Chapter 7:

Batch Distillation

A continuous distillation is thermody-

namically and economically efficient
method for producing large amounts of
material of constant composition

However, when small amounts of

products or varying product compositions
are required, a batch distillation provides
several advantages (over the continuous
distillation)

1
 Batch distillation is versatile and com-
monly employed for producing biochemical,
biomedical, and/or pharmaceutical
products, in which the production amounts
are small but a very high purity and/or an
ultra clean product is required

Equipment for batch distillation can be

arranged in a wide variety of configurations

In the simple batch distillation (see the

Figure on the next Page), the vapour (i.e.
the product) is withdrawn from the top of
the re-boiler (or the “still pot”) continuously
(while the liquid level is decreasing
continuously)

2
 Simple Batch Distillation

This is different from the flash

distillation for the fact that
• there is no continuous feed input
into the pot
• the remaining liquid is drained out
of the re-boiler only at the end of the
distillation

3
 An alternative, but still simple, batch
distillation is a constant-level batch
distillation, whose configuration is similar to
the simple batch distillation, but the liquid
(or the mixture to be separated) is continu-
ously fed into the pot (or the re-boiler) to
keep the liquid level in the pot constant

Another type, but more complex than

the first 2 batch distillation techniques, is a
multi-stage batch distillation, in which a
staged or packed distillation column is
placed above the re-boiler (see the Figure
on the next Page)

4
 Multi-stage Batch Distillation

In the usual operation (of the multi-

stage distillation), distillate is withdrawn
continuously until the distillation is ended

5
 An alternative technique of multi-stage
batch distillation is the batch distillation in
which no distillate is withdrawn from the
column, thus resulting in a continuous
change in the concentration or
composition of liquid in the pot (or the re-
boiler)

When pure bottom product is needed,

an inverted batch distillation is employed

In this method (the inverted batch

distillation), the bottom product or the
liquid in the re-boiler is withdrawn
continuously while the distillate is
withdrawn only at the end of the
distillation

6
Binary Batch Distillation: Rayleigh
Equation

The material balances for batch distil-

lation are different from those for
continuous distillation

In batch distillation, the main focus is

at the total amounts of input(s) [i.e.
feed(s)] and outputs (e.g., distillate or
bottom product) collected at the end of the
distillation, rather than the rates of such
inputs and outputs

The material balances around the batch

distillation system for the entire operating
time are as follows

7
 Overall:
F = Wfinal + Dtotal (7.1)
where
F = the feed fed into the distillation
column for the whole operating
period
Wfinal = the final amount of liquid
in the re-boiler (the notation W
is used because the remaining
liquid in the still pot is
normally a waste)
Dtotal = the total amount of distillate
withdrawn from the distillation
column (in some textbooks, the
notation Dfinal is used)

8
 Species balance (for more volatile
component: MVC):
x F F = x w ,finalWfinal + x D ,avg Dtotal (7.2)

where
xF = mole fraction of a more volatile
species in the feed
x w ,final = the mole fraction of an MVC

of the remaining liquid in the

re-boiler
x D ,avg = an average concentration of

an MVC in the distillate

(in some textbooks the
notation of x D , final is used

instead of x D ,avg )

9
 Note that some modifications/minor
changes in the definitions of some variables
are required for inverted batch distillation
(will appear in the Problem Set)

Normally, F and x F are specified (given),

and the value of either x w ,final or x D ,avg is also

specified (given)

Thus, there are 3 unknowns:

• Wfinal
• Dtotal
• either x D ,avg or x w ,final

but we have only 2 equations (i.e. Eqs. 7.1

& 7.2)

Hence, another or additional equation is

needed
10
 The additional equation for solving
batch distillation problems is commonly
known as the Rayleigh equation

To derive this equation, Lord Rayleigh

(1902) employed the facts that (see the
Figure on Page 3 or 5), at any instant of
time,
1) the rate of distillate flowing out of
the batch distillation system dD is
equal to the decreasing rate of the
liquid in the still pot −dW
2) the rate of species i in the distillate
flowing out of the batch distillation
system x D dD is equal to the
decreasing rate of species i the
liquid in the still pot −d (Wx w )

11
 From the fact # 2, the following equation
can be established:
x D dD = −d (Wx w ) (7.3)

Note that x D can be pulled out of the

differentiation because it is assumed to be
constant (for any instant of time)

From the fact # 1, we obtain the following

equation:
dD = −dW (7.4)

Combining Eq. 7.3 with Eq. 7.4 and re-

arranging gives
− x D dW = −d (Wx w )
− x D dW = −Wdx w − x wdW (7.5)

12
 Re-arranging Eq. 7.5 and integrating the
resulting equation yields
Wfinal xw , final
dW dx w
∫
W =F
W
= ∫
xF
xD − xw

or
xw , final
⎛Wfinal ⎞ dx w
ln ⎜
⎝ F ⎠
⎟= ∫
xF
xD − xw
(7.6a)

or
xF
⎛W ⎞ dx w
ln ⎜ final ⎟ = − ∫ (7.6b)
⎝ F ⎠ x − xw
x w , final D

In order to perform an integration of the

right hand side (RHS) of Eq. 7.6 (a & b), x D
must have a relationship with x w in the
following form:
xD = y = f ( xw )

13
 In a simple batch distillation as shown
as the Figure on Page 3, it is reasonable to
assume that the vapour ( D ) is in equilibrium
with the liquid (W ) in the re-boiler

Thus, if the total condenser is used,

y = xD
and x D and x w can be related by an equili-
brium curve or equilibrium equation

Accordingly, Eq. 7.6b can be re-written

as follows
xF xF
⎛W ⎞ dx dx
ln ⎜ final ⎟ = − ∫ =− ∫
⎝ F ⎠ x w , final
y−x x w , final
f (x ) − x

(7.7)
Note that y = x D and x = x w

14
 The integration of the RHS of Eq. 7.7
can be done sequentially as follows
1) Plot or fit the y-x (equilibrium) curve
2) At each value of x (from x F to x w ,final ),

determine the value of y from the

equilibrium curve/equation
1
3) Plot (Y-axis) against x (X-axis)
y−x
or fit it to an equation
4) Graphically or numerically integrate
from x F to x w ,final (the graphical

integration is illustrated on the next

Page)

15
5) The value of Wfinal can be obtained
from manipulating Eq. 7.7 as follows
⎛ xF dx ⎞
Wfinal = F exp ⎜ − ∫ ⎟ (7.8a)
⎜ x y − x ⎟
⎝ w , final ⎠
or
Wfinal = F exp ( −area under the curve )

(7.8b)

16
6) The value of the average distillate
concentration, x D , and the total
amount of distillate, Dtotal , can be
obtained by solving Eqs. 7.1 & 7.2
simultaneously
x F F − x w , finalWfinal
x D ,avg = (7.9)
F − Wfinal
and
Dtotal = F − Wfinal (7.10)

17
 If the equilibrium relationship between y
( x D ) and x ( x w ) is given as
αx
y=
1 + (α − 1) x
the RHS of Eq. 7.7 can be integrated analy-
tically as follows
x
⎛Wfinal ⎞ dx
F

ln ⎜ ⎟=− ∫
⎝ F ⎠ x w , final
y−x

1 ⎡ x w, final (1 − x F ) ⎤
= ln ⎢ ⎥
(α − 1) ⎢⎣ x F (1 − x w, final ) ⎥⎦
⎡ (1 − x F ) ⎤
+ ln ⎢ ⎥
⎣⎢ (1 − x w ,final ) ⎦⎥
(7.11)

18
 In the case that the value of x D is
specified, and the value of x w ,final is to be

determined, a trial & error technique is

required

This can be done by the following

procedure:
1) Make a 1st guess for the value of x w , final

and calculate the value of the inte-

gration or determine the area under
the curve, according to the guessed
value of x w ,final

2) Then, the value of Wfinal can be

calculated from Eq. 7.8 (either a or b)

19
3) Use the value of Wfinal obtained from 2
and the guessed value of x w,final made

in 1, combined with the given values

of F and x F , to compute the values of
Dcalc and x D ,calc using the following

equations:
Dcalc = F − Wfinal (7.12)
x F F − x w , finalWfinal
and x D ,calc = (7.13)
Dcalc
4) Compare the value of x D ,calc with the

given value of x D : if x D ,calc = x D , the

trial & error procedure is ended;

however, if x D ,calc ≠ x D , the trial &

error procedure has to be repeated,

until we obtain the guessed value of
x w ,final that makes x D ,calc = x D

20
Example Use the given equilibrium data of
methanol (MeOH) and water for solving the
simple batch distillation problem with the
following description:
A one-equilibrium-stage (or a simple) batch
still pot is used to separate MeOH from water
The feed with the total amount of 50 moles
of an 80 mol% MeOH is charged into the still
pot operated at 1 atm
The desired distillate concentration ( x D ) is
89.2 mol% MeOH
Determine:
a) the total amount of distillate collected
( Dtotal )
b) the amount of material remained in the
pot after the distillation (Wfinal ) and its

corresponding concentration ( x w , final )

21
 The equilibrium ( y − x ) data of MeOH is as
summarised in the following Table
Methanol liquid Methanol vapour
(mol%) (mol%)
0 0
2.0 13.4
4.0 23.0
6.0 30.4
8.0 36.5
10.0 41.8
15.0 51.7
20.0 57.9
30.0 66.5
40.0 72.9
50.0 77.9
60.0 82.5
70.0 87.0
80.0 91.5
90.0 95.8
95.0 97.9
100.0 100.0

22
 Given data:
• F = 50 moles
• x F = 0.80
• x D ,avg = 0.892

Unknowns:
• Dtotal
• Wfinal
• x w ,final

Since x w ,final , one of the integral bounda-

ries, is NOT known, a trial & error technique

must be employed to compute the answer
xF
dx
for the integral ∫
x w , final
y−x

23
 To start the calculations, the 1st guess of
x w ,final of 0.70 is used

From the given equilibrium data, the

1
value of for each value of x can be
y−x
summarised in the following Table (note
that interpolations are needed to obtain the
1
values of y and when the values of x
y−x
are, e.g., 0.75, 0.65):
1
x y y−x
y−x

0.80 0.915 0.115 8.69

0.75 0.895 0.145 6.89
0.70 0.871 0.171 5.85
0.65 0.845 0.195 5.13
0.60 0.825 0.225 4.44
0.50 0.780 0.280 3.57

24
 Plotting a graph between x (X-axis) and
1
(Y-axis) using the data in the Table on
y−x
the previous Page, from x w ,final = 0.70 to

x F = 0.80 , gives

0.80
dx
From the graph above, ∫ = area
0.70
y−x
under the curve from x = 0.70 to x = 0.80
is equal to 0.7044
25
 The value of Wfinal can then be calculated,
using Eq. 7.8b, as follows
Wfinal = F exp ( −area under the curve )
= ( 50 ) exp ( −0.7044 )
Wfinal = 24.72 mol

Thus, from Eq. 7.12,

Dcalc = F − Wfinal
= 50 − 24.72
Dcalc = 25.28

The value of x D ,calc can then be computed,

using Eq. 7.13, as follows

x F F − x w ,finalWfinal
x D ,calc =
Dcalc

=
[ ( 0.80 )( 50 )] − [( 0.70 )( 24.72)]
( 25.28 )
x D ,calc = 0.898
26
 However, the desired value of x D or x D ,avg

is 0.892 (89.2%)

Thus, a new guess of x w ,final is needed

When the guessed x w ,final is 0.60, we

obtain the following (by performing similar

calculations as above):
• area = 1.2084
• Wfinal = 14.93
• Dcalc = 35.07
• x D ,calc = 0.885 (too low)

Hence, we need to make a new guess of

x w ,final

27
 When the guessed x w ,final is 0.65, we

obtain the following:

• area = 0.9710
• Wfinal = 18.94
• Dcalc = 31.06
• x D ,calc = 0.891 (O.K.!)

28
Constant-level Batch Distillation

In a constant-level batch distillation, a

solvent is added to the re-boiler (or the still
pot) to keep the level of the liquid constant

Note that, during the addition of the

solvent, the total number of moles of all
species in the still pot is kept constant

The total mole balance is

⎡ Accumulation ⎤
[In] − [Out ] = ⎢ ⎥ (7.14)
⎣ in the still pot ⎦

Since the total number of moles is

constant, Eq. 7.14 becomes
[In] − [Out ] = 0

29
 Also, since this is a constant-level
batch distillation, the amount of solvent
evaporated ( −dV ) must be equal to the
amount of solvent added to the still pot
(dS ) — note that the solvent added into the
still pot is called “the second solvent”
−dV = dS (7.15)
(or dV = −dS )

When performing a species (or compo-

nent) balance on the evaporated solvent
(called “the original solvent”), we obtain
−ydV = −d (Wx w )
ydV = d (Wx w )
ydV = Wdx w + x wdW
but W is kept constant

30
 Thus, the above equation becomes
ydV = Wdx w (7.16)

Substituting Eq. 7.15 into Eq. 7.16 gives

−ydS = Wdx w (7.17)

Re-arranging Eq. 7.17 and integrating

the resulting equation yields
x w , initial
S dx w
W
= ∫
xw , final
y
(7.18)

Once again, the vapour and the liquid

in the system are assumed to be in
equilibrium with each other

31
 Hence, the value of y can be related to
the value of x w using either equilibrium
curve or equation

If the relationship between y and x w can

be expressed in the following form:
α xw
y=
1 + (α − 1) x w

the integral of the RHS of Eq. 7.18 is

S 1 ⎡ x w , initial ⎤ α − 1
= ln ⎢
W α ⎣ x w , final ⎦
⎥+
α
( x w , initial − x w , final )

(7.19)

32
Batch Steam Distillation

In the batch steam distillation, steam is

purged directly into the still pot (see the
Figure below)

Batch Steam Distillation

Normally, the direct addition of steam is

done with the system that is NOT miscible
(or immiscible) with water

33
 Like the steam distillation in the conti-
nuous operation, the principal purpose of
adding steam directly into the still pot is to
keep the temperature of the system
below the boiling point of water, while
eliminating the need of a heat transfer
device (as steam can provide heat/energy to
the system)

By employing the material-balance

equations (Eqs 7.1 & 7.2) to the system (see
the Figure on the previous Page), we obtain
the following equation:
⎛ xD − xF ⎞
Wfinal =F⎜ ⎟ (7.20)
x
⎝ D − x w ,final ⎠

34
 Since the more volatile component (i.e.
the volatile organics) is much more volatile
than water, its concentration (i .e. x D ) is
the vapour phase can be considered pure

Thus, Eq. 7.20 becomes

⎛ 1 − xF ⎞
Wfinal =F⎜ ⎟ (7.21)
⎝ 1 − x w , final ⎠

The amount of the distillate, D , can, thus,

be calculated using the following equation:
D = F − Wfinal (7.22)

The total amount of water obtained in

the separating tank (see the Figure on Page
33) can be calculated from the following
equation:
35
 D
Ptotal − x volatilePvolatile
*

∫
in org
nw = *
dn org
0
x volatile in org Pvolatile

(7.23)
and the total amount of steam required is
n w plus the moles of steam condensed to
heat the feed and to vaporise the volatile
organics, or
(n steam )total = n w + (n steam )used to heat the feed

36
Multi-stage Batch Distillation

If higher purity of product (either distil-

late or bottom product) is needed, a multi-
stage column is added to the batch distilla-
tion system (as shown in the Figure on Page
5)

The complexity with the multi-stage

equilibrium is attributed to the fact that x D
and x w are no longer in equilibrium with
each other

Thus, the Rayleigh equation, Eq. 7.6 (a

or b), cannot be integrated until the rela-
tionship between x D and x w can be esta-
blished

37
 Normally, this relationship can be
obtained by performing stage-by-stage
calculations

Performing material and energy balances

from stage j to the top of the column for the
Figure on Page 5, with the assumption that
the accumulation is negligible everywhere
except the re-boiler, results in
V j +1 = L j + D (7.24)

y j +1V j +1 = x j L j + x D D (7.25)

QC + V j +1H j +1 = L j h j + Dh D (7.26)

38
 Once again, in order to simplify the
calculations, CMO is assumed, and the
operating equation of the multi-stage batch
distillation can be written as follows
L ⎛ L⎞
y j +1 = x j + ⎜1 − ⎟ x D (7.27)
V ⎝ V⎠

However, the difficulty of using Eq. 7.27

in the batch distillation is that either x D or
L
is varying during the batch distillation,
V
thus resulting in the fact that the operating
line will also be changing (or it is not cons-
tant) throughout the operation

39
Constant Reflux Ratio

One of the most common multi-stage

batch distillation techniques is the
operation with a constant reflux ratio

In this kind of operation, x D is allowed to

vary (change), while the values of L and V
are kept constant (by fixing the reflux ratio)

The operating lines obtained from the

operating equation or Eq. 7.27 are plotted
for various values of x D on the McCabe-
Thiele diagram – note that the slopes of all
lines are identical as the reflux ratio is
kept constant

40
 To step off the specified number of
equilibrium stages/contacts on each value
of x D , we start with the given x D to find x w
corresponding to that x D at the given
number of equilibrium stages

Once x w values are obtained from the

McCabe-Thiele diagram for the given values
of x D , we can perform the integration for
Eq. 7.6 (the Rayleigh equation), and the
subsequent values of
• Wfinal
• Dtotal
• x D ,avg

will be obtained

41
 Note, once again, that if the value of
x D ,avg is specified, the trial & error technique

is to be employed to calculate the value of

x w ,final

42
Example The 50-mole feed comprising
32% EtOH and 68% water is to be distilled
in the multi-stage batch distillation with
the additional 2 equilibrium stages
Reflux is returned to the column as a
⎛ L⎞
saturated liquid with the reflux ratio ⎜ ⎟ of
⎝D ⎠
2/3
It is desired that the solvent remained in
the still pot has the concentration of EtOH
of 4.5 mol%
Determine the average distillate compo-
sition ( x D ,avg ) , the final amount (in moles) of

liquid in the still pot (Wfinal ) , and the total

amount of distillate collected ( Dtotal )
The operation is at 1 atm

43
 The schematic diagram of the batch
distillation in this Example is as shown
below

Since the operation is at 1 atm, the

equilibrium of EtOH can be obtained from
the x-y equilibrium data of the EtOH-water
mixture at 1 atm

44
 The values of x D is selected and the
corresponding values of x w is obtained for
the number of stages of 3 as shown in the
following McCabe-Thiele diagram

45
 From each pair of x D and x w , the value
1
of is calculated
xD − xw

1
Eventually, a graph between and
xD − xw
x w is plotted from x w ,final = 0.045 to x F =

0.32 as shown on the next Page, and the

area under the curve is found to be 0.608

Thus,
Wfinal = F exp ( −area under the curve )
= ( 50 ) exp ( −0.608 )
Wfinal = 27.21
and
Dtotal = F − Wfinal = 50 − 27.21 = 22.79

46
 The value of x D ,avg can also be calculated

as follows
x F F − x w , finalWfinal
x D ,avg =
Dtotal

=
[ ( 0.32)( 50 )] − [( 0.045 )( 27.21)]
( 22.79 )
x D ,avg = 0.648

47
Variable Reflux Ratio

In this batch distillation technique,

batch distillation is carried out in such the
way that the value of x D is fixed, while the
L
reflux ratio ⎛⎜ ⎞⎟ is varied (changing)
⎝D ⎠

As the reflux ratio is varied, the slope of

the operating line keeps changing as shown
in the Figure on the next Page

Note that the point where the operating

line intersects with the y = x line at x D is
fixed (as the value of x D is fixed)

48
 Similar to the constant reflux ratio batch
distillation (but not exactly the same), in
this technique, we start stepping off stages
from the given value of x D for a given
number of stages to find the value of the
corresponding x w for each slope of the
operating line
49
Operating Time for Batch Distillation

The overall time for batch distillation is

t batch = t op + t down (7.28)

The operating time for batch distillation

(t ) can be computed from the following
op

equation:
Dtotal
t op = (7.29)
D
where D = the flow rate of the distillate

The value of D cannot be set arbitrarily

(อยางไรก็ได or ตามใจชอบ), as the column
must be designed to accommodate a limited
amount of vapour flow rate (Vmax ) , to avoid
the flooding phenomenon
50
 The maximum distillate flow rate ( Dmax )
can be computed as follows

 Vmax
Dmax = (7.30)
L
1+
D

Normally, the optimal value of D is

0.75 D max

The down time (t down ) includes

1) the dumping time for the remaining
bottom product
2) the clean-up time
3) the loading time for the next batch
4) the heat-up time until the reflux
starts to appear

51
 The energy requirements for a condenser
and a re-boiler can be calculated from the
energy balance equations around the con-
denser and the whole system, respectively

If the reflux is a saturated liquid,

QC = −V1 ( H1 − h D ) = −V1λ (7.31)
where λ is the latent heat of vaporisation

The energy balance around the entire

system yields the heating load (QR ) as
follows
QR = −QC + Dh D (7.32)

52