling water inlet temperature = 32°C
Cooling water outlet temperature = 37°C
Heat duty, ϕ = heat duty of external heat exchanger during cooling period =
236.67 kW
Cooling water flow rate,
236.67 × 103 = m· CL Δt = m· × (4.1868 × 103) (37 – 32)
m· = 11.3055 kg/s ≡ 40 700 kg/h ≡ 40.7 m3/h
Mean temperature difference:
ΔTm = ΔTlm × Ft

ΔTlm = = 40.4485 °C

Let number of tube side passes = two

R =

R = = 2, S = = 0.1042

Ft = 0.99 (From Fig. 6.11)

ΔTm = 40.4485 × 0.99 = 40.044 °C
Evaluation of tube side heat transfer coefficient, hi:
Let tube side velocity, ut = 1.5 m/s
Density of water at 34.5°C = 994.202 kg/m3
Volumetric flow rate of water = = 0.011 37 m3/s

ut = = 1.5 m/s

Hence, tube side flow area, at = 7.58 × 10–3 m2 =

Np = 2 (selected)
Let do = 19.05 mm, di = 15.748 mm
7.58 × 10–3 = (0.015 748)2

Nt = 78
From Table 6.1 (d), for 25.4 mm triangular pitch, Np = 2, Shell ID = 305 mm

Re =

= 32 171.3
Pr = = 4.867

(Viscosity of water, μ = 0.73 cP, Thermal conductivity of water, k = 0.628 W/(m

· °C)

= 0.023 Re0.8 Pr 0.33 (6.19)

hi = 0.023 × × 32 171.30.8 × 4.867 0.33

= 6240.7 W/(m2 · °C)

Evaluation of shell side heat transfer coefficient, ho:
Shell side flow area,
As = × Ds × Bs (6.33)

Pt = 25.4 mm, do = 19.05 mm, Ds = 305 mm

Let Bs = 125 mm

As = × 0.305 × 0.125 = 9.531 25 × 10–3 m2

Shell side mass velocity, Gs = (6.34)

Gs = = 841.757 kg/(m2 · s)
us = (Gs/ρo) = (841.757/710) = 1.1856 m/s(6.35)
Shell side equivalent diameter,
de = (Pt2 – 0.907 do2)(6.37)

= (25.42 – 0.907 × 19.052)

de = 18.25 mm

Re = = 30 724

Pr = 14.75

= 0.36 Re0.55 Pr0.33 (6.39)

ho = × (30 724)0.55 × (14.75)0.33

= 1408.8 W/(m2 · °C)

Evaluation of overall heat transfer coefficient, Uo:
Thermic fluid (oil) side fouling coefficient, hod = 5000 W/(m2 · °C)
Tube material = mild steel
kW = 50 W/(m · °C)

Uo = 723.66 W/(m2 · °C)

Heat transfer area required,

Ar = = 8.167 m2

Tube length required, Lr = = 1.7495 m

Let tube length, L = 2 m

Heat transfer area, A = Nt π do L = 78 × π × 0.019 05 × 2 = 9.336 m2

% excess heat transfer area = × 100 = 14.31% (adequate)

Tube side pressure drop, Δpt:

For Re = 32 171.3, Jf = 3.5 × 10–3 from Fig. 6.13

Δpt = 2 ×

Δpt = 13 547 Pa ≡ 13.547 kPa (acceptable)

Shell side pressure drop, ΔPs:
For Re = 30 724, Jf = 0.041 from Fig. 6.15 for 25% baffle cut

Δps = 8 × 0.041 × 1

= 45 060 Pa ≡ 45.06 kPa (acceptable)

Resulting data for oil cooler (HE-2) of oil cycle are presented in TEMA format
as Fig. 10.8.
Notes: (i) Circulation rate of soybean oil is fixed at 71 m3/h. However,
differential pressure (DPIC) is unknown. It will depend on desired hydrogen
circulation rate in the reactor. To begin with Δp = 3 to 4 bar may be fixed and
by actual experience this can be optimised. Reactor pressure and differential
pressure together will decide discharge pressure and power requirement of the
gear pump.
(ii) Jet reactor shown in Fig. 10.6 is basically a slurry reactor in which a small
amount of fine nickel catalyst particles are in motion. Such a reactor can also
be called three-phase fluidised bed reactor or a suspended bed reactor. Such a
reactor has advantages of high capacity to provide good temperature control,
adoptability to batch processing and potentially high rate of reaction per unit
volume of reactor (depending on the catalyst activity). However, its
disadvantages are the residence time distribution close to that in a CSTR,
difficulty of removal of fine catalyst particles by filtration and possibly allow
more side reactions due to high ratio
 Fig. 10.8 Thermic Fluid Cooler Specifications

of liquid to solid. For example, during hydrogenation of edible oil by the slurry
process, hydrogenated fat might contain high trans fatty acids (due to
isomerisation) although the edible oil may hardly contain any trans fatty acid.
At times, cost of filtration system may be substantially high. In contrast, fixed
bed catalyst reactor results in less side reactions and freedom from difficulty of
filtration.

10.7 HEAT TRANSFER IN BATCH REACTORS

In earlier Examples 10.1 and 10.5, diameter and height of batch reactors were
fixed by calculations. However, heat transfer in batch reactors is effected in
many ways.
If jacket side fluid is cooling water, chilled water, brine, hot oil, etc., or if
convective heat transfer through the jacket is required then spiral baffled jacket
is preferred. Use of spirals in the plain jacket decreases flow area considerably,
increases velocity and Re of jacket side fluid and consequently increases jacket
side heat transfer coefficient considerably. In this case, flow area for jacket side
fluid is area between two successive metal strips which forms the spiral. (Refer
Fig. 10.9(c)). Spirals also act as stiffeners and decrease thickness required by the
shell to withstand external pressure and thereby decrease the conductive
resistance of the shell.
For the exothermic reaction (e.g., chlorination of acetic acid) saturated steam is
fed into the jacket to initiate the reaction but then after cooling water is
circulated through the same jacket to remove the heat of reaction. For such a
case plain jacket is selected as it is preferred for saturated steam. Plain jacket
avoids the accumulation of condensate over heat transfer surface. For cooling
water circulation or hot oil circulation spiral jacket, channel jacket, limpet (half)
coil, etc., provide considerably higher heat transfer coefficient as compared to
plain jacket. Refer to Fig. 10.9 for common arrangement of jackets, limpet coil,
and internal coils.
(e) Internal Helical Coil
 (f) Vessel with Helical Coils as Baffles
Fig. 10.9 Common Arrangements of Heat Transfer Jackets and Internal Coils

10.8 BUBBLE COLUMN REACTOR8, 9

In a bubble column reactor (Fig. 10.10), gas is dispersed in liquid phase at the
bottom of column by a suitable distributor. In most of the cases this distributor is
a sparger. Bubble column can be operated in a semibatch, countercurrent or
cocurrent manner. Bubble columns are used as reactors, absorbers, and strippers.
Bubble column reactor is preferred for slow or very slow gas–liquid reactions.
As overall rate of this reaction is governed by rate of chemical reaction, the rate
of consumption of limiting reactant in chemical reaction (–dNA/dθ) is directly
proportional to Vl (volume of liquid phase inside the reactor or liquid holdup).
Rate of consumption of limiting reactant in reaction (–dNA/dθ), in case of slow
or very slow reaction, does not depend on interfacial surface area of contact
between gas and liquid phase.
 Fig. 10.10 Bubble Column Reactor

Bubble column reactor does not use any moving part like agitator. Construction
of this reactor is simple and inexpensive.
10.8.1 Various Factors Affecting the Performance of
Bubble Column Reactor
1. Superficial gas velocity: Increase in superficial gas velocity increases gas
holdup, the effective interfacial area and the overall mass transfer rate. Increase
in superficial gas velocity decreases the size of bubbles hence, increases the
interfacial area of contact between gas and liquid. Effect of superficial gas
velocity is negligible on mass transfer coefficient (KL) but significant on mass
transfer area (a). For chemical reaction controlled gas–liquid reaction (slow or
very slow gas–liquid reaction) also, certain minimum KLa must be achieved to
overcome the effect of mass transfer. It is recommended to keep superficial
velocity of gas less than or equal to 10 m/s.
2. Properties of gas phase and liquid phase: Physical properties of gas phase
have practically no effect on the performance of the column. However, the
physical properties of liquid phase like surface tension, viscosity, etc., have a
profound effect on the performance of column. Increase in the viscosity of liquid
phase or decrease in the surface tension increases the effective interfacial area
and hence increases mass transfer rate. Presence of electrolyte in the liquid phase
effects greatly the performance. The electrolyte solution gives smaller bubble
size and consequently higher effective interfacial area and higher rate of mass
transfer.
3. Back mixing: Based on a study, it is found that gas flow in bubble column
operates in plug flow manner without any back mixing while considerable
amount of back mixing in the liquid phase is observed.
Back mixing in liquid phase decreases the concentration of liquid reactant and
hence the rate of chemical reaction. Consequently, it decreases the overall rate of
reaction. Use of packings, trays or baffles reduces liquid back mixing and
thereby reduces dilution of reactants by products. Hence, they provide the higher
concentration of reactant and hence the higher rate of chemical reaction; –rA = (–
1/Vl) (–dNA/dθ). But they decrease the liquid hold-up (Vl). Hence, combined
effect on rate of consumption of limiting reactant in reaction (–dNA/dθ) must be
checked before using any internal in the bubble column. In case of highly
exothermic gas–liquid reaction, back mixing in liquid phase is desirable to
control the exothermicity of reaction or to control the temperature of reaction. In
such a case, placing packings to improve the overall rate makes the temperature
control very difficult because it increases the heat duty required and decreases
heat transfer coefficient.
4. Mode of operation: For very slow gas–liquid reaction, semi-batch operation
is preferred in which liquid is charged to the reactor in batchwise manner while
gas is continuously passed through the column. For moderately fast reaction,
continuous operation is preferred. Both countercurrent and cocurrent operation is
used. But cocurrent is more common. In countercurrent contact, velocities of gas
and liquid through the tower are limited by flooding conditions while cocurrent
contact permits the higher velocities of gas and liquid through the tower
compared to countercurrent contact. Also, concentration of reactants, rate of
chemical reaction and overall rate of reaction do not depend on the mode of
operation (whether it is cocurrent or countercurrent) becuase bubble column
reactors are ideal mixed flow reactors.
If superficial liquid velocity is greater than 30 cm/s and superficial gas velocity
is less than 1 to 3 cm/s then it is better to use cocurrent down flow column. (Ref.
8)
5. Gas expansion and shrinkage: Inside the bubble column, there may be a
dramatic increase in superficial velocity of gas from gas inlet to outlet and is
called gas expansion. Decrease in superficial velocity of gas from inlet to outlet
is called gas shrinkage.
Example of gas expansion:
Chlorination of benzene gives chlorobenzene and hydrogen chloride gas. In this
case theoretically total number of moles of gas phase remains constant and
hence, flow rate of gas and superficial velocity of gas remain unchanged. But
actually this reaction is exothermic, resulting in vaporisation of organic liquids.
Hydrogen chloride and unconverted chlorine, leaving from the top, are saturated
with organic vapours at the outlet conditions. In addition to that hydrostatic head
of liquid decreases in upper part of column. Because of these combined effects,
there is a considerable increase in superficial velocity of gas from inlet to outlet.
Example of gas shrinkage:
Alkylation of benzene with ethylene in presence of AlCl3 (aluminium chloride)
solution as catalyst. Because of the consumption of gas (ethylene) in the
reaction, there is a substantial decrease in superficial velocity of gas in upper
part of column.
Since superficial gas velocity affects overall reaction rate, mass transfer rate and
heat transfer rate. This factor is important and should be carefully considered in
designing of bubble column reactor.
6. Pulsation: The performance of bubble column can be improved by pulsations.
For very low superficial velocity of gas (0.8 to 2.4 cm/s) the value of KLa can be
increased by factor as much as 3 by pulsation.
7. Addition of packing or packed bubble column: Adding packing reduces
axial mixing of liquid. Hence, packed bubble column is used where liquid back
mixing is undesirable. For gas–liquid reaction in which substantial gas shrinkage
is taking place in the reactor due to reaction, this results in very low superficial
gas velocity which in turn results in poor rate of mass transfer in the upper part
of reactor. Hence, in such a case placing of packings in upper part is beneficial.
Packing increases effective interfacial area and gas hold-up to an extent. Hence,
it provides higher rate of mass transfer. Packed bubble column is not used if fine
solid particles are present in the system as catalyst or as reactant. Packing is also
not preferred for highly exothermic or endothermic reaction as packed tower
provides poor heat transfer coefficient.
10.8.2 Industrial Examples of Bubble Column Reactor
1. Production of protein from methanol is carried out in a fermentor which is a
bubble column reactor. Largest size of bubble column reactor having capacity
of 3000 m3 is used for this application (Ref. 3).
2. Absorption of carbon dioxide in ammoniated brine for the manufacture of
soda ash.
3. Liquid phase air oxidation of a variety of organic compounds. For example,
(a) Air oxidation of acetaldehyde in presence of fine KMnO4 particles as
catalyst, and (b) Air oxidation of p–nitrotoluene sulphonic acid.
4. Air oxidation of black liquor containing Na2S (sodium sulphide) in pulp (in a
paper mill)
5. Air oxidation of ammonium sulphide [(NH4)2S]
6. Liquid phase chlorination of variety of organic compounds. For example, (a)
Chlorination of benzene to chlorobenzene, and (b) Chlorination of acetic acid
to monochloroacetic acid, etc.
7. (a) Carbonylation of methanol to acetic acid.
(b) Carbonylation of ethanol to propionic acid.
8. Liquid phase oxychlorination of ethylene for the production of vinyl chloride.
9. Hydration of propylene with sulphuric acid for the production of vinyl
chloride.
10. Reaction between ethylene (C2H4) and liquid hydrogen fluoride for the
manufacture of difluoroethane.
10.8.3 Advantages and Disadvantages of Bubble
Column Reactor Over Stirred Tank Reactor (Agitated
Vessel-type Reactor)
Advantages
1. Sealing problem is negligible in bubble column while the same is severe in
agitated vessel-type reactor. Sealing of the shaft of an agitator for high
pressure reactor is not only an initial design problem but also a continuing
maintenance problem. This problem is very important for highly toxic, high
pressure reaction system, e.g., ozonation reaction. Above 40 atm operating
pressure, agitated vessel-type reactor is not recommended for use.
2. Bubble column reactor provides more liquid hold-up than agitated vessel-type
reactor. Hence, it requires less volume for chemical reaction controlled gas-
liquid reaction (for slow and very slow gas-liquid reaction).
3. Bubble column reactor requires less maintenance as compared to agitated
vessel-type reactor.
4. Bubble column reactor requires less power consumption. Because of the
agitator, power consumption is higher with agitated vessel-type reactor.
5. It requires lesser floor space.
Disadvantages
1. Agitated vessel-type reactor provides higher mass transfer coefficient. Hence,
for mass transfer controlled gas-liquid reaction, agitated vessel-type reactor
requires lesser volume for the given extent of reaction. It may require lesser
fixed capital investment than bubble column reactor for such applications.
2. Agitated vessel-type reactor provides higher heat transfer coefficient
compared to bubble column reactor. Hence, temperature control is easier with
agitated vessel-type reactor than the same for bubble column reactor.
3. Higher interfacial area is obtained with agitated vessel-type reactor. Agitator
produces bubbles in smaller sizes and thereby increases interfacial area of
contact. Higher interfacial area provides higher mass transfer rate.
4. Bubble column reactor requires more height.
10.8.4 Criteria of Selection for Different Types of Gas–
Liquid Reactors
Table 10.9 Typical Ratios for Gas-Liquid Reactors (Ref. 1)
Type of reactor S/Vl S/VR Vl/VR Vl/Vfilm

Packed column 1200100 0.08 10 to 100

Tray tower 1000150 0.15 40 to 100
Agitated vessel 200 200 0.9 150 to 800
Bubble column 20 20 0.98 4000 to10 000

where, S = Interfacial surface area, m2

Vl = Volume of liquid, m3
VR = Volume of reactor, m3
Vfilm = Volume of liquid film, m3
Overall rate of any gas-liquid reaction is a function of rate of chemical reaction
and rate of mass transfer.
For very fast reaction overall rate of reaction is governed by rate of mass
transfer, i.e., rate of chemical reaction >> rate of mass transfer, or
Overall rate of reaction ≈ Rate of mass transfer.
To get the higher rate of mass transfer, packed tower is the best choice. Packed
tower provides higher mass transfer coefficient as well as higher interfacial area
of contact. But packed bed provides poor heat transfer coefficient.
For slow and very slow gas liquid reaction overall rate is governed by rate of
chemical reaction. Rate of consumption of reactant in a chemical reaction does
not depend on interfacial area of contact (S), but it depends on liquid hold-up
Vl = Vlk f(C)  (10.16)

Bubble column reactor provides maximum liquid hold-up. Vl/VR ratio is

maximum for bubble column reactor while the same is very low for packed bed
reactor. Bubble column reactor provides low heat transfer coefficient.
For the gas-liquid reactions having intermediate rates, overall rate of reaction
depends on both; rate of chemical reaction and rate of mass transfer. Hence, for
these reactions agitated vessel-type reactor is the best choice because it provides
higher S/VR ratio and higher Vl/VR ratio. It provides higher heat transfer
coefficient but it consumes more power. For high pressure gas-liquid reaction if
gaseous reactant is toxic then bubble column reactor is preferred against agitated
vessel-type reactor because of the sealing problem. For example, carbonylation
of methanol gives acetic acid. It is an intermediate gas-liquid reaction. Agitated
vessel-type reactor is the best choice. But, operating pressure of reactor is 50 atm
and carbon monoxide is a toxic gas. So, to avoid the use of agitator shaft sealing,
for the same reaction, bubble column reactor is preferred.
For highly exothermic gas-liquid reaction, loop reactor or jet reactor (with
recirculation, recirculating pump and external heat exchanger) can be
considered. Loop reactor provides higher heat transfer coefficient and hence
better temperature control than bubble column reactor. But, it consumes more
power. For hydrogenation of edible and non-edible oils, loop reactor or jet
reactor or gas induced agitated reactor is found to be an excellent choice as it
gives desired product pattern in less time than an agitated-type batch reactor with
sparger mechanism.
10.8.5 Process Design of Bubble Column Reactor
It can be divided in the following steps:
1. Find the working volume of reactor. Working volume of reactor means
volume of liquid in the reactor in running condition which also includes gas
hold-up.
Bubble column reactors are selected for slow or very slow gas-liquid reaction for
which overall rate of reaction is totally controlled by the rate of chemical
reaction. Hence, with most of the bubble column reactors (bubble column
reactors are used for different applications), it is very easy to overcome the
effect of mass transfer. Superficial velocity of gas in bubble column reactor is
fixed in such a way that effect of KLa or mass transfer is eliminated. If the
superficial velocity of the gas is fixed above the limiting value Sgm then
overall rate of gas-liquid reaction is equal to the rate of chemical reaction. Sgm
is the minimum superficial velocity of gas required to overcome the effect of
mass transfer.
For example, for the reaction:
A(g) + b B(l) → Product
If Sg >> Sgm

Overall rate = –rAl = = k CA CB  (10.17)

In bubble column reactor, liquid is flowing in mixed flow manner. Hence,

composition of reactants in liquid phase is almost uniform throughout and
 working volume of reactor can be determined by using the performance
equation of ideal mixed flow reactor.
=   (10.12)

If for the given gas-liquid reaction it is not possible to overcome the effect of
mass transfer, then form of overall rate of gas-liquid reaction must be
developed. It can be developed by laboratory scale and pilot plant scale
experiments. Different forms of rate equation are assumed and then verified
by experimental data. In such a case overall rate is the function of both mass
transfer coefficient (KLa) and reaction rate constant k.
For example, for the gas-liquid reaction
A(g) + B(l) → Product,
following form of rate equation can be assumed for verification:

–rA = (Typical)  (10.18)

where, KLa = Mass transfer coefficient

a = Interfacial area per unit volume
k = Reaction rate constant
HA = Henry’s law constant
pA = Partial pressure of reactant A in gas phase
CA = Concentration of reactant A in liquid phase
In such a case also, since in bubble column reactor liquid phase is flowing in
mixed flow manner, working volume Vl can be determined by the following
equation:
=   (10.12)

where, – rA = Overall rate of reaction = f (KLa, k)

2. Find or fix the value of superficial velocity of gas. Superficial velocity of gas
in bubble column reactor should be such that overall rate of reaction becomes
independent of mass transfer coefficient KLa. Minimum superficial velocity
of gas required to overcome the effect of mass transfer should be determined
by actual experiment in small scale (labscale or pilot plant reactor). Many
 different correlations are available which relate the superficial velocity of gas
and mass transfer coefficients; KLa. However, they are not reliable and
change from system to system. If it is not possible to overcome the effect of
mass transfer, then correlation of KLa must be developed in small scale
reactor for the given system and the same can be used to determine the
diameter of a commercial reactor. In scale-up, influence of wall effect (which
is significant in small scale reactor while it is negligible in commercial scale
reactor) must be considered.
=   (10.19)

where, qv = Volumetric flow rate of gas, m3/s

Di = Inside diameter of reactor, m
Sg = Superficial velocity of gas, m/s
Based on the value of Di, height of reactor H, can be determined.
3. Calculate the heat transfer area required. Heat duty required for bubble
column reactor can be determined from the heat of reaction ΔHr at the
reaction temperature.
ϕt = ΔHr kJ/mol × of limiting reactant

consumed ± ϕ ′
where, ϕ ′ = Heat utilised for other purpose

hi = 21 766.5   (10.20)

where, hi = Reacting fluid side heat transfer coefficient, W/(m2 · C)

Sg = Superficial velocity of gas, m/s
Prw = Prandtl number of water at room temperature
PrL = Prandtl number of liquid phase of reactor at reaction conditions.
4. Sparger design is not important if superficial gas velocity is more than or
equal to 10 cm/s.
In the continuous process for the manufacturing of
monochloroacetic acid (MCA) conversion of acetic acid is restricted to 50% to
avoid the formation of dichloroacetic acid. Reaction is carried out in a bubble
column reactor.
Determine the following:
(a) Working volume of reactor
(b) Diameter of reactor
(c) Height of liquid inside the reactor during reaction
(d) Heat duty of overhead condenser in which exist gas-vapour mixture is cooled
down to 40°C by cooling with water
(e) Heat transfer area required for bubble column reactor
Data
(i) Reaction: CH3COOH(l) + Cl2(g) CH2ClCOOH(l) + HCl(g)
(ii) Heat of reaction at reaction temperature, i.e., at 100°C
ΔHr = –87.92 kJ/mol (Exothermic)
(iii) Chlorine supply rate is 20% higher than that consumed in reaction.
(iv) Cooling water is available in plant at 32°C.
(v) Mass transfer coefficient data
Sg = 1 to 30 cm/s KLa = 0.25 × 10–2 to 0.4 s–1
(vi) Rate of chemical reaction
–rA = = kCA, k = 2.777 × 10–5 s–1

Density of acetic acid, ρ = 1048 kg/m3

(vii) Production rate of monochloroacetic acid = 1 t/h
(viii) Operating pressure in reactor = 0.1 atm g
Solution
(a) Working volume of reactor:
Bubble column reactor can be assumed as ideal mixed flow reactor.
=   (10.12)

Vl =

where, Vo = Volumetric flow rate of acetic acid at inlet

Stoichiometric equation:
CH3COOH(l) + Cl2(g) = CH2ClCOOH(l) + HCl(g)
60 71 94.5 36.5
Molar mass (M)
Conversion of acetic acid is restricted to 50% to avoid the DCA formation.
Acetic acid required to produce 1 t/h of MCA:
m·AA = 1000 ×

= 1269.84 kg/h
Vo = = 1.2117 m3/h

Vo = 3.3658 × 10–4 m3/s

Vl = = 12.12 m3

Working volume of reactor, Vl = 12.12 m3 Ans. (a)

(b) For superficial velocity of gas
Sg ∈ {1 cm/s, …, 30 cm/s}, KLa >> k (as given in data)
Let Sg = 10 cm/s
Overall rate of reaction = Rate of chemical reaction
Mass flow rate of chlorine at inlet
= 1.2 × × 1000

= 901.59 kg/h
Density of chlorine gas
= = 2.55 kg/m3

Volumetric flow rate of chlorine = = 353.56 m3/h ≡ 0.098 21 m3/s

=   (10.19)

= = 0.9821 m2

Di = 1.118 m Ans. (b)

(c) Let hL = Height of liquid during reaction
Vl =

12.12 = (1.118)2 × hL

hL = 12.346 m
Minimum distance required to facilitate gas-liquid separation is recommended to
be Di.
Let H = height of bubble column reactor = hL + Di
H ≅ 12.346 + 1.118 = 13.464 m
Let H = 14 m Ans. (c)
(d) Let ϕc = Heat duty of overhead condenser
In overhead condenser gas-vapour mixture is cooled down to 40°C.
Let n·t = Total molar flow rate of gas-vapour mixture leaving the reactor.
Exit gas from the reactor is saturated with liquid vapour at outlet conditions of
the reactor. Molar flow rate of chlorine at outlet of reactor.
= 0.2 × = 2.116 kmol/h

Molar flow rate of hydrogen chloride gas at outlet of reactor

= = 10.582 kmol/h

n·t =
and pt =
pt = + pvAA × xAA + xMCA × pvMCA
At 100°C vapour pressure of acetic acid pvAA = 400 Torr
At 100°C vapour pressure of MCA, pVMCA = 30 Torr
pt = 1.1 × 760 = + 400 × 0.5 + 30 × 0.5 = 836 Torr
= 621 Torr

n·t = 17.095 kmol/h

At 100°C or in the exit gas-vapour mixture from the reactor
= × 17.095 = 4.09 kmol/h

= × 17.095 = 0.307 kmol/h

This gas-vapour mixture is cooled to 40°C by cooling water in the overhead

condenser. At outlet of the overhead condenser, gas mixture is saturated with
acetic acid and MCA vapour.
To find out the composition of gas-vapour mixture at the outlet of overhead
condenser, for the first trial calculations assume that almost total condensation is
taking place in condenser.
Condensate composition at outlet of condenser
xAA = = 0.93

xMCA = 0.07
At 40°C, p′t = pt – ΔpHE
= 1.1 – 0.04 = 1.06 atm
The pressure drop in heat exchanger for gas-vapour mixture is assumed to be
0.04 atm.
At 40°C or at outlet of overhead condenser
p′t = 1.06 × 760 = + 0.93 × pvAA + 0.07 × pvMCA
At 40°C, pvAA = 38 Torr, pvMCA = 1 Torr
805.6 = + 0.93 × 38 + 0.07 × 1
= 770.19 Torr

n·t = 13.28 kmol/h

At 40°C, PvAA = 38 Torr, pvMCA = 1 Torr

= = 0.5826 kmol/h

= × 13.28 = 1.154 × 10–3 kmol/h

For second trial calculations, composition of condensate at outlet of overhead
condenser is calculated.
xAA = = 0.9198

xMCA = 1 – xAA = 0.0802

pt = 805.6 = + 0.9198 × pvAA + 0.0802 × pvMCA
805.6 = + 0.9198 × 38 + 0.0802 × 1
= 770.5674 Torr

= 0.9565 =

n·t = 13.276 kmol/h

= = 0.576 kmol/h

= = 1.32 × 10–3 kmol/h

For the new values of and

xAA = 0.919 97, xMCA = 0.08 or xAA = 0.92, xMCA = 0.08
These values are very close to previous values, Hence, third trial is not required.
Mass of acetic acid condensed = (4.09 – 0.576) × 60 = 210.84 kg/h
Mass of MCA condensed = (0.307 – 1.32 × 10–3) × 94.5 = 28.887 kg/h
Heat duty of overhead condenser
ϕtc = Sensible heat transfer of gas-vapour mixture + Latent heat transfer for
condensation of vapours + Subcooling of condensate
(Subcooling of condensate from inlet to outlet temperature is necessary in
multicomponent condensation.)
ϕtc =
+ +
where, = 2.1164 × 71 = 150.26 kg/h
= 10.582 × 36.5 = 386.24 kg/h
= × 60 = 139.98 kg/h
 = × 94.5 = 14.568 kg/h

Table 10.10 Specific Heat of Gases/Vapours and Liquids at 70°C

Component Cpi, kJ/(kg · °C)CLi, kJ/(kg · °C)

Chlorine 0.5024 –.–

Hydrogen chloride 0.7955 —
Acetic acid 1.2267 2.22
MCA 1.0467 1.9

m·AA = 210.84 kg/h, = 28.887 kg/h

λAA = 110 kcal/kg = 460.55 kJ/kg at 70°C temperature
λMCA = 85 kcal/kg = 355.88 kJ/kg at 70°C temperature
= = 105.42 kg/h

= = 14.44 kg/h

ϕtc = (150.26 × 0.5024 + 386.24 × 0.7955 + 139.98 × 1.2267 + 14.568

× 1.0467) × (100 – 40) + 210.84 × 460.55 + 28.887 × 355.88
+ (105.42 × 2.22 + 14.44 × 1.9) × (100 – 40)
= 157 253.15 kJ/h ≡ 43.6814 kW Ans. (d)
(e) Heat transfer area required for bubble column reactor:
Energy balance around reactor
Let 25°C be the reference temperature and liquid feed is at 25°C.
Heat must be removed by cooling medium circulated through the jacket around
reactor
ϕt = Heat produced during reaction – Heat removed in overhead condenser
– Enthalpy of product stream + Enthalpy of feed stream
= 87.92 of Acetic acid consumed – 157 253.15

– [(1000 × 1.9) + (634.92 × 2.2)] (100 – 25)

= 87.92 × 1000 × 10.582 – 157 253.15 – 248 214.18
= 524 903 kJ/h ≡ 145.806 kW
Let outlet temperature of cooling water from heat exchanger = 40°C
Mass flow rate of cooling water, m·w =
= 15 671 kg/h ≡ 15.7 m3/h
If cooling water is circulated through plain jacket then jacket side heat transfer
coefficient can be calculated by considering plain jacket as outside pipe of
double pipe heat exchanger.
ho:
where, de = Equivalent diameter = 4 × rH   (10.21)
rH =

rH =   (10.22)

where, d2 = Inside diameter of jacket, m

d1 = Outside diameter of reactor shell, m
Let d2 – d1 = 100 mm
de = 4rH = d2 – d1 = 0.10 m

G =

d1 = Reactor inside diameter + 2 × Thickness of shell (ts)

Let thickness of reactor shell, ts = 8 mm (assumption)
This thickness is actually determined based on mechanical design of shell.
d1 = 1118 + 2 × 8 = 1134
d2 = 1134 + 100 = 1234 mm
G = = 23.45 kg/(m2 · s)

Re = = 3256.9

(Viscosity of water at 36°C = 0.72 mPa · s)

Pr = =

= 4.84
ho = 0.023 × × (3256.9)0.8 × (4.84)1/3

ho = 156.53 W/(m2 · °C)

Value of ho is very low. To improve or to increase its value, let outlet
temperature of cooling water = 34°C
m·w = = 62 685 kg/h

G = 23.45 × = 93.65 kg/(m2 · s)

Re = = 13 006.9

ho = 156.53 × = 473.9 W/(m2 · °C)

hi: Eq. (10.20)

where, Sg = Superficial velocity of gas, m/s
Sg = 10 cm/s = 0.1 m/s
where, CL, μL and kL are the properties of liquid phase of reactor at 100°C
(reaction temperature).
Table 10.11 Properties at 100°C
Acetic acidMCA Water
CL, kJ/(kg · °C)2.3 1.884 4.1868
μ, mPa · s 0.49 0.55 0.28
k, W/(m · °C) 0.173 0.143 0.68

CL mix = Σ wi CLi

wAA = = 0.3883

wMCA = 1 – wAA = 0.6117

CLmix = Σ CLi wi = 0.6117 × 1.884 + 0.3883 × 2.3 = 2.046 kJ/(kg · °C)

μL = 0.525 mPa · s
kL = Σ ki wi = 0.3883 × 0.173 + 0.6117 × 0.143
= 0.1546 W/(m · °C)
PrL =
= 6.9479
Prandtl number of water at 100°C:
PrW = ×

= 1.724
hi = 21 766.5 × (0.1)0.22 ×

= 6533.27 W/(m2 · °C)

For MCA reactor, lead lined vessel is normally used.
Let thickness of lead lining = 5 mm
do' = OD of lead lining = 1.118 m
di' = ID of lead lining = 1.108 m
At 100°C, klead = 34 W/(m · °C),
ksteel = 45 W/(m · °C)
Conductive resistance offered by lead lining

= =

=   (10.23)
 Fig. 10.11 Proposed MCA Reactor Design

Overall heat transfer coefficient Uo is calculated by Eq. (6.46).

Let fouling coefficients
hod = hid = 5000 W/(m2 · °C)

+ = 333.3 W/(m2 · °C)

ΔTm = ΔTlm = = 66.995 °C

Aoreq = = = 6.53 m2
Let Aoavai = 6.53 × 1.2 = 7.84 m2 (Area provided)
= πdo L′ = π × (1.134) L′
where, L′ = Length of cylinder that must be covered by jacket = 2.2 m
Note: However, entire height of liquid pool (hL = 12.346 m) must be covered
by plain jacket as shown in Fig. 10.11 to keep the uniform temperature of entire
liquid pool. Reaction temperature can be controlled by controlling the flow rate
of cooling water as shown in Fig. 10.11.

10.9 THIN FILM ABSORBER CUM REACTOR

A thin film absorber can be used as a continuous reactor for gas-liquid reaction.
As shown in Fig. 10.12, it is basically a shell and tube heat exchanger. Gas and
liquid enter at top and traveling cocurrent manner. Liquid is introduced as a thin
film (2–3 mm thick) in the tubes (all around periphery) while gas flows in the
middle. Product mass leaves from bottom. Unreacted gas, if any, escapes from a
separate nozzle from bottom.
Thin film absorber cum reactor is strongly recommended for exothermic
reactions in which reaction rate is high and the reaction is controlled by heat
removal. Cooling medium (cooling water, chilled water, brine, etc.) can be used
in the shell to remove heat of reaction.
Absorption of hydrogen chloride in an alkali, sulphonation of organic
compounds with sulphur trioxide gas, ozonation, etc., can be carried out in such
reactors.
10.9.1 Thin Film Absorber-cum-Reactor for
Sulphonation
Use of sulphur trioxide (SO3) as a sulphonating agent in place of oleum offers
several advantages. It reduces reaction time from few hours to few seconds or
minutes and avoids the problem of the disposal of spent acid. Thus it forms a
part of green engineering. One of the most widely used systems for reacting
sulphur trioxide with organic compound is a tubular thin film sulphonation
reactor. The sulphonation reaction with SO3 as agent is instantaneous and
exothermic. Heat of reaction is ΔHr = –2094 kJ/kg of SO3 reacted for most
aromatic compounds. Liquid SO3 cannot be added directly to the organic
material to be sulphonated because of safety and possible product degradation
due to high exothermic heat of reaction. Hence SO3 is vaporised and diluted with
dry air/nitrogen.
In thin film sulphonation reactor SO3-air mixture is contacted with organic liquid
inside the tubes in cocurrent manner. Annular flow should be obtained inside the
tube in which liquid forms a thin film on inside of the tube wall and gas mixture
travels down the centre of the tube. Cooling water or chilled water is circulated
on shell side as a cooling medium.

Fig. 10.12 Falling Film Absorber-cum-Reactor

Design a thin film sulphonation reactor for the sulphonation of

dodecylbenzene (also known as linear alkylbenzene) with sulphur trioxide based
on following data:
(i) Heat of reaction: ΔHr = –2094 kJ/kg of SO3
(ii) Residence time of organic liquid in reactor, θ = 10 s
(iii) Dodecylbenzene feed rate = 1000 kg/h
(iv) Liquid film temperatures:
At inlet = 60°C, at outlet = 50°C
(v) Liquid flow rate inside the tubes
= 6.25 × 10–5 = m3/(s · m of periphery)
Volumetric fraction of liquid film = 0.052
(vi) SO3/Dodecylbenzene mole ratio = 1.05
(vii) Density of dodecylbenzene liquid at 50°C = 800 kg/m3
(viii) SO3 in air – SO3 mixture = 6% (by mole)
(ix) Reaction pressure = 140 kPa a
(x) Material of construction of reactor = SS 316
(xi) Cooling water supply temperature, t′1 = 32°C
(xii) Percent conversion of dodecylbenzene = 100% in single pass
Stoichiometry:

Solution
Mass flow rate of organic liquid, m·1= 1000 kg/h
Volumetric flow rate of liquid,

Residence time of organic liquid in reactor = 10 s

Liquid flow rate inside the tubes = 6.25 × 10–5 m3/(s · m of periphery)
6.25 × 10–5 × Nt × π di =

Let tube OD, do = 10.05 mm, tube ID, di = 15.748 mm (16 BWG)
6.25 × 10–5 × Ni × π × 0.015 748 =

Nt ≈ 122
Volume fraction of liquid film = 0.052
Residence time of liquid film in reactor, θ =

10 =

L = 2.8 m
Let length of tube, L′ = 3.048 m (10 ft)
Heat transfer area provided by shell and tube heat exchanger or reactor:
Apr = Nt π do L′ = 122 × π × 0.019 05 × 3.048
= 22.254 m2
Heat duty
ϕR = 2094 ×

SO3 consumed in reaction = × 80 = 325.2 kg/h

ϕR = 2094 × 325.2 = 680 968.8 kJ/h

≡ 189.16 kW
For triangular pitch arrangement Pt = 23.8125 mm (= 1.25 do)
Tube bundle diameter by Eq. (6.1):
k1 = 0.319, n1 = 2.142 (From Table 6.2)

Db = 19.05

= 306 mm
Di = shell ID = Db + C = 306 + 14 = 320 mm
Let baffle spacing, Bs = 100 mm
Type of baffle = 25% cut segmental
Cooling medium = cooling water
Let outlet temperature of cooling water from reactor, t′2 = 40°C
Mass flow rate of cooling water,

m·W =

m·W = 5.6475 kg/s ≡ 20.33 m3/h

Calculations for hi or hL:
Tube side heat transfer coefficient10 can be calculated using Eq. (10.24).

hL =

where,
kL = Average thermal conductivity of liquid in the film, W/(m · °C)
ρL = Average density of liquid in film, kg/m3
ρG = Average density of total gas stream, kg/m3
μL = Average viscosity of liquid in film, kg/(m · s)
Gt = Total gas and liquid mass velocity, kg/(m2 · s)
wg = Mass fraction of gas
Viscosity of dodecyl benzene at 55°C,
μ1 = 0.010 3174 exp [1934.3/(t + 273.15] cP
= 3.746 cP
Viscosity of liquid dodecyl benzene sulphonic acid at 55°C,
μ2 = 0.022 9282 exp [6643.4/(t + 273.15] × 10–6 cP
= 14.21 cP
Assume average mole fraction of alkyl benzene = 0.5 in the absorber cum
reactor.
Average mass fraction of alkyl benzene
wAB =

ln μL = wAB ln μ1 + (l – wAB) ln μ2
ln μL = 0.43 ln (3.746) + (1 – 0.43) ln (14.21)
μL = 8 cP or mPa · s
Thermal conductivity of organic liquid, kL = 0.15 W/(m · °C)
Specific heat of liquid, CL = 2 kJ/(kg · °C)
Prandtl Number for liquid
PrL =

= 106.67
Average flow rate of liquid in absorber

m·av =

SO3 at inlet =

SO3 at outlet =

Air flow rate =

Average flow rate of gas mixture, m·2

= 66.865 × 29 +

= 2117.9 kg/h
Mass fraction of gas =

Density of gas mixture ρG =

Mav =

ρG =

Gt =

at =

at = 0.023 76 m2
Gt =

Eq. (10.24): hL =

× (106.67)0.4
hi = hL = 617.66 W/(m2 · °C)
Calculations of ho
Shell side flow area, Eq. (6.33)
As =

= 0.0064 m2
Shell side mass velocity,
Gs =

Shell side velocity

us =

Shell side equivalent diameter; Eq. (6.37):

de = (Pt2 – 0.907 do2)

= (23.81252 – 0.907 × 19.052)

= 13.736 mm
Reo =

Pro = = 4.7849

= 0.36 Reo0.55 Pro 0.33 (6.39)

ho = 0.36 × × (16 834.6)0.55 × (4.7489)0.33 × 1

ho = 5827.33 W/(m2 · i)
Overall heat transfer coefficient: Eq. (6.46)

Uo = 406.1 W/(m2 · °C)

Mean temperature difference ΔTm = Ft × ΔTm
Ft = 1 for 1–1 shell and tube heat exchanger

ΔTlm = = ΔTm

Heat transfer area required

Ar =

Area provided Apr = 22.254 m2 < Ar

Revised tube length, L = 12 ft = 3.6576 m
Apr = 122 × π × 0.019 05 × 3.6576 = 26.71 m2
Hence excess heat transfer area
=

= 8.84 % (adequate)
Note: Sulphonation of aromatic compounds is a very well-known unit
operation in chemical industry. Either sulphuric acid (H2SO4) or sulphur
trioxide (SO3) is used as a sulphonating agent. When SO3 (liquid or gas) is
used, it is observed that its diffusion into the aromatic compound and the rate
of reaction are quite high. Since the exothermic heat of reaction is very high,
the design problem gets centred to effective design for heat transfer so that heat
is removed from the reaction mass at a rate equal to the product formation rate.
If this heat transfer rate is not effective, it is observed that side reactions take
place and eventually charing may take place.
To control the heat transfer, various solvents are being used. In one case,
sulphuric acid itself acts as a solvent. For the purpose, oleum of different
strengths is used by the industry, among which oleum of 30% strength is
common. In a jacketed batch reactor, aromatic compound is charged and oleum
is added at a controlled rate. Cooling medium (mostly cooling water) is
circulated in the jacket to cool the reaction mass. Presence of sulphuric acid
provides dilution effect thereby temperature control becomes easier. With
continuous stirring, oleum addition is carried out. The product mass is tested
and filtered through a filter. Mother liquor from the filter is contaminated
sulphuric acid. Excess sulphuric acid (with wash water) is filtered out. Thus
highly corrosive effluent generates during this operation which is contaminated
with organic impurities. Disposal of this waste acid is a serious problem. It is
either sold to some other manufacturers who can use for the manufacture of
FeSO4, Al2(SO4)3, etc., or the spent acid is neutralised with lime solution and
disposed as CaSO4 after leaching and drying. Final leached out effluent from
neutralisation is highly contaminated with organic impurities which can be
measured in terms of COD. Disposal of leached liquor is not an easy task and
calls for its extensive treatments, depending on local stipulations by the
pollution control authority. Apart from the waste disposal problems, it is
observed that if addition of oleum is not controlled effectively or temperature is
not controlled, byproducts do form and overall recovery of the product is rather
poor.
Use of a falling film absorber for the sulphonation reaction could be a relief. As
seen in this example, heat can be easily controlled by circulating the cooling
medium outside of the tubes. SO3 can be relatively easily generated by gentle
heating of liquefied SO3 in a separate vessel and passed in the absorber. As
addition of 100% SO3 is not desired, normally a mixture of SO3 and air or
nitrogen is preferred and concentration of SO3 in the mixture is controlled in
the range of 5 to 8 mole %. Reaction in this absorber produces the product with
less byproducts. Only constraint in use of an absorber for this reaction is that
the aromatic compound is required to be introduced as a solution (liquid) and
hence, if the reactant is a solid, a suitable solvent will have to be found which
has good solubility for the reactant as well as for the products and it does not
react with SO3. Thus use of falling film absorber acts as means for minimising
the waste generation and forms a part of Green Engineering practice. Several
solvents are in use, e.g., n-hexane, carbon tetrachloride, methanol, ethyl
acetate, pyridine, etc., for the purpose. For selection of a solvent and also
developing a film coefficient correlation [such as Eq. (10.24)], laboratory and
pilot experiments are needed. Such experiments can also provide useful
information on various other design parameters such as desired liquid film
thickness in the tubes, optimum concentration of SO3 in the ingoing gas
mixture, minimum reaction path (length) required, etc. It is thus possible to
develop a continuous or semi-continuous plant with a falling film absorber for
such a sulphonation unit operation without appreciable waste generation.

10.10 CATALYSIS
Catalysis plays an important role in reaction engineering. A preamble on
catalysts is therefore a necessity to process engineers who design catalytic
reactors for the process industry.
It is a known fact that a catalyst is a substance which increases the rate of a
reaction without being consumed in it. Catalysis can be described as being
homogeneous or heterogeneous. Heterogeneous catalysis involves the use of a
catalyst in a different phase from the reactants. Example is the use of a solid
catalyst with the liquid or gaseous reactants. Solid catalyst can be in the
suspended form (as a slurry) in liquid or in the fixed bed form on which liquid or
gas can react. In Example 10.5, nickel in slurry form was used as a catalyst.
In homogeneous catalysis, catalyst is in the same phase as that of reactant.
Commenly cited example in this category is the use of NaI or MnO2 as a catalyst
for decomposition of H2O2. Acid (such as sulphuric acid, phosphoric cid, etc.)
catalyzed reactions are well known in process industry. Manufacture of amino
resins, esterification reactions and other examples can be cited in this category.
Enzymes are organic catalysts with high molar mass. In human body, sugar
(glucose) is decomposed at 37°C to produce carbon dioxide and water which
releases energy with the help of a biological catalyst (a protein). In a similar
manner, enzymes allow reactions to take place under ordinary conditions (say
atmospheric pressure and ambient temperature). Esterification of acetic acid and
butyl alcohol, fermentation of molasses, manufacture of many antibiotics (such
as penicillin), etc., are examples of enzymatic reactions.
The fundamental mechanism of catalysed reaction relates to activation energy
requirement of a chemical reaction. Barring some spontaneous reactions, most
reactions need activation energy (Ea) for initiation. This energy can be supplied
in various forms; such as heat, electricity, ultraviolet light, friction, vibration, etc.
A catalyst lowers the activation energy required for the reaction. It is known that
most chemical reactions are exothermic (release heat to environment) while
some reactions are endothermic in nature and energy is supplied to the system.
Potential energy diagrams for both types of reactions are given in Fig. 10.13.

Fig. 10.13 Potential Energy Diagrams for Chemical Reactions

It can be observed in Fig. 10.13 that the catalyst has no effect on relative
(potential) energies of the reactants and products or it does not change
equilibrium constant (Kc). Role of a catalyst is thus to speed up the reaction
allowing equilibrium to be achieved faster. Catalyst activity is thus judged by
order of reduction in the activation energy.
With above introduction to the subject, focus will be on heterogeneous catalysis.
Solid catalysts are used in the process industry in different shapes and materials.
Figure 10.14 gives different shapes in which solid catalysts are available.
 Fig. 10.14 Some Shapes of Solid Catalysts

Ceramic support materials are well known for impregnation of noble metals (Pt,
Pd, Rh). Supports can take shape of Raschig and other ring forms, monolith
(honeycomb) structures, etc., so that large surface area is available for reaction
cites with high void for low pressure drop.
Zeolite is a class of materials which are made up of oxides of aluminium and
silicon. These materials are excellent for ion exchange applications, offering
themselves as catalysts for many reactions.
In a few applications, catalyst can be in powder form. Catalysts in wire gauge or
wire net form are also used in the industry.
10.10.1 Examples of Solid Catalysed Reactions
1. Reforming of natural gas or naphtha with steam. Raschig rings, partition rings
or wagon wheels, impregnated with nickel, are commonly used for this
application.
2. Shift conversion, ammonia/methanol synthesis, oxidation/hydrogenation
 reactions are carried out in catalytically packed bed reactors. Pellets, irregular
shapes and extrudates are some forms on which these reactions are carried
out.
3. Oxidation of ammonia to nitric acid is carried out on Pt/Rh wire net.
4. Palladium impregnated monolith catalysts are commonly used for
hydrogenation applications. For example, aromatic nitro compounds are
hydrogenated to corresponding amino compounds by hydrogenation.
5. Oxidation of toluene to benzaldehyde and subsequently to benzoic acid can be
carried out on a bed, packed with solid catalyst.
6. Reactive distillation unit can be designed for esterification reaction in which a
packed bed of strong cation exchange resin or zeolith material can be used for
the reaction.
7. In Bechamp process, nitro aromatic compounds are reduced to corresponding
amino compounds using iron powder as a catalyst.
10.10.2 Catalysis – A key to Waste Minimisation
Waste is generated in chemical synthesis on two accounts; one by formation of
byproducts and another due to use of excess reactants that are not always
recovered fully for recycle. Solvents are used in many reactions which also add
to the problem as solvents are not fully recovered on completion of the reaction
in most cases. An examples can be cited to the generation of the waste.
Consider reduction of para- or ortho-nitro toluene to corresponding amines.
Classically, this process is carried out in water as a solvent. To this, fine iron
powder is added and a small amount of acid (say HCOOH) is also added. At
least 10 to 20% excess iron powder is added to ensure completion of reaction
and also to account for purity of the powder. Reaction mass is continuously
stirred and heated to about 80oC at atmospheric pressure. After a few hours, the
reaction is considered completed (based on thin layer chromatographic results –
TLC). Stoichiometry of the reaction can be written as under.

After the completion of the reaction, stirring is stopped and mass is allowed to
separate in two layers. Organic (top) layer is decanted to another vessel. Vacuum
distillation is carried out at about 10 Torr a to recover the amine product. Bottom
aqueous layer is filtered and recycled as solvent for the next batch. However,
every time recycle may not be possible and it may become necessary to discard
contaminated aqueous layer (having high chemical oxygen demand – COD) after
a few cycles. Solids are removed from filter which is a mixture of iron oxides,
unreacted iron and other impurities. On drying (in open sky) the waste has to be
discarded.
In a typical run, a batch is charged with 700 kg of ortho-nitro toluene, 800 kg
iron powder (90% Fe), 6 kg acid (as catalyst) and 1000 kg water.
Stoichiometrically, the reaction should produce 547 kg ortho-amino toluene
(also known as ortho-toluidine) but, in actual practice, approximately 500 kg
ortho-amino toluene is produced for the same input of the reactant. Wet solid
waste generation (on filtration) is more than 1200 kg which may reduce to about
1000 kg on drying. This solid waste is contaminated with organic impurities and
cannot be used as such. Some attempts are made to convert this mass into a
usable product as a red oxide for mixing with paints but they have met with
limited success. Some cement manufacturers use this waste in formation of
clinker. If no disposal is realised, solid waste must be buried as a hazardous
waste. Considering 90% purity of iron powder, excess reactant is 11.8%. It is
reported that byproduct formation is quite low.
Above example clearly gives a picture of waste generation; both solid and liquid.
Also actual product realisation is 92%.
The key to waste minimisation is selectivity in organic synthesis. Unfortunately,
the standard definition of selectivity is the yield of product formed divided by
the amount of substrate converted. Hence, it is necessary to consider the overall
stoichiometry of the process by examining the atom utilisation. The atom
utilisation or atom selectivity is defined as the ratio of the molar mass of the
desired product to the sum of the molar mass of all materials produced in the
process. It is an useful tool for quickly evaluating the amount of waste generated
in alternative routes to a particular product. Alternative processes to the
particular product are compared on the basis of their atom utilisations and E
factor (kg byproduct per kg product). In order to calculate the E factor, the
amount of solvent consumed is also taken into account. A further refinement
could also take the energy consumption of alternative routes into consideration.
The next level of sophistication is to design catalytic processes which lend
themselves to facilitate recovery and recycling of the catalyst. In the foregoing
case, catalytic hydrogenation processes are developed. Stoichiometry of the
reaction can be written as given below

In one alternate, Rainey nickel is used as a slurry catalyst. Its presence in the
amount of say 5% is required and nearly 100% conversion with nearly 100%
yield are achieved. In this case, solid waste generation could be just 45 kg per
batch. Product realisation could be 98%. Spent nickel can be processed to
recover its metal value to produce fresh catalyst.
In another alternative, monolith catalyst, coated with palladium, can be
considered. Refer Fig. 10.1(j) or 10.1(k) in which an arrangement is represented
to hydrogenate nitro aromatic compound. In this alternative, there is no waste
generation and conversion/yield can be nearly 100%. Same vessel can be utilised
to distils out the water (byproduct), leaving behind the product. Such an
arrangement also calls for less energy than that required in the classical process.
This process can be considered as a part of cleaner production and forms a part
of Green Engineering.
Several similar cases can be cited to show that catalytic processes reduce waste
generation and increase productivity.
Hence, it can be concluded that catalysis is the key to waste minimisation.

10.11 DESIGN OF FIXED CATALYST BED

REACTORS FOR GASEOUS REACTIONS10, 11, 12, 2
In many syntheses, fixed bed catalytic reactors are used. Water shift reaction to
produce hydrogen by reaction of carbon monoxide with steam is carried out in a
fixed catalyst bed. However, this reaction is exothermic with practically no effect
of pressure on the reaction rate. At high temperature, the rate of reaction is
relatively high but equilibrium conversion is low. In such a situation, a battery of
reactors in series is used. High temperature, medium temperature, and low
temperature shift converters are common in syngas production from
hydrocarbons. Each reactor is packed with different types of catalysts. Exit gas
stream from each converter is passed through heat exchangers to recover the heat
and to lower the gas temperature, entering the next reactor. In this way,
advantages of high rate of reaction at higher temperature and higher equilibrium
conversion at low temperature are achieved. In a typical natural gas based plant,
carbon monoxide content in the reformer outlet gas is about 15% (v/v) which is
reduced to 0.2% (v/v) in two or three different axial flow converters. Generally,
these converters are designed for low to moderate pressures.
For certain exothermic gaseous reactions, rate of reaction at low temperature is
so low that required volume of catalyst works out to be very large. Cost of such a
large reactor could be prohibitive, more so for high pressure reaction. Hence
certain exothermic gaseous reactions are carried out at high temperature.
Examples are ammonia synthesis, methanol synthesis, dissociation of nitrous
oxide to nitrogen, etc. While designing such converters, process engineer can
consider different options. In one option, reaction is carried out in multiple
reactors having the same or different catalysts. Exit gas from each reactor is
cooled in an external heat exchanger. Partially converted gas mixture from the
heat exchanger enters the next reactor and so on as shown in Fig. 10.15.

Fig. 10.15 Multiple Catalytic Reactors with External Heat Exchangers

Cost of number of reactors and heat exchangers is becoming prohibitive in many

cases. Also heat recovery from the exit gas mixture from a reactor may not be
possible in each case.
Quench bed reactors are developed for carrying out exothermic reactions at high
temperature. Figures 10.16(a) and (b) depict such designs. Entire reactor is one
vessel in which there are a number of catalyst beds (usually 3 or 4).
A known portion of incoming gas mixture (stream F1 which is more than 50% of
total feed) is introduced at the bottom at relatively low temperature. It passes
through the annulus around catalyst baskets, thereby keeping the shell of the
reactor cool and also gets some heat from the reacting gas mixture. F1 stream
enters the shell and tube heat exchanger, located at top and gets heated by
outgoing gas mixture. Hot F1 stream enters first catalyst bed and partial
conversion takes place thereby the gas mixture is heated up. Portion of the main
gas stream is so controlled that temperature of first bed exit gas stream is
controlled at predefined value. Partially reacted gas mixture from first bed is
cooled by quenching by a definite proportion of the fresh gas mixture (first
quench). Mixing of both gas mixtures results in lowering of the temperature of
the gas mixture, entering second catalyst bed. In this manner, temperature of gas
mixture, entering other beds is controlled by quenching, thereby, conversion in
each bed is increased. Gas mixture from final catalyst bed is taken out through a
central pipe and taken to the heat exchanger for exchanging heat with incoming
gas mixture. Overall effect is to get definite conversion per pass through the
converter without exchanging heat in external heat exchangers.
Since number of moles reduce during syntheses of ammonia and methanol, high
pressure is employed. Quench bed reactor for such syntheses is a multi-layer
wall vessel. As explained earlier, wall temperature of the reactor is kept low by
the flow of ingoing gas mixture in the annulus. This is advantageous as at low
temperature, lower wall thickness is required.
In Fig. 10.16(a), gas flow through the catalyst bed is axial while in Fig. 10.16(b),
it is radial. Design, shown in Fig. 10.16(b), results in lower pressure drop in the
gas mixture, resulting in energy saving. Volume of catalyst in each bed differs.
Volume of fourth bed is highest while that of first bed is lowest. Volumes are
decided by reaction rate, catalyst activity and safety margin against possible
deactivation due to catalyst poisons.
While dissociating nitrous oxide to nitrogen and oxygen, increase in moles takes
place. According to Le Chatélier’s principle, low pressure favours the reaction.
Therefore, a quench bed reactor with multiple catalyst beds for dissociation of
nitrous oxide is designed for low pressure.
Design of fixed catalyst bed reactors for gaseous reactions thus can be
innovative. Process Engineer can design an efficient system with best heat
recovery options. Some other designs of industrial fixed bed catalytic reactors
are as follows.
(i) Packed bed reactor, (ii) shell and tube heat exchanger in which catalysts are
placed inside the tubes, (iii) annulus flow-type reactor in which catalyst are
placed inside the inner tube. (Feed gas first pass through the annulus and then
enter the inner tube. Product gases are taken out from the bottom of inner tube),
(iv) reactor in which catalyst can be in form of mesh. Mesh of catalyst may be
placed inside tower. Feed gas mixture
 Fig. 10.16 Multiple Quench Bed Reactor

is introduced from the top of tower. Product gases leave from the bottom of
catalyst mesh. (For example, ammonia oxidation to nitric acid.) In some cases
catalyst mesh may be wound around the tubes of shell and tube heat exchanger.
(For example, reactor of acetaldehyde plant.) In partial oxidation of ethanol
vapour, silver catalyst in the form of mesh is wrapped around the tubes of shell
and tube heat exchanger.
10.11.1 Industrial Examples of Fixed Bed Catalytic
Reactor
Product Major reaction step Catalyst
1. Ammonia N2 + 3 H2 2 NH3 FeO/Fe2O3,
promoted by
Al2O3 and K2O
2. Nitric acid 4 NH3 + 5 O2 → 4 NO + 5 H2O Pt and Rh
2 NO + O2 → 2 NO2
3 NO2 + H2O → 2 HNO3 + NO
3. Sulphur trioxide SO2 + 1/2 O2 → SO3 V2O5 + K2O
(Sulphuric acid plant)
4. Methanol CO + 2 H2 → CH3OH CuO
5. Formaldehyde CH3OH + 1/2 O2 → HCHO +
H2O Ag
6. Aniline C6H5OH + NH3 → C6H5NH2 + Ni Supported
CO2 + H2O on Al2O3
7. Disintegration of N2O → N2 + 1/2 O2 Transition
nitrous oxide metal oxides
8. Ethyl benzene C6H6 + C2H4 → C6H5CH2CH3 Zeolite
Fixed bed catalytic reactors are operated in both adiabatic and non-adiabatic
manner. If the heat of reaction is small then adiabatic packed bed-type reactor is
used as fixed bed reactor. In such a type of reactor, there is no provision for heat
transfer.
But the reactions with a large heat of reaction as well as reactions that are
extremely temperature sensitive are carried out in non-adiabatic reactors in
which indirect heat exchange takes place by circulating heat transfer medium
integrated in the fixed bed. With large heat of reaction, shell and tube-type
design is preferred in which catalyst is inside the tubes and heat transfer medium
is circulated on shell side. Alternatively, multiple quench bed reactor can be used
(e.g., ammonia synthesis). When high pressure and high temperature reaction is
carried out with indirect heat exchangers, conversion per pass can be near
equilibrium. However, in multiple quench bed reactor, temperature of final exit
stream controls the conversion per pass.
The heart of a fixed-bed reactor and the site of the chemical reaction is the
catalyst. The overall reaction taking place on catalyst surface can be theoretically
divided into the following five separate steps (a well accepted mechanism in
chemical reaction engineering):
1. Diffusion of the reactants from the gas space through the outer gas particle
boundary larger, macropores and micropores.
2. Chemisorption on active centres
3. Surface reactions
4. Desorption of the products
5. Back-diffusion of the products into the gas space. Also deactivation of catalyst
due to fouling, poisoning and elevation in temperature must be considered.
In the development of rate equation only a few of these five steps are considered
which control the overall rate; other steps may be either neglected or combined.
These steps do not necessarily proceed in series or in parallel, frequently making
it impossible to combine them by simple means.
For different type of solid catalysed reactions and for different mechanisms, rate
equations are given in Table 4.8 of Ref. 2.
For example, if the chemical reaction is A R
Assumed mechanism for this reaction is
A + S AS (Absorption)
AS RS (Reaction)
RS R + S (Desorption)
If desorption of R controls the overall rate, then rate equation is

rR =   (10.25)

where, rR = Rate of formation of product R (overall)

k = Overall reaction rate constant
kA = Reaction rate constant for forward reaction
K = Overall equilibrium constant for the chemical reaction
CA, CR = Concentrations of A and R, respectively.
All fixed bed catalytic reactors are assumed to behave like ideal plug flow
reactor. Equation used for the design or sizing of fixed bed reactor is
 =   (10.12)

where, W = Mass of catalyst, kg

FAo = Molar flow rate of limiting reactant in feed, kmol/s
XA = Fractional conversion of limiting reactant A

or XA =

–rA′ = Overall rate of reaction in

S = Surface area of catalyst, m2/kg

Most difficult part is the development of reliable expression of overall rate of
reaction. It also requires detailed kinetic data.
Rearranging Eq. (10.12),

  (10.26)

For a real reactor with packed catalyst bed, Eq. (10.26) is modified as under.
  (10.27)

where η represents catalyst effectiveness while ζ represents deactivation

function. Both these factors are derived from actual experience or pilot plant
studies for a reasonable long period.
10.11.2 Scale-up Method
Any type of commercial scale reactor cannot be designed totally based on
theoretical equations. It requires at least laboratory and/or pilot plant data of the
reactions involved. A satisfactory scale-up procedure requires a stepwise
empirical approach in which the size of the reactor is increased successively.
Basically the rate of chemical reaction is independent of the size and structure of
a reactor. But the overall rate of reaction is influenced by rate of mass transfer
and rate of heat transfer which are usually controlled by the size and structure of
the reactor. For example, due to significant wall effect smaller size reactor
provides higher value of mass transfer coefficient compared to larger size
reactor. Shell and tube-type construction provides higher heat transfer
coefficient, compared to jacketted tower-type structure. Thus a chemical reaction
is indirectly affected by reactor type and scale. Based on only lab scale data, it is
not possible to predict the reliable values of overall rate of reaction, reaction
time, and product composition for commercial scale reactor.
Design the fixed bed catalytic reactor for 1000 t/a
diethylbenzene plant based on the following data obtained on a pilot plant
reactor:
Data of pilot plant reactor14
(i) Feed composition
Component Mass %
Ethyl benzene 89
Ethanol 11
Feed flow rate in pilot plant reactor = 100 kg/h, Feed temperature = 375°C
(ii) Chemical reaction
(a) Main reaction
C6H5C2H5(g) + C2H5OH(g) C6H4–(C2H5)2(g) + H2O(g)
Ethyl benzene Ethanol Diethyl benzene Water
Reaction temperature = 375° C
(b) Side reactions are taking place in the reactor.
(iii) Product composition:
Component mass %
Diethyl benzene 16
Benzene 6
Ethyl benzene 74
Water 4
Product rate is approximately equal to the feed rate.
(iv) Catalyst: Zeolite with binders
Density of catalyst particle: 630 kg/m3
Catalyst porosity: 0.3
Form of catalyst: 1.5 mm extrudates of 4 to 6 mm length
(v) Type of reactor: Shell and tube heat exchanger type in which catalyst is
placed inside the tube.
(vi) Reaction is endothermic and heat of reation is 29 kJ/mole ethanol. It is
carried out in isothermal manner. Dowtherm A is to be circulated in liquid
form on shell side to maintain the isothermal condition.
(vii) = 17.3

(viii) Superficial mass velocity of feed gas

G = 0.55 kg/(m2 · s)
Feed composition, feed temperature, reaction temperature, operating pressure,
catalyst, type of reactor, etc., are same in pilot plant, and commercial reactor.
Solution
Capacity of plant = 1000 t/a of diethyl benzene (DEB)
Let number of working days per annum = 330
Production rate of DEB = = 126.26 kg/h

Let design production rate of DEB = 150 kg/h

Table 10.12 Product Composition for Commercial Scale Plant
Component mass % kg/h

DEB 16 150.00
Benzene 6 56.25
Ethyl benzene 74 693.75
Water 4 37.50
Total 100 937.50

Feed rate ≅ Product rate

Table 10.13 Feed Composition for Commercial Scale Plant
Component mass % kg/h

Ethyl benzene 89 834.375

Ethanol 11 103.125
Total 100 937.500

Ethanol is the limiting reactant. It does not appear in the product mixture which
means conversion based on ethanol is 100%.
Mass of catalyst required in commercial scale plant = 17.3 × 103.125 = 1784 kg
(min.)
Volume of solid catalyst =

= = 2.8317 m3

Porosity of catalyst bed = 0.3

Bulk volume occupied by the catalyst = 2.8317/0.7 = 4.045 m3
Superficial mass velocity of feed gas G = 0.55 kg/(m2 · s)
Total cross-sectional area of catalyst (or of) tubes = = 0.4735 m2

Select 50 mm NB tubes of 14 BWG (3.76 mm) thickness.

MOC of tube = SS 316
Tube OD = 50.8 mm, Tube ID = 43.28 mm
Total number of tubes required nt = = 322

Length of tubes required

L = =

= 8.539 m
Select the tubes of 10 m length. Above calculations have fixed the value of heat
transfer area available for heat transfer. Temperatures and flow rate of Dowtherm
A should be fixed to satisfy the heat transfer requirement.
Aav = Nt π do L
= 322 × π × 0.0508 × 10
= 513.89 m2
Heat duty, θt = ΔHr × kmol/h of ethanol consumption = 29 × 1000 ×
= 65 013.6 kJ/h ≡ 18.06 kW
Calculations of fixed bed side film coefficient, hi:
dp = equivalent diameter of catalyst cylindrical particle, m

dp = 0.002 565 m

= = 0.059 < 0.35

= 0.813 (dpG/μ)0.9  (10.28)

G = 0.55 kg/(m2 · s)
k = 0.04 W/(m · K)
μ = 0.015 mPa · s = 0.015 × 10–3 Pa · s

hi =

= 31.49 W/(m2 · °C)

For heat transfer in packed beds2, heat transfer coefficient for the fluid passing
through the packed bed can be calculated by Eq. (10.28) or Eq. (10.29). Lesser
of the two is selected.
=   (10.29)

hi = 457.51 W/(m2 · °C)

Let hi = 31.49 W/(m2 · °C) (Lesser of two values)
Shell side heat transfer coefficient, ho:
Let tube pitch, Pt = 1.25 do = 1.25 × 0.0508 = 0.0635 m
Type of tube arrangement = Equilateral triangular
Equivalent diameter, de =

= (0.06352 – 0.907 × 0.05082) = 0.036 63 m

Shell side mass flow rate, m· :

ϕt = m· CL Δt
Table 10.14 Properties of Dowtherm A at 380°C15
Property Value

CL 2.5832 kJ/(kg · °C)

μ 0.135 mPa · s
k 0.098 W/(m · °C)

Let Δt = 2°C for Dowtherm–A

m· = = 3.4957 kg/s
Density of Dowtherm A at 380°C, ρl = 0.709 kg/L

Circulation rate, qv = = 17.775 m3/h

Shell side flow area, As can be calculated using Eq. (6.33).

Shell inside diameter, Ds

Db = = 50.8 (6.1)

(k1 = 0.319 and n1 = 2.285 from Table 6.2)

Db = 1048.46 mm
Let Di = 1065 mm
Baffle spacing, Bs = 0.4 Di = 426 mm

As = = 0.0893 m2

Gs = = 39.1456 kg/(m2 · s)

Reynolds number, Re:

Re = = 10 621.5

Prandtl number, Pr:

Pr =

From Fig. 6.14, Jh = 0.0058 for 25% baffle cut

Using Eq. (6.40),
= 0.0058 × 10621.5 × 3.5580.33

ho = 250.55 W/(m2 · °C)

Overall heat transfer coefficient, Uo:
hod = hid = 5000 W/(m2 · °C)

=
Uo = 23.8356 W/(m2 · °C)
ϕt = UoA ΔTln = 18.06 × 103 = 23.8356 × 513.89 × ΔTln
ΔTln = 1.474°C
Very low temperature difference is required for the required duty of heat transfer.
Hence, area available is adequate for heat transfer. Dowtherm-A will be
circulated under pressure above its saturation pressure (8.21 bar a)
corresponding to operating temperature of 380°C. This means shell will be under
pressure.
Note: Empty volume of catalyst tubes works out to be 4.7272 m3 (based on 10
m tube length) which is 17.11% in excess over the bulk volume of the catalyst
(4.045 m3). In reality, manufacturing tolerance of tubes (±12.5%) and support
arrangement at the bottom of the tubes will call for sufficient excess volume so
that required volume of the catalyst can be packed.
Pressure drop of the catalyst packed tubes is usually much higher than that of
clean tubes. Manufactures of catalysts provide data relating to pressure drop in
the catalyst bed and suggest right size of the tube (i.e., diameter).
Over a period of operation some catalyst may settle or crumble and pressure
drop may increase. All these aspects will have to be borne in mind while
designing a reactor with the use of a catalyst.

EXERCISES
10.1 Sulphuryl chloride (SO2Cl2) formation is carried out in a batch reactor at
0°C and 330 kPa a. Reaction time is 0.6 h.
Reaction:
SO2(l) + Cl2(g) → SO2Cl2(l)
Sulphur dioxide Chlorine Sulphuryl chloride
At the reaction conditions, sulphur dioxide is in liquid state and it is a limiting
reactant. Conversion of SO2 at the end of reaction (at equilibrium) is 60%.
Chlorine supply rate to reactor is 10% higher than that consumed. Reaction is
carried out in a glass lined reactor. Heat of reaction at 0°C is 75 kJ/mol
(exothermic). For the production of 1000 kg of SO2Cl2 per batch,
(a) decide the diameter and height of batch reactor.
(b) select the suitable agitator for this gas-liquid reaction and calculate the rpm
and power required by the agitator.
(c) calculate the heat duty of overhead condenser. Overhead condenser will
recycle almost all excess SO2 back to the reactor, and
(d) calculate the heat duty of jacket.
10.2 In Example 10.5, a jet reactor is used for hydrogenation of edible oil.
Instead of the jet reactor, a reactor with gas induction type stirrer6 and spiral
jacket is to be designed as shown in Fig. 10.17. About 200 kg catalyst (1 to 2
μm particle size) is used with 10 t soybean oil. Use spiral baffled jacket (Fig.
10.9(c)) for heat transfer. Following equation5 can be used to calculate jacket
side heat transfer coefficient.
 Fig. 10.17 Gas Induction-Type Tube Strirrer for Hydrogenation

For Re > 10 000

= 0.027 Re0.8 Pr0.33   (10.30)

where, De = Equivalent diameter for cross section, m

De = 4 W
W = Width of jacket, m
Re =

u =

p = pitch of baffle spiral, m

ṁ' = Effective mass flow rate through spiral, kg/s
ṁ' = 0.6 ṁ
ṁ = Actual mass flow rate through spiral jacket, kg/s.
Dj = , mean diameter of jacket, m

Inside heat transfer coefficient can be safely calculated by following equation6:

= 0.36 Re2/3 Pr1/3   (10.31)

where, Di = Inside diameter of shell of reactor, m

If heat transfer area provided hy spiral jacket is not sufficient for the given heat
duty, then use internal helical coil. Coil side heat transfer coefficient can be
calculated by Eq. (6.19).
For coil, De = dci, where, dci = Inside diameter of coil. Also velocity of fluid

through coil, υc = .

10.3 Acetic acid is manufactured via carbonylation of methanol. Reaction is

carried out at 180°C and 50 atm a in presence of liquid phase rhodium
promoted methyl iodide as catalyst. Reaction is carried out in a bubble
column reactor. Acetic acid, itself is used as solvent.
Determine the following:
(a) Working volume of reactor.
(b) Diameter of reactor.
(c) Height of liquid inside the reactor.
(d) Heat duty of overhead condenser in which exit gas-vapour mixture is cooled
down to 40°C.
(e) Design a spiral baffled jacket around bubble column reactor.
Data
(i) Reaction:
CH3OH(l) + CO(g) CH3COOH(l)
CH3OH(l) + CH3COOH(l) → CH3COOCH3(l) + H2O(l)
(ii) Bubble column reactor can be approximated as an ideal mixed flow reactor.
Its uniform composition is 80% acetic acid, 10% methanol, 5% methyl
iodide, 2.5% methyl acetate, and 2.5% water (by mass).
(iii) Feed rate of carbon monoxide is 10% higher than that consumed in the
reaction.
(iv) Cooling water is available in the plant at 32°C.
(v) Overall rate of reaction does not depend on KLa, if Sg > 2 cm/s.
(vi) Rate of chemical reaction
–rA = = k CA, k = 89.26 × 10–5 s–1

CA = Concentration of methanol, kmol/m3

(vii) Production rate of acetic acid = 50 tpd
(viii) Heat of reaction at reaction temperature: ΔHr = –137 kJ/mol
(ix) Conversion per pass = 70%
10.4 Design the fixed bed catalystic reactor for 50 kg/h ethyl chloride plant
based on the following data obtained on a pilot plant reactor16:
Data of pilot plant reactor:
(i) Feed flow rate = 0.025 53 kmol/h.
(ii) Feed composition
Component mole %
Methane 86.55
Ethylene 7.40
Hydrogen chloride 6.05
(iii) Feed temperature = 171.1°C
(iv) Reaction is taking place at 171.1°C and 2859.22 kPa g.
C2H4(g) + HCl(g) → C2H5Cl (g)
Ethylene Hydrogen Chloride Ethyl chloride
(v) Product composition
Component mole %
CH4 90.77
C2H4 2.88
HCl 1.47
C2H5Cl 4.88
(vi) Catalyst: Zirconium oxychloride supported on silica gel
Bulk density = 650 kg/m3
Porosity of catalyst bed = 0.32
(vii) = 177.62

(viii) Superficial mass velocity of feed gas, Gg = 58.93 kg/(m2 · h)

(ix) Rate of reaction
– rA = 2.62 × 10–4

(x) Reaction is exothermic and heat of reaction at reaction temperature is 74

kJ/mol C2H4. It can be carried out in a shell and tube heat exchanger type
reactor.
Feed composition, feed temperature, reaction temperature, operating pressure,
catalyst, etc., are assumed same in pilot plant and commercial scale reactors.
10.5 Design a jet loop reactor for chlorination of benzene.
Stoichiometry
C6H6(l) + Cl2(g) = C6H5Cl(l) + HCl(g) ΔHor1 = –131.5 kJ/mol
C6H5Cl(l) + Cl2(g) = C6H5Cl2(l) + HCl(g) ΔHor2 = –124.4 kJ/mol
Reaction is carried out in jet loop reactor (similar to one shown in Fig. 10.1 (i) at
2.4 bar a and 55°C. Rates of reactions are first order and are given by
-rB = k1CB, rM = k1CB –k2CM and rD = k2CM
where, k1 = 1.0 × 10–4 s–1 and k2 = 0.15 × 10–4 s–1
Subscripts B, M and D denote benzene, monochloro-benzene and
dichlorobenzene, respectively. CB and CM are concentrations in mol/L. Ferric
 chlorile (FeCl3) is used as a catalyst (0.1 g FeCl3/100 g benzene).
Overall conversion of 87% of benzene is to be considered. Assume negligible
presence of chlorine in HCl gas stream, escaping the reactor and negligible
leftover in the liquid phase.
It is desired to produe 1000 kg monochlorobenzene per batch together with
byproduct dichlorobenzene and unreacted benzene. Calculate the following:
(a) Product composition.
(b) Working volume of the reactor.
(c) Considering 20% space for gas over liquid, find diameter and height of
reactor.
(d) Heat duty of external heat exchanger.
(e) Capacity of circulation pump.
10.6 Hydrodenitrogenation of straight run heavy gas oil is to be carried out in a
refinery in a trickle bed reactor18,19. Gas oil contains 900 mg/L nitrogen (as
N2). It is to be reduced to 15 mg/L by passing through a trickle bed reactor.
Hydrogen (much in excess) and gas oil are passed cocurrently over a fixed
bed of catalyst. Catalyst pellets are made of alumina, coated with nickel and
tungsten. Reaction is carried out at 415 bar a and 370°C. Flow rate of gas oil
is 5000 kg/h. Liquid Hourly Space Velocity (LHSV) is to be kept at 1000 h–1.
LHSV is defined as volume of liquid fed per hour per unit volume of catalyst.
Porosity of the catalyst bed (ε) is 0.3. Assume that hydrodenitrogenation
reaction is isothermal, irreversible and of first order with respect to reactant
concentration.
Rate constant kL = 1.624 74 × 10–3 cm3 of liquid/(cm3 of catalyst volume)
Density of gas oil is 800 kg/m3 at the operating conditions.
Calculate the volume of catalyst bed to be packed in the reactor.
Following assumptions are made:
(i) Plug flow of liquid, i.e., no despersion in the axial direction in the catalyst
bed.
(ii) There is no mass transfer or heat transfer limitations between gas and liquid,
between liquid and solid catalyst or inside catalyst pellets.
(iii) Catalyst pellets are completely wet (i.e., bathed) with the liquid.
(iv) No vaporisation or condensation takes place.

REFERENCES
1. Leavenspiel, O., Chemical Reaction Engineering, 2nd Ed., John Wiley &
 Sons, Inc., USA, 1972.
2. Perry, R. H., and D. Green, Perry’s Chemical Engineers’ Handbook, 6th Ed.,
McGraw-Hill Education, USA, 1984.
3. Elvers, B., Hawking, S. and Schulz, G. (Editors), Ullmann’s Encyclopedia of
Industrial Chemistry, Vol. B4, 5th Ed. VCH Verlagsgesellschaft, GmbH,
Germany, 1992.
4. Smith, J. M., Chemical Engineering Kinetics, 3rd Ed., McGraw-Hill
Education, USA, 1981.
5. Bondy, F., and S. Lippa, Chem. Engg, 90 (7), 1983, p. 62.
6. Elvers, B., Hawking, S. and Schulz, G. (Ed.), Ullman’s Encyclopedia of
Industrial Chemistry, Vol. B2, 5th Ed., Germany VCH Verlagsgesellschaft
GmbH, Germany 1988, p. 25–15, p. 25–17 and 25–21 to 25–22.
7. Moore, S., and S. Samdani, Chem. Engg., 101 (3), March, 1994, p. 32.
8. Doraiswamy, L. K., and M. M. Sharma, Heterogeneous Reactions; Analysis,
Examples and Reactor Design, Vol.: 1 & 2, John Wiley & Sons, USA, 1983.
9. Rase, H. F., Chemical Reactor Design for Process Plants, Vol. 1 & 2, John
Wiley & Sons, Inc., USA, 1977.
10. Johnson, G. R., and B. L. Crynes, Ind. Engg. Chem., Process Des. & Dev.,
13 (1), 1974, p. 6.
11. Denbich, K. G., and Turner, T. C., Chemical Reactor Theory–An
Introduction, 3rd Ed., Cambridge University Press, 1984.
12. Van den Hark S. M, Harrod, Applied Catalysis A: General, 210, 2001, p .
207.
13. Fogler, H.S., Elements of Chemical Reaction Engineering, 3rd Ed., Prentice-
Hall of India Pvt. Ltd., New Delhi, 2003.
14. Bokade, V. V. and R. Joshi, Chemical Engineering World, 37 (6) June, 2002,
p. 59.
15. Dowtherm A, Product Technical Data, The Dow Chemical, Co., USA 1991.
16. Thodos, G., and L. Stutzman, Ind. Engg. Chem., 50 (3), March, 1958, p. 413.
17. Davis, M. S., and R. J. Davis; Fundamentals of Chemical Reaction
Engineering, McGraw-Hill Education, USA, 2003.
18. Satterfield, C. N., AIChE J., 21 (2), March, 1975. p. 209.
19. Flinn, I. A., O. A. Larson, and H. Beuther, Hydrocarbon Processing &
Petroleum Refiner, 42 (9), 1963, p. 129.
11
Process Design of Utility Systems
11.1 PREAMBLE
Every manufacturing industry is dependent on one or more utilities. Role of
process utilities in chemical industry is of great importance. While it is observed
that most large chemical complexes have well-designed process utility systems,
small and medium scale chemical industry rather undermine the importance of
utilities and they suffer; at times heavily, in operation in terms of production loss
and product cost. Steam, heating by thermic fluid, cooling water, refrigeration,
compressed air, inert gas, etc., are needed at various stages of production and
careful design of these systems can pay rich dividends. This chapter is devoted
to process engineering aspects of the utility systems. This will include the major
parameters that are required for preparation of specifications of a utility system
and its operation thereof. Equipments, required to provide these utilities, are
described at suitable locations without going into the mechanical design aspects
of the same.

11.2 PROCESS HEATING SYSTEMS

Unit operations as well as unit processes require heating. It is carried out by
direct firing, supplying vapour condensation systems (steam and Dowtherm®
type), thermic fluid (hot oil), hot water and hot air. While hot air and hot water
systems are less common, steam and thermic fluid heating are most commonly
employed in chemical industry. Spray dryers and furnaces for metal processing
use hot air or hot flue gas for heating. Many heat sensitive product processing
call for hot water heating. Except for such specialised applications, steam or
thermic fluid are common heating mediums. Fig. 4.9(a), (b) and (c) show such
heating applications in a continuous stirred tank reactor (CSTR).
11.2.1 Steam Heating
Use of steam is by far the most common in industry. Reboilers of distillation
columns, jackets/limpet coils of CSTR or other reactors, desorption columns,
evaporators, dryers, indirect heating applications, activation of carbon and
numerous other applications can be listed where steam is used as a heating
medium. Some use steam indirectly while a few use live steam. Steam is non-
toxic and inflammable and its leak in saturated state can be seen visually so that
steps can be taken to stop the leakages.
Study of properties of steam will reveal some interesting facts. Two important
facts are worth noting. Saturation temperature increases and latent heat of
vaporisation is reducing with increase in steam pressure. These are rather
contradictory parameters for heating applications. Hence, selection of steam
temperature for heating application is critical. Unnecessary high temperature of
steam may not be preferred in design as it will call for higher design pressure of
the equipment. Also in heating applications, mostly latent heat of vapourisation
(λ) is utilised while sensible heat is lost in condensate. Thus at higher pressure,
more energy is wasted in the form of hot condensate. In most steam heating
applications, condensing steam temperature is selected 20 to 25°C higher than
the temperature requirement of the process fluid.
Consider manufacture of amino and alkyd resins. In the manufacture of amino
resins temperature to be maintained in the reactor is in the range of 95 to 120°C
while in the manufacture of alkyd resins, temperature requirement varies in the
range of 220 to 240°C. Thus for amino resin manufacture one would try to use
steam at 145 to 150°C (i.e., steam at saturation pressure in the range of 3.7 to 4.2
bar a). This selection is considered appropriate and one would design steam
distribution system for about 5 bar a. If the same considerations are followed for
alkyd resin manufacturing, steam temperature in the range 240 to 260°C would
be required which will mean saturated steam in the pressure range 34 to 50 bar a
will be required, Clearly design of steam equipment will be therefore at much
higher pressure and its cost would be high. Steam at 4 bar a has λ = 2133 kJ/kg
while at 50 bar a, λ = 1639.7 kJ/kg. This means energy usage will be lower by at
least 23% at higher pressure. Thus for the manufacture of alkyd resin, use of
thermic fluid or other condensable vapour should be considered rather than
steam.
Consider a distillation system in which methanol-water is to be separated. It is
clear that bottom temperature in the distillation column could be around 105°C,
depending on pressure drop in the column. Therefore steam temperature in the
reboiler should be kept at about 125 to 130°C (i.e., in the pressure range of 2.3 to
2.7 bar a). This is a common experience in industry. In vacuum creating systems,
steam at 3.5 to 4 bar a with a superheat of about 10°C is commonly used as a
primary fluid stream. These examples would suffice to justify the importance of
steam pressure. Process engineer should therefore study the heating temperature
requirement and select the right steam pressure. Saturated steam up to about 20
bar a (saturation temperature: 212.4°C) is recommended for process heating
beyond which equipment cost becomes high due to high design pressure.
Steam in superheated condition is not preferred for indirect heating system. This
is because superheated steam acts as a gas and heat transfer (film) coefficient of
the superheated phase would be much lower than that for condensing. Much of
the heat transfer surface would therefore be required for first desuperheating
steam. This will increase the cost of heating equipment. However, live
superheated steam at near atmospheric pressure is used for heating and
activation of carbon as condensation of steam on carbon is not desired in the
equipment. Similarly, regeneration of desiccant beds can also be carried out with
live superheated steam. Thus an application decides the steam conditions for use.
Use of wet steam for heating is not desired. This is because sensible heat (h) of
condensate is much lower than latent heat of vaporisation (λ). For example, at 3
bar a, h = 561.4 kJ/kg while λ = 2163.2 kJ/kg. When steam is conveyed over a
long distance, there could be condensate formation in the pipe and steam may
have significant wetness. Under such circumstance, first moisture of steam
should be removed before introducing to the heater. When steam pressure
reduction is desired, a separator is provided on upstream pipe to ensure that
saturated steam enters the pressure reducing valve. In this case in addition to the
possibility of carry-over of moisture to the heater, erosion of internals of the
pressure reducing valve can be experienced. A typical pressure reducing station
is shown in Fig. 11.1.

Fig. 11.1 Typical Steam Pressure Reduction System

Steam is generated in a boiler. In small and medium scale chemical industry,

steam generation is carried out up to a pressure of 12 to 15 bar g. Treated water
is used in the boiler to avoid build up of solids in the steam drum which can
cause fouling and/or corrosion. Requirements of boiler water quality are
specified in IS:10 392 for such boilers. Blowdown from the boiler is maintained
to meet these requirements. Scale build-up in the boiler will reduce heat transfer,
leading to increase in flue gas temperature thereby waste of energy. Even
explosions of boilers are reported due to excessive fouling. Combustion of fuels
and steam generation are dealt in many books and reader is advised to refer
them. Stoichiometric aspects of combustion are dealt at a length in Ref. 1.
Generation of steam at high pressures call for high quality boiler feed water.
Quality requirements for boilers generating steam for turbines (i.e., for power
generation) are extremely stringent as carry-over of solids (in particular silica)
mean heavy erosion of steam turbine rotors.
Insulation of steam system is highly recommended. Apart from safety of the
personnel, condensation of steam without utilisation in the heater can be
avoided. Variety of insulating materials are in use; such as calcium silicate,
mineral wool, etc.
Cascading of steam system is found very useful in its efficient utilisation.
Superheated steam at high pressure and temperature is first utilised for power
generation. Back pressure steam from the turbines can be used for heating
applications. Different flow sheets can be developed for cascading which can
result in optimum steam generation. Reference 1 describes several case studies
of steam cascading systems. Such systems are common in heavy chemical
plants, refinery, sugar mills, paper industry, etc. A captive power plant based on
steam turbines, followed by use of saturated steam at lower pressures for heating
can be ideal in many chemical plants to improve overall economy. Process
engineer should explore the possibility of a cascading steam balance to utilise
steam efficiently. Figure 4.4 is an example of a typical cascade steam system.
Use of a thermocompressor can be advantageous in reducing steam
consumption. Ejectors are commonly used as thermocompressors. However,
mechanical vapour compressors are also employed in industry. Twin lobe
blowers are popular for use as mechanical vapour compressors. Since
mechanical compressor discharges steam at higher temperature, compression
ratio is usually low; say 1.2 to 1.3. However, even this compression of steam can
be useful in a creating sufficient temperature difference across the heat transfer
surface for the desired heat duty. In Sec. 8.9.1(f), applications of an ejector for
reboiler steam are presented.
In multiple effect evaporators, thermocompressors are effectively used to
improve steam economy. Concentration of effluents, concentration of sodium
sulphate solution and many other application can be listed for which multiple
effect evaporators are used with low pressure steam and thermocompressor.
Following example will demonstrate the application.
Sodium sulphate solution is produced in sizable quantity in a
plant manufacturing sebacic acid. Strength of the feed solution is 8% by mass. In
one plant, triple effect evaporator system is to be installed for concentrating
solution to its saturation level. Slurry of Na2SO4 solution is obtained which is
passed through a Oslo crystalliser where crystals of Na2SO4 having 10%
moisture (by mass) is obtained. Saturated mother liquor from the crystalliser at
the rate of 315 kg/h, containing 30% (by mass) Na2SO4, from the last effect is
recycled and mixed with the fresh feed. Proposed system is shown in Fig. 11.2
below.
Two designs are being considered. In one design, saturated steam at 1.3 bar a is
used in the first effect for heating. In second design, part of the vapour from first
effect is compressed to 1.3 bar a in a thermocompressor (an ejector) with the
help of motive saturated steam at 7 bar a. Last effect operates at 0.125 bar a
(under vacuum) which is maintained by steam ejectors and a surface condenser.
The condenser uses cooling water at 32oC which limits the pressure in the last
effect. Anticipated operating conditions for both designs are listed in the
following table:
 Fig. 11.2 Triple Effect Evaporator System for concerntrating Aqueous Sodium Sulphate Solution
Table 11.1 Anticipated Operating Conditions of Evaporator Effects
System without ejector System with ejector
Average heat capacity of solution, Cp Temperature of Solution, Temperature of Solution,
Parameter Pressure kPa Pressure
kJ/(kg·°C) °C °C
a kPa
Liquid* Cond. Vapour Liquid* Cond. Vapour
1st Effect 0.95 103 100 101.325 97 95 84
2nd Effect 0.9 81 78 44 77 74 37
3rd Effect 0.7 56 50 12.5 56 50 12.5

*with elevation of boiling point

Thermocompressor uses motive saturated steam at 7 bar a and compresses 0.725
kg water vapour (at 84 kPa a) from the first effect per kg motive steam to 1.3 bar
a. Assume heat of crystallisation in the last effect to be 20 kJ/kg Na2SO4
(exothermic). Make stoichiometric calculations for a feed rate of 5400 kg/h of
aqueous solution containing 8% Na2SO4 at 35°C and find steam consumptions
for both the designs.
Solution
Basis: Aqueous solution feed rate of 5400 kg/h containing 8% Na2SO4
Let m.s = Fresh steam supply at 1.3 bar a, kg/h
m.1 = Solution flow rate leaving 1st effect, kg/h
m.2 = Solution flow rate leaving 2nd effect, kg/h
Na2SO4 content of feed solution = 5400 × 0.08 = 432 kg/h
Crystals of Na2SO4 are produced with 10% moisture.
Moist crystal flow rate from crystalliser = 432/0.9 = 480 kg/h
Mother liquor rate = 315 kg/h with 30% or 94.5 kg/h Na2SO4
Mother liquor is recycled to the fresh feed and combined feed enters first effect.
Combined feed rate = 5400 + 315 = 5715 kg/h
Total solid in combined feed = 432 + 94.5 = 526.5 kg/h
Average concentration of Na2SO4 in the combined feed = 526.5 × 100/5715 =
9.21%
Total evaporation in all three effects = 5715 – 480 – 315 = 4920 kg/h
Table 11.2 Properties of Steam1, 2
Enthalpy*, kJ/kg
Temperature, °C Satd. steam pressure p, kPa a
Sensible (h) Latent (λ)

165 700 697.1 2064.9

107.1 130 449.2 2237.8
100 101.325 419.0 2256.9
95 84 397.3 2270.7
78 44 327.3 2313.4
74 37 309.8 2329.9
50 12.5 210.3 2382.3

*Reference state: Enthalpy of saturated liquid water at 273.16 K (0.01°C) = 0 kJ/kg

A. Case without Thermocompressor
Enthalpy balance equations for three effects can be written as under.
ṁs × 2237.8 = 5400 × 0.95(103 – 35) + 315 × 0.7(103 – 56)
+ (5715 – ṁ1)2256.9(I)
(5715 – ṁ1)2256.9 = ṁ1 × 0.9(81 – 103) + (ṁ1 – ṁ2)2313.4(II)
(ṁ1 – ṁ2)2313.4 = ṁ2 × 0.7(56 – 81) + (ṁ2 – 480 – 315 )2382.3
– 480 × 0.9 × 20(III)
Solving Eq. (I), (II) and (III),
ṁs = 1827.10, ṁ1 = 4062.52 and ṁ2 = 2415.63 all in kg/h
Average concentration of Na2SO4 in the solution, leaving first effect will be
12.96% and leaving second effect will be 21.8%.
Specific steam consumption = 1827.10/4920 = 0.3714 kg/kg evaporation
Steam economy = 4920/1827.10 = 2.6928 kg evaporation/kg steam
B. Case with Thermocompressor
In this case, vapour amounting 0.725 ṁs is recycled through thermocompressor.
ṁs × (2064.9 + 697.1 – 449.2) + 0.725ṁs × 2270.7
= 5400 × 0.95(97 – 35) + 315 × 0.7(97 – 56) + (5715 – ṁ1)2270.7(IV)
(5715 – ṁ1 – 0.725ṁs)2270.7 = ṁ1 × 0.9(77 – 97) + (ṁ1 – ṁ2)2329.9(V)
(ṁ1 – ṁ2)2329.9 = ṁ2 × 0.7(56 – 77) + (ṁ2 – 480 – 315)2382.3 – 480 × 0.9 ×
20(VI)
Solving Eq. (IV), (V) and (VI),
ms = 1305.37, m1 = 3583.08 and m2 = 2400.00 all in kg/h
Vapour compressed = 0.725 ms = 0.725 × 1305.37 = 946.4 kg/h from first effect
Average concentration of Na2SO4 in the solution, leaving first effect will be
14.69% and leaving second effect will be 21.94%.
Specific steam consumption = 1305.37/4920 = 0.2653 kg/kg evaporation
Steam economy = 4920/1305.37 = 3.769 kg evaporation/kg steam
Reduction in steam consumption = 1827.10 – 1305.37 = 521.73 kg/h or 28.56%
Note that incorporation of the thermocompressor results in significant reduction
of fresh steam input. If low pressure (at 1.3 bar a) steam is used as fresh steam, a
mechanical vapour compressor can be provided to compress vapour amounting
946.7 kg/h from the first effect (from 84 kPa a to 130 kPa a) and similar results
can be achieved. Compression of vapour can be achieved in a twin lobe
compressor which may consume about 40 kW power.
Saturated steam condensate when discharged from a steam trap results in flash
steam due to pressure reduction. Quantum of flash steam depends on pressure
reduction across the trap. Therefore a two phase flow is realised after the trap. It
is a practice to convey both the phases simultaneously to a deaerator or to the
boiler feed tank in the pipe size correctly sized and correct size trap is installed.
Use of Fig. 5.17 can be made to size the pipe. Steam trap manufacturers provide
trap selection chart for a desired pressure drop across the trap and flow rate of
condensate. Amount of flash steam for a given pressure drop for saturated
condensate can be easily calculated by a simple energy balance. Figure 11.3 is
the graph to be used as a guideline for flash steam estimation.

Fig. 11.3 Flash Steam Production from Condensate

11.2.2 Combined Heat and Power Systems

Competitive pressures to reduce cost of production has led to think of Combined
Power and Heat (CHP) systems. This is also known as Cogeneration. When
effectively applied, a CHP system can make a facility less dependent on the
electrical grid or, more commonly, provide increased reliability for a strategic
portion of the electric load at the site. Such a system produces less emissions per
delivered unit of output which is a part of green engineering.
CHP or cogeneration is the sequential production of two forms of useful energy
from a single fuel source. In most CHP applications, chemical energy in fuel is
converted to both mechanical and thermal energy. The mechanical energy is
generally used to generate electricity, while the thermal energy or heat is used to
produce steam, hot water, or hot air. The principal technical advantage of a CHP
system is its ability to extract more useful energy from fuel compared to
traditional energy systems such as conventional power plants that only generate
electricity and industrial boiler systems that only produce steam or hot water for
process applications. Several CHPs can be thought of depending on
thermodynamic possibilities and the need of the user. One typical generalized
system is shown in Fig. 11.4.
Carnot engine3 for power generation is the ideal machine and assumes two
adiabatic steps. In the isothermal step at high temperature TH, heat QH is
absorbed by the working fluid in the engine and in the isothermal step at lower
temperature TC, heat amounting QC is discarded by the fluid. Work produced by
the Carnot engine is
W = QH – QC   (11.1)
 Fig. 11.4 Typical CHP System Options

Thermal efficiency (η) of the Carnot engine is

η = W/QC = 1 – (TC/TH)  (11.2)
Since Carnot engine assumes both steps reversible, its efficiency is highest
among all thermodynamic cycles of power generation.
For a Carnot engine, operating at TH = 773.15 K and TC = 318.98 K, engine
efficiency (η) can be calculated at
η = 1 – (318.98/773.15) = 0.594 or 59.4%
No actual thermodynamic cycle of power generation is fully reversible. For a
typical Rankine cycle3, operating between these temperatures and other specific
steam parameters and turbine efficiency of 100%, cycle efficiency is just 39.6%.
For a steam turbine of 75% efficiency, the cycle efficiency drops to only 29.6%.
Rest 70.4% energy is lost to atmosphere. It can be noticed that actual efficiency
of the power cycle based on Rankine cycle is just 49.8% of that of Carnot cycle.
Similarly, in boilers generating steam, overall efficiencies in the range of 60 to
70% (on GCV basic) are achieved. This is the basis of CHP system where energy
is efficiently utilised by extracting energy in two forms. One can get as high as
80+% overall efficiency of CHP with proper design which means overall energy
loss will be just 20% or less (mostly in flue gas and radiation).
Most CHP applications use gas turbines or reciprocating engines coupled to an
electrical generator. Heat recovered from these prime movers is then used to
produce steam or hot water to meet space heating or cooling requirements or to
control humidity. These systems may include heat exchangers, absorption
chillers, desiccant dehumidifiers, or other types of heat-recovery equipment. In
another common CHP application, boiler steam passes through a steam turbine
to generate electricity. Heat rejected from the steam turbine is then used for
process applications. This suggests that one of the simplest strategies for
implementing cogeneration technology is to replace steam pressure-reducing
valves with backpressure steam turbines and electrical generators. Earlier
mention of Fig. 4.4 was made which itself is an example of cogeneration. Newer
technologies are being developed to improve overall efficiency of cogeneration.
Use of low grade heat (i.e., primarily at low temperature) for power generation
by Kalina cycle4, extracting more heat from flue gas of a boiler by providing
recuperators of suitable material of construction, use of hot water in process
(instead of steam), etc., can be named under this category. From a practical
viewpoint, CHP systems can be of any size and there is no limit as to the ratio in
which the two types of energy are proportioned. In addition, power generated by
a CHP system can be used on-site or exported, and the thermal energy can be
used for heating, cooling, or both. Technical feasibility involves more than
applying appropriate engineering principles in the design of a CHP system to
meet performance criteria. Safety and reliability of the system are considered
most important while selecting a CHP system. Next parameter is the availability
of a fuel which should suit the requirements of the selected CHP system.
Nevertheless, CHP system is very important for chemical process industry and
its use must be critically evaluated by the process engineer in design.
11.2.3 Thermic Fluid Heating Systems
Hot oil heating system is next popular system in small- and medium-scale
chemical industry in India. In earlier section, it was noted that steam is rarely
used for heating beyond 200oC due to (i) high pressure equipment design
requirements, and (ii) boilers generating steam above 24 bar a are expensive.
Commonly, a hot oil system is used in plants where there are many relatively
small high-temperature heating requirements. Instead of building several small
fired heaters or providing small oil heating systems, it can be more economical
to supply heat to the process from circulating hot oil streams and build a single
fired heater that heats the hot oil. Use of hot oil reduces the risk of process
streams being exposed to high tube-wall temperatures that might be experienced
in a fired heater. Such heating systems are preferred even for relatively low
temperature heating in small and medium-scale chemical industry where
applications involve heating at low as well as at high temperatures. These
systems do not come under the purview of Indian Boiler Regulations. Desired
properties of thermic fluid are as given below.
(a) Fluid should be non-toxic, inflammable and non-corrosive.
(b) Fluid should be operable at high temperature without degradation in the
operating temperature range.
(c) Fluid should offer good film coefficient so that heat transfer area is limited.
(d) Fluid should have low vapour pressure.
(e) Fluid should have low pour point and high auto-ignition temperature.
Earlier it was seen that steam heating is unsuitable for alkyd resin
manufacturing. For such a system, hot oil system can suit the requirements. As
an alternate, an eutectic mixture of diphenyl-diphenyl oxide, abbreviated as
DP/DPO (and popularly known as Dowtherm A®), condensing at 257.1°C at
atmospheric pressure, can be considered without any pressure problem. DP/DPO
is thermally stable and is usually operated in a vaporisation-condensation cycle,
similar to the steam system. Condensed liquid DP/DPO is sometimes used for
intermediate temperature heating requirements. However, λ of the condensing
organic DP/DPO vapour is 296.2 kJ/kg at atmospheric pressure. This will mean
condensing rate of the vapour would be about 5.6 times that of steam at 50 bar a.
Cost of organic mixture is substantially high and any leakage of the mixture
from the system would be a costly affair. DP/DPO mixture can be safely
considered up to 400oC at which saturation pressure is 10.63 bar a and latent
heat of vapourisation (λ) is 206.3 kJ/kg. It can be realised that steam is the
preferred heating medium because of its high latent heat of vapourisation (λ) and
low cost.
Properties of some of the thermic fluid in use in India are given in Table 11.3.
Both mineral based and synthetic oils are in use.
Table 11.3 Properties of Thermic Fluids
General High temperature Food Heating/cooling (combined) DP/DPO
purpose applications grade* applications mixture

Operating temperature, 0 to 305 0 to 345 –39 to 345 –80 to 260 13 to 400

°C

Density (ρ), kg/m3

at 20°C 0.87 1.01 0.97 0.76 1.06
at 300°C 0.67 0.81 0.76 0.56 at 250°C 0.85

Kinematic viscosity (γ), mm2/s

at 20°C 108 122 7 1.7 1.02

at 300°C 0.6 0.5 0.33 0.26 at 250°C 0.27
Heat capacity, (Cp) kJ/(kg·°C)

at 20°C 1.92 1.55 1.68 2.10 1.55

at 300°C 2.96 2.57 2.62 3.05 at 250°C 2.32
Thermal Conductivity, (k) [W/(m·K)]

at 20°C 0.137 0.111 0.121 0.110 0.14

at 300°C 0.111 0.095 0.091 0.065 at 250°C 0.1
Fire point, °C 238 216 154 71 127
Auto-ignition 368 400 405 277 621
temperature, °C
Pour point, °C –48 –32 –48 < –85 12
Normal boiling point, °C 356 369 289 192 257.1
Average molar mass (M) 318 252 207 180 156

* US FDA approved
Properties listed in Table 11.3 are for synthetic thermic fluids. However, mineral
oils are also available for use as thermic fluids. In a petroleum refinery, mineral
oil from the operating stream is drawn and used for heating.
A typical P&I diagram of the oil heating system in use is given as Fig. 11.5.
Thermic fluid is pumped and is heated in fuel fired heater. Hot oil is sent to
various heaters and its flow through each heater is controlled individually by
throttling a valve on return line. Return oil is sent to a degassing tank and returns
to the circulation pump. Location of the degassing tank is at an elevation above
all heating equipments. This tank allows removal of any vapour formation along
with non-condensables. All pipes are well insulated. Quality of thermic fluid
must be checked periodically for degraded products and replacement is planned,
if required.
 Fig. 11.5 P&I Diagram for Thermic Fluid Heating System

In Fig. 4.9(a) thermic fluid is proposed for heating and cooling in CSTR. For this
purpose, oil, proposed in the fourth column of Table 11.3 can be recommended.
Solid, liquid or gaseous fuel is used in the fired heater. Firing is controlled by
fuel flow to the heater. Generally, the system is known as ‘Thermo Pack’ and is
rated in terms of heat release rate in kW. Usually thermo packs are operated up
to 80% duty of the design heat rating. Over a period, thermal efficiency of the
heater get reduced due to fouling in the heater which raises flue gas temperature.
It is recommended that flue gas temperature be controlled below 300oC.
11.2.4 Hot Air Systems
Hot air is produced by heating air at positive pressure in a fired heater. Hot air,
so produced is used in spray dryers, tray dryers, in buildings for heating, etc. Air
has low heat capacity and therefore large circulation for heating application is
required which calls for large ducts to convey so that pressure drops are
minimal. It is not recommended for indirect heating applications as its heat
transfer coefficient is low and large heat transfer area will be required for the
heater. A typical P&I diagram of air heating system is shown in Fig. 11.6.
Spray dryers are in use for producing dye powder, powders of pharmaceutical
products, food products (such as tomato powder, catechu powder, etc.). Filtered
hot air is extensively used in this equipment.
11.2.5 Hot Water Systems
As noticed earlier, hot water at high temperature is less preferred over steam.
However, in some special cases, hot water system is employed. To exemplify, in
extraction of herbs with an organic solvent, hot water is recommended. It is
considered ‘soft heating’. Hot water is also used in heating and ventilation
equipment for air conditioning. Heating system is similar to the oil heating
system. However, water pressure is normally kept about 10 bar higher than
saturation pressure at the maximum operating temperature. Multistage
centrifugal pumps, similar to boiler feed water pumps, but usually of stainless
steel construction, are used for circulation. This system is prone to corrosion and
fouling and so quality of water to be used for circulation must be very good. Use
of demineralised water is recommended and dissolved oxygen (air) is removed
by degassing or by chemical means.
Crude mentha oil is fractionated below 100°C under vacuum. Reboiler for the
fractionating equipment is a falling film heater in which either hot thermic fluid
or hot water can be used as heating medium. It is claimed that reboiler, heated
with hot water, gives better recovery of products with less reject at the bottom.
11.2.6 Thermal Insulation for Hot Surfaces
Energy cost is one of the major concerns to industry. Various measures are
adopted by the industry to reduce energy consumption. One of the important
measures is provision of adequate thermal insulation to all hot surfaces. Consider
a case of 50 mm (2 in) NB bare pipe carrying steam at 10 bar a (saturation
temperature = 180°C). Assume ambient temperature as 30°C which means a
temperature difference of 150°C. For this case, heat loss from bare pipe is
estimated at 465 to 475 W/m. If this pipe
 Fig. 11.6 P&I Diagram for Solid Fuel Fired Hot Air System

is provided with good insulation, the loss can be reduced to say 50 W/m. In
simple terms, energy equivalent to 420 W/m can be saved by providing
insulation. This loss could be as high as 2000 W/m for a 300 NB (12 in) pipe.
Several insulating materials are used in industry. For hot surfaces, calcium
silicate, mineral wool, glass wool, saw dust, wood fibre, etc., are in use. Among
these, mineral wool is the most commonly used insulating material in industry.
Rock materials are blended, melted at about 1600°C and fibres are spun of the
blend. Fibres are resin bonded and a felt is produced of desired density and
resilience. Metallic wire mesh is used to stitch to produce mats of desired
thicknesses. Typical thermal conductivities of mineral wool having 150 kg/m3
density at 50°C and 200°C are 0.046 and 0.075 W/(m · °C), respectively. At first
the surface is cleaned to make it free from oil, dust, etc. If the surface is of
stainless steel, aluminium foil of about 0.1 mm thickness is applied. Over the
pipe resin bonded mineral wool mattresses are wrapped. Finally, aluminium
sheet (22 to 24 SWG, i.e., 0.61 to 0.71 mm thickness) with self-screws is
wrapped over the mats.
As explained above increasing thickness of insulation can reduce heat loss.
However, cost considerations restrict insulation thickness. Table 11.4 gives
recommended economic insulation thickness for mineral wool mats.
Table 11.4 Economic Insulation Thickness for Rockwool Insulation*
Nominal pipe size Recommended thickness, mm Temperature range, °C

mm in 50–100 101–200 201–250 251–300 301–400

15/20 ½/¾ 25 40 40 50 80
25 1 23 40 50 50 80
40 1½ 25 40 50 65 80
50 2 25 40 50 65 80
80 3 25 40 50 80 100
100 4 40 50 65 80 100
150 6 40 50 65 80 100
200 8 40 50 75 100 125
300 12 40 50 75 125 150
400 16 50 75 75 125 150
500 20 50 75 100 125 150

600 24 50 75 100 125 175

* Ambient temperature is assumed to be 30–35°C.

For curved surface above 600 mm nominal diameter and flat surfaces, mattresses
of thickness, mentioned for 600 mm NB, are recommended but metallic clits are
recommended on the surface to hold them in place. Process engineer should
provide for economic insulation requirements in design for all heating
applications.

11.3 PROCESS COOLING SYSTEMS

Cooling of process fluids is carried out with the help of different fluids. When a
process stream is to be cooled, its heat can be utilised by exchanging heat with
other colder streams which may form a heat exchanger network. Such utilisation
results in energy conservation, also known as energy integration. However, such
an exchange is possible up to a level to recover heat economically. These aspects
were discussed in Section 6.6.2 in which it was indicated that an external coolant
will be needed beyond a pinch point to cool a stream. Where heat exchanger
network is non-existent, stream cooling below less than 100°C to near ambient
temperature, cooling medium is used in most cases. Cooling water or ambient air
can be considered for the purpose; the latter can be an expensive proposition.
11.3.1 Air Cooling
Where water scarcity is experienced, air cooling is considered. Design of heat
exchangers with air and water was dealt in detail in Chapter 6 where it was
observed that heat transfer area requirement of an air cooler is higher than that
with cooling water. Air coolers are usually finned tube heat exchangers with
induced or forced draft air fans.
There are certain advantages of cooling with air:
1. Use of air cooler enables plant location independent of a water source.
2. Cost of treating cooling water is a factor that is usually high and it requires
intricate knowledge of water chemistry. Air cooling does not need such
hassle.
3. Air cooler can be placed near the process equipment which avoids cooling
piping network.
4. Maintenance requirements are also relatively less for the air coolers.
5. During a power failure, although fan will stop air supply to the cooler but still
some cooling will continue because of ambient air. However, in case of cooler
with cooling water, supply of cooling water stops immediately and cooling
effectiveness will totally be lost.
6. Fouling of a heat exchanger using cooling water is a perennial problem which
calls for frequent cleaning of tubes. For this reason, excess heat transfer area
of the cooler is provided. In comparison, air cooler fins are less prone to
fouling and calls for less cleaning requirement.
7. Air is passed over the fins and process fluid (could be at higher pressure) is
passed through the tubes which call for lower thickness of tubes as well as of
tube sheets.
8. Size of an air cooler is usually is much bigger than a water cooler, calling for
 more ground area. However, air cooler can be placed at an elevation (say
above the building) to save the space. For placing the air cooler at an
elevation, however, civil structure should carefully be designed.
Despite above advantages, cooling with air is less preferred than water cooling
in a tropical country mainly because the process fluid can be cooled to a higher
temperature with air than with cooling water. Consider a location in India where
air attains 42 to 44°C dry bulb temperature and say 28°C wet bulb temperature
(max.) in summer. A cooling tower at this location can deliver cooling water at
32°C. Process fluid at such a location can be cooled up to 55°C with air and up
to 40oC with cooling water, a difference of 15°C in cooling. This is a significant
advantage in most cooling systems. Besides, air cooler need be located in an
open area so that required quantum of ambient air can be made available.
Several designs of air coolers are available (Fig. 6.36). Amongst these, cooler
with forced draft air fan are more common. These fans take relatively cold
ambient air in suction and therefore require somewhat less power than that is
required for induced draft fans. Refrigerant condenser (Exercise 6.10), surface
condenser for a steam turbine, compressed air/gas cooling for small
compressors, etc., are some industrial applications of air coolers.
11.3.2 Water Cooling
Once through cooling systems are less prevalent in industry than a closed
cooling water circuit. In most laboratories, cooling in glassware is carried by raw
water flowing in once through manner. Sea water is used as a cooling medium in
once through manner in large industrial plants (petroleum refinery, power plant,
etc.) located near sea. Due to high salinity of sea water, it is highly corrosive and
hence heat exchangers with tubes of special material of construction (say
Admiralty brass or Cu-Ni alloys) making such exchangers expensive. Special
care is also required to ensure clean sea water (free from fish and other external
materials) pumping to such exchangers. Slime formation in tubes of a heat
exchanger is a common problem when sea water is used. Despite all these
problems, sea as large heat sink attracts industry to use sea water for cooling.
Closed cooling water circuits are more common in industry. Natural and forced-
draft cooling towers are generally used to provide the cooling water required on
a site. Cooling water is used both; in direct contact equipment and in heat
exchangers in indirect manner. In an isolated case, cooling water can also be
used for heating; e.g., evaporation of liquid ammonia at –33oC can be carried out
with cooling water (refer to Example 2.10). Except for such isolated cases,
cooling water is mainly used for cooling a process stream. A typical cooling
circuit is represented in Fig. 11.7.
The minimum temperature that can be reached of cooling water depends on the
local climate. Cooling towers function by evaporating a part of the circulating
water to ambient air, causing the remaining water to cool. If the ambient
temperature and humidity are high, then a cooling water system will be less
effective. Most cooling towers have induced draft fans but cooling tower with
forced draft fans or no fans are not uncommon. Spray ponds are also in use for
circulating water in some air conditioning systems.
Difference in inlet cooling water temperature to and outlet cooling water
temperature from cooling tower is called a range of cooling tower. Difference
between outlet cooling water from cooling tower and ambient wet bulb
temperature is called approach. Usually, cooling towers are designed for 5 to
8°C range. Design approach of a cooling tower is important and it should not be
selected low. Usually, 4 to 5°C approach is specified for the design. Cooling
tower supplier can furnish a graph in which predicted cooling water temperature
can be read for different wet bulb temperature (than design) for the supplied
model.
 Fig. 11.7 Typical Cooling Water Circuit

Considering importance of this utility, usually two cooling water pumps are
installed, each of full capacity of which one is standby. Suction pipe size of the
pump is kept at least one size higher than that of suction nozzle size of the pump.
Similarly, discharge size pipe be kept at least same size that of discharge nozzle
size of the pump or higher. Non-return valves are provided in the discharge of
each pump.
It is a good practice to provide a side stream filter for filtering cooling water. Its
capacity could be 3 to 5% of design recirculation flow rate. Different types of
filters are provided. Amongst these, a stratified sand bed filter and a bag filter
(20 to 40 micron) are common. Clarified water from outlet of the filter is
recycled to cooling tower sump. Turbidity in the circulating water should be
checked once a day. It should preferably be below 5 mg/L on silica (as SiO2)
scale.
Cooling water is usually made up of clarified water with dissolved solids (DS)
less than 500 mg/L. Because of evaporation of some water from the cooling
tower, concentration of DS in cooling water increases. Ratio of DS in cooling
water to that in make up water is known as cycles of concentration. For
calculation of this parameter, normally chloride (as Cl) or silica (as SiO2) ions
are selected. It is desirable to keep cycle of concentration less than 3.5 so that
fouling can be controlled. While some water droplets leave with evaporated
vapour (known as drift loss), intentional blow down is maintained to keep cycles
of concentration under control. Apart from cycles of concentration, other
parameters; such as pH, hardness (as CaCO3), chlorides, silica and sulphates (as
SO4) need be controlled in the cooling water. Need for control for these
parameters is based on corrosion control and scale formation prevention.
Cooling water chemistry is quite complex and it is suggested that standard
reference books7,8 on the subject may be referred. Chemicals such as
phosphonates are being added to cooling water for control of fouling. Biocides
are added to control bacterial growth in tubes Dichromates and/or phosphates are
added to control corrosion, However, addition of these chemicals cause problems
while handling blow down. For example, destruction of Cr+3 or Cr+6 must be
carried out in the blow down water before disposing it as effluents. Biocides
should preferably be biodegradable.
It is a common practice to select cooling water on tube side while designing a
shell and tube heat exchanger. Desired velocity of cooling water be kept 1 m/s
(min.) so as to get good film coefficient and also reduce possibility of fouling.
Number of passes on tube side is based on this parameter. Cleaning of tubes is
carried out periodically so that fouling resistance of cooling water is under
control. Cleaning of tubes with a wire brush, attached to a rod, is used when
mechanical cleaning is resorted to. Chemical cleaning is another method in
which sulphamic acid is used with an inhibitor. Proprietary chemicals are also
available from certain manufacturers for chemical cleaning which may not
contain acidic chemicals. In another method, hydro-jetting technique is used for
cleaning the tubes. High pressure (in excess of 150 bar) water is used in lances,
attached to the tube, which loosen almost all types of scales and remove from the
tubes without their damage. This technique is found quite effective and ensures
clean tubes for the cooling service for a longer period.
A good practice is to provide a temperature indicator on cooling water line on
exit of each cooler so that flow of cooling water flow can be effectively
controlled. Also cooling water circuit should be designed such as no heat
exchanger starves of it. This is particularly important for heat exchangers located
at elevations and at farthest locations where enough cooling water flow should
be ensured.
In a special case a tube bundle is installed in a dedicated cooling tower in which
condensation takes place. Such an equipment is called an evaporative condenser.
Since evaporation of water takes place simultaneously with heat transfer,
constant cooling water temperature is assumed for heat exchanger design
calculations (refer to Exercise 6.10). Such an equipment can be used for
condensation of a refrigerant, steam (such as turbine exhaust), etc.
In some cases, cooling water is used in two or more heat exchangers in series as
each of them is designed for low temperature rise (say 2 to 3°C). While
designing such a circuit, care should be exercised that pressure drop in cooler is
low so that adequate flows are available in the coolers. Cooling in compressor
jackets, cooling in a battery of glass condensers above a glass lined reactor, etc.,
are some examples which fall in this category.
Both carbon steel and stainless steel (SS) tubes are in common use in the
coolers. When ERW SS tubes are used, chlorides in circulating water must be
controlled, preferably below 300 mg/L (as Cl), as stress corrosion is a possibility
in such coolers with chloride ions.
Special care is required while designing a closed circuit cooling water when
cooling water is to be used in direct contact equipment such as a barometric
condenser, venturi scrubber, etc. Cooling water could be contaminated with
carry-over material in the barometric condenser which may call for special
treatment of cooling water. Alternately excessive blow down may have to be
resorted to control the concentration of the contaminant(s).
In general, water cooling is cheaper than air cooling.
11.3.3 Refrigeration and Associated Cooling Systems
When cooling of a process stream is required below ambient temperature,
cooling medium must be at a lower temperature. Secondary coolant; chilled
water or brine is used for this purpose. Term refrigeration is used for achieving
temperature up to –50°C below which use of liquefied gases (such as liquid
nitrogen) is required and such cooling is known as cryogenics. It is possible to
achieve temperatures up to –80°C with cascade refrigeration but such systems
are rather uncommon. Secondary coolant is cooled by a refrigerant and used for
process heat transfer applications.
Chilled water systems In industry, chilled water is produced up to 7°C. Rise in
chilled water is usually restricted to say 7°C in the chiller. Both absorption and
mechanical refrigeration systems are in use in industry. A typical absorption
refrigeration system, commonly used in industry, is shown in Fig. 11.8. It uses
water (R-718) as a refrigerant and aqueous lithium bromide solution as an
absorbent. Low pressure steam (say at 2 bar g) is used for evaporating water
from strong aqueous LiBr solution.
 Fig. 11.8 Absorption Refrigeration Unit with LiBr Solution

Extraction of heat in the evaporator is carried out at 5°C (min.) and 6 Torr a.
Circulating chilled water is cooled by evaporating water. Steam condensate or
water from a reverse osmosis plant (having low DS) is used as circulating water
to avoid fouling problem. Secondary condensers for recovery of solvents such as
methanol, ethyl acetate, i-propyl alcohol (IPA), etc., use chilled water as a
coolant.
Mechanical refrigeration use ammonia (R-717) or a fluoro-chloro carbon
refrigerant. Considering ozone depletion potential of various refrigerants,
currently tetrafluoroethane (R-134a) is preferably used for chilled water
production although use of chlorodifluoromethane (R-22) is also made in certain
plants. In a conventional refrigeration cycle, refrigerant gas is compressed and
condensed in condenser with help of cooling water. Liquid refrigerant is
expanded to a lower pressure in an evaporator where latent heat of evaporation
(λ) of the refrigerant takes heat from circulating chilled water. Ammonia is a low
cost refrigerant but has toxic effects in comparison to fluoro-chloro carbon
refrigerants. Leak of ammonia is a concern and special care is required to control
the leaks.
Reciprocating, screw and scroll compressors are in use in industry. For chilled
water applications, single stage reciprocating compressor is used. Screw
compressor with jacket cooling provides compressed refrigerant at relatively
lower temperature at the discharge than reciprocating compressor. In general,
screw compressors are more costly than reciprocating machines but consume
less power and hence screw compressors are preferred in current designs. Scroll
compressors are used for relatively small chilling plants.
A process engineer is expected to specify the refrigeration duty (in kW), desired
chilled water circulation rate and its temperature. If low cost steam (or waste
steam is available), vapour absorption (VA) refrigeration may be selected.
Cooling water temperature and design steam pressure are important parameters
while designing VA system. System of mechanical refrigeration is selected on
the basis of coefficient of performance and cost of the refrigerant. Currently,
many evaporative condensers are in use in refrigerants systems and are claimed
to consume considerably low power consumption than a shell- and tube-type
condenser with cooling water as a coolant. However, selection of evaporative
condenser should carefully be made such that cleaning of external surface of the
tube bundle can be made with ease.
Brine systems Brines are secondary heat transfer fluids that are used for process
heat transfer applications below 5°C. Most commonly used brine in industry was
aqueous solution of calcium chloride. Corrosion was a serious problem faced
with CaCl2 brine in addition to other handling problems such as its slimy nature.
Over a period, use of aqueous ethylene glycol (MEG) solution, aqueous
methanol solution and other aqueous solutions have found applications as brines.
Properties needed for calculation of a film coefficient; such as specific gravity,
specific heat, freezing point, viscosity, etc., of brines are available in Refs. 2 and
9. Study of the properties help process engineer in selecting most suited brine for
a given application. For cooling of brines both; ammonia absorption
refrigeration and mechanical refrigeration are in use.
Ammonia absorption refrigeration system is used for achieving brine
temperature up to –30°C. Both reciprocating and screw compressors are used for
mechanical refrigeration; the latter is now preferred due to reliability and less
power consumption. Usually reciprocating compressor has two or more stages.
While aqueous MEG brines are used up to –35°C, aqueous methanol brines are
used up to –50°C. Ammonia (R-717), tetrafluoroethane (R-134a) and
chlorodifluoromethane (R-22) are used, depending on the desired brine
temperature, coefficient of performance and cost per unit of refrigeration.
Design of refrigeration system involves brine handling system. Process engineer
can design a single tank (3 to 6 m3 capacity) for small refrigeration system; say
for 15 kW capacity. Return brine from plant is returned to an insulated tank. A
recirculation pump takes suction from the tank and passes the chilled water or
brine through a chiller in which it exchanges heat with the refrigerant. Chilled
water or brine is sent to various heat exchangers in the plant. In larger
refrigeration plants, two tank (each of 3 to 7 m3 capacity) system is provided.
Also when storage of refrigerated brine is required, two tank system is preferred.
Usually a bigger tank is provided with a partition and thus two tanks are made.
In the event of internal circulation, overflow from cold well to hot well is
permitted. One part is known as hot well in which return brine is stored. A
primary pump takes suction from hot well and pass the brine through a chiller
and colder brine is stored in cold well. Another pump takes suction from the cold
well and circulates brine in the plant. Main advantage of the system is that the
secondary pump is not required to operate continuously. Primary pump pumps
brine through the chiller and thus provides sufficient storage of brine at desired
temperature. Figure 11.9(a) and (b) depict P&I diagrams for brine handling
systems.
Capacity of tanks are decided on the basis of supply of brine (i.e., flow rate)
required to number of equipments at a time. Thus these tanks act as buffer. When
large refrigeration duty is required in a small period, particularly at a low
temperature, required storage (or buffer) capacity is high, e.g., manufacture of
antibiotics. A side stream filter (3 to 5% of main circulation flow) is
recommended in the discharge of secondary pump and cold brine is recycled to
the cold well. Many a times, tank size becomes very large and heat gain from
atmosphere is excessive, adding to the basic load of refrigeration. A novel
innovative system is developed to have balls in which a known aqueous solution
is sealed. The solution freezes at somewhat higher temperature (say 3 to 5°C)
than actual temperature at which brine is required. In practice, the balls freeze by
transferring latent heat of fusion of the solution during chilling operation. When
actual cold brine circulation is started, heat transfer takes place in plant’s heat
exchangers. To provide sufficient refrigeration, fluid in the balls starts melting
and thus the brine temperature is maintained. Such a system requires relatively
smaller tanks and heat gain from atmosphere can be kept under control.
In a small chemical plant, a common brine plant may suffice refrigeration
requirements at process temperatures above 10°C and for process temperatures
below say 5°C. However, in larger plants, more than one refrigeration plants are
provided to cater refrigeration demands at different temperatures. When
distances are long for brine travel in the plant, it may be economical to provide
more than one refrigeration plants, located nearer to process requirements. For a
specific requirement of refrigeration temperature (say below –35°C), a dedicated
brine system is recommended.
Comparison of performance of various commonly used brines (secondary
coolants) should prove interesting. In Table 11.5, some parameters are compared
which may help in selecting a brine for the desired application.
Table 11.5 Brine Performance Comparison9
Brine (Secondary Concentration % by Recommended Freeze Pressure dropa Heat transfer coefficientb, Energy
coolant) mass circulation rate point °C kPa W/(m2∙°C) factorc
m3/(h ∙ kW)
Aq. calcium chloride 22 0.1800 –22.1 16.685 3214 1.447
Aq. sodium chloride 23 0.1652 –20.6 15.858 3169 1.295
Aq. ethylene glycol 38 0.1782 –21.6 16.410 2305 1.250
Aq. propylene glycol 39 0.1652 –20.6 20.064 1164 1.142
Aq. methanol 26 0.1685 –20.7 14.134 2686 1.078
Aqua ammonia 14 0.1602 –21.7 16.823 3072 1.000
Methylene chloride 100 0.4126 –96.7 12.824 3322 3.735

a
Pressure drop in one length of 4.9 m tube having 26.8 mm ID and 2.134 m/s brine velocity
b Heat transfer coefficient based on Sieder & Tate equation [Eq. (6.18)] in tubes having 4.9 m length, 26.8

mm ID and brine velocity of 2.134 m/s

c Energy factor represents relative energy requirement for same suction/discharge pump conditions, –6.7°C

average temperature of pumping, 6oC range in HE and freezing point 12oC below the lowest brine
temperature.
Fig. 11.9(a) Brine Handling System Incorporating Single Tank
 Fig. 11.9(b) Brine Handling System Incorporating Two Tanks

Direct expansion of refrigerants When refrigeration is required at low

temperature and heat transfer area (HTA) availability is restricted, direct
expansion of refrigerant is found useful. In fact, use of direct expansion of
refrigerant in various applications is in practice since long time. Expansion of
ammonia (R-717) in cooling coils (of a serpentine cooler) of a cold storage room
for preservation of vegetables, fruits, etc., is a common practice since decades.
Similarly in a fishery, ammonia expansion in coils at different temperatures is
carried out to preserve fish for long period. In some fisheries, different
compressors (single or multistage) are provided and compressed ammonia is
condensed in a common condenser. Liquid ammonia is stored in a common
storage tank. A liquid ammonia pump then supplies liquid ammonia for direct
expansion in different coils. Direct expansion of refrigerant is also commonly
carried out in dairy and icecream plants.
Despite this experience, use of direct expansion of refrigerant is not a common
practice in chemical plants. Use of chilled water/brine is more common in
chemical industry. However, in many applications, use of brine does not provide
sufficient refrigeration (in terms of kW at a definite temperature) due to
restricted HTA. Consider a CSTR in which a chemical (say a dye intermediate or
a bulk drug) reaction is to be carried out at low temperature but required
refrigeration duty is high (say in excess of 10 kW). Jacket or limpet coil attached
to CSTR may not provide sufficient HTA for heat exchange with a brine. For
such cases. direct expansion of refrigerant is carried out at a specified pressure
and liquid refrigerant is evaporated. Film coefficient of vaporising refrigerant is
quite high in comparison to convective coefficient of brine. Also since
evaporation takes place nearly at a constant temperature, temperature difference
across hot and cold streams is higher than with brine circulation system. Both the
factors result in high heat transfer rate in relatively less HTA. Such a system also
eliminates a need for brine handling system. However, the heat transfer
equipment is required to be designed for high pressure. For equipment in which
chilled water/brine is used for chilling, operating pressure is usually 3 to 3.5 bar
g while in equipment where direct expansion is carried out design pressure range
from 15 to 20 bar g although actual expansion may take place to low pressure in
normal run. This calls for thicker shell as well as jackets which means costly
equipment. Also additional safety measures such as provision of safety relief
valves/rupture discs on jackets are necessary. Despite this cost factor, direct
expansion is found economical, particularly where a dedicated refrigeration
system is provided for a single reactor.
Properties of refrigerants are extensively covered in Refs. 2 and 9. Process
engineer can design an efficient refrigeration system with the help of these
properties. Refrigeration cycles are well described in Refs. 3 and 9 and reader is
advised to refer to them.
11.3.4 Thermal Insulation for Cold Surfaces
Heat absorption from atmosphere of cold surfaces is a concern. This could be a
very expensive particularly for temperatures below 0°C. Insulation is provided
on all cold surfaces for which different materials are used. Amongst these, rigid
cellular polyurethane is the one which is commonly used for insulation of cold
surfaces. This material finds application for temperatures up to well below –
100°C. Its thermal conductivity is 0.05 W/(m · K) in the temperature range –20
to 10°C. Table 11.6 may be used as a guide for cold insulation
Table 11.6 Economic Insulation Thickness for Polyurathane Insulation*
Nominal pipe size Recommended thickness, mm Temperature range, °C

mm in Amb. to –4 –5 to –20 –21 to –30 –30 to –40 –40 to –50

15/20 ½/¾ 25 25 40 40 50
25 1 23 40 40 50 50
40 1½ 25 40 40 50 65
50 2 25 40 50 50 65
80 3 40 50 50 50 65
100 4 40 50 50 65 65
150 6 40 50 50 65 75
200 8 40 50 50 65 75
300 12 40 50 50 75 100
400 16 50 50 65 75 100
500 20 50 65 75 100 125
600 24 50 65 75 100 125

* Ambient temperature is assumed to be 30–35°C.

11.4 COMPRESSED AIR SYSTEMS

In chemical plants, compressed air is used mainly for two purposes; for
instrument air for control system and plant air for general use. For both the
services, air at 6 to 7 bar g is stored and distributed in the plant. For small
capacity, single stage reciprocating compressor is used. For larger capacity two-
stage reciprocating compressor or a screw compressor is used. While oil-free
reciprocating compressors provide oil-free air, efficient oil removal system is
necessary for a screw compressor. At the suction of compressor, an air filter is
provided to get rid of dust and other coarse foreign particles. In the discharge
also a micron (fine) filter is provided to filter air. Compressed air is cooled in an
aftercooler and stored in a tank. Piping network for compressed air is designed in
accordance with the principles, described in Chapter 5. When compressed air
requirement is cyclic or erratic, more than one compressor is provided with a
‘start’ signal based on pressure (on the tank) to start the standby compressor.
Such an arrangement also results in energy saving. For more than two
compressors in parallel a complex instrument control system is used for
optimum energy use.
11.4.1 Plant Air System
Compressed air is used in pneumatic tools, breathing apparatus, in acid egg
pumps, diaphragm pumps, during maintenance, etc. Literature provides data on
air consumption of pneumatic tools, breathing apparatus, etc. Based on these
data, requirement of plant air is estimated. Plant air is not dried. In fact, it is a
necessity to have some moisture in air used in breathing apparatus.
11.4.2 Instrument Air System
A drying system is needed for the instrument air. Pneumatic transmitters and
control valves need instrument air for operation. With the adoption of electronic
instruments, air requirement for transmitters is not required. However, final
control element is a control valve with a diaphragm and instrument air is
inevitable for them. For normal control valves, dry air in the range of 0.2 to 1 bar
g is needed. For on-off control valves, air at higher pressure is used. Selection of
operating pressure for on-off valves is important and it is recommended that dry
air pressure of 5 bar g (max.) may be specified when air receiver is maintained at
6 to 7 bar g. Instrument suppliers specify dry air requirement for the entire
system. However, as a thumb rule, dry air consumption rate of 3 Nm3/h per
control loop may be assumed.
Drying systems for instrument air Drying of compressed air is carried out in
different ways:
1. A system incorporating two desiccant dryers is a common system in the
plants. A flow diagram for such a system is shown in Fig. 11.10.
Compressed air enters one dryer from top in which desiccant is filled. Activated
alumina, silica gel and molecular sieves can be used as desiccants. Dry air
leaves from bottom and taken to a receiver after passing through a micron
filter. It is recommended that final air should have foreign particles less than
0.1 gm/Nm3 with particle sizes less than 3 μm. Recent trend is to use stainless
steel piping from receiver outlet to usage point. For use of dry instrument air
for control valves, 6 mm dia stainless steel (316 grade) tubes with flared
fittings are used.
After a service period of the dryer, it is isolated for regeneration. Normally, a
service cycle of 8 h is preferred for design. Simultaneously the standby dryer
comes in line. Isolated dryer is depressurised. A portion of dry air from
common outlet of the dryer is passed through a heater in which air is heated
either by electric heater or steam to say about 110 to 120°C. Hot air is passed
through the exhausted dryer (after depressurisation) from bottom and let out
to atmosphere from top. After a definite period, heater is cut off and cool dry
air passing is continued till the desiccant is cooled to say 50°C. Dryer is now
ready for the next cycle. Regeneration period is usually 4 h for all the steps.
Regeneration air requirement ranges from 8 to 12% of total air output of
service cycle period.
2. As seen in (a), air requirement for regeneration could be up to 12% which is
considered high. For reducing this requirement, a refrigerant condenser can
be provided in an alternate arrangement on compressed air line. Chilled water
or brine is used for chilling air and water is separated from air. Condensed
water is drained out through an automatic trap. Such a system does not
require twin dryer system but calls for refrigeration.
3. In another system, a dryer is provided with a special membrane. This
membrane is designed to permeate moisture with some air to low pressure
(say atmospheric) side. This system can provide air with relatively high
moisture content than previous two designs and also lose up to 10% air in the
permeate. This design is not common in India.
 Fig. 11.10 Desiccant-Type Air Drying System

Dew point of instrument air Dew point (DP) of instrument air is an important
parameter in designing a drying system. It is a common practice in industry to
specify –40°C at atmospheric pressure for the design. Air with this dew point
means moisture content of approx. 90 mg/Nm3. However, drying up to this
temperature calls for a desiccant dryer of much larger size. It is not necessary to
dry air up to –40°C DP for all locations. For production of instrument air with –
40°C DP at atmospheric pressure, compressed air 7 bar a needs to be dried up to
–21.4°C under pressure. For dryer with refrigeration, brine at –30oC would be
necessary for this purpose. Since high cost is involved in achieving this DP, it is
important that right DP may be specified at the design stage. As a guideline, DP
of instrument at atmospheric pressure may be specified 10°C less than minimum
ambient temperature experienced at the desired location. Figure 11.11 gives DP
of air (gas) at atmospheric pressure corresponding to DP at 7 bar a. A point on
the line represents same moisture content (on molar basis) at both pressures.
 Fig. 11.11 Dew Points of Gas

For example, at a particular location in India, minimum ambient temperature

experienced in a year is 5°C. For this location, DP at atmospheric pressure may
be specified –5°C which means moisture content of approx. 2850 mg/Nm3 air.
This corresponds to +23°C DP at 7 bar a for which chilled water at 7oC should
be adequate in the refrigerated dryer. This represents substantial reduction of
drying load on the dryer. Process engineer should therefore specify correct DP
for design of the dryer. Online instruments are available for measurement of DP
of air (on for any gas) under operating pressure. In laboratory, DP is evaluated at
atmospheric pressure. Figure 11.10 can be useful in correlating both values.
Figure 11.11 is universal and can be used for any gas but specific humidity (in
mg/Nm3) of the gas depends on its molar mass. Following equation represents
relationship satisfactorily between DP at 7 bar a (T′) and DP at atmospheric
pressure (T) in the range of 5 to –60oC of DP at atmospheric pressure:
T′ = 2.7355 × 10–3T2 – 9.983 × 10–2T + 125.77   (11.3)
where, T′ = DP at 7 bar a, K
and T = DP at atmospheric pressure. K
Temperature of compressed air entering dryer Another important parameter is
inlet temperature of compressed air entering the dryer. As noted earlier,
compressed air is cooled in an aftercooler and stored in a receiver. If cooling
water at 32°C is used for cooling in the aftercooler, it is possible to cool air up to
say 40°C. Compressed air will be fully saturated at the operating pressure (say 7
bar a). This value may be specified for design of the air dryer. However, if lower
temperature than 40°C is specified; say 35°C, it will result in substantially higher
moisture content (at 40°C) in ingoing air in actual operation. In such a case,
dryer will be under designed. Study of vapour pressure data of water indicates
that vapour pressure of water increases 6+% for each degree rise in the
temperature range 20°C up to 45°C. This means moisture removal load also
increases in the same order. Hence, if the dryer is designed for inlet temperature
of 35°C and fed with saturated compressed air at 40°C in actual situation,
moisture load on the dryer will increase by say 31%. This will call for higher
desiccant requirement in the dryer. Process engineer should therefore specify
correct inlet air temperature to the dryer. Also this discussion leads to
importance of cleaning compressed air aftercooler periodically. If an air cooler is
used for cooling compressed air, temperature of air after the cooler could be still
high (say up to 60°C).
Due to any reason, poor drying of air is achieved (i.e., high DP than the
ambient), condensation of water is feared on the diaphragm of the control valve
and it may malfunction.

11.5 INERT GAS SYSTEMS

Oxygen is required by all living animals for survival. Also oxygen is used for
combustion of a fuel and in chemical industry for oxidation reactions. However,
presence of oxygen is objectionable in many instances. For this reason, certain
gases are used as inert gases (i.e., non-reactive gases). Nitrogen is used in
majority of cases. However, use of carbon dioxide and argon are also being
practised but they are relatively costly to nitrogen. Following are some of the
reasons for use of inert gas:
1. Blanketing is required in many instances. For removal of a flammable solvent
from a product in a dryer, presence of oxygen is objectionable as it can result
in flammable (hazardous) mixture in the exhaust. As an example, in a bulk
drug plant, a product with acetone (highly flammable solvent) is dried in a
fluid bed dryer. If air is used for this drying operation, there is a probability of
explosion due to a spark or static electricity at the outlet. For such a
hazardous operation, inert gas (e.g., nitrogen) is used as a carrier gas in
drying.
Presence of oxygen over certain oxygen sensitive products can result in its
degradation over a period. Consider transportation of mixed tocopherols, β-
carotene concentrate, etc., over long distances in drums or in an isotank. If air
is present in contact with such a product for a longer period, it will
deteriorate. However, if oxygen concentration is kept below say 1% in the
environment above such a product, its concentration does not get reduced for
long periods. For this reason, blanketing of the product with an inert gas in a
container is carried out at a slightly higher pressure than atmospheric.
Many food products like potato chips, garlic powder, fruit juices, etc., use
nitrogen during packing for blanketing so that its taste is maintained for
longer duration.
2. In hazardous chemical plants, inert gas is used for different reasons. Consider
a liquefied petroleum gas bullet which requires inspection of empty vessel.
For man entry in such vessels, presence of oxygen is required. After
depressurisation, for safety reasons direct purging of vessel with air is not
carried out. Instead, the vessel is first purged with an inert gas and
subsequently purged with air. Only after ensuring that all flammable gases are
purged and desired oxygen concentration is achieved, man entry is permitted.
Pressure purging with inert gas is commonly employed in industry. It is preferred
over vacuum purging or sweep purging. Material balance calcutations1 can be
easily carried out to determine number of cycles of purging and quantum of
inert gas required.
In another case, consider replacement of catalyst in a methanator in ammonia
plant. Presence of lethal gas; carbon monoxide, is anticipated in the vessel.
Direct introduction of air is likely to result in explosion in the vessel. Here
 again, vessel is first purged with an inert gas and then with air.
During a shutdown in large chemical plant like petroleum refinery, ammonia
plant, methanol plant, etc., catalyst vessels are isolated and kept under inert
gas blanket under pressure.
3. Dry inert gas is used for deriming (defrosting) of a cold box. In Chapter 6, use
of plate-fin heat exchangers was mentioned for various applications,
including in cryogenic gas separation. In this heat exchanger, gap between the
plates is kept very less; say 2 mm. After a prolonged use of this equipment, a
necessity is felt to flush out thin gaps of the heat exchanger due to increase in
pressure drop. Flushing operation is carried out with an inert gas.
4. In gas chromatography, inert gas is used as a carrier gas.
5. Electronic industry and nuclear industry use inert gas for various purposes.
Apart from blanketing properties, nitrogen and carbon dioxide (as dry ice) are
used for refrigeration applications below –50°C. Dry ice sublimes at –78.46°C
by absorbing latent heat of sublimation (571.0 kJ/kg). Dry ice is useful in
laboratory for achieving low temperatures as it can be conveniently handled in
solid form. In plants, liquid nitrogen at –195.8°C is used for chilling applications
by absorbing latent heat of vaporisation (198.7 kJ/kg).
Transportation of liquid nitrogen at temperature less than –195°C in the plant is
carried out in special steel pipes (meant for cryogenic applications). Inner pipe is
covered with another pipe and annulus is evacuated and sealed. Outer pipe is
insulated by at least 150 mm thick polyurethane foam blocks. Despite this care,
there can be a substantial vaporisation of liquid nitrogen (up to 25 to 30%) from
storage to application point due to its low latent heat of vaporisation (λ). This
refrigeration loss also must be taken into account while calculating liquid
nitrogen requirements for any cryogenic application in the plant. Since liquid
nitrogen is transported at very low temperature, expansion/compression study of
the piping system must be carried out and suitable supports, expansion joints,
etc., must be provided.
As mentioned earlier, nitrogen is used as an inert gas for most operations. Use of
argon is restricted in limited applications because of its cost. Properties of
common inert gases are listed in Table 11.7.
Table 11.7 Properties of Common Inert Gases10
Property Nitrogen Argon Carbon dioxide

Symbol N2 Ar CO2

Molar mass 28.0134 39.948 44.0095

Normal boiling point/
–195.80 –185.85 –78.46
 Sublimation temperature, °C

Specific volume at 1 atm and 35°C, m3/kg 0.9025 0.6326 0.5701

Heat capacity at constant pressure (Cp) at 1 atm and 35°C, kJ/(kg · °C) 1.0414 0.5215 0.8595

Critical temperature (tc), °C –147.05 –122.45 30.96

Critical pressure, (pc), bar 33.94 48.63 73.75

Quality of nitrogen varies from an application to application. For blanketing,

drying or purging purpose, O2 content up to 1% could be permitted. For
production of nitrogen of this quality, Pressure Swing Adsorption (PSA) plant is
adequate. Air is first compressed, dried and passed through carbon molecular
sieve beds for separation of N2.
For catalyst blanketing, O2 content in nitrogen is desired below 100 ppm (v/v).
In electronics industry, nitrogen with O2 content less than 1 ppm (v/v) is
demanded. For achieving high purity N2, PSA plant with trace oxygen removal
(by oxidising with H2) facility is required. Alternately, cryogenic air separation
technique is used which can produce either liquid or gaseous nitrogen of any
desired quality. Nitrogen can also be generated onsite by cracking ammonia.
Nitrogen so generated may have H2 as an impurity which is acceptable for
catalyst preservation. However, same nitrogen when required to be used for
blanketing, H2 must be removed by oxidising H2 with O2 which is added (in the
form of air) in calculated amount.
Purge gas from ammonia plant is processed in a cryogenic plant to recover
hydrogen from it. Reject stream of this plant is normally 95 to 97% pure N2 with
some impurities of hydrogen, methane and argon. This reject stream can be used
for blanketing of catalyst or purging of cold box. However, this stream cannot be
used for purging vessels or for applications where impurities are not acceptable.
Thus selection of a nitrogen generation facility must be carefully made after
ascertaining the quality requirements for the desired application. Liquid nitrogen
is supplied in tankers at about 6 bar g and –195.8°C. For small requirements,
liquid nitrogen is supplied in a stainless steel flask. Gaseous nitrogen is supplied
at ambient temperature in cylinders (approx. 6 Nm3) at about 150 bar g.
For deriming applications, nitrogen gas with –60°C DP at line pressure (approx.
1 mg/Nm3 moisture by volume at 7 bar a) is required. For this purpose, nitrogen
is dried on molecular sieves. For chromatographic applications also, impurity
level must be very low and it is specified by the instrument supplier.
Carbon dioxide is available from ammonia plants which can be liquefied by
compression and refrigeration at about 20 bar g (corresponding saturation
temperature of –15°C). This corresponds to –44.2°C water dew point at
atmospheric pressure and saturation humidity of approx. 60 mg/Nm3 gas.
Therefore when compressed (and liquefied) CO2 is used as inert gas at low
pressure, there is a possibility of ice (of water) formation unless gas DP is below
–44.2°C at atmospheric pressure. This point is important while designing CO2
liquefaction facility.
Carbon dioxide can also be produced from flue gas. At first flue gas is treated in
an absorber with some solution (e.g., aqueous monoethanol amine solution) to
remove CO2 from the gas and then CO2 is stripped off from the solution in a
stripper by supplying heat. Purified gas is compressed and dried before
liquefaction. Fermentation of molasses also produces CO2 which can be washed,
compressed and dried before liquefaction.
Liquid carbon dioxide is supplied in a tanker at about 20 bar g while gaseous
CO2 is supplied at ambient temperature in cylinders (approx. 6 Nm3) at about
150 bar g. Dry ice blocks, each weighing 15 to 20 kg, are supplied. During
transportation of dry ice, there could be substantial loss of CO2 by sublimation
unless transported in well-insulated container.

11.6 OTHER UTILITIES

There are a few other utilities that are of interest to chemical industry. Raw water
usage, steam generation and utilisation, fire protection system, effluent
treatment, electricity, heating, ventilation and air conditioning, etc., also need
due attention at the design stage. Due to their prime dependence on the location,
design aspects significantly vary from place and place. For example, raw water
could be a surface water source or an underground water source. Qualities of
theses sources vary significantly at different locations and hence their treatments
are also specific for general purpose use (including sanitation), boiler feed,
cooling tower make-up, etc. Similarly, extent of effluent treatment depends on
stipulations by local effluent control authority, and so on. For design of systems
related to these utilities, standard books on the related subjects may be referred.

EXERCISES
11.1 Name two chemicals used for removal of final traces of oxygen from boiler
feed water or (compressed) water used for heating application.
11.2 Consider process heating requirement at 110°C. Suggest steam pressure for
indirect heating application.
11.3 An indirect heater uses hot water at 130°C for heating a heat sensitive
organic solution. Suggest pressure of water to be circulated in the system.
11.4 A distillation column separates an aqueous solution of acetic acid (refer to
Example 8.5). Assume that column operates essentially at atmospheric
pressure. Suggest steam pressure in the reboiler.
11.5 A vacuum system consists of two steam ejectors in series. Both use steam at
4.5 bar a with 10°C superheat as motive steam. What will be the steam
temperature of motive steam?
11.6 A high pressure boiler, generating steam for power generation, uses water
having conductivity equal to 5 μS/cm and silica content of 0.02 mg/L.
Suggest water treatment steps for an underground water source having 1200
mg/L dissolved solids.
11.7 A horizontal pipe of 200 mm (8 in) carries steam at 200°C (ts). It is
insulated with rockwool insulation (having 150 kg/m3 density) of 50 mm
thick mattress which is expected to have a surface temperature of 50°C (tc)
above the aluminium sheet with ambient temperature of 35°C.
Heat transfer coefficient (h) for horizontal bare pipe can be calculated using the
following equation:
h = [0.85 + 0.005(ts – ta)] × 10 W/(m2 · °C)
where, ts = Bare surface temperature, °C
ta = Ambient temperature, °C
Heat transfer coefficient (h′) for insulated pipe can be calculated by using the
following equation from insulated surface:
h′ = [0.31 + 0.005(tm)] × 10 W/(m2 · °C)
where, tm = (ts + tc)/2, °C
tc = Cold surface (over insulation) temperature, °C
Calculate energy saving for 100 m steam pipe by providing insulation. If cost of
energy of saturated steam at 10 bar a (from a solid fuel fired boiler) is Rs.
1.25/kg, calculate annual saving.
Assume that energy requirement is 2475 kJ/kg for 10 bar a steam generation and
overall boiler efficiency (η) is 75%.
11.8 In a particular plant, CO2 is liquefied at 35 bar a and (ts =) 0°C. What should
be the water DP (max.) of compressed CO2 at atmospheric pressure before
liquefaction?
Vapour (sublimation) pressure of ice can be calculated from the following
equation in the temperature range 273.16 K to 193.15 K.
= –13.928[1 – (T/273.16)–1.5] + 34.7078[1 – (T/273.16)–1.25]

where, p = vapour (sublimation) pressure of ice, Pa

T = temperature at which vapour pressure is desired, K

REFERENCES
1. Bhatt, B. I., and S. B. Thakore, Stoichiometry, 5th Edition, McGraw-Hill
Education (India), New Delhi, 2010.
2. Bhatt, B. I., Design Data Book, Properties of Steam, Selected Refrigerants, n-
Hexane and Brines, CBS Publishers and Distributors, New Delhi, 2007.
3. Smith, J. M., H. C. Van Ness, M. M. Abbott, and B. I. Bhatt, Introduction to
Chemical Engineering Thermodynamics, (SI Units), 7th Edition, McGraw-
Hill Education (India), New Delhi, 2010.
4. Madhawa Hettiarachchi H. D., M. Golubovic, W. M. Worek and Y. Ikegami,
J. of Energy Resourcces Technology (ASME publication), 129, Sept. 2000, p.
243.
5. Broughton, J, (Ed.), Process Utility Systems – Introduction to Design,
Operation and Maintenance, Institution of Chemical Engineers (UK), 1994.
6. Towler, G. P., and R. K. Sinnott, Chemical Engineering Design: Principles,
Practice and Economics of Plant and Process Design, Butterworth-
Heinemann, UK, 2008.
7. Betz Handbook of Industrial Water Conditioning. 9th Ed., published by Betz,
Laboratories, USA, 1991.
8. Flynn, D. The Nalco Water Handbook, USA.
9. AHEARE Handbook (SI Edition), published by the American Society of
Heating, Refrigeration and Air-Conditioning Engineers, Inc. USA, 2010.
10. Reference Fluid Thermodynamic and Transport Properties (Version 8.0),
National Institute of Standards and Technology (NIST), USA, 2007.
 1
Appendix
Table A1.1 Length Units
Metres Centimetres Inches Feet Yards
(m) (cm) (in) (ft) (yd)

m 1 100 39.370 08 3.280 84 1.093 613

cm 0.01 1 0.393 701 3.280 84 × 10–2 1.093 613 × 10–2
in 0.0254 2.54 1 8.333 333 × 10–2 2.777 78 × 10–2
ft 0.3048 30.48 12 1 0.333 333
yd 0.9144 91.44 36 3 1

Table A1.2 Area Units

Square Square Square Square Square
metres centimetres inches feet yards
(m2) (cm2) (in2) (ft2) (yd2)

m2 1 104 1550.003 10.763 91 1.195 99

cm2 10–4 1 0.155 1.076 391 × 10–3 1.195 99 × 10–4

in2 6.4516 × 10–4 6.4516 1 6.944 444 × 10–3 7.716 049 × 10–4
ft2 0.092 903 929.0304 144 1 0.111 111

yd2 0.836 127 8361.273 1296 9 1

Table A1.3 Density and Concentration Units

Kilograms per Grams per cubic Pound per cubic Pound per UK Pound per US
cubic meter centimetre foot gallon gallon
(kg/m3) (g/cm3) (lb/ft3) (lb/UK gal) (lb/US gal)

kg/m3 1 10–3 6.242 795 × 10–2 1.002 2 × 10–2 8.345 402 × 10–3

g/cm3 1000 1 62.427 95 10.0224 8.345 402

Ib/ft3 16.018 462 1.601 846 × 10–2 1 0.160 54 0.133 681

Ib/UKgal 99.776 37 9.9776 × 10–2 6.228 842 1 0.832 675

Ib/USgal 119.8264 0.119 826 7.480 519 1.200 95 1

*1g/cm3 = 1 t/m3 = 1.000 028 g/mL = 1.000 028 kg/L

Table A1.4 Volume and Capacity Units
Cubic Metres Cubic Centimetres Cubic Inches Cubic Feet UK (Imperial) gallons US gallons
(m3) (cm3) (in3) (ft3) (UK gal) (US gal)
(also called psi)

m3 1 106 6.102 374 × 104 35.314 67 219.9625 264.172

cm3 10–6 1 6.102 374 × 10–2 3.531 467 × 10–5 2.199 69 × 10–4 2.641 722 × 10–4

in3 1.638 706 × 10–5 16.387 06 1 5.787 037 × 10–4 3.604 651 × 10–3 4.329 007 × 10–3
ft3 2.831 685 × 10–2 28.316 85 × 103 1728 1 6.2289 7.480 52
UK gal 4.546 095 × 10–3 4.546 095 × 103 2.774 194 × 102 0.160 54 1 1.200 95
 US gal 3.785 412 × 10–3 37.854 12 × 102 231 0.133 681 0.832 674 1

Table A1. 5 Mass Units

Kilograms Grams Tonnes Pounds (av.) Tons Tons (short)
(kg) (g) (t) (lb) (T) (Ts)

kg 1 1000 10–3 2.204 623 9.8421 × 10–4 1.102 312 × 10–3

g 10–3 1 10–6 2.204 623 × 10–3 9.8421 × 10–7 1.102 312 × 10–6
t 1000 106 1 2204.623 0.984 21 1.102 312

lb 453.592 37 × 10–3 453.592 37 4.535 924 × 10–4 1 4.4643 × 10–4 5 × 10–4

T 1016.047 1.016 047 × 106 1.016 271 2240 1 1.12

Ts 907.185 9.071 847 × 105 0.907 185 2000 0.892 86 1

Table A1.6 Force Units

Newtons Kilograms force Dynes Pounds force Poundals
(N) (kgf) (dyn) (lbf) (pdl)

N 1 0.101 972 105 0.224 809 7.233

kgf 9.806 65 1 9.806 65 × 105 2.204 62 70.932

dyn 10–5 1.019 72 × 10–6 1 2.2046 × 10–6 7.233 × 10–5

Ibf 4.448 22 0.453 592 4.448 22 × 105 1 32.174

pdl 0.138 255 1.4098 × 10–2 1.382 25 × 104 3.1081 × 10–2 1

Table A 1.7.1 Pressure or Stress Units

Newtons per Bars Standard Kilograms-force per Dynes per square
square metre (bar) atmospheres square centimetre centimetre
N/m2 (Pa) (atm) (kgf/cm2 or at) (dyn/cm2)

N/m2 (Pa)1 10–5 9.869 233 × 10–61.019 716 × 10–5 10

bar 105 1 0.986 923 1.019 716 106
atm 101.325 × 103 1.013 25 1 1.033 227 101.325 × 104
2 2 0.980 65
kgf/cm 908.665 × 10 0.967 841 1 9.806 65 × 105
dyn/cm2 0.1 10–6 9.869 233 × 10–71.019 716 × 10–6 1

Torr 133.3224 1.333 224 × 10–31.315 79 × 10–3 1.359 51 × 10–3 13.332 24 × 102
in Hg 3386.388 3.386 395 × 10–23.342 105 × 10–23.453 155 × 10–2 33.863 886 × 103
m H2O 9806.65 9.8067 × 10–2 9.6784 × 10–2 0.1 9.806 65 × 104

ft H2O 2989.063 2.989 068 × 10–22.949 89 × 10–2 3.048 × 10–2 2.989 063 × 104

Ibf/in2 6894.757 6.894 757 × 10–26.804 596 × 10–27.030 696 × 10–2 6.894 753 × 104

Table A 1.7.2 Pressure or Stress Units (Contd.)

Torr or Barometric millimetres of mercury (Torr or mm Barometric Pounds-force
H*g) inches Head of water** per
of mercury m H2O ft H2O square inch
(in H*g) (lbf/in2)
(also called psi)

N/m2 7.500 616 × 10–3 2.952 999 × 10–4 1.019 72 × 10– 3.345 53 × 10– 1.450 377 × 10–4
(Pa) 4 4

bar 750.061 6 29.529 99 10.1972 33.4553 14.503 77

atm 760 29.921 26 10.332 27 33.8985 14.695 95
 kgf/cm2 735.5592 28.959 03 10 32.8084 14.223 34

dyn/cm2 7.500 617 × 10–4 2.953 × 10–5 1.019 72 × 10– 3.345 53 × 10– 1.450 377 × 10–5
5 5

Torr 1 3.937 008 × 10–2 1.359 51 × 10– 4.460 35 × 10– 1.933 678 × 10–2
2 2

in Hg 25.4 1 0.345 316 1.132 92 0.491 154

m H2O 73.5556 2.8959 1 3.280 84 1.422 334

ft H2O 22.4198 0.882 67 0.3048 1 0.433 526

Ibf/in2 51.714 92 2.036 021 0.703 072 2.306 67 1

*at 0°C **at 4°C

 Table A 1.8.1 Energy and Heat Units
Joules Kilowatt hours Kilocalories Kilogram–force
(J) (kWh) (kcalIT) metres (kgf · m)

T 1 2.777 778 × 10–7 2.388 459 × 10–4 0.101 972

kWh 3.6 × 106 1 859.845 2 3.670 98 × 105
kcalIT 4186.8 1.163 × 10–3 1 426.935

kgf · m 9.806 65 2.724 07 × 10–6 2.3423 × 10–3 1

L · atm 101.325 2.814 583 × 10–5 2.420 107 × 10–2 10.332 31

BtuIT 1055.056 2.930 711 × 10–4 0.251 996 107.586

lbf · ft 1.355 818 3.766 161 × 10–7 3.238 315 × 10–4 0.138 255

Table A 1.8.2 Energy and Heat Units (Contd.)

Litre atmospheres British thermal units Pound-force feet
(L · atm) (BtuIT) (lbf · ft)

J 9.869 233 × 10–3 9.478 17 × 10–4 0.737 562

kWh 3.552 924 × 104 3.412 142 × 103 2.655 224 × 106
kcalIT 41.320 5 3.968 32 3088.96

kgf · m 9.678 38 × 10–2 9.294 88 × 10–3 7.233

L · atm 1 9.603 759 × 10–2 74.733 51

BtuIT 10.412 59 1 778.1693

lbf · ft 1.338 088 × 10–2 1.285 067 × 10–3 1

Table A 1.9 Specific Energy Units

Joules per kilogram Kilocalories per British thermal unit
(J/kg) kilogram (kcalIT/kg) per pound (BtuIT/lb)

J/kg 1 2.388 459 × 10–4 4.299 226 × 10–4

kclIT/kg 4186.8 1 1.8

BtuIT/lb 2326 0.555 556 1

Table A 1.10 Power Units

Kilowatts Kilogram Metric horse- Pound- Horsepowers
(kW) force metre powers force feet (hp)
per second (mhp) per second
(kgf · m/s) (lbf · ft/s)
kW 1 101.971 62 1.359 62 737.561 2 1.341 022
kgf · m/s 9.806 65 × 10–3 1 1.3333 × 10–2 7.233 1.3151 × 10–2
mhp 0.7355 75 1 542.4761 0.986 32
lbf · ft/s 1.355 82 × 10–3 0.138 25 1.8434 × 10–3 1 1.8182 × 10–3
hp 0.7457 76.040 2 1.013 87 550 1

Table A 1.11 Temperature Units

Degrees Kelvin Degrees Celcius Degrees Fahrenheit Degrees Rankine
(K) (°C) (°F) (°R)
K T t + 273.15 5/9 (t ′ + 459.67) 5/9 T ′
°C T – 273.15 t 5/9 (t ′ – 32) 5/9 (T ′ – 459.67)
 °F 9/5 T – 459.67 9/5 t + 32 t ′ T ′ – 459.67
°R 9/5 T 9/5 t + 491.67 t ′ + 459.67 T ′

Table A 1.12 Absolute Viscosity Units

kg/(m · s) or Poise (P) or lb/(ft · s)
N · s/m2 g/(cm · s) or dyn · s/cm2

kg/(m · s) 1 10 0.671 97
Poise 0.1 1 0.067 197
Ib/(ft · s) 1.488 164 14.881 64 1

1 Poise = 100 certipoises (cP) = 100 mPa · s

Table A 1.13 Kinematic Viscosity Units
m2/s Stoke (St) ft2/s

m2/s 1 104 10.763 91

Stoke 10–4 1 1.076 391 × 10–3
ft2/s 0.092 903 929.03 1

1 St = 100 cSt
Table A 1.14 Thermal Conductivity Units
W/(m · K) kcalIT/(h · m · °C) BtuIT/(h · ft · °F)

W/(m · K) 1 0.859 854 0.577 789

kcalIT/(h · m · °C) 1.163 1 0.671 969
BtuIT/(h · ft · °F) 1.730 735 1.488 164 1

Table A 1.15 Heat Transfer Coefficient Units

W/(m2 · °C) kcalIT/(h · m2 · °C) BtuIT/(h · ft2 · °F)

W/(m2 · C) 1 0.859 845 0.176 11

kcalIT/(h · m2 · °C) 1.163 1 0.204 816

BtuIT/(h · ft2 · °F) 5.678 264 4.882 43 1

Reproduced with the permissions of

(i) Thermodynamics Research Centre (TRC), USA from its Publication: Reprint of the Introduction to the
TRC Thermodynamic Tables: Non–Hydrocarbons, December 31, 1991 and
(ii) International Union of Pure and Applied Chemistry (IUPAC), UK from its Publication: Quantities,
Units and Symbols in Physical Chemistry, Edited by I. Mills, T. Cvitas, K. Homann, N. Kallay and K.
Kuchitsu, 2nd Ed., Blackwell Science Ltd., UK, 1993.
(iii) Guide for the Use of the International System of Units (SI), NIST Special Publication 811, 2008,
National Institute of Standards and Technology, USA.

Conversion Tables
 Appendix
2
Viscometer Conversion Chart
 Index
Absorber
 criteria of selection
 process design of
Absorption
 exothermic
 of HCl gas in
 falling film type absorber
 in a packed tower
 in a spray tower
 of SO2 in spray chamber
 in venturi scrubber
 physical
 with chemical reaction
Active (bubbling) area of a sieve tray
Activity coefficient
 its determination by
 NRTL method
 UNIFAC method
 UNIQUAC method
 van Laar equation
Activity of a component
Adiabatic compressor
Adjacent keys
Adsorptive distillation
Advantages and disadvantages of
 air cooler over water cooler
 batch reactors over continuous reactors
 brazed aluminium plate fin heat exchanger
 bubble type reactor over stirred tank type reactor
 centrifugal extractors
 fixed tube heat exchangers
 gas induction type agitator
 jet reactors over agitated vessel type reactors
 mixed flow reactors
 mixer settler
 PHE over shell and tube heat exchanger
 plug flow reactor
 pulse column
 rotating disc contactors
 supercritical extraction over liquid-liquid extraction
 U-tube heat exchangers
 vacuum distillation over atmospheric distillation
Agitator
 anchor type
 crossbeam and blade type
 gas induction type
 helical ribbon type
 high shear mixer type
 paddle type
 Pfaudler
 pitch blade type
 power correlations for
 propeller type
 turbine type
AIChE method
 for efficiency in distillation
Air cooled heat exchanger
 induced draft type
 forced draft type
Air cooling
Air heater
Anchor agitator (stirrer)
Annular flow
ANSI standards for PFD and P&ID
°API
Atomic Mass
Avogadro’s number
Antoine equation
Approach to wet bulb temperature
Artificial vapour pressure
ASME Code
Auxiliary heat transfer medium
Azeotropes
 its compositions
 separation thereof
 heterogeneous
 homogeneous
 maximum boiling
 minimum boiling
Azeotropic distillation
Back-mix reactor
Back mixing
Bare tube area
Baffles
 dam type
 disc and doughnut type
 for condensers
 orifice type
 nest type
 segmental type
Baffle cut
Bale packing
Batch distillation
 binary
 multicomponent
 settlers
 simple
 with rectification
Batch operations
Batch reactors
 performance equation
Batch system
Batch type supercritical extractor
oBaume
Beam type packing support
Bell cap liquid redistributor
Bell’s method
Binary distillation
Binodal curve
Blower
 centrifugal
 Lobe type
Block diagram
Bogart equation
Boiling point
 normal (at atmospheric pressure)
Boiling of liquid mixture
Boyko-Kruzhilin equation
Brazed aluminium plate-fin exchanger
Brines
 systems
Bubble cap tray
Bubble column reactor
 factors affecting its performance
 process design of
Bubble point
 calculations of
Bypassing stream
Calibration of
 orifice meter
 rotameter
Cap type packing support
Capacity ratio curve
Carbon dioxide (CO2)
 critical properties of
 p-t  diagram
Catalysis
 a key to waste minimization
Catalytic distillation
 contacting devices for
 bale packing for
 structured packing for
Centrifugal extractors
Centrifugal pump
 characteristics of
 viscosity correction charts for
Ceramic packing
Channel cover of shell & tube type heat exchanger
Channeling in packed tower
Checking for
 liquid entrainment
 weeping
Characteristics of random packing
Chemical technology
Chilled water system
Coefficient of heat transfer
 boiling
 condensation
 with no phase change
Coefficient of performance
Cogeneration
Combined heat and power systems
Common arrangements for jackets and coils
Compressed air systems
 instrument air
 plant air
Conversion in a chemical reaction
Conversion tables
Cooler
 air type
 of solids
Cooling medium
 air
 brine
 cooling water
 chilled water
 oil (thermic fluid)
Compressor
 adiabatic
 polytropic
Compressible fluid
Comparison of pervaporation, adsorption and distillation
Condensation coefficient
 shell side
 tube side
Condensation with noncondensables
Condensation with subcooling
Continuous extractor
Continuous flow reactor
Continuous stirred tank reactor (CSTR)
Continuous phase
Conversion
Conveyor cooler
Coriolis flow meter
Cornell’s method
Cosolvents and surfactants
Cost cutting measures
Countercurrent type packing support
 mesh type
 welded ring type
Crawford-Wilke correlation
Criteria of selection of condensers
Criteria of selection of extractors
Critical heat flux
Critical pressure
Critical properties
 pseudo
Critical surface tension of packing Material
Critical temperature
Critical temperature drop
Critical volume
Cross beam and blade stirrer
Crystallization
 freeze
Dam baffle
Data sheets of equipments
Decaffeination of coffee
Decantation
Decanter
 design calculations of
 horizontal type
 vertical type
Degree of completion of reaction
Degree of freedom
Delaware method
Density
Derived dimensions (units)
Design of a hydrogenation reactor
Desired properties of solvent
 for absorption
 for extraction
Determination of
 activity coefficient
 LLE data
 VLE data
Determination of transfer units
 for absorption
 for extraction
Determination of theoretical stages
 for absorption
 for distillation
 for extraction
Determination of tower diameter
 for sieve tray tower
 for packed tower
 for absorber/distillation
 for extractor
Dew point
 calculations of
 of instrument air
 atmospheric
 pressure
Different type of agitators
Different type of reactors
Differential continuous contact extractor
Differential pressure type volumetric flow meters
Differential thermal expansion
Diluent
Dimensions
 fundamental
 derived
DIN standards for PFD and P&ID
Dispersed phase
Distillation
 adsorptive
 azeotropic
 batch
 binary
 catalytic
 extractive
 multicomponent
 pressure swing
 reactive
 short path
Distillation columns
 process design steps
 selection criteria for
Distributed components
Distribution coefficient
Dittus-Bolter equation
Divided flow shell
Divided wall column
Downcomer
 checking of residence time
 flooding in
 pressure drop in
 various types
Dummy tubes
Eduljee’s equation
Efficiency of pumps
Efficiency of sieve tray
Ejectors for distillation system
Electrical units
 quantity of
Electromagnetic flow meters
Energy
Energy balances
Energy conservation in distillation
 energy efficient distillation by design
 efficient operation of column
 by advanced process control
 by heat integration
 by use of high efficiency packing/tray
 with the use of heat pump
Enriching section
Enthalpy
Entrainer
Entrainment
Estimation of
 density of liquid mixture
 density of vapour/gas mixture
 molar heat capacity of liquid and gas mixtures
 surface tension of liquid
 surface tension of liquid mixture
 thermal conductivity of gas mixture
 thermal conductivity of binary liquid mixture
 thermal conductivity of multicomponent liquid mixture
 viscosity of liquid mixture
 viscosity of multicomponent gas mixture
 viscosity of binary gas mixture
 mean particle diameter of solids
Equation of state
 ideal gas law
Equation tearing procedure for distillation
Equilibrium constant
Equilibrium curve
Equilibrium line
Equipment data sheet
Equivalent diameter
 for square pitch
 for triangular pitch
 of fin
 of flow channel of plate HE
Equivalent length
 of fittings and valves
Equivalent number of velocity head
 for fittings and valves
ERW pipes
ETBE manufacture
 catalytic distillation process
 effect of various parameters on
 conventional process
Evaporative condenser
Exothermic absorption of gases
 of HCl and NH3  
Expansion joint
Extract phase
Extractive distillation
Extractor
 agitated column type
 batch type
 centrifugal type
 continuous
 design of
 mixer settler
 packed tower type
 rotating disc type
Fans
 axial flow
 centrifugal
Fanning or Darcy equation
Fanning friction factor
 chart
Falling film absorber
 advantages an disadvantages of
 of HCl and NH3 (its design)
Fensky-Underwood-Gillalnad equation
Film wise condensation
Fin efficiency
Finned tube heat exchanger
Fixed bed catalytic reactor
 design of
 examples of
 scale-up
Fixed tube sheet shell & tube heat
 exchanger
 advantages and disadvantages of
Flash steam from steam condensate
Floating head shell & tube heat exchanger
 advantages and disadvantages of
 pull through type
 split ring type
Flooding
 in packed tower type absorber
 in packed tower type extractor
 in tray tower
 downcomer flooding
 jet flooding
Flooding velocity
 in falling film absorber
 in packed tower
 in absorber/distillation column
 in extractor
 in sieve tray tower
Flow arrangements in PHE
Flow meters
 electromagnetic type
 Coriolis
 flow nozzles
 orifice type
 rotameters
 ultrasonic
 venturi type
 volumetric
 vortex
 weirs
Flow sheet of batch supercritical extraction
Fluid
 compressible
 incompressible
Fluid allocation
Fluid bed cooler
Fluid moving devices
Force
 Newton
Forced circulation reboiler
Forced draft air cooled heat exchanger
Fouling coefficient
 for shell and tube heat exchangers
 for plate heat exchangers
Francis formula
Freeze crystallization
Friction factor
 for air side
 for plate heat exchangers
 for pipes
 shell side
 tube side
Freezing point
Froude number
FUG method
Fugacity coefficient
Fundamental quantities
Gas absorption
Gas expansion and shrinkage
Gas film controlled absorption
Gas injection type packing support
 beam type
 cap type
Gas induced agitator
Gas phase transfer unit
Gas phase mass transfer coefficient
Gases
Gasket material for PHE
Generalized flooding and pressure drop correlation
Gilliland’s correlation
Gilmore’s equation
Grand composite curve
Green engineering
Gross calorific value
Heavy key component
Heat
Heat capacity
 molar (isobaric/isochoric)
 of ideal gas mixtures
 of liquid mixtures
Heat duty
Heat of absorption
Heat of combustion
Heat of formation
Heat of mixing
Heat exchanger design software
Heat exchangers
 brazed aluminium film type
 finned tube type
 plate type
 shell & tube type
 spiral type
Heat exchanger networking
Heat of reaction
 endothermic
 exothermic
 standard
Heat of mixing
Heat of solution
Heat integration in distillation column
Heat pump in distillation
Heat transfer area
 calculations thereof
Heat transfer coefficient
 overall, estimation of
 calculations of
 for air cooler/heaters
 for agitated vessel
 for bubble column reactor
 for condensation
 for condensation with noncondensables
 for falling film absorber
 for no phase change
 for nucleate boiling
 for plain jacket
 for plate heat exchanger
 for spiral baffle jacket
 for spiral heat exchanger
 overall
 typical values
 shell side
 tube side
Heat transfer factor
 shell side
 tube side
 for air side
Heater
 air type
Heat transfer in batch reactors
Heat transfer in solids
Heating medium
 auxiliary medium
 condensing fluids
 flue gas/hot air
 molten salt
 oil (thermic fluid)
 saturated steam
Hegstebeck-Geddes equation
HETP of packings
Height of equivalent overall transfer
 unit for absorber
 based on gas phase
 based on liquid phase
Height of equivalent overall transfer unit for extractor
 based on raffinate phase
 based on continuous phase
 determination by experiment
Helical ribbon impeller
Hess’s law
HETS values fo extractors
High efficiency packing/trays
High shear mixer
High speed tray for absorption
Heterogeneous azeotrope
Hold down plate
Homogeneous azeotrope
 maximum boiling
 minimum boiling
Horizontal condenser
Horizontal spray chamber
Hot air systems
Hot oil (thermic fluid) cycle
Hot water systems
Hydraulic flooding
Ideal batch reactor
 advantages and disadvantages of
 its performance equation
Ideal gas law
Ideal mixed flow reactor
Ideal plug flow reactor
Inclined downcomer
Incompressible fluid
Induced draft air cooled heat exchanger
Industrial applications of liquid-liquid extraction
Industrial examples
 of batch reactor
 of bubble column reactor
 of mixed flow reactor
 of plug flow reactor
Industrial extractors
Inert gas systems
Inline mixer
Internal coil in a vessel
International System of Units
Jackets
 common arrangements of
 limpet coil type
 plain
 spiral baffle type
 heat transfer coefficients for
 calculations thereof
 selection criteria of
Jet flooding
Jet reactor
Joint efficiency
Joule
Kay’s rule
KATMAX structured packing
Kettle type reboiler
 process design steps
 selection criteria of
Key components
 light key
 heavy key
 split key
Ki  values for hydrocarbons
Kirk Bride equation
Lacy’s equation
Latent enthalpy
 of fusion
 of sublimation
 of vaporization
Le Chatelier’s principle
Lewis-Matheson method
Li method
Light key component
Limitations of heat exchanger networking
Limiting reactant
Limpet coil
Linde trays
Liquid back-up in downcomer
Liquid crest over weir
Liquid distributors
 orifice type
 perforated pipe type
 trough type
 weir type
Liquid redistributors
Liquid entrainment
Liquid flow patterns in sieve tray tower
Liquid phase transfer unit
Liquid phase mass transfer coefficient
Liquid redistributors
 bell cap
 wiper type
Locations of pressure taps for orifice meter
Logarithmic mean temperature difference (LMTD)
Loop reactor
LMTD correction factors
 for shell and tube heat exchangers
 for air coolers/heaters
 for plate heat exchangers
McAdams, Drew and Bay’s equation
Malokanov equation
Mass
 atomic
 molar
Mass fraction
Mass flow meter
 Coriolis type
Mass flow rate
 through orifice meter
 through rotameter
 through venturi meter
Mass transfer coefficients
 of gas phase
 of liquid phase
Mass velocity
Material balances
 without chemical reactions
 involving chemical reactions
MATHCAD®  
McCabe-Thiele method
McAdams, drew and Bay’s equation
Mean temperature difference (MTD)
 logarithmic
 weighted average
MESH equations
Minimum amount of solvent
 for gas absorption
 for liquid-liquid extraction
Minimum reflux ratio,
 for batch distillation
 for binary distillation
 for multicomponent distillation
Minimum wetting rate for different packing
Mist eliminator
Mixed flow reactor
Mixer settler
Mixing of reaction systems
 gas-liquid mixture
 homogeneous liquid mixture
 immiscible liquid mixture
 slurry
MKS system of units
Molar heat capacity
 isobaric
 isochoric
Molar mass
Molar volume of gas at NTP
Mole
Mole fraction
Molecular distillation
Molecular sieve drying
Monolithic reactor
Morris and Jackson equation
Mostinski equation
Multicomponent condensation
Multicomponent distillation
Multiple quench bed reactor
 axial flow
 radial flow
Multistage countercurrent extraction
Net area for vapour-liquid disengagement on a tray
Net positive suction head (NPSH)
 available
 for saturated liquid with dissolved gases
 required
 for calculations thereof
 correction factors for high temperature
Non-distributed components
Normal boiling point
Normal temperature and pressure (NTP)
NRTL equation for determination of activity coefficient
Number of overall gas phase transfer units
 of packed tower
 of spray tower
 of venturi scrubber
Number of transfer units (NTU)
 for gas absorption
 for liquid-liquid extraction
Nucleate boiling
Nusselt equation
 for shell side coefficient
 for tube side coefficient
Nusselt number
Onda, Takeuchi and Okumoto correlation
Open steam
 use in distillation
Operating instructions manual (OIM)
Operating range of trays
Operating flooding
Optimum amount of solvent for gas absorption
Optimum pipe size
Optimum pressure drop
 in shell and tube heat exchanger
Optimum reflux ratio
Optimum solvent to feed ratio for extraction
Optimum tray spacing
Orifice coefficient
 for orifice meter
 for sieve tray
Orifice meter
 advantages and disadvantages
 process design steps
 various tap locations
Orifice type liquid distributor
Overall heat transfer coefficients
 calculations thereof
 for shell & tube heat exchangers
 for plate heat exchangers
 for spiral heat exchangers
 typical values for SHE
p-V-T  relationship
p-t diagram for CO2  
p-t diagram for H2O
Packed tower for extraction
 internals thereof
Packing
 different types
 random
 structured
 selection criteria for
Packing factor
Packing supports
 beam type
 cap type
 gas injection type
 mesh type
 welded ring type
Paddle type agitator
Pall rings
Partial pressure
Partition coefficient
Partition plate
 shell side
 tube side
Partitioned distillation column
Pascal
Pass partition plate
 shell side
 tube side
Pelen and Small equation
Peng-Robinson equation
Percent void space in packing
Perforated pipe type distributor
Performance of centrifugal pump
 while handling viscous liquids
Performance test run
Pervaporation
Petyluk column
Pfaudler agitator
Phases
 continuous
 dispersed
 extract
 raffinate
Phase diagram
 of carbon dioxide (CO2)
 of water (H2O)
Phase equilibrium
 liquid-liquid
 of a ternary system of liquids
 vapour-liquid
Physical absorption
 isothermal
 non-isothermal
Pilot studies
Pinch point
Pinch technology
Pinch temperature difference
Piping & instrumentation diagram (P&ID)
Plait point
Plate heat exchanger
Plug flow reactor
Point efficiency of sieve tray
Ponchon-Savarit Method
Polar solvents
Positive displacement pump
Power
Power correlations for agitators
Power number
Power requirement of
 agitators
 blower/adiabatic compressor
 fan
 pump
Prandtl number
Pressure
Pressure drop in a sieve tray
Pressure swing adsorption
Pressure swing distillation
Process intensification
Projected perimeter of fin
Propeller
Properties of common inert gases
Properties of solvents
 for absorption
 for extraction
Properties of supercritical fluids
Pressure
 absolute
 critical
 gauge
 partial
Pressure drop calculations for
 fluid bed cooler
 orifice meters
 packed tower
 piping system
 in fittings and valves
 plate heat exchanger
 shell side
 sieve tray tower
 spiral heat exchanger
 tube side
 two phase system
 venturi scrubber
Process design
 alternate designs
Process flow diagram (PFD)
Process engineering
 alternate routes
Process cooling systems
Process heating systems
Process intensification
Process research
Projected perimeter of finned tube
Pseudo critical properties
Pulsed column for extraction
Pump
 different types
Quench bed reactor
 axial
 radial
q-line
Raffinate phase
Rankine coefficient of performance
Random packing
Raoult’s law
Raschig rings
Reaction completion
 degree of completion
Reactive distillation
 advantages and disadvantages of
 applications
 different ways of
Reactors
 different types
 batch type
 batch loop type
 baffle tank type
 bubble column type
 continuous flow reactors
 continuous stirred type
 fixed bed catalytic type
 fluid bed catalytic
 gas induction type
 heterogeneous continuous type
 jet reactor
 mixed flow reactor
 monolith loop type
 plug flow type
 thin film absorber type
 tubular type
 quench bed type
Reactive and catalytic distillation
 advantages and disadvantages
 applications
 contact devices for catalytic distillation
 design of experiment for
Reboiler
 forced circulation type
 kettle type
 thermosyphon type
Recirculation ratio in reboiler
Recommended fluid velocities
Rectification
Recycle stream
Refrigeration
Reflux ratio
 minimum
 optimum
Regular packing
 grids
 stacked rings
 structured packing
Relative volatility
Research & development
Residence time in downcomer
Residual pressure drop
Reynolds number
Rigorous methods for multicomponent distillation
 equation tearing procedure
 Lewis-Matheson method
 relaxation methods
 Thiele-Geddes method
Riedel equation
Roller wiper system
Rotary cooler for solids
ROSE process
Rotameter
 calibration of
 magnetic type
 standard
Rotating disc columns for extraction
Rotating distributor plate
Saddles
 berl type
 Intalox type
Shapes of solid catalysts
Scale-up
 of agitated vessels
 of fixed bed catalytic reactor
Schedule number
Scrubbing
Sealing strip
Seamless pipes
Segmental baffle
Selection criteria for condensers
Selection criteria for different types of absorbers
Selection criteria for different types of extractors
Selection criteria for different types of reboilers
Selection criteria for tray tower and packed tower
Selection criteria for solvents
 for absorption
 for extraction
 for extractive distillation
Selection criteria for various types of gas-liquid reactors
Selection criteria for various types of trays
Selection of cooling/heating medium
Selection of key component
Selection of operating pressure for distillation column
Selectivity in extraction
Selectivity in reaction
Sensible heat changes
Settler
Shear mixer
Shell
 divided flow type
 single pass
 split flow
Shell side pass partition plate
Shell & tube heat exchangers
 as absorber
 as reactors
 different parts of
 different types
Sherwood-Leva=Eckert correlation chart
Short path distillation
 applications
 design and working of
SI prefixes
SI system of units
 base units
 derived units
Sieder-Tate equation
Sieve tray
 efficiency
 in extraction
 process design of
Simple batch distillation
Slurry reactions
Solute
Solvent
Smith-Brinkley method
Smoker’s equation
Spacer
Sparged vessel type reactor
Specific gravity
Split keys
Spiral plate heat exchanger
 advantages and disadvantages
 design of
 flow arrangement
 standard dimensions of
Spray chamber/tower
 horizontal
 vertical
Spray chamber design
 for SO2  scrubbing
Spray cooling in rotary cooler
Square pitch arrangement
Stagewise extractor
Standard heat of combustion
Standard heat of formation
Standard heat of reaction
Standard pipes
Standards for PFD
Standards for P&ID
Steam heating
Stratified flow
Static mixers
Stationary cooler for solids
Steam ejector
Steam stripping
Stripping section
Stirred tank as absorber
Stoichiomtric factor
Stoichiometry
Stoke’s law
Structured packing
Subcooling
Sudgen equation
Supercritical extraction
 advantages and disadvantages
 applications
Supercritical fluids
 critical properties of
 properties
 reactions in
Superficial gas velocity
Superficial liquid velocity
Supplementary units
Surface area of packing
Surface roughness
 of various materials
Surface tension
 critical of packing material
Symbols
 of equipments
 of piping, instruments and controls
Synthesis of ETBE
Tail gas scrubber
TEMA
 designation
Tellerettes
Temperature
 absolute
 critical
Theoretical stages in distillation
Thermal conductivity
Thermal energy
Thermal insulation for cold surfaces
Thermal insulation for hot surfaces
Thermally coupled
 distillation
 direct sequence
 indirect sequence
 side column rectifier
 side stream stripper
Thermic fluid heating systems
Thermocompressor for distillation system
Thermosyphon reboiler
 horizontal
 vertical
Thickness of pipe
Thiele-Geddes method
Tie line
Tie rod
Tinker’s flow model
Tray efficiency
 Actual
 AIChE method
 Van Winkle’s correlation
Tray spacing
Tray tower
 different types
 bubble cap
 high capacity
 sieve tray
 valve tray
Tray tower as absorber
Triangular pitch arrangement
Tridiagonal matrix method
Triple point
 of CO2  
 of H2O
Trouble shooting
Trough type distributor
Tube
Tube bundle diameter
Tube hole count
Tube to tubesheet joint
 expanded
 welded
Tubeside pass partition plate
Tube sheet (tube plate)
 tube hole count
Turbine type agitator
 curved blade
 flat blade
 pitched blade
Turndown ratio
oTwaddell
Two phase flow
 estimation of pressure drop in steam-condensate system
Ultrasonic flow meter
Unagitated static columns for extraction
Underwood’s method
Units
 conversion factors
 different systems
UNIQUAC method for determination of activity coefficient
Universal gas constant (R)
US customary units
Use of open steam in distillation
U-tube shell & tube heat exchanger
 advantages and disadvantages
Vacuum distillation
 advantages and disadvantages
Valves & fittings
 equivalents lengths of pipe for
 equivalent number of velocity heads for
Valve tray
van Larr equations
Vapour–liquid equilibrium data
Vapour–liquid equilibrium coefficients
 for hydrocarbons
Vapour recompression system for distillation system
Valve trays
Vapour pressure
 Antoine equation
 artificial
Vaporizer
van Laar equation
Variable area flow meter
Vena contracta
Vertical condenser
Vertical spray tower
Vertical thermosyphon reboiler
VLE data
Volume
Vortex meter
Venturi meter
Venturi scrubber
 for SO2  scrubbing
Viscosity
 conversion diagram
 correction for centrifugal pump
Water (H2O)
 critical properties of
 p-t  diagram
Water cooled heat exchanger
Water cooling
Water hammer
Watson equation
Weeping
 Froude number for checking weeping
Weep point correlation
Weir on a tray
Welded ring type packing support
Wetted surface of packing
Wetting rate
Weirs
 for distillation tower
 for flow measurement
Weir type distributor
Weighted temperature difference
Wiped film evaporator
Wiper type redistributors
Wiremesh type packing support
Work
Yield
Zeolites
Zuber’s equation