Prof. K. V.

Rao
Programme Director
Petroleum Courses
JNTUK
Why should you know process modeling and simulation of
distillation columns?
In recent years the use of commercial process
simulation packages for distillation column design has
increased dramatically.

Process simulation packages have made performing
design calculations for distillation column significantly
easier.

To arrive at a viable design, the design engineer should
have basic understanding of column modeling and
simulation procedures.
What do you expect from this presentation?
Modeling and simulation based on nesting of equations
of Thiele Geddes method.

Modeling and simulation based on the method of
Holland.
THIELE-GEDDES METHOD
Figure1: Configuration of a
conventional distillation
column
with liquid and vapor streams
including product streams for
the programs
T. G. method is a simulation method
to predict Product distribution for a
given multi-component Distillation
column.
1. Specify the following variables
i. Number of components in the liquid mixture for
separation, nc
ii. Feed rate, F: composition of each component, x
i,F
,
saturated liquid feed Q=1
iii. Product recovery, D/F or B/F;
iv. Number of stages in rectifying section, n
r
and in
stripping section n
s
. Feed plate location, nt= ns+nr+2
v. Reflux ratio, R. The reflux is assumed to be saturated
liquid.
vi. Liquid and vapor flow rates in the rectifying and
stripping section are assumed (RL and RV in rectifying
section, RL=D*R; RV=L+D; SL & SV in SL=RL+Q*F ;
SV= RV-(1-Q) *F )
vii. Type of condenser :total
viii. Top pressure of the column and allowable pressure
drop per plate,
ix. K values as function of T for all the components at the
given pressure
2. Assume a linear temperature and pressure profile of the
column.
i. Calculate bubble point and dew point temperature for
the assumed compositions of the bottom and top
product. The temperatures may be taken as the initial
bottom temperature (Reboiler temperature or Nth plate
temperature) and top temperatures. Alternatively, if the
key components are significant (Heavy key in the
bottom product and light key component in the top
product),the boiling temperatures of these components
at the given pressure may be taken as the initial
bottom (T
bott
) and top temperature(T
top
).

ii. Calculate T
T= (T
bot
-T
top
) /(np-1)

iii. Calculate each plate temperature in the column,
T
j
= T
j-1
+T (j=2,np)

iv. Calculate the pressure of each plate for given
condenser pressure, top plate pressure and P
P
j
= P
j-1
+P (j=2,np)
3. Calculate the values of equilibrium constant, K
i,j
for each
component on each plate.(If K
i
is function of both
temperature and pressure, then correlation for each
component may be used)

K
i,j
= exp[ A
i
B
i
/ (T
j
+ C
j
) ] /P
j
(j=1,np ; i=1,nc)

4. Initialize the ratios of the compositions y
i,1
/x
i,D
and y
i,N
/x
i,B


(i) y
i,1
/ x
i,D
= 1 ( i=1,nc)

(ii) y
i,N+1
/ x
i,B
= K
i,N+1
( i=1,nc)

5. Calculate the values of the ratios of the components for
each plate in the rectifying section.

(i) x
i,j
/x
i,D
= (y
i,j
/x
i,D
) / K
i,j
( j=1,nr ; I=1, nc)

(ii) y
i,j+1
/ x
i,D
= (RL* (x
i,j
/ x
i,D
) + D) / RV
(j=1,nr ; i=1to nc)
6. Calculate the ratios of compositions, x
i,j
/ x
i,B
and y
i,j
/ x
i,B
for
each component on each plate in the stripping section.

(i) x
i,(np+1-j)
/ x
i,B
= (SV* y
i
,
(np+1-j+1)
/ x
i,B
+ B) /SL (j= 1,ns+1;
i=1,nc)

(ii) y
i,(np+1-j)
/x
i,B
= K
i,(np+1-j)
* x
i,(np+1-j)
/ x
i,B
(j= 1,ns+1 ; i=1,nc)

7. Calculation of the ratio compositions, (x
i,B
/x
i,D
)

x
i,B
/x
i,D
= (y
i,NF
/ x
i,D
) /( y
i,NF
/ x
i,B
) (i=1,nc)

8. Calculate the top product composition x
i,D
from the overall
component balance of the column and then x
i,B
.

x
i,D
= F.x
i,F
/ (D+B (x
i,B
/ x
i,D
)),
x
i,B
= (x
i,B
/ x
i,D
) x
i,D

9. Normalize the product compositions, x
i,D
and x
i,B

x
i,D,cal
=x
i,D
/ x
i,D

x
i,B,cal
= x
i,B
/ x
i,B

10. Obtain the liquid compositions of all the components on
each plate .

(i) Rectifying section
x
i,j
= (x
i,j
/x
i,D
) x
i,d,cal
(j=1,nr ; i=1,nc)
(ii) Stripping section
x
i,np-j
= (x
i, np-j
/x
i,B
) x
i,B,cal
(j=1,ns+1; i=1,nc)

11. Normalization of the liquid compositions on each plate in the
column
x
i,j
/ x
i,j
(j=1,np; i=1,nc)
12. Correct the temperature profile using bubble method.
i. K
i,j
= exp[ A
i
B
i
/ (T
j
+ C
j
) ] /P
j
(j=1,np ; i=1,nc)
ii. Calculate the y
i,j
=K
i,j
* x
i,j
(j=1,np; i=1,nc)
iii. Calculate sum of y
ij
(j=1,np; i=1,nc)
iv. F = sumy-1
v. If Absolute value of F is less than a error tolerance of
10
-4
or 10
-6
, T[j] is corrected by using Newtons
method.
vi. T[j] = T[j] F/DF. Where DF is derivative of F in
Temperature.
vii. Then the calculation is repeated from (i) to (vi) until
the condition in (v) is satisfied.
13. Change the iteration number.
iter =iter+1;

14. Repeat the calculations from the step3. No convergence
procedure used except for 20 or more iterations. After
convergence, plate compositions and temperature are
obtained along with the values of x
i,D
and x
i,B
.

Should it be necessary to study the effect of variables like
reflux ratio, feed plate location, feed composition on product
profile, one variable at a time, may be changed and the
results may be obtained. Thus, several simulations will lead
to decide upon the best possible separation.

Hollands Method
Tridiagonal Matrix for material balance in terms of
component vapor flow rates and absorption factors.
Thomas Algorithm to solve the tridiagonal matrix.
KB method for correcting temperature profile.
method of convergence.
The Method of Convergence
To achieve fast convergence of the problems dealing with
the simulation of distillation columns, the method
combined with K
b
method was proposed by Holland .

In the iterative procedure, the improved sets of liquid
mole fractions on each plate in a column required for the
calculation of a new temperature profile are obtained
using the corrected product compositions (top and
bottom).

These product compositions should satisfy two
conditions. i) the overall component material balance and
ii. the criterion of the sum of the top product
compositions, which is equal to unity.
Thus for all the components,
Fx
i,F
= D(x
i,D
)
cor
+ B(x
i,B
)
cor

x
i,D
=1

The corrected values of (x
i,D
/x
i,B
)
cor
and the calculated
values of (x
i,D
/x
i,B
)
cal
are related by a multiplier
( x
i,B
/ x
i,D
)
cor
= ( x
i,B
/ x
i,D
)
cal

The value of becomes unity when the convergence is
reached.
The K
B
method
The new temperature profile for the next iteration is
calculated on the basis of the corrected liquid mole
fractions and the temperature profile of the previous
iteration by using the K
B
method.

This method is considered as a modified bubble point
method. For any plate, it is applied as







Where,
i,j =
K
i,j
/ K
j,b
, the relative volatility of component i
at the temperature of plate j , and K
j,b
is the K value of
base component ( in the calculations , heavy key
component is considered as the base component).

As the value of K
j,b
is function of temperature, the new
temperature of each plate can be calculated using K
equation rewritten for T.
Tridiagonal Matrix Equation
A set of variables-the component-flow rates in the vapor
and liquid phases-are introduced, namely,



Also, the flow rates of component i in the distillate and
bottoms are represented by



and the flow rates of component i in the vapor and liquid
parts of the feed by


ji j ji
y V v
ji j ji
x L l

and
Di i
x D d
Bi i
x B b

and
Fi F Fi
y V v
Fi F Fi
x L l

and
The equilibrium relationship may be restated in
an equivalent form in terms of the component-flow rates
and as follows. Then, expression may be
restated in the form

ji ji ji
x K y
ji
v
ji
l
ji ji ji
x K y
Where the absorption factor and the stripping are
defined as follows
ji
A
ji
S
j
ji
ji
j
ji
L
l
K
V
v

ji ji ji
l S v
ji ji ji
v A l and
=
=
ji
A
ji
S
1
j ji
j
V K
L
An equivalent set of component-material balances is
obtained by enclosing each stage (j=1, 2, .., N, N+1) by
a component-material balance.

Corresponding set of material balance for each
component i are as follows
0
1 0

i i i
v d l
0
2 1 1 0

i i i i
v l v l
) 2 ,......., 4 , 3 , 2 ( , 0
, 1 , 1


f j v l v l
i j ji ji i j
Fi fi i f i f i f
v v l v l
, 1 , 1 , 2
Continued in next slide
Fi i f fi fi i f
l v l v l
, 1 , 1
) ,...., 2 , 1 ( , 0
, 1 , 1
N f f j v l v l
i j ji ji i j


0
1 ,

i i N i N
b v l
The s may be eliminated by use of the equilibrium
relationship. For the case of a total condenser, and
have the same composition, and thus
ji
l
i
l
0 i
d
i i
d
D
L
l ) (
0
0

For a partial condenser, and hence
may be restated as follows
Di i
x y
0
i i i
x K y
0 0 0

i
i
Di
x L
L
DK
Dx
0 0
0
0
) (
or
i i i
d A l
0 0

D K
L
A
i
i
0
0
0
Where
Continued in next slide
The expression given above may be used to represent
both a partial condenser and a total condenser, provided
that A
0i
is set equal to for a total condenser.

Also, the form of A
N+1i
differs slightly from that for A
ji

because of the double representation of the reboiler by
the subscripts N+1 and B.

Thus, the equilibrium relationship may
be restated in the form
D
L
0
i N i N i N
x K y
1 1 1

Bi
N i N
i N N
Bx
B
V K
y V ) (
1 1
1 1



i N i N i
v A b
1 1

1 1
1

N i N
i N
V K
B
A
or
Where
When the s and are eliminated from the above
MB equations, the following result is obtained
ji
l
i
b
0 ) 1 (
1 0

i i i
v d A
0 ) 1 (
2 1 1 0

i i i i i
v v A d A
) 2 ,......, 3 , 2 ( , 0 ) 1 (
, 1 , 1 , 1


f j v v A v A
i j ji ji i j i j
Fi fi i f i f i f i f
v v v A v A
, 1 , 1 , 2 , 2
) 1 (
Fi i f fi fi i f i f
l v v A v A
, 1 , 1 , 1
) 1 (
) ,......, 2 , 1 ( 0 ) 1 (
, 1 , 1 , 1
N f f j v v A v A
i j ji ji i j i j


0 ) 1 (
1 1 , ,

i N i N i N i N
v A v A
(23)

This set of equations may be stated in the matrix form
i i i
f v A
Where
Continued in next slide
T
i n ni fi i f i i i i
v v v v v v d v ] ... ... [
1 1 2 1

T
Fi Fi i
l v f ] 00 ... 0 0 ... 00 [
) 1 (
ji ji
A
The procedure proposed by Boston and Sullivan for
Thomas algorithm is used here in the following form
Continued in next slide
Again after the and have been computed, the
values of flow rates, are computed
by use of the following equations for all components on
each plate



After the recurrence formulas have been applied for each
component i and the set of component vapor rates
have been found, the corresponding set of liquid
rates
are then calculated.

These sets of calculated flow rates are used in
conjunction with the method of the convergence and
the K
b
method in the determination of an improved set of
temperatures.
s f ' s g'
1 2 1
, ,......, ,
i i iN iN
v v v v

Thiele-Geddes Method
Example 1: Separation of Phenol-Cresol mixture
Feed rate, F: 100 moles/hr; Feed composition, x
iF
: Phenol 0.35;
o-cresol 0.15; m-cresol 0.30; xylenol 0.20
Feed enthalpy, Q: 1; Distillate, D: 33.1;Reflux, R: 10.0;
Theoretical plates in rectifying section, NR: 16
Theoretical plates in stripping section, NS: 8; Pressure drop
per plate = 4mmHg
Pressure in the condenser = 150mmHg; Total Condenser and
Partial Reboiler
INPUT DATA FILE FOR EXAMPLE 1
PHENOL-CRESOL-SEPARATION
100 1 33.1 10.0 4 16 8 410 426 150 4 0 1 1 1
0.35 16.4279 3490.89 -98.59
0.15 15.9148 3305.37 -108.0
0.30 17.2878 4274.42 -74.09
0.20 16.2424 3724.58 -102.4
Output Data File
PHENOL-CRESOL-SEPARATION
number of components= 4
Iterations=21
TOP AND BOTTOM PRODUCT COMPOSITIONS
-----------------------------------------------
XD XB
-----------------------------------------------
0.955389 0.050273
0.044574 0.202196
0.000037 0.448511
0.000000 0.299020
---------------------------------------------
------------------------------------------------------------------
plate liquid compositons
no T X1 X2 X3 X4
------------------------------------------------------------------
1 404.68 0.943438 0.056491 0.000071 0.000000
2 405.47 0.929879 0.069991 0.000130 0.000000
3 406.26 0.914587 0.085182 0.000232 0.000000
4 407.04 0.897452 0.102142 0.000405 0.000001
5 407.83 0.878386 0.120914 0.000698 0.000002
6 408.62 0.857319 0.141483 0.001193 0.000005
7 409.41 0.834203 0.163766 0.002019 0.000013
8 410.21 0.809000 0.187581 0.003387 0.000032
9 411.02 0.781665 0.212623 0.005630 0.000082
10 411.85 0.752106 0.238422 0.009266 0.000207
11 412.71 0.720131 0.264270 0.015084 0.000515
12 413.62 0.685366 0.289131 0.024239 0.001264
13 414.59 0.647136 0.311472 0.038341 0.003051
14 415.69 0.604331 0.329041 0.059424 0.007204
15 416.99 0.555308 0.338599 0.089576 0.016516
16 418.61 0.498040 0.335812 0.129817 0.036331
17 420.71 0.430971 0.315923 0.177779 0.075327
18 421.69 0.394371 0.339072 0.189805 0.076751
19 422.76 0.354032 0.360260 0.206830 0.078878
20 423.95 0.310334 0.376802 0.230557 0.082307
21 425.30 0.263991 0.385213 0.262571 0.088224
22 426.86 0.216162 0.381477 0.303462 0.098898
23 428.70 0.168544 0.361785 0.351308 0.118363
24 430.88 0.123365 0.323877 0.399796 0.152962
25 433.45 0.083167 0.268800 0.437203 0.210830
26 436.43 0.050273 0.202196 0.448511 0.299020
-------------------------------------------------------------------
EXAMPLE 2
Separation of Toluene-Ethyl benzene
Feed rate, F: 100 moles/hr
Feed composition, x
iF
: Benzene 0.022; Toluene 0.074; Ethyl
Benzene 0.434; Styrene 0.470
Feed enthalpy, Q: 1;Distillate, D: 5.58;Reflux, R: 4.36;Theoretical
plates in rectifying section, NR: 5
Theoretical plates in stripping section, NS: 3;Pressure drop per
plate = 3mmHg
Pressure in the condenser = 160mmHg; Total Condenser and
Partial Reboiler
INPUT DATA FILE FOR EXAMPLE 2
SEPERATION-OF-TOLUENE_ETHYL-BENZENE
100 1 5.58 4.36 4 5 3 330 360 160 3 0 1 3 1
0.022 15.9008 2788.51 -52.36
0.074 16.0137 3096.52 -53.67
0.434 16.0195 3279.47 -59.95
0.470 16.0193 3328.57 -63.72
Output Data File
SEPERATION-OF-TOLUENE_ETHYL-BENZENE
number of components= 4
Iterations=21
TOP AND BOTTOM PRODUCT COMPOSITIONS
----------------------------------------------------
XD XB
----------------------------------------------------
0.364715 0.001747
0.485727 0.049669
0.126900 0.452148
0.022658 0.496435
----------------------------------------------------
---------------------------------------------------------------
plate liquid compositons
no T X1 X2 X3 X4
---------------------------------------------------------------
1 336.98 0.130411 0.481448 0.310103 0.078039
2 346.73 0.045858 0.338564 0.459399 0.156179
3 352.98 0.023200 0.209843 0.527835 0.239123
4 356.78 0.017361 0.134006 0.531047 0.317586
5 359.14 0.015566 0.096276 0.498194 0.389964
6 360.76 0.014831 0.078579 0.449805 0.456785
7 361.89 0.010152 0.077442 0.453027 0.459379
8 362.98 0.006537 0.073897 0.456444 0.463123
9 364.14 0.003793 0.065885 0.458979 0.471344
10 365.66 0.001747 0.049669 0.452148 0.496435
-----------------------------------------------------------------------------------------------------------------------------
Hollands Method
Output Data File for Example 1
PHENOL-CRESOL-SEPARATION
Number of Components=4
ITERATIONS=7
THETA=0.999997
TOP AND BOTTOM PRODUCT COMPOSITIONS
----------------------------------------------
XD XB
--------------------------------------------
0.955610 0.050363
0.044354 0.202270
0.000036 0.448413
0.000000 0.298954
----------------------------------------------
------------------------------------------------------------------
plate liquid compositons
no T X1 X2 X3 X4
------------------------------------------------------------------
1 403.96 0.943704 0.056227 0.000070 0.000000
2 404.76 0.930191 0.069681 0.000127 0.000000
3 405.56 0.914948 0.084825 0.000227 0.000000
4 406.36 0.897865 0.101738 0.000397 0.000001
5 407.16 0.878850 0.120463 0.000686 0.000002
6 407.96 0.857834 0.140989 0.001173 0.000005
7 408.77 0.834768 0.163233 0.001987 0.000012
8 409.58 0.809614 0.187016 0.003338 0.000032
9 410.40 0.782325 0.212038 0.005556 0.000081
10 411.24 0.752809 0.237830 0.009158 0.000204
11 412.11 0.720874 0.263690 0.014927 0.000508
12 413.02 0.686145 0.288584 0.024021 0.001250
13 414.00 0.647945 0.310985 0.038049 0.003022
14 415.11 0.605156 0.328641 0.059052 0.007151
15 416.41 0.556122 0.338306 0.089142 0.016430
16 418.04 0.498790 0.335628 0.129365 0.036218
17 420.15 0.431574 0.315809 0.177380 0.075237
18 421.13 0.394997 0.339009 0.189341 0.076653
19 422.21 0.354670 0.360271 0.206291 0.078768
20 423.40 0.310965 0.376910 0.229946 0.082179
21 424.76 0.264590 0.385430 0.261910 0.088070
22 426.33 0.216697 0.381790 0.302800 0.098713
23 428.17 0.168982 0.362149 0.350721 0.118148
24 430.35 0.123684 0.324217 0.399362 0.152737
25 432.93 0.083361 0.269033 0.436958 0.210648
26 435.92 0.050363 0.202270 0.448413 0.298954
-------------------------------------------------------------------
Output Data File for Example-2
SEPERATION-OF-TOLUENE_ETHYL-BENZENE
Number of Components=4
ITERATIONS=12
THETA=0.999995
TOP AND BOTTOM PRODUCT COMPOSITIONS
-----------------------------
XD XB
----------------------------
0.365042 0.001727
0.486792 0.049605
0.125836 0.452212
0.022330 0.496456
-----------------------------
--------------------------------------------------------------
plate liquid compositons
no T X1 X2 X3 X4
--------------------------------------------------------------
1 336.44 0.130484 0.483426 0.308783 0.077307
2 346.19 0.045840 0.340179 0.458695 0.155286
3 352.46 0.023165 0.210704 0.527798 0.238333
4 356.27 0.017322 0.134361 0.531299 0.317018
5 358.64 0.015526 0.096394 0.498429 0.389652
6 360.28 0.014790 0.078602 0.449871 0.456737
7 361.42 0.010100 0.077454 0.453105 0.459340
8 362.51 0.006488 0.073888 0.456532 0.463091
9 363.69 0.003757 0.065848 0.459070 0.471325
10 365.20 0.001727 0.049605 0.452212 0.496456
---------------------------------------------------------------
REFERENCE
S
1. Amundson, N.R. and Pontinen, A.J. (1958) Ind. Eng. Chem. 50,
730.
2. Boston, J.F., and S.L. Sullivan, Jr., Can. J. Chem. Eng., 50,
p.663, 1972.
3. Billingsley, D.S., Am. Inst. Chem. Engrs. J., 16, p.441, 1970.
4. Brian, P.L.Thibaut, Staged Cascades in Chemical Processing.
Prentice-Hall, 1972.
5. Carnahan, B., H.A. Luther, and J.O.Wilkes: Applied Numerical
Methods. John Wiley and Sons, Inc., New York, 1964.
6. Couper, J.R. (2005), Chemical Process Equipment: Selection
and Design, 2
nd
Edition, Elsevier, Oxford, UK.
7. Deshpande, P.B. (1985) Distillation Dynamics and Control
(Arnold).
8. Friday, J.R. And B.D. Smith, Am. Inst. Chem. Engrs. J., 10, p.
689, 1964.
9. Hanson, D.N., Duffin, J.H. and Somerville, G.E. (1962)
Computation of Multistage Separation Processes (Reinhold).
Continued in next slide
10.Henley, E.J., and J.D. Seader, Equilibrium-Stage Separation
Operations in Chemical Engineering, Wiley, New York, 1980.
11.Holland, C.D. (1963) Multicomponent Distillation (Prentice-
Hall).
12.Holland, C.D., Fundamentals of Multicomponent Distillation,
(McGraw-Hill), New York, 1981.
13.Kister, H.Z., Distillation Design, McGraw-Hill, New York, 1992.
14.Kooijman, H.A and Taylor, R. (2000), The Chemsep Book,
Printed by Books on Demand, www.bod.de
15.Lewis, W.K. and Matheson, G.L., Ind. Eng. Chem., 24, 494
(1932).
16.Lyster, W.N., Sullivan, S.L. Billingsey, D.S. and Holland, C.D.
(1959) Pet. Ref.38 (June) 221 (July) 151 (Oct). 139 and 39
(Aug) 121 (in four parts). Figure distillation this way.

Continued in next slide
17. McKetta, J.J., Unit Operations Handbook, 1993, CRC Press,
Florida, USA.
18. McKetta, J.J. and Cunningham (1982), Encyclopedia of
Chemical Processing and Design, CRC Press, Florida, USA.
19. Napthali, L.M. and Sandholm, D.P. (1971) AICHE J, 17, 148.
20. Oliver, E.D. (1966) Diffusional Separation Processes (Wiley).
21. Perry R.H and Don W.Green (1999), Chemical Engineers
Handbook 6
th
edition, McGrawHill publishers.
22. Reid, R.C., Prausnitz, J.M., Sherwood, T.K.(1976) Properties
of gases an Liquids, third edition, McGraw-Hill Book Co., New
York.
23. Rousseau, R.W. (1987), Handbook of Separation Process
Technology, New York, (Wiley).
Continued in next slide
24. Smith, B.D (1963) Design of Equilibrium Stage Processes
(McGraw-Hill), New York
25. Thiele, E.W. and Geddes, R.L. (1933) Ind. Eng Chem. 25,
289.
26. Tomich, J.F., Am. Inst. Chem. Engrs. J., 16, p.229, 1970.
27. Treybal, R. (1980) Mass Transfer Operations, 3
rd
edn
(McGraw-Hill).
28. Wang. J.C., and G.E., Henke, Hydrocarb. Proc., 45(8), p. 663,
1972.
29. Van Winkle, M. (1967) Distillation (McGraw-Hill).