CHAPTER VIII

CHEMICAL AND MECHANICAL DESIGN OF PALMITIC ACID DISTILLATION

COLUMN (T-100)

8.1 INTRODUCTION

Continuous distillation column uses variation of temperature and pressure conditions

along the height of the column to get more volatile component at the top of the column
and less volatile component at the bottom of the column. The mixture to be separated is
introduced at somewhere along the height of a vertical column. If the mixture contains
more of lighter component, then the inlet is closer to top of the column. Conversely if the
mixture has more of the heavy component, then inlet is closer to bottom of distillation
column.

The column is provided with heat by a reboiler which is continuously boiling the
liquid from the bottom of the columns. And the vapors from the top of the column are
continuously cooled and condensed by an overhead condenser provided at the top of
the column. The heating and cooling actions of these heat exchangers are responsible
for vapor liquid equilibrium conditions in the column. Also due the action of reboiler the
bottom of a distillation column has highest temperature and pressure conditions.
The condenser is responsible for lowest temperature and pressure conditions at the top
of the column.

A distillation column consists of a number of stages. Each of these stages is

formed by a perforated tray. Liquid from the top condenser flows down from tray to tray
to the column bottom. Vapors flow from bottom to top through the perforations in each of
these trays. Thus trays provide the interface for vapor liquid contact and depending on
the residence time on each tray the vapors and liquids tend to form vapor liquid
equilibrium conditions. Hence each tray can be ideally thought of as vapor liquid
equilibrium at different temperature and pressure conditions. The temperature and
pressure decrease from tray to tray as we move from bottom to the top of the column,
due to action of reboiler at bottom and condenser at top of column.
 8-2

The vapor-liquid contact at each tray enhances separation. Heavy component

vapors get condensed when they are in contact with cold liquid from the top. Conversely
lighter component gets vaporized and stripped away from the liquid phase. Hence high
number of separation implies higher degree of separation. A very high number of trays
are often required to separate very closely boiling liquids or to get very high purity which
is required for pharmaceutical substances (McGraw-Hill Science & Technology
Encyclopedia).

8.2 PROCESS BACKGROUND

Distillation is a commonly known method for separation of liquid components with

different boiling points. The ease of separation through distillation usually depends on
the difference between boiling points. The boiling point for Palmitic Acid, Tetradecene
and Acetic Acid are 351°C, 251°C and 118°C respectively. Stream 8 is the inlet to the
distillation column and Stream 9 and 10 are the outlet from the distillation column. The
flowrate entering the distillation column is 44600 kg/h. The function of this distillation
column is to separate Palmitic Acid from the multicomponent mixture.

As the Palmitic Acid is the main component to be separated from the

multicomponent mixture, the temperature at the bottom of the distillation column must be
high enough to vaporize Tetradecene and Acetic Acid but must be lower than the
Palmitic Acid boiling point. So that, Palmitic Acid will remain at the bottom of the
distillation column. Palmitic Acid has the highest boiling point and it is easy to separate
as the boiling point difference from other components is quite significant. Thus, it can be
separated through the bottom of the distillation column. On the other hand, Tetradecene
and Acetic Acid will be separated through the top of the distillation column. The
Tetradecene and Acetic Acid are then sent to another distillation column for further
separation. While, the Palmitic Acid will be recycled back into the process. Simple
schematic diagram on the Palmitic Acid distillation column process is shown in Figure
8.1.
 8-3

S9 Flowrate = 14720 kg/h

Components:
Tetradecene (T)
Acetic Acid (AA)

Flowrate = 44600 kg/h

Components: S8
Palmitic Acid (PA)
Tetradecene (T)
Acetic Acid (AA)

Flowrate = 29870 kg/h

Components:
S10 Palmitic Acid (PA)

Figure 8.1: Simple schematic diagram on the Palmitic Acid distillation column process

8.3 CHEMICAL DESIGN OF PALMITIC ACID DISTILLATION COLUMN

8.3.1 Estimation of Dew Point and Bubble Point

By definition, a saturated liquid is at its bubble point where any rise in temperature will
cause a bubble of vapor to form, and a saturated vapor is at its dew point where any
drop in temperature will cause a drop of liquid to form.

Dew points and bubble points can be calculated from the vapor-liquid equilibrium
for the system. In terms of equilibrium constants, they are defined by the equations

bubble point:  y  K x
i i i  1.0

yi
and dew point: x   K
i  1.0
i
 8-4

For multicomponent mixtures, the temperature that satisfies these equations, at a

given system pressure, must be found by iteration. Calculation for dew point and bubble
point can be calculated by using Antoine equation as below:

 D  ln T ( K )  E  T ( K ) F
B
ln Psat (kPa)  A 
T (K )  C

With related at equilibrium, constant K,

Psat
Ki 
P

Table 8.1 below is the value of the constant for Antoine equation, the data from Table
8.1 below is obtained from HYSYS simulation. The components involved are Palmitic
Acid (PA), Tetradecene (T), Acetic Acid (AA) and Gum (G).

Table 8.1: Constant value for Antoine equation

Component Ant A Ant B Ant C Ant D Ant E Ant F

PA 6.20E+01 -1.17E+04 0.00E+00 -6.12E+00 1.66E-06 2.00E+00
T 2.06E+02 -1.72E+04 0.00E+00 -2.75E+01 1.20E-05 2.00E+00
AA 6.13E+01 -6.77E+03 0.00E+00 -6.73E+00 4.84E-06 2.00E+00
G 2.33E+02 -2.29E+04 0.00E+00 -3.04E+01 8.53E-06 2.00E+00

The composition for inlet and outlet of the distillation column is obtained from HYSYS
simulation is shown in Table 8.2.

Table 8.2: Composition for Palmitic Acid distillation column

Component Feed XF Distillate XD Bottom XB
PA 117.0682 0.5073 0.5706 0.0050 116.4975 1.0000
T 56.8456 0.2463 56.8453 0.4975 0.0003 0.0000
AA 56.8469 0.2463 56.8469 0.4975 0.0000 0.0000
G 0.0011 0.0000 0.0000 0.0000 0.0011 0.0000
Total 230.7617 1.0000 114.2628 1.0000 116.4989 1.0000
 8-5

Calculation for the relative volatile, α,

  K LK K HK

Where KLK = Light key component.

KHK = Heavy key component.

The separation between the top and bottom products usually specified by setting
limits on two key components. The light key will be the component that it is desired to
keep out of the bottom product, and the heavy key is the component to be kept out of the
top product. So that, in this process Tetradecene will be the light key while Palmitic Acid
is the heavy key.

By using goal seek in the Excel programmed the value of the bubble point at
bottom column = 340.66 °C and dew point at the top column is 217.04 °C. Data from the
goal seek programmed is shown in Table 8.3 and 8.4 below.

Bubble point calculation, Bottom Column:

P 80 kPa
Tbub 613.81 K 340.66 deg C

Table 8.3: Bubble point calculation data

Component XB Psat (kPa) K KXB

PA 1.000 79.989 1.000 1.000
T 0.000 540.862 6.761 0.000
AA 0.000 7746.258 96.828 0.000
G 0.000 20.142 0.252 0.000
Total 1.000 1.000
 8-6

Dew Point Calculation, Top Column:

P 70 kPa
Tdew 490.19 K 217.04 deg C

Table 8.4: Dew point calculation data

Component XD Psat K XD/K

PA 0.005 2.097 0.030 0.167
T 0.497 43.394 0.620 0.803
AA 0.498 1125.349 16.076 0.031
G 0.000 0.480 0.007 0.000
Total 1.000 1.000

Table 8.5 shows the value of the relative volatile for every component at top and bottom
of distillation column.

Table 8.5: Relative volatilities for each component.

α
Top Bottom Average
Temp (°C) 217.04 340.66
PA 1.000 1.000 1.000
T 20.696 6.762 11.830
AA 536.714 96.841 227.983
G 0.229 0.252 0.240
 8-7

8.3.2 Calculation for Minimum Reflux Ratio, Rmin

Colburn (1941) and Underwood (1948) have derived equations for estimating the
minimum reflux ratio for multicomponent distillations. Underwood equation is used to find
the minimum reflux ratio.

The equation can be stated in the form:

 i xi , d
  Rmin  1
i 

And θ is the root of the equation

 i xi , f
  1 q
i 

As the feed is at its boiling point, q = 1.

 i xi , f
  11  0
i 

From equation above Table 8.6 is tabulated.

Table 8.6: Data by using Underwood equation

Component XF α αXF
PA 5.07E-01 1.000 5.07E-01
T 2.46E-01 11.830 2.91E+00
AA 2.46E-01 227.983 5.62E+01
G 4.66E-06 0.240 1.12E-06

By using trial and error, the value of θ must lie between the values of the relative
volatility of the light and heavy keys. Table 8.7 shows the iteration value for θ.
 8-8

Table 8.7: Iteration value to find θ

θ αxF/α-θ θ αxF/α-θ
1 - 21 -0.0718
2 0.0377 22 -0.0380
3 0.3260 23 -0.0099
4 0.4538 24 0.0138
5 0.5517 25 0.0343
6 0.6514 26 0.0521
7 0.7730 27 0.0678
8 0.9438 28 0.0818
9 1.2229 29 0.0944
10 1.7940 30 0.1058
11 3.7208 31 0.1162
12 -16.8867 32 0.1257
13 -2.2708 33 0.1345
14 -1.1192 34 0.1427
15 -0.6917 35 0.1503
16 -0.4676 36 0.1575
17 -0.3291 37 0.1642
18 -0.2346 38 0.1706
19 -0.1658 39 0.1766
20 -0.1133 40 0.1823

So, θ = 2 that resulted Σ(αXF/α-θ) near enough to zero.

 i xi , d
Substitute θ = 2 into equation   Rmin  1
i 

Table 8.8 shows substitution of above equation with θ value to find Rmin.

Table 8.8: Data after substitution of θ value

Component XD α αXD αXD/(α-θ)
PA 0.0050 1.0000 0.0050 -0.0050
T 0.4975 11.8296 5.8852 0.5987
AA 0.4975 227.9825 113.4236 0.5019
G 0.0000 0.2400 0.0000 0.0000
Total 1.0956
 8-9

Rmin  1  1.10

 Rmin  1.10  1.00  0.10

For many systems, the optimum reflux ratio will lie between 1.2 to 1.5 times the
minimum reflux ratio. Optimum reflux ratio, 0.1 x 1.5 = 0.15.

8.3.3 Determination of Plate Number Using Fenske Equation

The Fenske equation can be used to estimate the minimum stages require at total reflux.
The equation applies equally to multicomponent systems.

 x  x 
log  LK   HK 
 xHK  d  xLK  b
N min 
log  LK

Where αLK = average relative volatility of the light key with respect to the heavy
key
xLK = light key concentration
xHK = heavy key concentration

  56.845   116.498  
log   4 

  0.571   2.93  10  
N min   10.43  11 stages
log 11.830

Normally after using the Fenske equation, the value of Nmin is given by the equation
below to get the number of stages, NT.

N T  2 N m in   2  11

N T  22 stages

To get the real number of stage, the efficiency of the process must be considered.

From HYSYS simulation, viscosity, µa = 0.5369 mNs/m2

 a  a  0.5369  11 .830  6.35

 8-10

From Coulson and Richardson (2008), Figure 11.19, Eo = 35%.

N T  1 22  1
N   60 stages
Eo 0.35

Kirkbride (1944) has devised an approximate method to estimate the number of

theoretical stages above and below the feed which can be used to estimate the feed-
stage location.

 B x  xb , LK 
2

N   f , HK
log r  0.206  log     
Ns  D  x f , LK  x  
  d , HK  

Where Nr = number of stages above the feed

Ns = number of stages below the feed

xf, HK = concentration of the heavy key in the feed

xf, LK = concentration of light key in the feed

xd, HK = concentration of the heavy key in the top product

xb, LK = concentration of the light key if in the bottom product

Nr  117.068  116.499  2.56  10 6  2 

log  0.206  log     
3 
Ns  56.846  114.263  4.99  10  

Nr
log  1.67  10 8
Ns

Nr
 1.000
Ns
 8-11

Number of stages, excluding the reboiler = 59 stages.

N r  N s  59

N s  59  N r  59  N s

59
Ns   29.5  30
2

8.3.4 Estimate the Physical Properties

The properties considered in this design are liquid flow rate, vapor flow rate, liquid
surface tension, liquid density and vapor density. The physical properties evaluated at
the system temperature by using HYSYS generated data or estimate manually from
mass and energy balance data.

From HYSYS simulation:

Feed, F = 12.39 kg/s

Distillate, D = 4.09 kg/s
Bottom, B = 8.30 kg/s

8.3.4.1 Relative Molar Mass and Density

Table 8.9 below shows liquid density for each component.

Table 8.9: Calculation data for relative molar mass and density
Liq
Mole Fraction
Comp MW density
Feed Distillate Bottom kg/m3
PA 256.43 0.5073 0.0050 1.0000 881.58
T 196.38 0.2463 0.4975 0.0000 774.10
AA 60.05 0.2463 0.4975 0.0000 1051.50
 8-12

From table above relative molar mass, RMM can be calculated for feed, distillate and
bottom.

 RMM  feed 0.5073 256.431  0.2463196.378  0.2463 60.052

  RMM  feed   193.2595

So that, the above step is repeated for distillate and bottom.

 RMMdistillate  128.8544

 RMMbottom  256.4285

Liquid density and vapor density for top and bottom distillation column can be calculated
by using the following equations.

Density at the top of distillation column

Condition: Operating pressure, POP = 0.7 atm
Operating temperature, TOP = 414.25 K
Standard volume, VSTP = 22.4 m3
Standard temperature, TSTP = 273.15 K
Standard pressure, PSTP = 1 atm

Liquid density:

 L   x D,i  i  (0.005  881.58)  0.498  774.10  0.498  1051.50

 L  912.6450 kg / m 3

Vapor density:

MW TSTP POP 128.8544 273.15 0.7

v         2.6537 kg / m 3
VSTP TOP PSTP 22.4 414.25 1
 8-13

Density at the bottom of distillation column

Condition: Operating pressure, POP = 0.8 atm
Operating temperature, TOP = 613.85 K

Liquid density:

 L   x B ,i  i  (0.999  881.58)  2.51 10 6  774.10  1.13  10 10  1051.50

 L  881.5696 kg / m 3

Vapor density:
MW TSTP POP 128.8544 273.15 0.8
v         4.0729 kg / m 3
VSTP TOP PSTP 22.4 613.85 1

8.3.4.2 Determination of Liquid and Vapor Mass Flow Rate

Mass balance on top of distillation column (Feed condition = liquid)

Vm
R D

Lm

Figure 8.2: Top of distillation column

Liquid mass flow rate, Lm = RxD = 0.61 kg/s

Vapor mass flow rate, Vm = D + Lm = 4.70 kg/s
 8-14

Mass balance on bottom of distillation column (Feed condition = liquid)

Vn Ln

Figure 8.3: Bottom of distillation column

Liquid mass flow rate, Ln = (R x D)+F = 13.00 kg/s

Vapor mass flow rate, Vn = Ln - B = 4.71 kg/s

8.3.5 Calculation for Column Diameter

Neglecting differences in molecular weight between vapor and liquid:

Lm v
FLV top   0.01
Vm L

Ln v
FLV bottom   0.19
Vn L

Take plate spacing as 0.5 m

From Coulson and Richardson (2008), Figure 11.34:

Top K1 = 0.09

Bottom K1 = 0.07
 8-15

Correction for surface tensions

From HYSYS simulation results:

Surface tension, σtop = 0.0166 N/m

Surface tension, σbottom = 0.0049 N/m

  
0.2

Top K correction    K 1  0.0847

 0.02 

  
0.2

Bottom K correction    K 1  0.0528

 0.02 

From Coulson and Richardson (2008),

 L  V
Top flooding velocity, u f  K1  1.57 m / s
V

 L  V
Bottom flooding velocity, u f  K1  0.78 m / s
V

Design for 85% flooding at maximum flow rate

Top u n  0.85  u f  1.33 m / s

Bottom u n  0.85  u f  0.66 m / s

Maximum volumetric flow rate

Vm
Top   1.77 m 3 / s
v

Vn
Bottom   1.16 m 3 / s
v
 8-16

Net area required

Maximum volumetric flow rate

Top   1.33 m 2
un

Maximum volumetric flow rate

Bottom   1.75 m 2
un

As first trial take downcomer area as 12% of total.

Liquid is transferred from plate to plate through vertical channels called downcomers.

Column cross sectional area

Net area
Top   1.51 m 2
0.8

Net area
Bottom   1.99 m 2
0.8

Column diameter

Column cross sec tion area  4

Top   1.39 m


Column cross sec tion area  4

Bottom   1.59 m

 8-17

8.3.6 Liquid Flow Pattern

Liquid flow pattern can be calculated using same diameter above and below feed,
reducing the perforated area for plates above the feed.

Ln
Maximum volumetric liquid rate   0.01 m 3 / s
L

The plate diameter is outside the range of Coulson and Richardson (2008), Figure
11.35, but it is clear that a single-pass plate can be used.

8.3.7 Provisional Plate Design

Column diameter, Dc = 1.592 m

Column area, Ac = 1.991 m2

Downcomer area, Ad = 0.12 x Ac = 0.239 m2

Net area, An = Ac – Ad = 1.752 m2

Active area, Aa = Ac – 2Ad = 1.513 m2

Hole area, Ah = 0.1 x Aa = 0.106 m2 (take 7% as first trial)

From Figure 11.39, Coulson and Richardson (2008),

Iw
 0.76
Dc

Weir length, Iw = 0.76 x Dc = 1.210 m

Take weir height, hw = 50 mm (normally 40 – 90 mm)

Hole diameter = 6 mm (twice the plate thickness)
Plate thickness = 3 mm (stainless steel)
 8-18

8.3.8 Check Weeping

Maximum liquid rate, Lw = 13.00 kg/s

Minimum liquid rate at 70% turn-down = 0.7 x 13.00 = 9.10 kg/s

From Coulson and Richardson (2008),

2/3
 Lw 
how  750   
 L  Iw 

Maximum weir crest, how = 39.7 mm liquid

Minimum weir crest, how = 31.3 mm liquid

At minimum rate, hw + how = 50 + 31.3 = 81.31 mm liquid

From Coulson and Richardson (2008), Figure 11.37:

K2 @ hw + how= 30.8

uˆ h (min) 
K 2  0.9025.4  d h   6.6 m/ s
 v 1 / 2

min vapor rate 0.7  1.16

Actual min . vapor velocity  
Ah 0.106

Actual min . vapor velocity  7.6 m / s

Ua > Ŭh (min), therefore minimum operating rate will be well above weep point. The
vapor velocity at the weep point is the minimum value for stable operation.
 8-19

4.2.9 Plate Pressure Drop

Dry plate drop

Maximum vapor velocity through holes

uˆ h max 
max . volumetric flow rate
 10.9 m / s
Ah

From Coulson and Richardson (2008), Figure 11.42, for plate thickness/hole
diameter=0.5.

Ah Ah
Perforated area, 
A p Aa

Ah A
Percent perforated area,  100%  h  100%
Ap Aa

Percent perforated area = 7%

Orifice coefficient, Co = 0.71

Dry plate drop, hd

2
u  
hd  51 h  v  55.6 mm liquid
 Co   L

Residual head, hr

12.5  103
hr   14.2 mm liquid
L

Total plate pressure drop, ht

ht  hd  hw  how   hr  159 mm liquid

 8-20

8.3.10 Downcomer Liquid Back-up

Downcomer pressure loss

Take

hap  hw  10  50  10  40 mm

Clearance area under the downcomer, Aap

Aap  hap  I w  0.048 mm2

As this is less than Ad, use Aap in equation 11.60 Coulson and Richardson (2008).

Head loss in downcomer, hdc

2
 L 
hdc  166 wd   15.41 mm
  L Aap 

Back-up in downcomer, hb

hb  hw  how   ht  hdc  264 .61 mm

By checking the criterion,

hb < 1/2(plate spacing + weir height)

264.61 < 275.00

Plate spacing is acceptable.

 8-21

8.3.11 Check Residence Time

Residence time in downcomer, tr

Ad hbc  L
tr   4.29 s
Lwd

Greater than 3 s, satisfactory.

8.3.12 Check Entrainment

Actual velocity based on net area

u max
un   0.66 m / s
An

un
Percent flooding   100%  85%
uf

From Coulson and Richardson (2008),

When FLV,bottom = 0.19, so from Figure 11.36, ψ = 0.014, well below 0.1.

8.3.13 Trial Layout

Use cartridge-type construction. Allow 50 mm unperforated strip round plate edge; 50

mm wide calming zones.
 8-22

8.3.14 Perforated Area

From Coulson and Richardson (2008),

At Iw/Dc = 1.210/1.592 = 0.76, θc = 99°

Angle subtended by the edge of the plate = 180 – 99 = 81°

Mean length, unperforated edge strips = (1.592 – 50 x 10-3)π x 81/180 = 2.81 m

Area of unperforated edge strips = 50 x 10-3 x 2.18 = 0.109 m2

Mean length of calming zone, approx = 1.210 + 50 x 10-3 = 1.260 m

Area of calming zone = 2(1.260 x 50 x 10-3) = 0.1260 m2

Total area for perforations, Ap = 1.513 – 0.109 – 0.126 = 1.278 m2

Ah 0.106
  0.083
Ap 1.278

From Coulson and Richardson (2008), Figure 11.41:

Ip
 3.15
dh

2.85; satisfactory, within 2.5 to 4.0.

8.3.15 Number of Holes

d2
Area of one hole    2.83  10 5 m 2
4

Ah
Number of holes   3747 holes
2.83  10 5
 8-23

8.3.16 Column Height

Follow the plate spacing and the plate number. The column height will be calculated
based on equation given below. The equation determined the height of the column
without taking the skirt or any support as consideration. It’s determined based on
condition in the column.

Number of stages, N = 60

Plate thickness = 0.003 m (for stainless steel)

Tray spacing = 0.5 m

Column height  ( No stage  1)(tray spacing)  2(tray spacing)  ( No stage  1)( plate thickness)

Column height  (60  1)(0.5)  (2  0.5)  (60  1)(0.003)  30.7 m

The overall height from the calculation is 30.7 m, but in a real construction it will add
slightly more because of vapor and liquid area at top and bottom column. The space for
vapor and liquid are required if uncertain condition occur in the column, such as over
flooding, over vapor pressure or upset in reaction situation.
 8-24

8.4 MECHANICAL DESIGN OF PALMITIC ACID DISTILLATION COLUMN

In the mechanical design, the temperature and pressure are important properties in
evaluating the thickness and the stress of material. Therefore, the safety factor must be
added as precaution and determined by certain consideration such as corrosion factor,
location and process characteristic.

From the HYSYS simulation, the operating pressure is 1.97 atm and the safety
factor is 10% above operating pressure. Based on the simulation result, the temperature
of column operated in 141.1 °C at top of column and 340.7 °C at bottom of the column.

Stainless steel 316 is used in construction of the palmitic acid distillation column,
carbon steel is used in skirt support material and insulation material used is mineral
wool. All the selection named above is based on chemical and mechanical design.

8.4.1 Column Design

For safety reason, take design pressure 10% above the operating pressure.

101325 N m2
Design pressure, Pi  1.97atm    1.1  0.2196 N / mm 2
atm  m 2
(1000 mm) 2

For safety reason, take design temperature 10% above the operating temperature.

Design temperature, Ti  340 .66  1.1  374 .73 o C

8.4.2 Material Construction

The material of construction used is Stainless steel (18Cr/8Ni/Mo-2.5%, 316). For this
material, the design stress at 374.73 °C is obtained from Table 13.2, Coulson and
Richardson (2008).

Design stress, f = 105 N/mm2

Tensile strength = 520 N/mm2
Vessel diameter, Di = 1592 mm
 8-25

8.4.3 Vessel Thickness

The minimum thickness of column and other design are calculated based on equation
below:

Pi Di 0.2196  1592
e   1.67 mm
2 f  Pi 2(105)  0.2196

Add corrosion allowance = 4 mm

Thickness of the column = 1.67 mm + 4 mm

= 5.67 mm
≈ 6 mm

As a first trial, divide the column into five sections, with the thickness increasing by 2 mm
per section. Try 6, 8, 10, 12 and 14 mm. The average wall thickness is 10 mm.

8.4.4 Heads and Closures

Standard torispherical heads are the most commonly used end closure for vessels up to
operating pressure of 15 bar. It can be used for higher pressure, but above 10 bar titcost
should be compared with that of an equivalent ellipsoidal head. Above 15 bar an
ellipsoidal will usually prove to be the most economical closure to use. Hemispherical
heads are used for high pressure.

As the cost of construction material is directly proportional to the thickness of the

pressure vessel. The minimum the thickness the cost will be reduced. Since the
operating pressure is 1.97 atm, torispherical and ellipsoidal head are used as choice for
comparison.
 8-26

8.4.4.1 Calculation for Torispherical Dome Head

Crown radius, Rc = Dc = 1.592 m

Knuckle radius, Rk = 6%Rc = 0.0955 m

A head of this size would be formed by pressing: no joints, so J = 1.

1  Rc  1 
   3  1.592   1.7706
Cs  3
4  Rk  4 0.0955 
 

Therefore, minimum thickness, e:

Pi Rc C s 0.2196  1592  1.7706

e 
2 fJ  Pi (C s  0.2) (2  105  1)  0.2196(1.7706  0.2)

e  2.94 mm

4.3.4.2 Calculation for ellipsoidal dome head

By taking J = 1 (no joints), minimum thickness, e:

Pi Di 0.2196  1592
e 
2 fJ  0.2 Pi (2  105  1)  (0.2  0.2196)

e  1.67 mm

So, an ellipsoidal head would probably be the most economical. Take as same thickness
as wall, 6 mm.
 8-27

8.4.5 Design of Column Subject to Combined Loading

8.4.5.1 Dead Weight of Vessel, Wv

For a steel vessel, below equation is used:

Wv  240 C w Dm ( H v  0.8Dm )t

Where
t = average wall thickness, mm = 10 mm
-3
Dm = mean diameter (Di + t x 10 ) m = 1.602 m
Cw = a factor; 1.15 for distillation columns

Dead weight of vessel, W v

Wv  240  1.15  1.602  30.7  0.81.602 10

Wv  141335 N  141.33 kN

8.4.5.2 Weight of Plates, Wp

From Nelson Guide, in Coulson and Richardson (2008), it takes contacting plates, 1.2
kN/m2. This estimation based on the empirical value done by Nelson.

Plate area, Ac

(1.592 ) 2
Ac    1.99 m 2
4

Weight of plate = 1.2 x 1.99 = 2.39 kN

Weight of 60 plates, W p = 2.39 x 60 = 143.38 kN
 8-28

8.4.5.3 Weight of Insulation, Wi

Insulation material is important related with the heat transfer. Assume insulation material
is mineral wool with 75 mm thick.

Density of mineral wool = 130 kg/m3

Volume of insulation = π x Dm x Hv x thickness of insulation

= π x 1.602 x 30.7 x 0.075
= 11.6 m3

Weight of insulation, W i = Volume of insulation x ρwool x g

= 11.6 x 130 x 9.81
= 14770.18 N
= 14.77 kN

Double this value to allow fittings, so W i = 29.54 kN.

8.4.5.4 Total Weight, WT

Therefore, total weight of vessel

WT  Wv  W p  Wi

WT  141.33  143.38  29.54  314.25 kN

 8-29

8.4.5.5 Wind Loads

The wind loads is calculated based on location and the weather of surrounding.
Therefore, the value of wind speed assumes as below and wind load is calculated.

Wind speed, Uw = 160 km/h

For a smooth cylindrical column stack, the following semi-empirical equation can be
used to estimate wind pressure, Pw.

Pw  0.05u w  0.05  160  1280 N / m2

2

Effective column diameter, Deff = Dc + 2(tshell + tinsulation)

= 1.592 + [2 x (10 + 75) x 10-3]
= 1.76 m

Loading per unit length of column, Fw = Pw x Deff

= 1280 x 1.76
= 2256 N/m

8.4.5.6 Bending Moment

Mx with X = Distance measure from free end ( 30.7 m)

Fw ( X ) 2 2256  30.7 2
Mx  
2 2

M x  1061425.88 N

M x  1061.43 kN
 8-30

8.4.5.7 Analysis of Stress

Stress from bottom tangent line ( take t = 14 mm, which is a thickness at the bottom).

Longitudinal pressure stress, σh

Pi Deff 0.2196  1.76  1000

h    13 .82 N / mm 2
2t 2  14

Circumferential pressure stress, σL

Pi Deff 0.2196  1.76  1000

L    6.91 N / mm 2
4t 4  14

Dead weight stress, σw

W 141.33  1000
w  
 Dc  t t  1592  14  14

 w  2.00 N / mm 2

Bending stress, σb

Do = 1592 + 2(14) = 1620 mm

Di = 1592 mm

Second moment of area, Iv

Iv 

  Do 4  Di 4     1620  15924
4

64 64

I v  2.28  1010 mm 4
 8-31

Therefore, bending stress, σb

 M  D   1061425.88  1592 
 b   x  c  t    10 
 14   0.6892 N / mm 2
 I v  2   2.28  10  2 

The resulted longitudinal stress, σz is

σw is compressive and therefore negative.

σz (upwind) = σL – σw + σb = 6.91 - 2.00 + 0.6892 = 5.5986 N/mm2

σz (downwind) = σL – σw – σb = 6.91 - 2.00 - 0.6892 = 4.2202 N/mm2

As there is no torsional shear stress, the principle stresses will be σz and σh. The radial
stress is negligible.

The greatest difference between the principle stresses will be on the down-wind side.

13 .82  (4.2202 )  18 .04 N / mm 2

well below the maximum allowable design stress.

Check elastic stability

Critical bulking stress, σc

 t 
  2  10 4 
14 
 c  2  10 4  
 Dm   1602 

 c  174.75 N / mm 2

Maximum compressive stress will occur while the vessel not under pressure

  w   b  2.00  0.6892

 2.690 N / mm 2

Well below the critical bulking stress, so design is acceptable.

 8-32

8.4.6 Design of Vessel Support (Skirt Design)

Type of support : Straight cylindrical skirt

θs : 90°
Material construction : Carbon steel
Skirt height : 10 m
Design stress, fs : 135 N/mm2 at ambient temp. 20°C
Young modulus, E : 200000 N/mm2

At ambient temperature, the maximum dead weight load on the skirt will occur when the
vessel is full of the mixture.

Approximate weight, Wapp

Dc 2 H v  L g   1.592 2  10  1000  9.81

Wapp  
4 4

Wapp  194241.84 N

Wapp  194.24 kN

Weight of vessel = 312.94 kN

Therefore, total weight = 194.24 + 312.94 = 507.18 kN

Wind load, Fw = 2256 N/m = 2.256 kN/m

Bending moment at skirt base, Ms

  
M s  Fw H v  H skirt  / 2  2.256  30 .7  10 2 / 2
2
 
M s  1861 .42 kNm
 8-33

8.4.6.1 Bending Stress In Skirt, σbs

Take the skirt thickness as the same as that of the bottom section of the vessel, 14 mm.

Bending stress in skirt, σbs

4M s 4  1861.42  10 6
 bs  
 ( Ds  t s )t s Ds  1592  1414  1592

 bs  66.56 N / mm 2

8.4.6.2 Dead Weight Stress In The Skirt, σws

For test,

W 507.18  1000
 ws  
 ( Ds  t s )t s  1592  1414

 ws  7.20 N / mm 2

For operating,

Wapp 507.18  1000

 ws  
 ( Ds  t s )t s  1592  1414

 ws  4.44 N / mm 2

Thus, the resulting stress in the skirt, σs

Maximum σs (compressive) = σbs + σws (test)

= 66.56 + 7.20
= 73.76 N/mm2

Maximum σs (tensile) = σbs - σws (operating)

= 66.56 - 4.44
= 62.12 N/mm2
 8-34

8.4.7 General Consideration for Design

Take the joint factor, J as 0.85.

Criteria for design:

σs (tensile) < fs J sinθ

62.12 < 0.85 x 135 x sin 90°
62.12 < 102.6

σs (compressive) < 0.125EY(tsk/Ds)sinθ

73.76 < 0.125 x 200000 x (17/3968) sin 90°
73.76 < 197.1

Both criteria are satisfied; adding 2 mm for corrosion gives a design thickness of 16 mm.

8.4.8 Base Rings and Anchor Bolts

Assume pitch circle diameter = 2.092 m

Circumference of bolt circle = 2092π
Bolt stress design, fb = 123 N/mm2
Recommended spacing between bolts = 600 mm
Bending moment at base skirt, Ms = 1861.42 kNm

Total weight of vessel, Wt = 507.18 kN

Minimum number bolt required, Nb ; Closest multiple of 4,

2092
Nb   10.96  8
600
 8-35

Area of bolt, Ab

1  4M s  1  4  1241.48  1000 
Ab   W     363.09
N b f b  Db  8  123  2.092 

Ab  3108.86 mm 2

Bolt root diameter, d

Ab  4 3108.86  4
d 
 

d  62.92 mm

Total compressive load on the base ring per unit length,

 4M s
Fb  
  W
  
  4  1861.42  103   507.18  103 
  
  D    
 Ds   1.592 2     1.592 
2
  s  

Fb  1041767 N / m

8.4.8.1 Base Ring Width

Taking the bearing pressure, fc as 5 N/mm2.

Minimum width of the base ring, Lb

 F   1   1041767   1 
Lb   b    3     3 
 f c   10   5   10 

Lb  208.35 mm

Actual width can be calculated from this minimum width.

 8-36

Use M64 bolts (BS 4190:1967) root area = 2680 mm, Figure 13.30 Chemical
Engineering Volume6.

A = 83, B = 152, C = 102, D = 25, E = 50, F = 70, G = 76

Actual width required

Lb  L r t s  50  152  14  50

Lb  216 mm

Actual bearing pressure on concrete foundation

Fb 1041767
fc  
actual width 194

f c  4.823 N / mm 2

8.4.8.2 Base Ring Thickness

Actual minimum base thickness, tb

0.5
3f 
t b  Lr  c 
 fr 

Where fc = actual bearing pressure on base, N/mm2

fr = allowable design stress in the ring material, typically 140 N/mm2

0.5
3f   3  4.823 
t b  Lr  c   152 
 fr   140 

t b  48.87 mm
 8-37

8.4.9 Design of Piping

In order to connect the distillation column with pipeline, flange need to be installed to
ensure two compartments is well connected. It needs to consider the pressure and the
optimum duct diameter. Figure below shows the parameter in selecting the suitable
flange. Figure 8.4 below shows the parameters in selecting the suitable flange.

Figure 8.4: Steel welding neck flanges at 40 bar.

 8-38

Table 8.10 shows parameters needed for piping calculation

Table 8.10: Flow rate and density for each pipe inlet and outlet

Pipe no Location G (kg/s) ρ (kg/m3)

N1 Feed 12.39 672.40
N2 Top outlet 16.36 3.80
N3 Top inlet 12.27 749.60
N4 Bottom outlet 31.83 639.50
N5 Bottom inlet 23.53 4.24

Feed pipe sizing (liquid) N1

For stainless steel pipe, optimum pipe diameter, Dopt

Dopt  260G 0.52  0.37

Therefore,

Dopt  260G 0.52  0.37  260  12.39 0.52  672.40 0.37

Dopt  87 mm

From Figure 8.4, nominal pipe size = 100 mm.

Nozzle thickness, tn

Ps = Operating pressure = 1.97 atm = 0.22 N/mm2

σ = Design stress at working temperature = 105 N/mm2

Ps Dopt 0.22  87
tn    0.01 mm
20  Ps  20 105  0.22

So, thickness of nozzle = corrosion allowance + tn

= 4 + 0.01
= 4.01 mm
≈ 5.00 mm
 8-39

Product nozzle sizing (gas) N2

Optimum diameter, Dopt

Dopt  260G 0.52  0.37  260  16.36 0.52  3.80 0.37

Dopt  679 mm

From Figure 8.4, nominal pipe size = 700 mm.

Nozzle thickness, tn

Ps = Operating pressure = 1.97 atm = 0.22 N/mm2

σ = Design stress at working temperature = 105 N/mm2

Ps Dopt 0.22  679

tn    0.07 mm
20  Ps  20  105  0.22

So, thickness of nozzle = corrosion allowance + tn

= 4 + 0.07
= 4.07 mm
≈ 5.00 mm
 8-40

Reflux nozzle sizing (liquid) N3

Optimum diameter, Dopt

Dopt  260G 0.52  0.37  260  12.27 0.52  749.60 0.37

Dopt  83 mm

From Figure 8.4, nominal pipe size = 100 mm.

Nozzle thickness, tn

Ps = Operating pressure = 1.97 atm = 0.22 N/mm2

σ = Design stress at working temperature = 105 N/mm2

Ps Dopt 0.22  83
tn    0.01 mm
20  Ps  20 105  0.22

So, thickness of nozzle = corrosion allowance + tn

= 4 + 0.01
= 4.01 mm
≈ 5.00 mm
 8-41

Product nozzle sizing (liquid) N4

Optimum diameter, Dopt

Dopt  260G 0.52  0.37  260  31.830.52  639.50 0.37

Dopt  144 mm

From Figure 8.4, nominal pipe size = 150 mm.

Nozzle thickness, tn

Ps = Operating pressure = 1.97 atm = 0.22 N/mm2

σ = Design stress at working temperature = 105 N/mm2

Ps Dopt 0.22  144

tn    0.02 mm
20  Ps  20  105  0.22

So, thickness of nozzle = corrosion allowance + tn

= 4 + 0.02
= 4.02 mm
≈ 5.00 mm
 8-42

Reboiler nozzle sizing (gas) N5

Optimum diameter, Dopt

Dopt  260G 0.52  0.37  260  23.530.52  4.24 0.37

Dopt  787 mm

From Figure 8.4, nominal pipe size = 800 mm.

Nozzle thickness, tn

Ps = Operating pressure = 1.97 atm = 0.22 N/mm2

σ = Design stress at working temperature = 105 N/mm2

Ps Dopt 0.22  787

tn    0.08 mm
20  Ps  20  105  0.22

So, thickness of nozzle = corrosion allowance + tn

= 4 + 0.08
= 4.08 mm
≈ 5.00 mm

Table 8.11 below shows the summary of piping design.

Table 8.11: Summary of nominal pipe size and nozzle thickness

Nom. Size with corrosion

Pipe no Description Dopt (mm) tn (mm)
(mm) allowance (mm)
N1 Feed 87 100 0.01 4.01
N2 Top outlet 679 700 0.07 4.07
N3 Top inlet 83 100 0.01 4.01
N4 Bottom outlet 144 150 0.02 4.02
N5 Bottom inlet 787 800 0.08 4.08
 8-43

8.5 EQUIPMENT SPECIFICATION SHEET

EQUIPMENT SPECIFICATION SHEET

Identification:
Item: Palmitic Acid Distillation Column
Item code: T-100 Date: 14th May 2011
No. required: 1 By: NURSYUHADA RAMLE
Function: To separate Palmitic Acid from multicomponent mixture
Operation: Continuous
Specification Data Design Sizing Data
Design type Vertical Domed head wall thickness 6.00 mm
Material of construction Stainless steel Tray spacing 0.50 m
Material of insulation Mineral wool Column diameter 1.59 m
Column type Sieve plate Column wall thickness 10.00 mm
Column head type Ellipsoidal Total column height 30.70 m
Vessel thickness 1.66 mm
Operating Condition Data Downcomer Liquid Backup Data
Feed temperature 240 °C Height of the bottom edge
Top temperature 141 °C of the apron above plate
Bottom temperature 341 °C (apron clearance), hap 40.00 mm
Feed pressure 3.3 bar Height of liquid crest over
Top pressure 0.7 bar downcomer weir, how 31.33 mm
Bottom pressure 0.8 bar Weir height, hw 50.00 mm

Number of tray 60 Downcomer back-up, hb 264.67 mm

Feed tray 30
Mechanical Design Data Vessel Support Data
Dead weight 140.92 kN Type of support Skirt
Plates weight 142.56 kN Skirt height 10.00 m
Insulation weight 29.46 kN Material of construction Carbon Steel
Total vessel weight 312.94 kN Bolt size M64
Wind load 2.25 kN/m Root area 2680.00 mm
Bending moment (base) 1058.70 kNm Base ring thickness 39.99 mm
Pressure stress, σL 6.89 N/mm2 Skirt thickness 14.00 mm
 8-44

Pressure stress, σh 13.78 N/mm2

Dead weight stress 2.00 N/mm2
Bending stress ± 0.69 N/mm2
Longitudinal stress:
σz (upwind) 5.59 N/mm2

σz (downwind) 4.20 N/mm2

Elastic stability:
Critical bulk stress 175.24 N/mm2
Maximum compress stress 2.69 N/mm2
Plate Design Data
Plate diameter 1.59 m
Weir length, Iw 1.21 m
Wide calming zone 50.00 mm
Unperforated strip edge 50.00 mm
Hole diameter 6.00 mm
θ 99°
Plate material Stainless Steel
Downcomer material Stainless Steel
Plate thickness 3.00 mm
No. of holes 3726 holes