Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

04 Distillation Sequencing

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 36

IV

Distillation Sequencing
Outline
1. Basic Concepts of Distillation
Sequence Design
2. Choice of Sequence and its
Operating Pressure.
3. Performance of Distillation
Column (Sieve tray and packed
tower)
4. Separation and Recycle for
continues process

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.1.
BASIC CONCEPT OF
DISTILLATION SEQUENCING

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.1.1. Introduction
 Consider: Separation of a homogeneous
multi-component fluid mixture into a
number of products, whereas all
separations are carried out using
distillation only.
 If this is the case, how to choice the
distillation sequence?

 For example: two simple column


sequences (direct and indirect)
could be employed in separation
of a three-component mixture
into three relatively pure
products.
Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY
Direct Sequence vs Indirect Sequence

direct sequence indirect sequence


the lightest component is taken the heaviest component is taken as
overhead in each column bottom product in each column
requires less energy for both requires more energy for both
reboiling and condensation supplied reboiling and condensation supplied
by utilities by utilities
component A (light material) is only Component A (light material) is
vaporized once vaporized twice
can be more energy-efficient if the
feed to the sequence has a low
flowrate of the light material (A) and
a high flowrate of heavy material (C)

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Alternative sequences for the separation of a four-product mixture.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Number of possible distillation sequences using simple columns

 The problem is that there may be significant differences in the capital


and operating costs between different distillation sequences that can
produce the same products.
 In addition, heat integration may have a significant effect on operating
costs (would be discussed next).

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.1.2. Practical Constraints
1. Remove as early as possible:
a. A particularly hazardous component  safety consideration
b. Reactive or heat-sensitive component  to avoid problems of
product degradation
c. Corrosive component  to minimize the use expensive material of
construction
1. The main component that difficult to be condensed should be
removed as early as possible  using refrigeration system or high
pressure system
2. Don’t take the final product from the bottom of column if:
a. The component is decomposed in the reboilers (it can
contaminates the product)
b. Polymerization inhibitors are used to inhibit polymerization of
some components when distilled. These polymerization inhibitors
tend to be nonvolatile, ending up in the column bottoms

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.2.
CHOICE OF SEQUENCE AND
ITS OPERATING PRESSURE

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.2.1. Heuristics of Choice of Sequence (Smith, R., 2005)

1. Component with its relative volatility close to unity or that


exhibit azeotropic behavior should be removed last.
2. The lightest components should be removed alone one by
one in column overheads (use direct sequence).
3. A component composing a large fraction of the feed
should be removed first.
4. Favor splits in which the molar flow between top and
bottom products in individual columns is as near equal as
possible.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Example 4.2.1: Data for a mixture of alkanes to be separated by
distillation are as follows:
 The relative volatilities
have been calculated on
the basis of the feed
composition to the
sequence, assuming a
pressure of 6 barg using
the Peng–Robinson
Equation of State with
interaction parameters set
to zero.
 Different pressures can, in
practice, be used for
different columns in the
sequence

 Use the heuristics to identify potentially good sequences that are


candidates for further evaluation!

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Solution: Alternative-1
Heuristic 1: Do D/E split last since this separation has the
smallest relative volatility.
All four heuristics
are in conflict
here: Heuristic 1
Heuristic 2: Favor the direct sequenc e suggests doing the
: D/E split last, and
Heuristic 3
suggests it should
be done first.
Heuristic 3: Remove the most plentiful Heuristic 2
component first: suggests the A/B
split first and
Heuristic 4 the
C/D split first.
Heuristic 4: Favor near-equimolar
splits between top and
bottom products:
Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY
Solution: Alternative-2:
Take one of the candidates and accept, say, the A/B split first.
Again the heuristics are in conflict
Heuristic 1: Do D/E split last • Heuristic 1 again suggests doing
. the D/E split last, whereas again
Heuristic 3 suggests it should be
Heuristic 2: done first.
• Heuristic 2 suggests the B/C split
first and Heuristic 4 the C/D split
first.

Heuristic 3:
• There are 14 posible sequence
• This process could be continued
and possible sequences
identified for further
consideration.
Heuristic 4: • Some possible sequences would
be eliminated

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Quantitative measure as other consideration

 Since heuristics (qualitative procedure) can be in conflict, a


quantitative measure of the relative performance of different
sequences would be preferred
 The vapor flow up the column as a physical measure can be readily
calculated. This provides an indication of both capital and operating
cost.
 More vapor flow up the column, more heat duty required for
reboiler and condenser, increase the operating cost of hot utility
(steam) and cold utility (water or refrigerant)
 A high vapor rate leads to a large diameter column, and also
requires large reboilers and condensers, therefore the capital cost
increases
 Consequently, sequences with lower total vapor load would be
preferred to those with a high total vapor load.
 But how is the total vapor load predicted?

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Prediction of the total vapor load
Underwood Vmin  D1 Rmin  … (4.2.1)

Eq. (4.2.1) can also be written at finite reflux.


Defining RF to be the ratio R/Rmin (typicaly R/Rmin = 1.1):

V  D1 RF Rmin  … (4.2.2)

1  xDLK   xDHK 
Rmin can be calculated: Rmin    … (4.2.3)
  1  xFLK xFHK 

where:  = relative volatility between the key components


xDLK = mole fraction of light key in the distillate
xFLK = mole fraction of light key in the feed
xDHK = mole fraction of heavy key in the distillate
xFHK = mole fraction of heavy key in the feed
Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY
Assuming a sharp separation:
• only the light key and lighter than LK components in the overhead
• only the heavy key and heavier than HK components in the bottoms
1  xDLK  1 F
Rmin      … (4.2.4)
  1  xFLK    1  D 

where: F = feed flow rate


D = distillate flow rate
 RF F   RF 
Combining (4.2.4) and (4.2.2): V  D 1   D  F   … (4.2.5)
  1 D    1 
thus:
 RF 
V  FA  FB  ... FLK  FA  FB  ... FLK  FHK  ... FNC   … (4.2.6)
  1 

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Example 4.2.2:

Using the data (below) for a ternary separation of benzene, toluene,


and ethyl benzene. Based on the vapor flow-up, determine whether
the direct or indirect sequence should be used!

Relative
Volatility
betwwen
Flowrate Relative adjacent
Symbol Component (kmole/h) Volatility components
A Benzene 269 3.53
1.96
B Toluene 282 1.80
1.80
C Ethyl Benzene 57 1.00

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Solution of Example 4.2.2:

Direct Sequence: A/BC and B/C

 V  269 269  282  57


 1.1   282 82 57 1.1 
  
 1.96 1   1.80 1 
 965.7  748.1

 1713.8kmole/h Hence we should use the direct sequence

Indirect Sequence: AB/C and A/B

 V  269  282 269  282  57


 1.1   269 296  282 1.1 
  
 1.80 1   1.96 1 
 1387  900.4

 2287.4kmole/h

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Direct and Indirect Sequence
(example 4.2.2)

269 kmol/h
269 kmol/h

282 kmol/h
269 kmol/h
282 kmol/h

282 kmol/h

57 kmol/h 57 kmol/h

V =1713.8 kmol/h V =2287.4 kmol/h

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Example 4.2.3:

Using the Underwood Equations, determine the best distillation


sequence, in terms of overall vapor load, to separate the
mixture of alkanes in Example 4.2.1 into relatively pure
products. The recoveries are to be assumed to be 100%.
Assume the ratio of actual to minimum reflux ratio to be 1.1
and all columns are fed with a saturated liquid. Neglect
pressure drop across each column. Relative volatilities can be
calculated from the Peng–Robinson Equation of State with
interaction parameters assumed to be zero (see Chapter 4).
Determine the rank order of the distillation sequences
on the basis of total vapor load for all column pressures
fixed to 6 barg with relative volatility calculated from
the feed to the sequence.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Solution of Example 4.2.3:

Dr. Eng. Y. D. Hermawan – C hemEng - UPNVY


IV.3.
PERFORMANCE OF
DISTILLATION COLUMN
(Plate/Tray Column and Packed Column)

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


Distillation Tray and Packing

Distillation Tray Distillation Packing


Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY
Plate/Tray Column vs Packed Column

Plate/Tray Column Packed Column


Contact of vapor-liquid relatively chanelling and backmixing could be
good happen
More liquid hold-up ---
Easy to be cleaned ---
--- Small Pressure drop, prefer to
vacuum operation
--- Cheaper for corrosive fluid
--- Prefer to small diameter
Can be used for liquid that contains Solid particle plugs the packed
solid particles
--- Foaming liquid
--- Lighter
Products can be taken from the ---
side-stream

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.4.
SEPARATION AND RECYCLE SYSTEM
FOR CONTINUES PROCESS

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.4.1. Introduction

 Do separation for some reasons:


1. to achieve product specification
2. to meet environment law
 Material to be separated:
1. reactants
2. main product
3. byproduct (could be sold
4. waste (could not be sold)

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


IV.4.2. Function of Process Recycles

1. Reactor conversion:
Consider FEED  PRODUCT with conversion of about 95%
Incomplete conversion in the reactor requires a recycle for
unconverted feed material.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


2. Byproduct formation
Consider:
1st : FEED  PRODUCT + BYPRODUCT
or
1st : FEED  PRODUCT

2nd : PRODUCT BYPRODUCT

Using a purge saves the cost of a separator


but incurs raw material losses, and possibly
waste treatment and disposal costs. This
might be worthwhile if the FEED-
Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY BYPRODUCT separation is expensive.
3. Recycling byproducts for improved selectivity
Consider: If a byproduct is formed via a
reversible secondary reaction then
recycling the byproduct can inhibit its
formation at source.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


4. Recycling byproducts or contaminants that damage
the reactor
 When recycling unconverted feed material, it is possible
that some byproducts or contaminants, such as products of
corrosion, can poison the catalyst in the reactor.
 It is clearly desirable to remove such damaging
components from the recycle stream.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


5. Feed impurities

If the impurity has


an adverse effect on
the reaction or
poisons the catalyst

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


5. Feed impurities (continued)

I if the impurity does


not have a significant
effect on the reaction,
then it could perhaps
be passed through
the reactor and be
separated

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


5. Feed impurities (continued)
 As with its use to
separate byproducts,
the purge saves the
cost of a separation,
but incurs raw
material losses.
 This might be
worthwhile if the
FEED-IMPURITY
separation is
expensive.
 Care should be taken to ensure that the resulting increase in
concentration of IMPURITY in the reactor does not have an
adverse effect on reactor performance.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


6. Reactor diluents and solvents.

 An inert diluent such as steam is sometimes needed in the


reactor to lower the partial pressure of reactants in the vapor
phase.
 Diluents and solvents are normally recycled.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


7. Reactor heat carrier.

The introduction of an extraneous component as a heat


carrier effects the recycle structure of the flowsheet.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY


7. Reactor heat carrier (continued)
 This figure illustrates the use of the product as the heat carrier.
 This simplifies the recycle structure of the flowsheet and removes the
need for one of the separators.
 The use of the product as heat carrier is obviously restricted to situations
where the product does not undergo secondary reactions to unwanted
byproducts.

Dr. Eng. Y. D. Hermawan – ChemEng -UPNVY

You might also like