Introduction to Multi-component

Distillation
Chapter 22
McCabe & Smith, “Unit Operations of
Chemical Engineering” 7th Edition.
Introduction
The calculations of equilibrium stages use mass
and enthalpy balances and vapor liquid equilibria.
Mass balance can be written for each component.
There will be only one enthalpy balance.
Phase equilibria is much more complex.
Since temperature and equilibria change from
stage to stage.
Same as in the case of binary distillation.
Then what’s the difference?
 Phase Equilibria in Multi-component
Distillation
The vapor liquid equilibria for a mixture are described
by distribution coefficients or K factors.
Where K for each component is the ratio of mole
fractions in the vapor and liquid phases at equilibrium
at given temperature.

K values can be calculated from the vapor pressure and

the total pressure making use of Raoult’s law and
Dalton’s law.
Raoult’s law is good approximation for a mixtures of
similar compounds, such as paraffins or aromatics.
However at high pressure K values do not vary
exactly inversely with total pressure because of
compressibility factor.
The ratio of K values is relative volatility.

Applying Raoult’s law:

Relative volatility becomes ratio of vapor pressures.

 Bubble-point and Dew-point
Bubble-point – initial boiling point of the liquid mixture.
Dew-point – initial condensation temperature.
Required for distillation. Basic equations are:

Where Nc is the number of component.

Usage: Assume temperature  Find Ki from published
data or obtained from vapor pressure data and known
total pressure  if summation exceed 1.0 the lower the
temperature.
If summation exactly equals 1.0, then, the
composition of the vapor in equilibrium with this
liquid is given directly by term Kixi.
When the summation is close to 1.0, the vapor
composition can be determined with little error from
the relative contribution of each term:

A similar procedure is use to determine the dew point

of a vapor mixture and composition of the liquid in
equilibrium with this mixture.
Flash Distillation
In which petroleum fractions are heated in pipe stills and
then heated fluid is flashed into vapor and residual liquid
streams, each containing many component.
If f be the molal fraction of feed that is vaporized and
withdraw continuously then based on one more by mass
balance for more volatile component:

where:
yD = mol. fr. in vapor
xB = mol. fr. in liquid
For multi-component mixture the equation can be
written for each component:

Since the distillate and bottom streams are in

equilibrium, the equation may be changed to:

Solving for xBi and summing over N component:

Fractionation of Multi-component Mixtures
As in binary mixtures, ideal plates are assumed and
number of stages is subsequently corrected for pate
efficiencies.
Two limiting conditions of total and minimum reflux
also helps to validate the design.
Calculation of distillation column are made either by:
Assuming desire separation and numbers of plates
above and below the feed are calculated.
Assuming number of plates and separation of
component is calculated using reflux from
condenser and vapor from re-boiler.
If α is temperature dependent then plate by plate
calculation are necessary.
The temperature and liquid composition for plate n+1 are
calculated by values already known for plate n.
The calculation proceeds from plate to plate up or down
the column.
Procedure will be:
Estimate of minimum number of plates at infinite reflux
and minimum reflux with an infinite number of plates are
made by assuming constant relative volatilities.
Two empirical relations for the number of plates at an
operating reflux are then described.
 Key component
In binary distillation, we choose light key component for
calculating theoretical number of stages.
In multi-component distillation there are three or more
components and specifying concentration of one component
in each stream dose not fully characterized the products.
If concentration of two out of three or three out of four
component are specified then it is impossible to meet these
specification exactly.
An increase in reflux ratio or number of plates would
increase the sharpness and desire concentration of one
component in each product could be achieved but it would
be coincidence if other concentrations exactly match those
specified beforehand.
The designer generally chooses whose concentration in
distillate and bottom product are a good index of the
separation achieved.
After these component are identified, they are called key
components.
The more volatile one is called light key component and
less volatile one is called heavy key component.
LKC in bottom and HKC in distillate are then assigned
small values.
The distillate may be nearly pure if the keys are two most
volatile component.
Any component heavier than the HKC will most often
recovered in bottom product.
Any component lighter than the LKC will most often
recovered in distillate product.
An exception to these generalizations are encountered in
distillation of very close boiling materials, such as isomers.
Unlike binary distillation the choice of two key component
does not give determinate mass balances and equilibrium
calculations are required to calculate concentration of dew
point vapor from top plate and bubble point liquid leaving
the re-boiler.
Although any two component can be nominated, usually
they are adjacent in rank of volatility. Such choice is called
sharp separation.
In sharp separation the keys are the only components that
appear in both products that is distillate and bottom in
appreciable concentration.
Minimum number of plates
Minimum number of plates can be found from Fenske
equation at total reflux :
Minimum reflux ratio
The minimum reflux ratio for multi-component distillation
has the same significance as for binary distillation.
At this reflux ratio, the desire separation is just barely
possible but with infinite number of stages.
The minimum reflux ratio is the guide for choosing optimum
reflux ratio for operating column and number of stages.
For multi-component system, the desired separation is
usually refers to the amount of light key recovered in
distillation and amount of heavy key recovered in bottom
product.
The actual mol. fraction in products is not specified.
 An approximate value of minimum reflux ratio can be
obtained by treating the mixture as pseudo-binary.
Taking only the moles of LK and HK to make a new
pseudo-feed.
Product composition can be calculated along vapor-liquid
equilibrium curve based of αLK-HK.
Then RDm can be determine in a conventional way as for
binary mixture.
An alternate way for saturated liquid feed gives the
minimum ratio of liquid rate to feed rate for a binary
mixture of A and B.

For multi-component distillation A – is LK component and

Distributed and undistributed component
To help estimate the products composition the concept of
distributed and undistributed components are introduced.
Distributed component is found in both distillate and
bottom products whereas undistributed component is
found in only one product.
LK and HK are always distributed as are any component
having volatilities between these two.
Non-key component are almost completely recovered in
one of the products (distillate or bottom) and therefore can
be considered as undistributed.
For the case of minimum reflux ratio the distinction
between distributed and non-distributed is clear because of
large number of plates are involved for separation.
Consider what is required for a heavy component to be
completely absent from distillate then xD is zero then
equation for upper part of the column becomes:

 and for ideal stage

And the ratio of vapor concentration for successive plates

is:

Similarly light component are undistributed at minimum

reflux then xB is assumed zero and the bottom operating
line becomes:
Calculation for minimum reflux ratio
At minimum reflux ratio, there are invariant zones above
and below the feed plate where the composition of liquid
and vapor do not change.
These zones are similar to pinch regions as in binary
distillation.
If there are undistributed components in the feed, their
concentrations change from plate to plate near the feed and
the reduced to zero when the invariant zone is reached.
Thus liquid in upper invariant zone will have all component
except heavy undistributed ones.
At lower invariant zone, all component except light
undistributed one will be present.
The two invariant zone will be at different temperature
and have different liquid and vapor composition because
of undistributed components.
If the undistributed components are a small fraction in
feed, temperature of two invariant zones are same and
calculation of minimum reflux ratio is relative easy.
When these temperature differ considerably, exact
calculation of minimum reflux ratio is difficult because of
relative volatilities in the two zones are different.
Material balance equation for each component in upper
section of column with yn/K in place of xn.

In invariant zone yn+1, i = yn, i.

Rearranging:

Summed for all component:

Similar treatment for lower section of the column:
To determine minimum reflux a value of RD is assumed.
Which gives L/V and D/V.
The temperature at which the equation for upper section is
satisfied with all term positive is determined by trail.
The flow rates in the lower section of column are then
calculated from feed condition: L’ = L + qF and V’ = V –
(1-q)F where q is number of moles of liquid entering the
stripping section per moles of feed.
Temperature that satisfied equation for lower section is
again found with all positive terms.
If the calculated temperature is same or in the wrong order
then the assumed RD is incorrect and calculation is
repeated for lower RD.
 Underwood method
An approximate but fairly accurate method for determining
RDm was developed by Underwood.
Relative volatilities for each component is taken to be
constant in upper and lower invariant zones and constant
molal flow is assumed.
The equation for invariant zone are written in terms of
relative volatility with heavy key generally taken as a
reference component.

The correct root φ of the equation must lies between values of

α for keys.
Self Reading
Calculation required for reflux ratio and concentration
profiles (Plate by plate calculation by Lewis –
Matheson method).
Rate-based models for multi-component distillation.
Number of ideal plates at operating reflux
Although the precise
calculation of the number of
plates in multi-component
distillation is best
accomplished by computer.
For preliminary calculation
a simple method due to
Gilliland is used.
The correlation requires the
knowledge of minimum
number of plates and
minimum reflux ratio.
Problems
Assignment
AZEOTROPIC AND EXTRACTIVE DISTILLATION
In the systems considered previously, the vapor becomes
steadily richer in the more volatile component on successive
plates.
There are two types of mixture where this steady increase in
the concentration of the more volatile component, either does
not take place, or it takes place so slowly that an uneconomic
number of plates is required.
If, for example, a mixture of ethanol and water is distilled, the
concentration of the alcohol steadily increases until it reaches
96 per cent by mass, when the composition of the vapor equals
that of the liquid, and no further enrichment occurs. This
mixture is called an azeotrope, and it cannot be separated by
straight forward method.
A large number of azeotropic mixtures have been found, some
of which are of great industrial importance, such as water-
 When the activity coefficient is
greater than unity, giving a
positive deviation from
Raoult’s law, the molecules of
the components in the system
repel each other and exert a
higher partial pressure than if
their behavior were ideal.
 This leads to the formation of a
“minimum boiling” azeotrope.
 For values of the activity
coefficient less than unity,
negative deviation from
Raoult’s law results in a lower
partial pressure and the
formation of a “maximum
boiling” azeotrope.
The second type of problem occurs where the relative volatility of a
binary mixture is very low, in which case continuous distillation of
the mixture to give nearly pure products will require high reflux
ratios with correspondingly high heat requirements. In addition, it
will necessitate a tower of large cross-section containing many trays.
An example of the second type of problem is the separation of n-
heptane from methyl cyclohexane in which the relative volatility is
only 1.08 and a large number of plates is required to achieve
separation.
The principle of azeotropic and of extractive distillation lies in the
addition of a new substance to the mixture so as to increase the
relative volatility of the two key components, and thus make the
separation relatively easy.
In azeotropic distillation the substance added forms an
azeotrope with one or more of the components in the mixture,
and as a result is present on most of the plates of the column
in appreciable concentrations.
With extractive distillation the substance added is relatively
non-volatile compared with the components to be separated,
and it is therefore fed continuously near the top of the
column.
This extractive agent runs down the column as reflux and is
present in appreciable concentrations on all the plates.
The third component added to the binary mixture is
sometimes known as the entrainer or the solvent.
Azeotropic distillation
Extractive distillation
Extractive distillation is a method of rectification similar
in purpose to azeotropic distillation.
To a binary mixture which is difficult or impossible to
separate by ordinary means, a third component, termed a
solvent, is added which alters the relative volatility of the
original constituents, thus permitting the separation.
The added solvent is, however, of low volatility and is
itself not appreciably vaporised in the fractionator.
The solvent added to the mixture in extractive distillation
differentially affects the activities of the two components,
and hence the relative volatility, α.
An example of extractive
distillation is the separation of
toluene, boiling point 384K,
from paraffin hydrocarbons of
approximately the same
molecular weight.
Using iso-octane of boiling
point 372.5 K, as an example
of a paraffin hydrocarbon.
In the presence of phenol,
boiling point 454.6 K,
however, the relative volatility
of iso-octane increases, so
that, with as much as 83 mole
per cent phenol in the liquid,
the separation from toluene is
relatively easy.
Azeotropic distillation
Separation of original mixture may also be enhanced by
adding a solvent that forms an azeotrope with one of the key
components.
The process is called azeotropic distillation.
The azetrope forms the distillate or bottom product from the
column and is later separated into solvent and key
component.
Usually the material added forms a low-boiling azeotrope
and is taken overhead, and such material are called
entrainers.
The azeotrope will of course contain some of all components
in the feed, but it will have a much different ratio of the keys
than the feed.
 Extractive distillation
The separation of component that have nearly same boiling points
is difficult by simple distillation because of azeotrope formation.
For such system the separation can often improved by adding a
third component to alter the relative volatilities of original
components.
The added component may be higher boiling point liquid or
“solvent” that is miscible with both of the key components but is
chemically more similar to one of them.
The key component that is more like a solvent will have lower
activity coefficient then the other, so separation is enhanced.
This process is called extractive distillation and is like liquid-
liquid extraction with added vapor phase.
 STEAM DISTILLATION
Where a material to be distilled has a high boiling point, and
particularly where decomposition might occur if direct
distillation is employed, the process of steam distillation
may be used.
Steam is passed directly into the liquid in the still and the
solubility of the steam in the liquid must be very low. Steam
distillation is perhaps the most common example of
differential distillation.
Two cases may be considered. The steam may be
superheated and so provide sufficient heat to vaporize the
material concerned, without itself condensing.
Alternatively, some of the steam may condense, producing a
liquid water phase.