29

Separation of Multicomponent Mixtures By Use of Conventional

Distillation Column with Multiple Stages
Keywords: Convention Distillation Column, Theta and Zedadda method, triangular matrix,

A conventional distillation column is defined as one that has one feed and two product
streams, the distillate D and the bottoms B. consider the case in which the following
specifications are made for a column at steady state operation:

 Number of plates in each section of the column

 Quality, composition and the thermal condition of the feed
 Column pressure
 Type of overhead condenser (total or partial)
 Reflux ratio, Lo/D or V1 or Lo
 One specification on the distillate such as the total flow rate D.

For this set of conditions the problem is to find the compositions of the top and bottom
products. The set of equations required to represent such a system is as follows:

1. Material balance equations

Vj+1.yj+1,i = Ljxj,i + DXD,i (1 ≤ i ≤ C) and (0 ≤ j ≤ f-2)
Vf.yf,i + Vf.yf,i = Lf-1xf-1,i + DXD,i (1 ≤ i ≤ C)
Vj+1.yj+1,i = Ljxj,i - BXB,i (1 ≤ i ≤ C) and (f ≤ j ≤ N)
FXi = DXD,i + BXB,i (1 ≤ i ≤ C)

2. Equilibrium Relations

y, = 1

x, = 1

Where i = 1 to C …29.1

and j = 0 to N + 1

3. Summation Equation
yj,i = kj,i.xj,i
 4. Enthalpy balance equations
Vj+1.Hj+1 = Ljhj + DHD + QC (0 ≤ j ≤ f -2)
Vf.Hf + Vf.Hf = Lf-1hf-1 + DHD +QC
Vj+1.Hj+1 = Ljhj – BhB + QR (f ≤ j ≤ N)
FH = B.hB + DHD + QC – QR

Fig. 29.1: Conventional Distillation Column

Inspection of this set of equations show that the equations are logical extension of those may
be stated for a binary system.

When it is assumed that the vapour and liquid streams form ideal solutions, the enthalpy per
mole of vapour and the enthalpy per mole of liquid leaving plate ‘j’ are given by following
expressions.
 C
H j   H j ,i y j ,i (Vapour) …29.2
i 1

C
h j   h j ,i x j ,i (Liquid) …29.3
i 1

Where, the enthalpy of each pure component ‘i’ in the vapour and liquid streams
leaving plate ‘j’ are represented by H j ,i & h j ,i respectively. These enthalpies are evaluated at
the temperature and pressure of j th plate, H D depends on condenser used. For a total
condenser (D is withdrawn from the accumulation as a liquid at its bubble point temperature
‘T’0 at the column pressure, and y1,i  xDi  X Di ),

C C
H D   h0,i X Di   h0,i x0i  h0 …29.4
i 1 i 1

For a partial condenser (D is withdrawn from the accumulation as a vapour at its dew
point temperature ‘T’0 at the column pressure, and y0,i  X Di ),

C C
H D   H 0,i X Di   H 0,i y0i  H 0 …29.5
i 1 i 1

The enthalpy per mole of bottom has double but equivalent representation, hB and
hN 1 i.e.

C C
hB   hBi xBi   hN 1,i x N 1,i  hN 1 …29.6
i 1 i 1

The symbols QC & QR represent the condenser & reboiler heat duties, respectively. The
calculation procedure consists

- An iterative technique which employs the θ method of convergence.

- The tridiagonal formulation of the component material balances & equilibrium
relationships.
- The Kb method for the determination of temperature.
- The constant composition method for the determination of the total flow rates.

Formulation of Tridiagonal Matrix:

In the case of the component material balances, a new set of variables, the component flow
rates in the vapour and liquid phases are introduced, namely,

v j ,i  V j y j ,i and l j ,i  L j x j ,i …29.7
Also the flow rates of component ‘i’ in the distillate and the bottoms are represented by

di  DX Di and bi  BX Bi …29.8

and the flow rates of component ‘i’ in the vapour and liquid parts of the feed are represented
by

v f ,i  V f y f ,i and l f ,i  L f X f ,i …29.9

The equilibrium relationship y j ,i  k j ,i x j ,i may be restated in an equivalent form in terms of

the component flow rates v j ,i & l j ,i as follows. First, observe that the expression y j ,i  k j ,i x j ,i
may be restated in the form:

V k 
V j y j ,i   j j ,i  L j x j ,i …29.10
 L 
 j 

v j ,i  S j ,i l j ,i & l j ,i  A j ,i v j ,i …29.11

1 Lj
Where, A j ,i  
S j ,i V j k j ,i

Instead of enclosing the end of the column and respective plates in each section of the column
as demonstrated by eq. (29.1), an equivalent set of component material balances is obtained
by enclosing each plate ( j = 0, 1, 2, ....., N, N+1) by a component material balance as
demonstrated ion following figure. The corresponding set of material balances for each
component ‘i’ are as follows:
 Fig. 29.2: Conventional Distillation Column

 l0,i  d i  v j ,i  0

 l j 1,i  v j ,i  l j ,i  v j 1,i  0 1  j  f  2
l f 2,i  v f 1,i  l f 1,i  v f ,i  v f ,i …29.12

l f 1,i  v f ,i  l f ,i  v f 1,i  l f ,i

l j 1,i  v j ,i  l j ,i  v j 1,i  0  f 1  j  N 
l N ,i  v N 1,i  bi  0

Except for the first expression of eq. (29.12), the l j ,i ’s may be eliminated from eq. (29.12) by
use of equilibrium relationship eq. (29.11).

For the case of total condenser, l0,i and di have the same composition and thus

L 
l0,i   0 d i …29.13
D

For a partial condenser, y0,i  X Di and hence

  Dk 
D. X Di   0,i  L0 x0,i …29.14
 L0 

l0,i  A0,i d i …29.15

L0
Where A0,i 
k0,i D

The expression given by eq.(29.15) may be used to represent both a partial condenser and a
total condenser, provided A0,i is set equal to L0 for a total condenser. Also, the form of
D
AN 1,i differs slightly from that of A j ,i because of the double representation of the reboiler by
the subscripts " N 1" and B, i.e.

V k 
VN 1 y N 1,i   N 1 N 1,i  BxBi
 B 

or, bi  AN 1,i v N 1,i …29.16

B
AN 1,i 
k N 1,iVN 1

Where the l j ,i ’s and bi are eliminated from eq. (29.12) by use of eqs. (29.11), (29.13),
(29.15), and (29.16), the following results are obtained:

  A0,i  1d i  v j ,i  0

A j 1,i v j 1,i  A j ,i  1v j ,i  v j 1,i  0 1  j  f  2

A f 2,i v f 2,i  A f 1,i  1v f 1,i  v f ,i  v f ,i

A f 1,i v f 1,i  A f ,i  1v f ,i  v f 1,i  l f ,i …29.17

A j 1,i v j 1,i  A j ,i  1v j ,i  v j 1,i  0  f 1  j  N 

AN ,i v N ,i   AN 1,i  1v N 1,i  0

This set of equations may be written in matrix form:

Ai vi   Li …29.18

Where,
Ai 

 P0,i 1 0 0
........... .......... ........ ......... 0 0
. . .
A0,i  P1,i 1 0 ........... .......... ........ ......... 0 0
. . .
........ .......... .......... ........... ........... .......... ........ ......... ........... ...........
. . . .
0 ........... ........... ........... ........... .......... 1 0 0 0
.
0 ........... ........... A f 2,i ........... ..........  Pf ,i 1 0 0
.
........ ........... ........... ........... ........... .......... ........ ......... ........... ...........
. . . . . .
0 ........... ........... ........... ........... .......... 0 AN 1,i  PN ,i 1
. . . . .
0 ........... ........... ........... ........... .......... 0 0 AN ,i  PN 1,i
. . . . .


vi  d i v1,i v2,i ........v f 1,i v f ,i .......vN ,i vN 1,i 
T


Li  000.............v f 1,i v f ,i ..............00 T

Pji  A ji  1

On the basis of the assumed temperatures and total flow rates, the absorption factors A ji 
appearing in eq. (29.18) may be evaluated for component i on each plate j, since Ai in eq.
(29.18) is of tridiagonal form, this matrix equation may be solved for the calculated values of
the vapour rates for the component i [denoted by v ji cal ] by use of the Thomas algorithm.
Consider the following set of linear equations in the variables x0 , x1 ,........., xN , xN 1 whose
coefficients, form a tridiagonal matrix.

B0 x0  C0 x1  D0

A1 x0  B1 x1  C1 x2  D1

A2 x1  B2 x2  C1 x2  D2

...................................

AN xN 1  BN xN  CN xN 1  DN

AN 1 xN  BN 1 xN 1  DN 1
The following recurrence formulas are applied in the order stated:

C0 D
f0  , g0  0
B0 B0

fK 
Ck
k  1,2,......, N  1
Bk  Ak f k 1

Dk  Ak g k 1
gK  k  1,2,......, N  1
Bk  Ak f k 1

After the f ’s and g ’s have been computed, the values of xN 1 , xN ,....... x1 , x0 are computed as
follows:

vN 1  g N 1

xk  g k  f k xk 1 k  N , N  1,......,2,1,0
After these recurrence formulas have been applied for each component i and the
 
complete set of vapor rates v ji ca has been calculate, the corresponding set of liquid rates
{(lji)ca} is then calculated by use of Eq. (29.7). These sets of calculated flow rates are used in
conjunction with  method of convergence and the Kb method in the determination of an
improved set of temperatures.