MULTICOMPONENT

DISTILLATION
INTRODUCTION
Multi-component Distillation involves
separation of more than two components
Theory and principles for multi-component
distillation are same as that for binary
systems
Equilibrium data are same. But the method
of analysis will be different
Design calculations involve stage to stage
and trial and error calculations
Therefore computers are generally used for
rigorous solutions
 Equilibrium data
Consider a mixture containing A,B,C and D
Raoult’s law and Dalton’s law are applicable
if considered as ideal.
Raoult’s law: 𝑝𝐴 = 𝑝𝐴𝑜 . 𝑥𝐴
Dalton’s law: 𝑝𝐴 = 𝑃𝑇 . 𝑦𝐴
𝑜
Therefore; 𝑝𝐴 . 𝑥𝐴 = 𝑃𝑇 . 𝑦𝐴
𝑜
𝑝𝐴
𝑦𝐴 = . 𝑥𝐴
𝑃𝑇
𝑦𝐴 = 𝑘𝐴 . 𝑥𝐴
K factor charts are available by Depriester,
hadden and Gorayson charts which is
sometimes called Depriester charts
Relative volatility
Volatility of a substance (A), relative to
another substance (B).
𝑘𝐴
∝𝐴,𝐵 =
𝑘𝐵
𝑝𝑜
𝐴
𝑃𝑇
∝𝐴,𝐵 = 𝑝𝑜
𝐵
𝑃𝑇
𝑜
𝑝𝐴
∝𝐴,𝐵 = 𝑜
𝑝𝐵
Bubble point and Dew point
calculations
Bubble-point
the temperature at which the liquid
starts to boil

Dew-point
the temperature at which the saturated
vapor starts to condense
 Bubble point calculation
At the bubble point
𝑝𝑖 = 𝑃𝑇 = 𝑃𝑆𝑦𝑠𝑡𝑒𝑚
𝑝𝐴𝑜 . 𝑥𝐴 + 𝑝𝐵𝑜 . 𝑥𝐵 + 𝑝𝐶𝑜 . 𝑥𝐶 + 𝑝𝐷𝑜 . 𝑥𝐷 = 𝑃𝑇 = 𝑃𝑆𝑦𝑠𝑡𝑒𝑚

𝑜 𝑜 𝑜
𝑝𝐴 𝑝𝐵 𝑝𝐶
. 𝑥𝐴 + . 𝑥𝐵 + . 𝑥𝐶 +
𝑃𝑆𝑦𝑠𝑡𝑒𝑚 𝑃𝑆𝑦𝑠𝑡𝑒𝑚 𝑃𝑆𝑦𝑠𝑡𝑒𝑚
𝑜
𝑝𝐷
. 𝑥𝐷 = 1
𝑃𝑆𝑦𝑠𝑡𝑒𝑚
𝑘𝐴 . 𝑥𝐴 + 𝑘𝐵 . 𝑥𝐵 + 𝑘𝐶 . 𝑥𝐶 + 𝑘𝐷 . 𝑥𝐷 = 1

Since 𝑦𝑖 = 𝑘𝑖 . 𝑥𝑖 ,
𝑦𝐴 + 𝑦𝐵 + 𝑦𝐶 + 𝑦𝐷 = 1

𝑦𝑖 = 1 [Condition]
 Bubble point calculation
using k values (Method I)
Assume a temperature for the bubble point
Obtain k values for all the components at
the assumed temperature and the system
pressure.
Calculate 𝑦𝑖 values for all the components
using 𝑦𝑖 = 𝑘𝑖 . 𝑥𝑖
If 𝑦𝑖 = 1, assumed temperature is the
correct temperature
If the assumed temperature is incorrect
select new temperature such that
𝑘𝑗,𝑜𝑙𝑑
𝑘𝑗,𝑛𝑒𝑤 =
𝑦𝑖,𝑐𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑
Find the temperature (Tnew) at which k
value of jth component is 𝑘𝑗,𝑛𝑒𝑤
Component 𝑥𝑖 𝑘𝑖 𝑦𝑖 = 𝑘𝑖 . 𝑥𝑖
A
B
C
D
S𝑦𝑖 =
Dew point calculation using
k values (Method I)
Assume a temperature for the dew point
Obtain k values for all the components at
the assumed temperature and the system
pressure
Calculate 𝑥𝑖 values for all the components
𝑦𝑖
using 𝑥𝑖 =
𝑘𝑖
If 𝑥𝑖 = 1, assumed temperature is the
correct temperature
If the assumed temperature is incorrect
select new temperature such that
𝑘𝑗,𝑛𝑒𝑤 = 𝑘𝑗,𝑜𝑙𝑑 × 𝑥𝑖,𝑐𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑
Find the temperature (Tnew) at which k
value of jth component is 𝑘𝑗,𝑛𝑒𝑤
Component 𝑦𝑖 𝑘𝑖 𝑦𝑖
𝑥𝑖 =
𝑘𝑖
A
B
C
D
S𝑥𝑖 =
 Bubble point calculation
using 𝜶 values (Method II)
𝑘𝑖
Relative volatility is given by, ∝𝑖,𝑗 =
𝑘𝑗

At the bubble point 𝑦𝑖 = 1

Since, 𝑦𝑖 = 𝑘𝑖 . 𝑥𝑖 𝑘𝑖 . 𝑥𝑖 = 1
𝑘𝑖 .𝑥𝑖 1
=
𝑘𝑗 𝑘𝑗
1
∝𝑖,𝑗 𝑥𝑖 = [Condition]
𝑘𝑗
Assume a temperature for the bubble point
Obtain k values for all the components at
the assumed temperature and the system
pressure.
Calculate relative volatility values with
respect to a selected component (j)
1
If ∝𝑖,𝑗 𝑥𝑖 = , assumed temperature is
𝑘𝑗
the correct temperature
If the assumed temperature is incorrect
select new temperature such that
1
𝑘𝑗,𝑛𝑒𝑤 =
∝𝑖,𝑗 𝑥𝑖
𝐶𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑
Find the temperature (Tnew) at which k
value of jth component is 𝑘𝑗,𝑛𝑒𝑤 and repeat
calculation
Component 𝑥𝑖 𝑘𝑖 ∝𝑖,𝑗 ∝𝑖,𝑗 𝑥𝑖
A
B
C
D
S ∝𝑖,𝑗 𝑥𝑖 =
Dew point calculation using
𝜶 values (Method II)
𝑘𝑖
Relative volatility is given by, ∝𝑖,𝑗 =
𝑘𝑗

At the dew point 𝑥𝑖 = 1

𝑦𝑖
Since, 𝑦𝑖 = 𝑘𝑖 . 𝑥𝑖 =1
𝑘𝑖
𝑦𝑖
𝑘𝑗
𝑘𝑖 =1
𝑘𝑗

𝑦𝑖
= 𝑘𝑗 [Condition]
∝𝑖,𝑗
Assume a temperature for the dew point
Obtain k values for all the components at
the assumed temperature and the system
pressure
Calculate relative volatility values with
respect to a selected component (j)
𝑦𝑖
If = 𝑘𝑗 , assumed temperature is
∝𝑖,𝑗
the correct temperature
If the assumed temperature is incorrect
select new temperature such that
𝑦𝑖
𝑘𝑗,𝑛𝑒𝑤 = ∝𝑖,𝑗
𝐶𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑

Find the temperature (Tnew) at which k

value of jth component is 𝑘𝑗,𝑛𝑒𝑤
Component 𝑦𝑖 𝑘𝑖 ∝𝑖,𝑗 𝑦𝑖
∝𝑖,𝑗
A
B
C
D
𝑦𝑖
S ∝𝑖,𝑗 =
 Flash Distillation
Equipment and the operation of multicomponent
distillation is similar to binary systems
Material balance and equilibrium relations are also
similar
However, higher the number of components in the
system larger the number of simultaneous
equations to be solved
Therefore, for multi-component systems a trial and
error procedure is used
Overall Material balance
𝐹 =𝐿+𝑉 -----(1)
Let us assume that the number of components in
the system is ‘n’
Therefore ‘n’ number of material balance
equations can be written
Material balance for the ith component
𝐹𝑥𝑖,𝐹 = 𝐿𝑥𝑖 + 𝑉𝑦𝑖 -----(2) ‘n’ equations
Vapor and liquid produced are in equilibrium
𝑦𝑖 = 𝑘𝑖 . 𝑥𝑖 ----- (3) ‘n’ equations
𝑦𝑖 = 1 ----- (4)
𝑥𝑖 = 1 ----- (5)
Number of equations : 2n+2
Unknowns to be calculated are;
𝑥𝑖 , 𝑦𝑖 , L and V values 2n+2 unknowns
Total number of equations 2n+2
Solving large number of simultaneous equations
are time consuming
Therefore, trial and error procedure is used to
solve multi-component flash distillation problems
From equations 1 and 2;
𝐿 =𝐹−𝑉
𝐹𝑥𝑖,𝐹 = 𝐹 − 𝑉 𝑥𝑖 + 𝑉𝑦𝑖
From equation (3) 𝑦𝑖 = 𝑘𝑖 . 𝑥𝑖
Therefore, 𝐹𝑥𝑖,𝐹 = 𝐹 − 𝑉 𝑥𝑖 + 𝑉 𝑘𝑖 . 𝑥𝑖
𝑥𝑖,𝐹
Simplifying, 𝑥𝑖 =
1+ 𝑘𝑖 −1 𝑉 𝐹
 𝑦𝑖
Similarly, 𝑥𝑖 =
𝑘𝑖
𝑦𝑖
𝐹𝑥𝑖,𝐹 = 𝐹 − 𝑉 + 𝑉𝑦𝑖
𝑘𝑖
𝑘𝑖 𝑥𝑖,𝐹
Simplifying, 𝑦𝑖 =
1+ 𝑘𝑖 −1 𝑉 𝐹
Calculation procedure
Assume a value for V/F (fraction vaporized).
Obtain 𝑘𝑖 values for the operating
temperature and pressure
𝑘𝑖 𝑥𝑖,𝐹
Calculate 𝑦𝑖 values using, 𝑦𝑖 =
1+ 𝑘𝑖 −1 𝑉 𝐹
Calculate 𝑦𝑖 and if 𝑦𝑖 = 1, assumed V/F
is correct
If 𝑦𝑖 ≠ 1, assume a new value for V/F and
repeat the calculation
Estimate new V/F using Newton’s convergence
method - suitable for calculations using computers
If 𝑦𝑖 = 1, calculate 𝑥𝑖 values using
𝑦𝑖
𝑥𝑖 =
𝑘𝑖
Calculate L and V using the correct V/F
MULTICOMPONENT
DISTILLATION
Column Distillation
Column Distillation
 Rigorous computer methods for solving
multi-component distillation problems are
available
The approximate or "shortcut" methods
are common for preliminary design
Fenske Underwood Gilliand method
Lewis and Matheson method
Matrix method
 FUG (Fenske, Underwood,
Gilliland) method
FUG method consists of four basic steps.
i. Identification of the key components
ii. Determination of minimum reflux ratio (Rm)
using Underwood equation
iii. Determination of minimum number of stages for
Total reflux using Fenske equation
iv. Estimation of actual number of plates required
for a finite reflux ratio (R) using Nm and Rm and
Gilliland correlation
 Example
The following feed at 82 °C and 1035 kN/m2
is to be fractionated at that pressure so that
the vapor distillate contains 98% of the C3H8
but only 1% of the C5H12. The q factor of the
feed is 0.667.
Comp. CH4 C2H6 C3H8 C4H10 C5H12 C6H14
𝐱 𝐢,𝐅 0.03 0.07 0.15 0.33 0.30 0.12
k80 21.0 5.9 2.49 0.95 0.395 0.18
Identify the key components
Determine the minimum reflux ratio
required and product compositions under
Rmin using Underwood equations
Determine the minimum number of plates
required and product compositions under
total reflux using Fensky equation
Estimate the number of trays required if
the reflux ratio is 0.8 using Gilliland
correlation.
Determine product compositions when R =
0.8
 Identification of the key
components
Heavy Key - The least volatile component
in the distillate but in a significant
concentration
Light Key - The most volatile component in
the bottoms but in a significant
concentration
Heavy Key - C5H12
Light Key - C3H8
Key components must be present in both
the overhead and the bottoms
Non-key components are classed as
distributed if they occur in both products, or
non-distributed if they appear in only one
product
In calculating the relative volatility (α),
heavy key is used as the reference
component
 At minimum Reflux
𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03
C2H6 0.07
C3H8 (LK) 0.15
C4H10 0.33
C5H12 (HK) 0.30
C6H14 0.12
 Assume F=100 kmol/h
𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03 3
C2H6 0.07 7
C3H8 (LK) 0.15 15
C4H10 0.33 33
C5H12 (HK) 0.30 30
C6H14 0.12 12
98% of the C3H8 splits into Distillate
𝐷. 𝑥C3H8,𝐷 = 15 × 0.98 = 14.7

1% of the C5H12 splits into Distillate

𝐷. 𝑥C5H12,𝐷 = 30 × 0.01 = 0.3
 𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03 3
C2H6 0.07 7
C3H8 (LK) 0.15 15 14.7
C4H10 0.33 33
C5H12 (HK) 0.30 30 0.3
C6H14 0.12 12
 𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03 3 Nil
C2H6 0.07 7 Nil
C3H8 (LK) 0.15 15 14.7
C4H10 0.33 33
C5H12 (HK) 0.30 30 0.3
C6H14 0.12 12 Nil
 𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03 3 3 Nil
C2H6 0.07 7 7 Nil
C3H8 (LK) 0.15 15 14.7 0.3
C4H10 0.33 33
C5H12 (HK) 0.30 30 0.3 29.7
C6H14 0.12 12 Nil 12
∝𝑖,80 is calculated by,
𝑘𝑖,80
∝𝑖,80 =
𝑘𝐻𝐾,80
𝑘CH ,80
∝CH4,80 = 4
𝑘C H ,80
5 12
21
∝CH4,80 =
0.395
∝CH4,80 = 53.2
 𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03 3 3 Nil 53.2
C2H6 0.07 7 7 Nil 14.94
C3H8 (LK) 0.15 15 14.7 0.3 6.3
C4H10 0.33 33 2.405
C5H12 (HK) 0.30 30 0.3 29.7 1
C6H14 0.12 12 Nil 12 0.456
Determination of minimum reflux
ratio using Underwood equation
Assumptions
i. α is a constant at average column
temperature (Assume a value for the
average column temperature. This should
be checked at the end of the calculation)
ii. Constant molar overflow
iii. Optimal feed location
iv. Reflux is saturated liquid
Solve Equation 1 for ∅ values, the values will
lie between α values.
∝𝑖 .𝑥𝑖,𝐹 .𝐹
=𝐹 1−𝑞 ------------------(1)
∝𝑖 −∅
∅ values calculated from equation (1) can be
used to solve Underwood equation and
determine D & Rm
∝𝑖 .𝑥𝑖,𝐷 .𝐷
= 𝐷 𝑅𝑚 + 1 ------------------(2)
∝𝑖 −∅
From equation (1) with q= 0.667,
53.2×3 14.94×7 6.3×15 2.405×33 1×30
+ + + + +
53.2−∅ 14.94−∅ 6.3−∅ 2.405−∅ 1−∅

0.456×12
= 100 1 − 0.667
0.456−∅
The ∅ values will lie between α values.
53.2 > ∅1 > 14.94 > ∅2 > 6.3 >
∅3 > 2.405 > ∅4 > 1 > ∅5 > 0.456
To solve the problem ∅ values between ∝𝐿𝐾
and ∝𝐻𝐾 is necessary. i.e ∅3 and ∅4 .
For hand calculations ∅3 and ∅4 can be
found by trial and error procedure
∅3 = 4.7760 and ∅4 = 1.4177
Now, ∅3 and ∅4 can be used to solve Underwood
equation

∝𝑖 .𝑥𝑖,𝐷 .𝐷
= 𝐷 𝑅𝑚 + 1
∝𝑖 −∅

53.2×3 14.94×7 6.3×14.7

+ + +
53.2−4.7760 14.94−4.7760 6.3−4.7760

2.405×𝑥C H ,𝐷 .𝐷 1×0.3
4 10
+ = 𝐷 𝑅𝑚 + 1 ---
2.405−4.7760 1−4.7760
 53.2×3 14.94×7 6.3×14.7
+ + +
53.2−1.4177 14.94−1.4177 6.3−1.4177

2.405×𝑥C H ,𝐷 .𝐷 1×0.3
4 10
+ = 𝐷 𝑅𝑚 + 1 ------ (B)
2.405−1.4177 1−1.4177
From (A) and (B),
𝑥C4H10,𝐷 . 𝐷 = 13.05
𝐷 𝑅𝑚 + 1 = 60.99
Applying material balance for C4H10,
𝑊. 𝑥C4H10,𝑊 = 𝐹. 𝑥C4H10,𝐹 − 𝐷. 𝑥C4H10,𝐷
𝑊. 𝑥C4H10,𝑊 = 33 − 13.05
𝑊. 𝑥C4H10,𝑊 = 19.95
 𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03 3 3 Nil 53.2
C2H6 0.07 7 7 Nil 14.94
C3H8 (LK) 0.15 15 14.7 0.3 6.3
C4H10 0.33 33 13.05 19.95 2.405
C5H12 (HK) 0.30 30 0.3 29.7 1
C6H14 0.12 12 Nil 12 0.456
S 38.05 61.9
From the table,
𝐷= 𝐷. 𝑥𝑖,𝐷 and 𝑊 = 𝑊. 𝑥𝑖,𝑊
Therefore, 𝐷 = 38.05 and 𝑊 = 61.9
But, 𝐷 𝑅𝑚 + 1 = 60.99
Therefore, 𝑅𝑚 = 0.6029
 𝑥𝑖,𝐷 and 𝑥𝑖,𝑊 columns can be completed
𝒙𝐢,𝑭 𝐅. 𝒙𝐢,𝑭 𝐃. 𝒙𝐢,𝑫 𝐖. 𝒙𝐢,𝑾 ∝𝒊,𝟖𝟎 𝒙𝐢,𝑫 𝒙𝐢,𝑾
CH4 0.03 3 3 Nil 53.2 0.0789 Nil
C2H6 0.07 7 7 Nil 14.94 0.183 Nil
C3H8 (LK) 0.15 15 14.7 0.3 6.3 0.386 4.8x10-3
C4H10 0.33 33 13.05 19.95 2.405 0.343 0.321
C5H12 (HK) 0.30 30 0.3 29.7 1 7.8x10-3 0.479
C6H14 0.12 12 Nil 12 0.456 Nil 0.193
S 38.05 61.9
Then the assumed average column temperature
should be checked
Distillate dew point (trial and error) 46°C
Residue bubble point (trial and error) 113°C
46+113
Average column temperature = = 79.5°C
2
Assumption Tcolumn avg = 80 °C is valid
If it is different, assume a new value for Tcolumn avg
and repeat the calculation
Determination of minimum number of
stages for Total reflux using Fenske
equation
Fenske equation

𝑥𝐿𝐾,𝐷 .𝐷 𝑥𝐻𝐾,𝑊 .𝑊
𝑙𝑛
𝑥𝐿𝐾,𝑊 .𝑊 𝑥𝐻𝐾,𝐷 .𝐷
𝑁𝑚 + 1 =
𝑙𝑛 ∝𝐿𝐾,𝑎𝑣𝑔
∝𝐿𝐾,𝑎𝑣𝑔 is the relative volatility of the light key (geometric
average between distillate dew point and residue bubble
point)
∝𝐿𝐾,𝑎𝑣𝑔 = ∝𝐿𝐾,𝑇𝐷𝑖𝑠𝑡𝑖𝑙𝑙𝑎𝑡𝑒 𝐷𝑒𝑤 𝑝𝑜𝑖𝑛𝑡 ×∝𝐿𝐾,𝑇𝑅𝑒𝑠𝑖𝑑𝑢𝑒 𝐵𝑢𝑏𝑏𝑙𝑒 𝑝𝑜𝑖𝑛𝑡
∝𝐿𝐾,46°𝐶 = 8.34 and ∝𝐿𝐾,113°𝐶 = 5.00
∝𝐿𝐾,𝑎𝑣𝑔 = 8.34 × 5.00 = 6.4576
Nm – Number of plates in the column excluding the
reboiler
 For Total Reflux
𝑥𝑖,𝐹 𝐹. 𝑥𝑖,𝐹 𝐷. 𝑥𝑖,𝐷 𝑊. 𝑥𝑖,𝑊 ∝𝑖,80 𝑥𝑖,𝐷 𝑥𝑖,𝑊
CH4 0.03 3 3 Nil 53.2
C2H6 0.07 7 7 Nil 14.94
C3H8 (LK) 0.15 15 14.7 0.3 6.3
C4H10 0.33 33 2.405
C5H12 (HK) 0.30 30 0.3 29.7 1
C6H14 0.12 12 Nil 12 0.456
 𝑥𝐿𝐾,𝐷 .𝐷 𝑥𝐻𝐾,𝑊 .𝑊
𝑙𝑛
𝑥𝐿𝐾,𝑊 .𝑊 𝑥𝐻𝐾,𝐷 .𝐷
𝑁𝑚 + 1 =
𝑙𝑛 ∝𝐿𝐾,𝑎𝑣𝑔
14.7 29.7
𝑙𝑛
0.3 0.3
𝑁𝑚 + 1 =
𝑙𝑛 6.4576

𝑁𝑚 + 1 = 4.55
𝑁𝑚 = 3.55 ≈ 4
Following equation can be used to determine the
distribution of the non-key components
𝑥𝑖,𝐷 .𝐷 𝑁𝑚 +1 𝑥𝐻𝐾,𝐷 .𝐷
=∝𝑖,80
𝑥𝑖,𝑊 .𝑊 𝑥𝐻𝐾,𝑊 .𝑊
For C4H10
𝑥C H ,𝐷 .𝐷 4.55 0.30
4 10
= 2.4
𝑥C H ,𝑊 .𝑊 29.7
4 10

𝑥C4H10,𝐷 . 𝐷 = 0.547 × 𝑥C4H10,𝑊 . 𝑊 ---------- (C)

Applying material balance for C4H10,
𝑊. 𝑥C4H10,𝑊 = 𝐹. 𝑥C4H10,𝐹 − 𝐷. 𝑥C4H10,𝐷
𝑊. 𝑥C4H10,𝑊 = 33 − 𝐷. 𝑥C4H10,𝐷 ---------- (D)

Solving (C) and (D),

𝑥C4H10,𝐷 . 𝐷 = 11.67 and 𝑥C4H10,𝑊 . 𝑊 = 21.33
 𝒙𝐢,𝑭 𝐅. 𝒙𝐢,𝑭 𝐃. 𝒙𝐢,𝑫 𝐖. 𝒙𝐢,𝑾 ∝𝒊,𝟖𝟎 𝒙𝐢,𝑫 𝒙𝐢,𝑾
CH4 0.03 3 3 Nil 53.2
C2H6 0.07 7 7 Nil 14.94
C3H8 (LK) 0.15 15 14.7 0.3 6.3
C4H10 0.33 33 11.67 21.33 2.405
C5H12 (HK) 0.30 30 0.3 29.7 1
C6H14 0.12 12 Nil 12 0.456
S 36.67 63.33
From the table,
𝐷= 𝐷. 𝑥𝑖,𝐷 and 𝑊 = 𝑊. 𝑥𝑖,𝑊
Therefore, 𝐷 = 36.67 and 𝑊 = 63.33
 𝑥𝑖,𝐷 and 𝑥𝑖,𝑊 columns can be completed
𝒙𝐢,𝑭 𝐅. 𝒙𝐢,𝑭 𝐃. 𝒙𝐢,𝑫 𝐖. 𝒙𝐢,𝑾 ∝𝒊,𝟖𝟎 𝒙𝐢,𝑫 𝒙𝐢,𝑾
CH4 0.03 3 3 Nil 53.2 0.081 Nil
C2H6 0.07 7 7 Nil 14.94 0.190 Nil
C3H8 (LK) 0.15 15 14.7 0.3 6.3 0.400 4.73x10-3
C4H10 0.33 33 11.67 21.33 2.405 0.318 0.336
C5H12 (HK) 0.30 30 0.3 29.7 1 8.18x10-3 0.468
C6H14 0.12 12 Nil 12 0.456 Nil 0.189
S 36.67 63.33
 Feed tray location
𝑥𝐿𝐾,𝐷
𝑥𝐻𝐾,𝐷
𝑙𝑛 𝑥𝐿𝐾,𝐹
𝑥𝐻𝐾,𝐹
𝑁𝐹,𝑚𝑖𝑛 = ∝𝐿𝐾
𝑙𝑛 ∝𝐻𝐾

Assume relative feed location is constant as we

change the reflux ratio from total reflux to a finite
value.
𝑁𝐹,𝑚𝑖𝑛 𝑁𝐹
=
𝑁𝑚𝑖𝑛 𝑁
𝑁𝑚𝑖𝑛 Number of stages at total reflux
𝑁 Number of stages at finite reflux (R)
𝑁𝐹,𝑚𝑖𝑛 Feed tray number at total reflux
 0.400
𝑙𝑛 8.18x10−3
0.15
0.3
𝑁𝐹,𝑚𝑖𝑛 =
𝑙𝑛 6.3 1

𝑁𝐹,𝑚𝑖𝑛 = 2.48
Estimation of actual number of plates
required for a finite reflux ratio (R)
Estimation of actual number of plates
a. Gilliland correlation
b. Erbar and Maddox correlation
c. Rule of Thumb
 Gilliland correlation
𝑅−𝑅𝑚
𝑥=
𝑅+1
For 0 ≤ 𝑥 ≤ 0.01
𝑁−𝑁𝑚
= 1 − 18.5715𝑥
𝑁+1
For 0.01 ≤ 𝑥 ≤ 0.9
𝑁−𝑁𝑚 0.002743
= 0.545827 − 0.591422𝑥 +
𝑁+1 𝑥
For 0.9 ≤ 𝑥 ≤ 1
𝑁−𝑁𝑚
= 0.16595 − 0.16595𝑥
𝑁+1
Erbar and Maddox correlation
Rule of Thumb
Suitable for very preliminary calculations
only

Approximately N =2.5 Nm
𝑅 = 0.8
0.8−0.608
𝑥= = 0.106
0.8+1
𝑁−𝑁𝑚
From Gilliland correlation, = 0.5
𝑁+1
𝑁𝑚 = 3.55
𝑁=8
Accuracy of the shortcut method by Gilliland
correlations is ±30%
 Determine the composition of
products when R = 0.8
𝑹 𝑹 𝑫. 𝒙𝐂𝟒𝐇𝟏𝟎,𝑫 𝑾. 𝒙𝐂𝟒𝐇𝟏𝟎,𝑾
𝑹+𝟏
∞ 1 11.67 21.33
𝑅 = 0.8 0.444
𝑅𝑚 = 0.608 0.378 13.05 19.95
By linear interpolation, 𝐷. 𝑥C4H10,𝐷 and 𝑊. 𝑥C4H10,𝑊 at
𝑅
𝑅+1 = 0.444 can be calculated.
𝐷. 𝑥C4H10,𝐷 at R=0.8 is given by,
13.05−11.67
13.05 + × 0.444 = 12.06
0.378−1
𝑊. 𝑥C4H10,𝑊 at R=0.8 is given by,
33 − 12.06 = 20.94
 𝐴𝑡 𝑅 = 0.8
𝒙𝐢,𝑭 𝐅. 𝒙𝐢,𝑭 𝐃. 𝒙𝐢,𝑫 𝐖. 𝒙𝐢,𝑾 ∝𝒊,𝟖𝟎 𝒙𝐢,𝑫 𝒙𝐢,𝑾
CH4 0.03 3 3 Nil 53.2
C2H6 0.07 7 7 Nil 14.94
C3H8 (LK) 0.15 15 14.7 0.3 6.3
C4H10 0.33 33 12.06 20.94 2.405
C5H12 (HK) 0.30 30 0.3 29.7 1
C6H14 0.12 12 Nil 12 0.456
S 37.06 62.94
From the table,
𝐷= 𝐷. 𝑥𝑖,𝐷 and 𝑊 = 𝑊. 𝑥𝑖,𝑊
Therefore, 𝐷 = 37.06 and 𝑊 = 62.94
 𝑥𝑖,𝐷 and 𝑥𝑖,𝑊 columns can be completed
𝒙𝐢,𝑭 𝐅. 𝒙𝐢,𝑭 𝐃. 𝒙𝐢,𝑫 𝐖. 𝒙𝐢,𝑾 ∝𝒊,𝟖𝟎 𝒙𝐢,𝑫 𝒙𝐢,𝑾
CH4 0.03 3 3 Nil 53.2 0.081 Nil
C2H6 0.07 7 7 Nil 14.94 0.189 Nil
C3H8 (LK) 0.15 15 14.7 0.3 6.3 0.397 4.77x10-3
C4H10 0.33 33 12.06 20.94 2.405 0.325 0.333
C5H12 (HK) 0.30 30 0.3 29.7 1 8.09x10-3 0.472
C6H14 0.12 12 Nil 12 0.456 Nil 0.191
S 37.06 62.94
MULTICOMPONENT
DISTILLATION
Lewis and Matheson
method
Top Operating Line (TOL)
𝐿 𝐷
𝑦𝑛+1 = 𝑥𝑛 + 𝑥𝐷
𝐺 𝐺
𝑅 𝑥𝐷
𝑦𝑛+1 = 𝑥𝑛 +
𝑅+1 𝑅+1
Bottom Operating Line (BOL)
𝐿′ 𝑊
𝑦𝑚 = ′ 𝑥𝑚−1 + ′ 𝑥𝑤
𝐺 𝐺
 Lewis and Matheson
Calculation Method
Exact calculation procedure to determine the
number of trays required for multi-component
distillation
Instead of the Gilliland correlation for the
short cut method
Product compositions should be known or
compositions calculated using approximate
method can be used
 Steps for the Rectifying section
Assume a temperature for the 1st plate (T1)
Plate temperature (T1)
= Dew point of the vapor leaving the tray (G1)
= Bubble point of the liquid leaving the tray (L1)
Determine k values at T1
𝑦1
Calculate 𝑥1 values from 𝑥1 = ; use 𝑦1 values
𝑘1
obtained from approximate calculation method
If xi,1 = 1 assumed T1 is correct
If xi,1 ≠ 1 assume a new value for T1 and
repeat the calculation
When 𝑛 = 1 Top Operating Line equation
𝑅 𝑥𝑖,𝐷
𝑦𝑖,2 = 𝑥𝑖,1 +
𝑅+1 𝑅+1
Use 𝑥𝑖,𝐷 values from approximate calculation
Calculate 𝑦𝑖,2 values for all the component
Repeat the procedure for all the trays
When each tray calculation is over that tray
should be checked for the feed tray (discussed
later)
Once the feed tray is decided calculation should
be performed to determine the number of trays in
the stripping section
 Steps for the Stripping section
Assume temperature for the partial re-boiler, Tm+1
(residue bubble point, already determined in
approximate calculation)
Calculate the composition of the vapor leaving
the partial re-boiler 𝑦𝑖,𝑚+1
𝑦𝑖,𝑚+1 = 𝑘𝑖 . 𝑥𝑖,𝑊
Calculate 𝑥𝑚 using the bottom operating line
equation
𝐿′ 𝑊
𝑦𝑚+1 = ′ 𝑥𝑚 + ′ 𝑥𝑤
𝐺 𝐺
Estimate a value for the last tray temperature, Tm
Calculate 𝑦𝑚
𝑦𝑚 = 𝑘 𝑇,𝑚 . 𝑥𝑚
Repeat the procedure for all the trays in the
stripping section
Check each tray for the feed tray
Then count the number of stages in the
rectifying and stripping sections
 Feed Tray Location
TOL for Light Key;
𝐿 𝐷
𝑦𝐿𝐾,𝑛+1 = 𝑥𝐿𝐾,𝑛 + 𝑥𝐿𝐾,𝐷 ------------------(a)
𝐺 𝐺

TOL for Heavy Key;

𝐿 𝐷
𝑦𝐻𝐾,𝑛+1 = 𝑥𝐻𝐾,𝑛 + 𝑥𝐻𝐾,𝐷 ------------------(b)
𝐺 𝐺
Eliminating 𝐿 from (a) and (b)

𝐷 𝐺 𝐷 𝐺
𝑦𝐿𝐾,𝑛+1 − 𝑥𝐿𝐾,𝐷 = 𝑦𝐻𝐾,𝑛+1 − 𝑥𝐻𝐾,𝐷
𝐺 𝑥𝐿𝐾,𝑛 𝐺 𝑥𝐻𝐾,𝑛

𝐷 𝑥𝐿𝐾,𝑛 𝐺.𝑦𝐻𝐾,𝑛+1 −𝐷.𝑥𝐻𝐾,𝐷

𝑦𝐿𝐾,𝑛+1 = 𝑥𝐿𝐾,𝐷 + -------(1)
𝐺 𝑥𝐻𝐾,𝑛 𝐺
BOL for Light Key;
𝐿′ 𝑊
𝑦𝐿𝐾,𝑚 = ′ 𝑥𝐿𝐾,𝑚−1 − ′ 𝑥𝐿𝐾,𝑤 ------------------(c)
𝐺 𝐺
BOL for Heavy Key;
𝐿′ 𝑊
𝑦𝐻𝐾,𝑚 = ′ 𝑥𝐻𝐾,𝑚−1 − ′ 𝑥𝐻𝐾,𝑤 ------------------(d)
𝐺 𝐺
Eliminating 𝐿 from (c) and (d)

𝑊 𝐺′ 𝑊 𝐺′
𝑦𝐿𝐾,𝑚 + 𝑥 = 𝑦𝐻𝐾,𝑚 + 𝑥
𝐺 ′ 𝐿𝐾,𝑤 𝑥𝐿𝐾,𝑚−1 𝐺 ′ 𝐻𝐾,𝑤 𝑥𝐻𝐾,𝑚−1

𝑊 𝐺 ′ .𝑦𝐻𝐾,𝑚 +𝑊.𝑥𝐻𝐾,𝑤 𝑥𝐿𝐾,𝑚−1

𝑦𝐿𝐾,𝑚 = − ′ 𝑥𝐿𝐾,𝑤 + -------(2)
𝐺 𝐺′ 𝑥𝐻𝐾,𝑚−1
At feed tray, (1)=(2)
𝐷 𝑥𝐿𝐾,𝑛 𝐺.𝑦𝐻𝐾,𝑛+1 −𝐷.𝑥𝐻𝐾,𝐷 𝑊 𝐺 ′ .𝑦𝐻𝐾,𝑚 +𝑊.𝑥𝐻𝐾,𝑤 𝑥𝐿𝐾,𝑚−1
𝑥 + = − ′ 𝑥𝐿𝐾,𝑤 +
𝐺 𝐿𝐾,𝐷 𝑥𝐻𝐾,𝑛 𝐺 𝐺 𝐺′ 𝑥𝐻𝐾,𝑚−1

𝐷 𝑥𝐿𝐾 𝐺.𝑦𝐻𝐾,𝑛+1 −𝐷.𝑥𝐻𝐾,𝐷 𝑊 𝐺 ′ .𝑦𝐻𝐾,𝑚 +𝑊.𝑥𝐻𝐾,𝑤 𝑥𝐿𝐾

𝑥 + = − ′ 𝑥𝐿𝐾,𝑤 +
𝐺 𝐿𝐾,𝐷 𝑥𝐻𝐾 𝑖𝑛𝑡 𝐺 𝐺 𝐺′ 𝑥𝐻𝐾 𝑖𝑛𝑡
𝑥𝐿𝐾 𝐺.𝑦𝐻𝐾 −𝐷.𝑥𝐻𝐾,𝐷 𝐺 ′ .𝑦𝐻𝐾 +𝑊.𝑥𝐻𝐾,𝑤 𝑊 𝐷
− = − 𝐺 ′ 𝑥𝐿𝐾,𝑤 − 𝑥
𝑥𝐻𝐾 𝑖𝑛𝑡 𝐺 𝐺′ 𝐺 𝐿𝐾,𝐷

𝑥𝐿𝐾 𝐷.𝑥𝐻𝐾,𝐷 𝑊.𝑥𝐻𝐾,𝑤 𝑊 𝐷

+ = 𝑥 + 𝑥𝐿𝐾,𝐷
𝑥𝐻𝐾 𝑖𝑛𝑡 𝐺 𝐺′ 𝐺 ′ 𝐿𝐾,𝑤 𝐺

𝑊 𝐷
𝑥𝐿𝐾 𝑥𝐿𝐾,𝑤 + 𝑥 𝑥𝐿𝐾
𝐺′ 𝐺 𝐿𝐾,𝐷
= 𝐷 𝑊 ≈
𝑥𝐻𝐾 𝑖𝑛𝑡 𝑥𝐻𝐾,𝐷 + ′ 𝑥𝐻𝐾,𝑤 𝑥𝐻𝐾 𝐹𝑒𝑒𝑑 𝑇𝑟𝑎𝑦
𝐺 𝐺
At feed tray,

𝑥𝐿𝐾 𝑥𝐿𝐾 𝑥𝐿𝐾

≥ ≥
𝑥𝐻𝐾 𝑓−1
𝑥𝐻𝐾 𝑖𝑛𝑡
𝑥𝐻𝐾 𝑓