CPE230 Dr Alasdair Campbell

4. Absorption and Stripping

4.1 Introduction

In the past few lectures you have looked in detail at the theory of vapour-liquid equilibrium.
Now we seek to establish how we may exploit the science of vapour-liquid equilibrium; this
is where we switch into ‘engineer mode’.

In the next few lectures we are going to look at gas absorption and stripping in staged
processes such as tray columns. We are going to introduce the concept of the operating line,
which is really just a mass balance. We are going to write and plot suitable equilibrium data
and construct a solution for the number of stages required to get from the required inlet to the
desired outlet concentrations.

4.2 Definitions

Gas absorption (scrubbing) is the process of removing one or more components from a gas
phase by causing them to dissolve in a liquid.

Gas stripping is the process of removing one or more solutes from a liquid phase by causing
them to join a gas phase. This can be used to remove volatile contaminants from waste water
streams.

Table 3. Summary of some key industrial systems

Absorption
Solute Absorbent
Sulphur Dioxide (power station combustion gases) Water
Carbon Dioxide NaOHaq
Ammonia Water
Benzene and Toluene Hydrocarbon Oil
Ethanol Water
Stripping
VOCs (in waste water) Air
PAH (from contaminated soil) Steam

In general, we assume that there is at least one component in the gas phase which is insoluble
in the liquid phase and that there is at least one component in the liquid phase that is non-
volatile and will not evaporate into the gas phase. The only component that moves between
streams is the solute. These inert components are the major clue as to how to perform the
analysis.

Both gas absorption and stripping can be performed in both co-current and counter-current
equipment but for our purposes in this module we will take almost no notice of co-current
operations.

Both gas absorption and stripping can be performed in packed towers, tray towers, bubble
columns and spray towers. The laws of gravity mean that tray towers must be counter-

Absorption and Stripping 58 Separation Processes

CPE230 Dr Alasdair Campbell

current but the other three can be either counter- or co-current. In this module we are going
to concentrate on tray towers exclusively; you have packed towers in the Mass Transfer
section of the module.

4.3 Gas – Liquid Solubility and Equilibria

We have developed the theory above to help us tackle equilibrium. Here, we begin to exploit
some of the key results. The solubility of gases in liquids is a complex topic however there
are two basic truisms which may be useful

1) At low concentrations, the relationship between gas and liquid phase compositions
becomes linear and this limiting behaviour is known as Henry’s Law.

2) Usually an increase in temperature causes gas to become less soluble in liquids.

(Think about all the bubbles that appear as you boil a pot of water)

We may be able to develop methods to decide how many ideal stages are required for a
process without resort to any theoretical description of the equilibrium, with luck we may be
able to work straight from the raw experimental data.

4.3.1 Henry’s Law

Dilute
Henry’s law is an empirical observation that, at low concentration, gas solubility becomes
linear, hence we will write

𝑦𝑦 = 𝐻𝐻𝑒𝑒 𝑥𝑥 (155)

where we expect the Henry’s Law constant, He, to be dependent on temperature and pressure
and to be independent of composition only at rather low values of x.

Rich Mixtures (Bulk Absorption)

When we have inert components present we work in terms of mass (or mole, it does not
matter) ratios and we will write Henry’s law as follows:

𝑌𝑌 = 𝐻𝐻𝐻𝐻 (156)

and again we expect this to apply to rather small values of X and Y only. We will consider
these molar ratios more below.

For a gas that is very strongly soluble in the liquid solvent, the value of He will be very small
and this is what we want for an absorber. For stripping, we want a solute which is less
strongly soluble in the liquid and we seek to increase the value of He, either by raising the
temperature or by reducing the pressure or both.

Table 4. Henry’s Law coefficients for gases in water at 1 bar, 20 °C

N2 O2 CO2 SO2 NH3
He 81 × 103 40 × 103 1420 12.2 0.63

Absorption and Stripping 59 Separation Processes

CPE230 Dr Alasdair Campbell

Equilibrium data or equations tell us how the stream LEAVING each stage relate to one
another.

4.3.2 Direct Empirical Data

Y or y
Measured data is often the only realistic way to
Notice they
manage complex systems over wide ranges of all get
composition. These data are often very system straighter
specific and therefore commercially sensitive. down here
General purpose methods to predict solubility of
gases in liquids are still unreliable despite
considerable research effort. In Figure 24 are
shown typical shapes of equilibrium data over a
rather wide range of x and y
X or x
Figure 23 Typical Equilibria
4.4 Enthalpy Balance & Non-Isothermal
Operation

Usually, it is possible to assume that dilute systems are also isothermal. This will be a good
approximation when the heat of absorption is small compared to the amounts of the inert
components present.

However, for bulk systems in which there are strong reactions, considerable heat may be
generated (or absorbed) and there may be substantial changes of temperature across the
system. The most extreme example usually encountered is the use of concentrated sulphuric
acid to absorb water from chlorine gas that has been generated by electrolysis and is therefore
initially saturated with water.

What to do…

1) Always calculate the overall heat balance to check that an isothermal assumption is
justified if one has been made.
2) Consider intermediate cooling (or heating) within the process to make it more
isothermal.
3) Consider increasing the flow rate of an inert liquid component to ‘drown out’ the
effect the heat of absorption might have on the temperature of the system.
4) Calculate the system as a series of smaller sub-systems which can be assumed to be
‘internally isothermal’ but which differ in temperature from each other as calculated
from the energy balance.

Absorption and Stripping 60 Separation Processes

CPE230 Dr Alasdair Campbell

4.5 Trays and Tray Numbering

4.5.1 Tray Configuration and Numbering

There are many different types and designs

of trays and we will look at how they work
in much greater detail in another lecture.

The objective is to bring the liquid and gas

into contact so that mass transfer processes
can occur, then to allow the two phases to
separate so that the gas can continue up the
column (without carrying too much liquid
with it) and the liquid can carry on down
the column without taking too many
bubbles of gas with it.

The most basic design is just a perforated

plate, called a sieve tray, where the
pressure drop for the gas being pushed up
through the holes is all that prevents the
liquid from dripping down through.
Figure 24. Schematic Representation of Trays

We will always number the trays from the top of the column, and the vapour and liquid
leaving the tray will be numbered from their tray of origin as shown in Figure 24. Beware:
textbooks vary!

4.5.2 The Ideal Stage

In this lecture we are going to look at the number of ideal stages required, where we define
an ideal stage as one where the gas and liquid leave the stage in equilibrium.

LF xF Ideal Stage 1 x1 L1
Liquid
Feed Equilibrium Stage 1 Subscripts tell us the stage being
G yF y1 G1
Gas Theoretical Tray 1

Figure 25. Schematic of an ideal stage

Assumptions

(i) Operation is at steady state

(ii) The carrier gas is insoluble in the liquid solvent
(iii) The liquid solvent is involatile
(iv) The stage is isobaric and isothermal

Absorption and Stripping 61 Separation Processes

CPE230 Dr Alasdair Campbell

Hence if tray i is an ideal stage, then xi and yi are in equilibrium.

4.6 Trays Versus Packing

This part of the module is about staged processes but we should remember that the majority
of absorption and stripping operations are performed in packed columns and not tray
columns.

The reverse is probably still true for distillation: there are more tray columns than packed
columns for distillation.

Why… and how do we choose?

There are three main reasons: cost, cost and cost!

Here are some rules of thumb:

1) Packing is more difficult and expensive in very tall or very fat systems – in tall
systems it can start to crush itself, and in fat systems it can be difficult to stop the gas
and liquid segregating and bypassing disastrously.
2) Trays are complicated and cannot be fully mass produced and tend to be more
expensive to purchase and install than packing.
3) Trays are too expensive to be eaten by corrosive fluids, usually packings are used in
extreme applications.
4) Trays are used in applications where large variations in gas or liquid rate or both have
to be survived. Hence the use of bubble cap or valve trays. Packing can dry out or
flood too easily over too small a range of L and G.
5) Packing is more common in gas absorption than in distillation because tray
efficiencies can be rather low in gas absorption or stripping. (but sometimes it is still
economic to use trays where the tray efficiency has dropped to as little as 20%)
6) Systems that form foam may have to be handled in tray columns even when the tray
spacing becomes rather large as a result. It is very difficult to handle foaming systems
in a packed column. (Contact your local supplier of foam inhibitors?)
7) In dirty systems, e.g. where there are contained solids, or solids can be formed or
precipitated to start fouling packings and trays, fouling is usually less of a problem for
packings than it is for trays.
8) In some systems, such as systems where it is necessary to operate at vacuum, pressure
drop can be a critical problem. Vacuum systems also suffer from vast differences
between the vapour and liquid volumetric flow rates. Modern, high-performance
packings (which can be every bit as expensive as trays) may be the only feasible
option.

4.7 Combined Absorption and Stripping Systems

It is common to combine absorption and stripping in a system as illustrated below. There is a

solvent circulated between the two columns which has been selected to remove selectively
the required component from the primary gas stream. The solvent, laden with this component
is then transferred to a stripping column where the solvent is regenerated and the removed

Absorption and Stripping 62 Separation Processes

CPE230 Dr Alasdair Campbell

component is released in concentrated form. Either column can be either packed or a tray
column.

Gas 1 A in Gas 2

Absorber

Stripper

A in Gas 1 Gas 2
Figure 26 Combined Absorber & Stripper

4.8 Operating Lines –Dilute vs Bulk

There are broadly two approaches that may need to be taken to the construction of the
operating lines for a system and the deciding factor concerns whether the liquid or vapour
flow rates change significantly as a result of the mass transfer between the phases.

A system is said to be ‘dilute’ if we can assume that the total liquid and vapour rates do not
change with position in the equipment. In this case we can formulate an operating line based
on the total flow and the mass or mole fractions of the absorbable species.

If the system cannot be described as dilute then we usually describe the process as bulk
absorption or bulk stripping. In ‘bulk’ processes we normally work in mass or mole ratios,
referenced to an inert component in each phase where the inter-phase transport is small or
negligible.

Absorption and Stripping 63 Separation Processes

CPE230 Dr Alasdair Campbell

Operating Lines for Dilute Counter-Current Staged Absorption or Stripping Systems

The principle for staged columns is the same as for packed columns but the notation may be
different. We assume here that there is a constant total liquid flow rate L (kmol/s or kg/s) and
a constant total gas rate G (kmol/s or kg/s). As usual we number the trays from the top.

Figure 27. Schematic for construction of operating line

Component Balance (on the whole column)

𝐿𝐿𝑥𝑥𝑇𝑇 + 𝐺𝐺𝑦𝑦𝐵𝐵 = 𝐿𝐿𝑥𝑥𝐵𝐵 + 𝐺𝐺𝑦𝑦𝑇𝑇 (157)

Taking the mass balance between the i th and the i + 1th stage and round the top of the
column gives us the familiar operating line where x and y are liquid and gas mole or mass
fractions:

𝐿𝐿𝑥𝑥𝑇𝑇 + 𝐺𝐺𝑦𝑦𝑖𝑖+1 = 𝐿𝐿𝑥𝑥𝑖𝑖 + 𝐺𝐺𝑦𝑦𝑇𝑇 (158)

Which can be rearranged:

Absorption and Stripping 64 Separation Processes

CPE230 Dr Alasdair Campbell

𝐿𝐿 𝐿𝐿
𝑦𝑦𝑖𝑖+1 = 𝑥𝑥𝑖𝑖 + �𝑦𝑦𝑇𝑇 − 𝑥𝑥𝑇𝑇 � (159)
𝐺𝐺 𝐺𝐺

This defines our operating line, which describes how PASSING STREAMS relate to one
another. It is simply a way of representing a MASS BALANCE.

This is clearly a straight line only provided L and G are constant.

Example 10:

A kerosene stream with a flow rate of 1500 kg/hr is used to remove nicotine from an aqueous
stream containing 0.01 kg of nicotine per kg of solution and flowing at 1000 kg/hr. Assuming
that at the temperature concerned Henry’s Law holds with He = 0.9 (x is the mass fraction of
nicotine in the kerosene stream, y is the mass fraction in the aqueous), calculate the
performance of two equilibrium stages arranged:

(a) Co-currently
(b) In cross flow, with uniform distribution of kerosene
(c) Counter-currently

Note that this is not a gas absorption problem (!), it’s actually liquid-liquid extraction, but we
can treat it the same as absorption because the aqueous and organic streams are immiscible.

Solution:

L = 1500 kg/hr, G = 1000 kg/hr, He = 0.9, xF = 0

(a) xF = 0 x1 x2
L
1 2
yF = 0.01 y1 y2 G

If equilibrium is reached in stage 1 then stage 2 does nothing! Only need to consider the first
stage.

𝐿𝐿𝑥𝑥𝐹𝐹 + 𝐺𝐺𝑦𝑦𝐹𝐹 = 𝐿𝐿𝑥𝑥1 + 𝐺𝐺𝑦𝑦1

𝐿𝐿
𝑦𝑦1 = − 𝑥𝑥 + 𝑦𝑦𝐹𝐹
𝐺𝐺 1
𝐿𝐿 𝑦𝑦1
𝑦𝑦1 = − + 𝑦𝑦𝐹𝐹
𝐺𝐺 𝐻𝐻𝑒𝑒
𝐿𝐿
𝑦𝑦1 �1 + � = 𝑦𝑦𝐹𝐹
𝐺𝐺𝐻𝐻𝑒𝑒

Absorption and Stripping 65 Separation Processes

CPE230 Dr Alasdair Campbell

𝑦𝑦𝐹𝐹 0.01
𝑦𝑦1 = = = 0.00375
𝐿𝐿 1500
�1 + 𝐺𝐺𝐻𝐻 � 1 + 0.9 × 1000
𝑒𝑒

L/2 L/2
(b)
xF = 0 xF = 0

yF = 0.01 y1 y2
1 2 G

x1 x2

Consider stage 1

𝐿𝐿 𝐿𝐿
𝑥𝑥𝐹𝐹 + 𝐺𝐺𝑦𝑦𝐹𝐹 = 𝑥𝑥1 + 𝐺𝐺𝑦𝑦1
2 2
𝐿𝐿
𝑦𝑦1 = − 𝑥𝑥 + 𝑦𝑦𝐹𝐹
2𝐺𝐺 1
𝐿𝐿 𝑦𝑦1
𝑦𝑦1 = − + 𝑦𝑦𝐹𝐹
2𝐺𝐺 𝐻𝐻𝑒𝑒
𝐿𝐿
𝑦𝑦1 �1 + � = 𝑦𝑦𝐹𝐹
2𝐺𝐺𝐻𝐻𝑒𝑒

𝑦𝑦𝐹𝐹 0.01
𝑦𝑦1 = = = 0.00545
𝐿𝐿 1500
�1 + 2𝐺𝐺𝐻𝐻 � 1 + 2 × 0.9 × 1000
𝑒𝑒

Similarly for stage 2:

𝑦𝑦1 0.00545
𝑦𝑦2 = = = 0.00298
𝐿𝐿 1500
�1 + 2𝐺𝐺𝐻𝐻 � 1 + 2 × 0.9 × 1000
𝑒𝑒

(c) xF = 0 x1 x2
L
y1 1 y2 2 yF = 0.01
G

Consider stage 1

Absorption and Stripping 66 Separation Processes

CPE230 Dr Alasdair Campbell

𝐿𝐿𝑥𝑥𝐹𝐹 + 𝐺𝐺𝑦𝑦2 = 𝐿𝐿𝑥𝑥1 + 𝐺𝐺𝑦𝑦1

𝐿𝐿
𝑦𝑦2 = 𝑥𝑥 + 𝑦𝑦1
𝐺𝐺 1
𝐿𝐿 𝑦𝑦1
𝑦𝑦2 = + 𝑦𝑦1
𝐺𝐺 𝐻𝐻𝑒𝑒
𝐿𝐿
𝑦𝑦2 = 𝑦𝑦1 �1 + �
𝐺𝐺𝐻𝐻𝑒𝑒

Consider stage 2

𝐿𝐿𝑥𝑥1 + 𝐺𝐺𝑦𝑦𝐹𝐹 = 𝐿𝐿𝑥𝑥2 + 𝐺𝐺𝑦𝑦2

𝐿𝐿 𝐿𝐿
𝑦𝑦2 = 𝑥𝑥1 + 𝑦𝑦𝐹𝐹 − 𝑥𝑥2
𝐺𝐺 𝐺𝐺
𝐿𝐿 𝐿𝐿 𝑦𝑦1
𝑦𝑦2 �1 + �= + 𝑦𝑦𝐹𝐹
𝐺𝐺𝐻𝐻𝑒𝑒 𝐺𝐺 𝐻𝐻𝑒𝑒

𝐿𝐿 𝐿𝐿 𝐿𝐿
𝑦𝑦1 �1 + � �1 + �− 𝑦𝑦 = 𝑦𝑦𝐹𝐹
𝐺𝐺𝐻𝐻𝑒𝑒 𝐺𝐺𝐻𝐻𝑒𝑒 𝐺𝐺𝐻𝐻𝑒𝑒 1
𝑦𝑦𝐹𝐹 0.01
𝑦𝑦1 = = 2 = 0.00184
𝐿𝐿 𝐿𝐿 2 1500 1500
�1 + 𝐺𝐺𝐻𝐻 + �𝐺𝐺𝐻𝐻 � � 1 +
𝑒𝑒 𝑒𝑒 0.9 × 1000 + �0.9 × 1000�

Why is the counter-current cascade superior? We use all of the purest solvent on the purer gas
stream (therefore maximising the driving force for transfer, because we’re as far from
equilibrium as possible) and then use the ‘pure enough’ solvent on the richer gas stream. This
generalises to N stages.

4.9 Graphical Construction for Design/Rating Calculations (McCabe-Thiele Method)

We derived an operating line which describes a mass balance on passing streams in an

absorber or a stripper. We now want to investigate how we can use this operating line to
design our column. We know that passing streams are described by the operating line and the
streams leaving each stage are given by the equilibrium line. We should therefore consider a
graphical solution to solve this problem. To see how many stages are required we therefore
need to step of between the operating and equilibrium lines. This is the McCabe-Thiele
method.

4.9.1 Absorption

The highest concentrations in the system appear at the bottom of the column and we see that
the operating line and equilibrium lines will be arranged as shown in Figure 28. We know
that the operating line has to be to the left of the equilibrium line because we put in solvent at

Absorption and Stripping 67 Separation Processes

CPE230 Dr Alasdair Campbell

the top of the column which has almost none of the solute in it (think about chemical
potentials, or fugacities…).

Bottom
yF
Tray 3
Operating Line

Tray 2

y1
T op Tray 1
≡m

xF xB x
Figure
Figure 15.1 Absorption
28. McCabe-Thiele constructionof
foraabsorption
Dilute Component
of a dilute component

4.9.2 Stripping

In stripping, the highest concentrations are at the top of the column and the operating line
must lie to the right of the equilibrium line because the liquid feed to the column is very rich
in solute.

Absorption and Stripping 68 Separation Processes

CPE230 Dr Alasdair Campbell

y
4
yout Top

yin Bottom
x out x in x
Figure 29 McCabe-Thiele construction for stripping of a dilute component

4.9.3 The Design Problem

In absorption we are usually given G, yF, yT and xF and our job is to specify L and the number
of trays required.

In stripping we are usually given L, xF, xB and yF and we have to specify G and the number of
trays required.

4.9.4 The Rating Problem

In the rating case, the column already exists, so we know the number of stages. In addition
we know G, L, yF, xF and our job for an absorber is to find yT and for a stripper is to find xB.

4.10 Analytical Solution For Henry’s Law Systems

For very dilute systems we may also be able to assume that the equilibrium is a straight line,
i.e. that Henry’s law applies

For absorption, assume we succeed with our separation, then

𝑦𝑦1 = 𝑦𝑦𝑇𝑇 (160)

(i.e. the gas stream emerging from the top stage (1) meets our exit criterion) and from the first
ideal stage the liquid in equilibrium with this gas is x1 where

𝑦𝑦1 = 𝐻𝐻𝑒𝑒 𝑥𝑥1 (161)

Absorption and Stripping 69 Separation Processes

CPE230 Dr Alasdair Campbell

and this liquid passes gas of composition y2 going up the column and according to the
operating line

𝐿𝐿 𝐿𝐿
𝑦𝑦2 = 𝑥𝑥1 + �𝑦𝑦1 − 𝑥𝑥𝑇𝑇 � (162)
𝐺𝐺 𝐺𝐺

and we can substitute for x1 in Equation (161) from Equation (162) to give

𝐿𝐿 𝐿𝐿
𝑦𝑦2 = 𝑦𝑦1 + �𝑦𝑦1 − 𝑥𝑥𝑇𝑇 � (163)
𝐺𝐺𝐻𝐻𝑒𝑒 𝐺𝐺

G, y1 L, xT
Bottom
yB

Operating Line:
slope = L/G

y1
Top
Equilibrium Line:
slope = He

xT xB
G, yB L, xB

Figure 30 Dilute System, Linear Equilibrium

At this point it is conventional to define an absorption factor, it is the ratio of the slopes of the
operating line to the equilibrium line.

𝐿𝐿
𝐴𝐴 ≝ (164)
𝐺𝐺𝐻𝐻𝑒𝑒

This absorption factor is analogous to the relative volatility, α.

Equation (163) becomes

𝐿𝐿 𝐿𝐿 𝐿𝐿
𝑦𝑦2 = 𝐴𝐴𝑦𝑦1 + �𝑦𝑦1 − 𝑥𝑥𝑇𝑇 � = 𝐴𝐴𝑦𝑦1 + 𝑦𝑦1 − 𝑥𝑥𝑇𝑇 = 𝐴𝐴𝑦𝑦1 + 𝑦𝑦1 − 𝑦𝑦𝑇𝑇∗
𝐺𝐺 𝐺𝐺 𝐺𝐺𝐻𝐻𝑒𝑒
(165)

Here we introduce the new term 𝑦𝑦𝑇𝑇∗ , which is the concentration in the gas phase which would
be in equilibrium with the liquid going in at the top. Note that 𝑦𝑦𝑇𝑇∗ and yT are NOT THE
SAME THING.

We can rearrange Equation (165) to give:

𝑦𝑦2 − 𝑦𝑦1 = 𝑦𝑦2 − 𝑦𝑦𝑇𝑇 = 𝐴𝐴(𝑦𝑦𝑇𝑇 − 𝑦𝑦𝑇𝑇∗ ) (166)

Absorption and Stripping 70 Separation Processes

CPE230 Dr Alasdair Campbell

We could continue using the operating line to link the gas composition after a given plate to
the conditions at the top of the column. It turns out, our lives get easier if instead we consider
a solute balance over a single plate, i.

𝐿𝐿
𝑦𝑦𝑖𝑖+1 − 𝑦𝑦𝑖𝑖 = (𝑥𝑥𝑖𝑖 − 𝑥𝑥𝑖𝑖−1 ) = 𝐴𝐴(𝑦𝑦𝑖𝑖 − 𝑦𝑦𝑖𝑖−1 ) (167)
𝐺𝐺

We can now consider Equation (167) for each plate in our column:

For i=1 𝑦𝑦2 − 𝑦𝑦𝑇𝑇 = 𝐴𝐴(𝑦𝑦𝑇𝑇 − 𝑦𝑦𝑇𝑇∗ )

i=2 𝑦𝑦3 − 𝑦𝑦2 = 𝐴𝐴(𝑦𝑦2 − 𝑦𝑦𝑇𝑇 ) = 𝐴𝐴2 (𝑦𝑦𝑇𝑇 − 𝑦𝑦𝑇𝑇∗ )

i=3 𝑦𝑦4 − 𝑦𝑦3 = 𝐴𝐴(𝑦𝑦3 − 𝑦𝑦2 ) = 𝐴𝐴3 (𝑦𝑦𝑇𝑇 − 𝑦𝑦𝑇𝑇∗ )

⋮ ⋮ ⋮ ⋮

i=N 𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑁𝑁 = 𝐴𝐴(𝑦𝑦𝑁𝑁 − 𝑦𝑦𝑁𝑁−1 ) = 𝐴𝐴𝑁𝑁 (𝑦𝑦𝑇𝑇 − 𝑦𝑦𝑇𝑇∗ )

Adding
𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇 = (𝑦𝑦𝑇𝑇 − 𝑦𝑦𝑇𝑇∗ ) ∑𝑁𝑁
1 𝐴𝐴
𝑛𝑛
(168)

Now a little maths, because the geometric progression 1 + A + A2… has a sum

𝐴𝐴−𝐴𝐴𝑁𝑁+1
∑𝑁𝑁 𝑛𝑛 2 3 𝑁𝑁
1 𝐴𝐴 = 𝐴𝐴 + 𝐴𝐴 + 𝐴𝐴 + ⋯ + 𝐴𝐴 = (169)
1−𝐴𝐴

Hence Equation (168) becomes:

𝑦𝑦𝑇𝑇 −𝑦𝑦𝑇𝑇∗ 1−𝐴𝐴

= (170)
𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇 𝐴𝐴−𝐴𝐴𝑁𝑁+1

which can be rearranged into the form

𝑦𝑦𝑇𝑇 −𝑦𝑦𝑇𝑇∗ 1−𝐴𝐴

= (171)
𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇∗ 1−𝐴𝐴𝑁𝑁+1

Equation (171) is the first of several useful forms of the KREMSER EQUATION.

If we are interested in designing a column, we need to solve Equation (171) for N, the
number of stages (assuming that we know our desired gas exit composition yT, and we have
specified the feed compositions xT and yB, as well as A)

𝑦𝑦 −𝑦𝑦 ∗
log�(1−𝐴𝐴−1 )� 𝐵𝐵 ∗𝑇𝑇 �+𝐴𝐴−1 �
𝑦𝑦𝑇𝑇 −𝑦𝑦𝑇𝑇
𝑁𝑁 = (172)
log 𝐴𝐴

Absorption and Stripping 71 Separation Processes

CPE230 Dr Alasdair Campbell

If we want to rate a column, i.e. find out the exit conditions for operation of an existing
column, equation (171) can be re-arranged into the form:

𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇 𝐴𝐴−𝐴𝐴𝑁𝑁+1

= (173)
𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇∗ 1−𝐴𝐴𝑁𝑁+1

If we consider the special case where the liquid is initially free from the solute (i.e. xT = 0 and
hence 𝑦𝑦𝑇𝑇∗ = 0) then Equation (171) becomes

𝑦𝑦𝑇𝑇 solute leaving in gas stream 1−𝐴𝐴

= = (174)
𝑦𝑦𝐵𝐵 solute entering in gas feed 1−𝐴𝐴𝑁𝑁+1

Similarly (173) becomes

𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇 solute transfered to liquid 𝐴𝐴−𝐴𝐴𝑁𝑁+1

= = (175)
𝑦𝑦𝐵𝐵 solute entering in gas feed 1−𝐴𝐴𝑁𝑁+1

The number of stages becomes

𝑦𝑦
log�(1−𝐴𝐴−1 )� 𝐵𝐵 �+𝐴𝐴−1 �
𝑦𝑦𝑇𝑇
𝑁𝑁 = (176)
log 𝐴𝐴

In Figure 31 we see the number of stages required to produce a reduction of y with the
parameter being the absorption factor, A. We note that A can be less than one but very large
numbers of stages are required to produce very modest separations. As A becomes
significantly larger than one very large separations can be achieved with relatively few
stages.

This reflects the fact that as we make the Henry’s law constant smaller and smaller, the gas is
getting more and more soluble in the liquid. Making Henry’s law constant small implies
larger and larger values of A.

We can also make A large by putting more and more liquid through the system and this again
we would expect to reduce the number of stages required.

Absorption and Stripping 72 Separation Processes

CPE230 Dr Alasdair Campbell

1
yT / yB, (assumes solute free feed)

0.1

0.01

0.001

0.0001
1 10 100
Number of Stages, N

Figure 31. Plot of the Kremser Equation for a solute free liquid, with A as a parameter.

Example 11

A kerosene stream with a flow rate of 1500 kg/hr is used to remove nicotine from an aqueous
stream containing 0.01 kg of nicotine per kg of solution and flowing at 1000 kg/hr. Assuming
that at the temperature concerned Henry’s Law holds with He = 0.9 (x is the mass fraction of
nicotine in the kerosene, y is the mass fraction in the aqueous), calculate the performance of
two equilibrium stages arranged counter-currently.

Note that this is not a gas absorption problem, it’s actually liquid-liquid extraction, but we
can treat it the same as absorption because the aqueous and organic streams are immiscible.

xF = 0 x1 x2
L
y1 1 y2 2 yF = 0.01
G

𝐿𝐿 1500
The absorption factor A is given by: 𝐴𝐴 ≝ 𝐺𝐺𝐻𝐻 = 0.9×1000 = 1.667
𝑒𝑒

We know the liquid is initially free of solute, hence: 𝑦𝑦𝑇𝑇∗ = 𝐻𝐻𝑒𝑒 𝑥𝑥𝐹𝐹 = 0

We can choose different forms of the Kremser equations to make sure they tell us the same
things!

𝑦𝑦𝑇𝑇 −𝑦𝑦𝑇𝑇∗ 1−𝐴𝐴

Let’s start with Equation (171): = 𝐴𝐴−𝐴𝐴𝑁𝑁+1
𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇

Absorption and Stripping 73 Separation Processes

CPE230 Dr Alasdair Campbell

𝑦𝑦𝑇𝑇 1 − 𝐴𝐴
=
𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇 𝐴𝐴 − 𝐴𝐴𝑁𝑁+1

Rearranging:

𝑦𝑦𝑇𝑇 (𝐴𝐴 − 𝐴𝐴𝑁𝑁+1 ) = (𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇 )(1 − 𝐴𝐴) = 𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴) − 𝑦𝑦𝑇𝑇 (1 − 𝐴𝐴)

𝑦𝑦𝑇𝑇 (𝐴𝐴 − 𝐴𝐴𝑁𝑁+1 ) + 𝑦𝑦𝑇𝑇 (1 − 𝐴𝐴) = 𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴)

𝑦𝑦𝑇𝑇 (𝐴𝐴 − 𝐴𝐴𝑁𝑁+1 + 1 − 𝐴𝐴) = 𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴)

𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴)
𝑦𝑦𝑇𝑇 =
1 − 𝐴𝐴𝑁𝑁+1

Starting from Equation (172) we have:

𝑦𝑦𝑇𝑇 − 𝑦𝑦𝑇𝑇∗ 1 − 𝐴𝐴
∗ =
𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇 1 − 𝐴𝐴𝑁𝑁+1

𝑦𝑦𝑇𝑇 1 − 𝐴𝐴
=
𝑦𝑦𝐵𝐵 1 − 𝐴𝐴𝑁𝑁+1

𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴)
𝑦𝑦𝑇𝑇 =
1 − 𝐴𝐴𝑁𝑁+1

as above.

Finally, Equation (173) gives us:

𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇 𝐴𝐴 − 𝐴𝐴𝑁𝑁+1
=
𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇∗ 1 − 𝐴𝐴𝑁𝑁+1

𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇 𝐴𝐴 − 𝐴𝐴𝑁𝑁+1

=
𝑦𝑦𝐵𝐵 1 − 𝐴𝐴𝑁𝑁+1

𝑦𝑦𝐵𝐵 (𝐴𝐴 − 𝐴𝐴𝑁𝑁+1 )

𝑦𝑦𝐵𝐵 − 𝑦𝑦𝑇𝑇 =
1 − 𝐴𝐴𝑁𝑁+1

𝑦𝑦𝐵𝐵 (𝐴𝐴 − 𝐴𝐴𝑁𝑁+1 )

𝑦𝑦𝑇𝑇 = 𝑦𝑦𝐵𝐵 −
1 − 𝐴𝐴𝑁𝑁+1

𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴𝑁𝑁+1 ) − 𝑦𝑦𝐵𝐵 (𝐴𝐴 − 𝐴𝐴𝑁𝑁+1 ) 𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴)

𝑦𝑦𝑇𝑇 = =
1 − 𝐴𝐴𝑁𝑁+1 1 − 𝐴𝐴𝑁𝑁+1

We can see that all three forms yield the same expression (which is reassuring!) but a
judicious choice of (172) would save us some algebraic toil.

𝑦𝑦𝐵𝐵 (1 − 𝐴𝐴) 0.01 × (1 − 1.667)

𝑦𝑦𝑇𝑇 = = = 0.00184
1 − 𝐴𝐴𝑁𝑁+1 1 − 1.6672+1

Absorption and Stripping 74 Separation Processes

CPE230 Dr Alasdair Campbell

4.11 Limiting Conditions

4.11.1 Minimum L for Absorption

We recall that usually we have a specified

G and yB to process and yT that we have to Bottom
achieve. The solvent recovery system yB
will also supply a specified xT.
(L/G)min
Figure 32 shows that conditions at the top
of the column are fixed and as we reduce Equilibrium
the L, then the slope of the operating line Line
is reduced and we approach a situation Top
where we need an infinite number of yT
stages. This is when the liquid flow rate
is reduced such that the operating line
touches the equilibrium curve at or xT
before the bottom of the column. Thus, Figure 32 Absorption of a Dilute Component for Lmin
the liquid leaving would be in
equilibrium with the gas entering.

For a system with linear, Henry’s Law equilibrium, then

𝑦𝑦 = 𝐻𝐻𝑒𝑒 𝑥𝑥 (177)

and the operating line is

𝐿𝐿 𝐿𝐿
𝑦𝑦 = 𝑥𝑥 + �𝑦𝑦𝑇𝑇 − 𝑥𝑥𝑇𝑇 � (178)
𝐺𝐺 𝐺𝐺

If these lines intersect at the bottom of the column, then

𝐿𝐿 𝐿𝐿
𝑦𝑦𝐵𝐵 = 𝑥𝑥𝐵𝐵 + �𝑦𝑦𝑇𝑇 − 𝑥𝑥𝑇𝑇 � (179)
𝐺𝐺 𝐺𝐺

and the equilibrium must apply, i.e.

𝑦𝑦𝐵𝐵 = 𝐻𝐻𝑒𝑒 𝑥𝑥𝐵𝐵 (180)

Substitution of Equation (180) into Equation (179) gives

𝐿𝐿 𝑦𝑦𝐵𝐵 𝐿𝐿
𝑦𝑦𝐵𝐵 = �𝐺𝐺 � + �𝑦𝑦𝑇𝑇 − �𝐺𝐺 � 𝑥𝑥𝑇𝑇 � (181)
𝑚𝑚𝑚𝑚𝑚𝑚 𝐻𝐻𝑒𝑒 𝑚𝑚𝑚𝑚𝑚𝑚

and solving for (L / G)min gives

𝐿𝐿 𝑦𝑦 −𝑦𝑦𝑇𝑇
�𝐺𝐺 � = 𝑦𝑦𝐵𝐵𝐵𝐵 (182)
𝑚𝑚𝑚𝑚𝑚𝑚 −𝑥𝑥𝑇𝑇
𝐻𝐻𝑒𝑒

In terms of the minimum absorption parameter we can write

Absorption and Stripping 75 Separation Processes

CPE230 Dr Alasdair Campbell

𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇
𝐴𝐴𝑚𝑚𝑚𝑚𝑚𝑚 = (183)
𝑦𝑦𝐵𝐵 −𝐻𝐻𝑒𝑒 𝑥𝑥𝑇𝑇

We could have also derived these equations by considering an overall balance on the solute,
whilst assuming that the streams at the bottom were at equilibrium i.e. 𝑥𝑥𝐵𝐵 = 𝑥𝑥𝐵𝐵∗ = 𝑦𝑦𝐵𝐵 /𝐻𝐻𝑒𝑒 . As
we approach the minimum value, we approach a PINCH, and hence the number of stages
required tends to infinity. This pinch will occur at the bottom of the column if the equilibrium
line is straight or concave. If the equilibrium line is convex, there exists the possibility of a
tangential pinch somewhere in the middle of the column.

Usually when we are designing an absorber, we are given the gas rate that we must process
and we are asked to select a liquid rate: Equation (182) effectively specifies the minimum
liquid rate.

4.11.2 Minimum Stages for Absorption

We can increase so much the liquid rate that yB

eventually only one ideal stage is required.
This situation is illustrated in Figure 33. It
represents the condition where the liquid out
of the bottom of the tower is in equilibrium
with the gas coming out of the top of the
tower.

Of course, it is possible to put even more

liquid down through the tower (tray
hydraulics permitting) but this would be a yT
waste, or would it?

We have already noted that the tray

efficiency can become rather low in gas xT XB
absorption. It would certainly not be worth
Figure 33 Absorption in a Single Ideal Stage
increasing the liquid rate until the number of
actual trays required was less than one!

If we accept an upper limit of one ideal stage, and if there is linear equilibrium, then a very
simple solution exists as follows

𝐿𝐿 𝐿𝐿
𝑦𝑦𝐵𝐵 = 𝑥𝑥𝐵𝐵 + �𝑦𝑦𝑇𝑇 − 𝑥𝑥𝑇𝑇 � (184)
𝐺𝐺 𝐺𝐺

where the equilibrium applies for one ideal stage

𝑦𝑦𝑇𝑇 = 𝐻𝐻𝑒𝑒 𝑥𝑥𝐵𝐵 (185)

Therefore, we can write

𝐿𝐿 𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇
�𝐺𝐺 � = 𝑦𝑦𝑇𝑇 (186)
𝑚𝑚𝑚𝑚𝑚𝑚 −𝑥𝑥𝑇𝑇
𝐻𝐻𝑒𝑒

Absorption and Stripping 76 Separation Processes

CPE230 Dr Alasdair Campbell

Example 12

Suppose xT = 0, yB = 0.05 and we require yT = 0.0001, and for our solvent He = 0.015

𝐿𝐿 𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇 0.05−0.0001

�𝐺𝐺� = 𝑦𝑦𝐵𝐵 = 0.05 = 0.015
𝑚𝑚𝑚𝑚𝑚𝑚 −𝑥𝑥𝑇𝑇 −0
𝐻𝐻𝑒𝑒 0.015

and

𝐿𝐿 𝑦𝑦𝐵𝐵 −𝑦𝑦𝑇𝑇 0.05−0.0001

�𝐺𝐺� = 𝑦𝑦𝑇𝑇 = 0.0001 = 7.5
𝑚𝑚𝑚𝑚𝑚𝑚 −𝑥𝑥𝑇𝑇 −0
𝐻𝐻𝑒𝑒 0.015

It is also instructive to calculate the number of stages required if we work at, say, twice the
minimum L/G. From equation (176) for a solute free liquid

1 0.05 1
𝑦𝑦
log�(1−𝐴𝐴−1 )� 𝐵𝐵 �+𝐴𝐴−1 � log��1− 2×0.015 �� �+ 2×0.015 �
0.0001
𝑦𝑦𝑇𝑇
𝑁𝑁 = = 0.015
2×0.015
0.015
=8
log 𝐴𝐴 log� 0.015 �

This may look OK unless the tray efficiency has dropped to 10% in which case, we have to
build a column with 80 actual stages. This is not completely out of the question, but your
boss may want to look at the reasons you elected not to use a packed tower very closely!

4.11.3 Minimum G for Stripping

Now the situation is reversed for stripping: yT Top

we are given L and xT, we have to get down
to some xB and we have been given a yB
Equilibrium
(often but not always zero). As we reduce Line
the gas flow rate we will require more and
(L/G)max
more ideal stages until the situation in
Figure 34 is reached and this represents the
minimum gas flow rate. yB Bottom

Again we can extract an analytical solution

for the case where the equilibrium is linear, xB xT
this time, for a change lets work from the Figure 34 Stripping of a Dilute Component for Gmin
geometry of the diagram directly

𝑦𝑦𝑇𝑇 = 𝑦𝑦𝑇𝑇∗ = 𝐻𝐻𝑒𝑒 𝑥𝑥𝑇𝑇 (187)

and

𝐿𝐿 𝑦𝑦𝑇𝑇 −𝑦𝑦𝐵𝐵
�𝐺𝐺 � = (188)
𝑚𝑚𝑚𝑚𝑚𝑚 𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵

Here, we neither know nor care about yT hence we eliminate it to give

Absorption and Stripping 77 Separation Processes

CPE230 Dr Alasdair Campbell

𝐿𝐿 𝐻𝐻𝑒𝑒 𝑥𝑥𝑇𝑇 −𝑦𝑦𝐵𝐵

�𝐺𝐺 � = (189)
𝑚𝑚𝑚𝑚𝑚𝑚 𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵

4.11.4 Minimum Stages for Stripping

Here, again, we will take the rather arbitrary

view that less than one ideal stage is not called
for.

This time, as we increase G, the number of

stages required falls and we seek (L/G)min as yT
illustrated in Figure 35.

Again, for a linear equilibrium, we can analyse

very easily yB
xB xT
Again, we can work straight from the diagram/. Figure 35 Stripping in a Single Ideal Stage

The equilibrium condition is now such that

𝑦𝑦𝑇𝑇 = 𝐻𝐻𝑒𝑒 𝑥𝑥𝐵𝐵 (190)

and the operating line slope is such that

𝐿𝐿 𝑦𝑦𝑇𝑇 −𝑦𝑦𝐵𝐵 𝐻𝐻𝑒𝑒 𝑥𝑥𝐵𝐵 −𝑦𝑦𝐵𝐵

�𝐺𝐺 � = = (191)
𝑚𝑚𝑚𝑚𝑚𝑚 𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵 𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵

Example 13

Consider a solvent where the solute is 10× less soluble than in the previous example, so that
He = 0.15 and xT = 0.02 and yB = 0 and we require a design that can achieve xB = 0.0002.

𝐿𝐿 𝐻𝐻𝑒𝑒 𝑥𝑥𝑇𝑇 −𝑦𝑦𝐵𝐵 0.15×0.02−0

�𝐺𝐺 � = = = 0.152
𝑚𝑚𝑚𝑚𝑚𝑚 𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵 0.02−0.0002

and

𝐿𝐿 𝐻𝐻𝑒𝑒 𝑥𝑥𝐵𝐵 −𝑦𝑦𝐵𝐵 0.15×0.0002−0

�𝐺𝐺 � = = = 0.00152
𝑚𝑚𝑚𝑚𝑚𝑚 𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵 0.02−0.0002

Suppose now we work at twice the minimum G, then we calculate the absorption parameter

𝐿𝐿 1 0.152
𝐴𝐴 = � � = = 0.505
𝐺𝐺 𝑚𝑚𝑚𝑚𝑚𝑚 2𝐻𝐻𝑒𝑒 2×0.15

For stripping, another parameter S which is the stripping factor is also used. This is defined as

𝐺𝐺𝐻𝐻𝑒𝑒 1
𝑆𝑆 ≝ = (192)
𝐿𝐿 𝐴𝐴

Absorption and Stripping 78 Separation Processes

CPE230 Dr Alasdair Campbell

For our example: S = 1.98.

The overall material balance gives us the yT

𝐿𝐿 0.1515
𝑦𝑦𝑇𝑇 = (𝑥𝑥𝑇𝑇 − 𝑥𝑥𝐵𝐵 ) + 𝑦𝑦𝐵𝐵 = (0.02 − 0.0002) + 0 = 0.0015
𝐺𝐺 2

To calculate the number of stages, we can use the Kremser equation for stripping. Rather than
rederive the Kremser Equation for stripping, we can recognise that if we turn the column upside
down and note the changed identities: T ↔ B, L ↔ G, y ↔ x, S ↔ A, then Equation (172)
becomes for stripping:
∗
𝑥𝑥𝐵𝐵 −𝑥𝑥𝐵𝐵 1−𝑆𝑆
∗ = (193)
𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵 1−𝑆𝑆 𝑁𝑁+1

and Equation (172) for the number of stages becomes:

𝑥𝑥 −𝑥𝑥 ∗
log��1−𝑆𝑆 −1 �� 𝑇𝑇 𝐵𝐵 −1
∗ �+𝑆𝑆 �
𝑥𝑥𝐵𝐵 −𝑥𝑥𝐵𝐵
𝑁𝑁 = (194)
log 𝑆𝑆

So for our example:

0.02
log�(1−0.505)� �+0.505�
0.0002
𝑁𝑁 = = 5.73
log 1.98

Just under 6 stages are required.

4.12 The Kremser Equation for Stripping a Dilute System

We begin by numbering the stages from the bottom up rather than the top down and proceed
in the same way as we did for absorption.

We consider a solute balance over a single plate, i.

𝐺𝐺
𝑥𝑥𝑖𝑖 − 𝑥𝑥𝑖𝑖+1 = (𝑦𝑦𝑖𝑖−1 − 𝑦𝑦𝑖𝑖 ) = 𝑆𝑆(𝑥𝑥𝑖𝑖−1 − 𝑥𝑥𝑖𝑖 ) (195)
𝐿𝐿

We can now consider Equation (195) for each plate in our column:

For i=1 𝑥𝑥𝐵𝐵 − 𝑥𝑥2 = 𝑆𝑆(𝑥𝑥𝐵𝐵∗ − 𝑥𝑥𝐵𝐵 )

i=2 𝑥𝑥2 − 𝑥𝑥3 = 𝑆𝑆(𝑥𝑥𝐵𝐵 − 𝑥𝑥2 ) = 𝑆𝑆 2 (𝑥𝑥𝐵𝐵∗ − 𝑥𝑥𝐵𝐵 )

i=3 𝑥𝑥3 − 𝑥𝑥4 = 𝑆𝑆(𝑥𝑥2 − 𝑥𝑥3 ) = 𝑆𝑆 3 (𝑥𝑥𝐵𝐵∗ − 𝑥𝑥𝐵𝐵 )

⋮ ⋮ ⋮ ⋮

i=N 𝑥𝑥𝑁𝑁 − 𝑥𝑥𝑇𝑇 = 𝑆𝑆(𝑥𝑥𝑁𝑁−1 − 𝑥𝑥𝑁𝑁 ) = 𝑆𝑆 𝑁𝑁 (𝑥𝑥𝐵𝐵∗ − 𝑥𝑥𝐵𝐵 )

Absorption and Stripping 79 Separation Processes
CPE230 Dr Alasdair Campbell

Adding 𝑥𝑥𝐵𝐵 − 𝑥𝑥𝑇𝑇 = (𝑥𝑥𝐵𝐵∗ − 𝑥𝑥𝐵𝐵 ) ∑𝑁𝑁

1 𝑆𝑆
𝑛𝑛
(196)

Hence Equation (196) becomes:

∗
𝑥𝑥𝐵𝐵 −𝑥𝑥𝐵𝐵 1−𝑆𝑆
= (197)
𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵 𝑆𝑆−𝑆𝑆 𝑁𝑁+1

Which can be rearranged into the form

∗
𝑥𝑥𝐵𝐵 −𝑥𝑥𝐵𝐵 1−𝑆𝑆
∗ = (198)
𝑥𝑥𝑇𝑇 −𝑥𝑥𝐵𝐵 1−𝑆𝑆 𝑁𝑁+1

Equations (197) and (198) can be compared with (170) and (171) for absorption to confirm
what was said above, that to get from the absorption equation to the stripping equation you
simply swap: T ↔ B, L ↔ G, y ↔ x, S ↔ A.

Solving Equation (198) for N,

𝑥𝑥 −𝑥𝑥 ∗
log�(1−𝑆𝑆 −1 )� 𝑇𝑇 𝐵𝐵 −1
∗ �+𝑆𝑆 �
𝑥𝑥𝐵𝐵 −𝑥𝑥𝐵𝐵
𝑁𝑁 = (199)
log 𝑆𝑆

4.13 ‘Rich’ Absorption and Stripping

We have so far considered the case of dilute absorption and stripping. We now need to remove
this restriction and consider systems that cannot be considered so dilute that the gas and the
liquid phase flow rates can be assumed to be constant. We will continue to assume that the
system is isothermal, even though this is rather less likely in a bulk system.

4.13.1 Basis for Flow Rate and Composition

We will assume that the gas phase has at least one insoluble component with molar flow rate
G´, the prime is to remind us that this is only the flow of the insoluble component.

Similarly, for the liquid phase we assume that there is one non-volatile component in the
solution and that the molar flow is L´.

We will denote the gas phase mole ratio of the distributed component as Y where

𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑖𝑖𝑖𝑖 𝑔𝑔𝑔𝑔𝑔𝑔 𝑦𝑦

𝑌𝑌 ≝ = (200)
𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑔𝑔𝑔𝑔𝑔𝑔 1−𝑦𝑦

and the mole ratio of the distributed component in the liquid phase will be denoted X, where

𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑖𝑖𝑖𝑖 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑥𝑥

𝑋𝑋 ≝ = (201)
𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑖𝑖𝑖𝑖 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 1−𝑥𝑥

Absorption and Stripping 80 Separation Processes

CPE230 Dr Alasdair Campbell

You should note carefully as we go through the analysis that the compositions and flow rates
must all be in either mass and mass ratio units or moles and mole ratio units, but provided they
are all one the same basis it does not otherwise matter which they are.

NB: X and Y can be higher than unity!

4.13.2 Henry’s Law for Bulk Systems

This is almost a contradiction in terms, because it is unlikely that the equilibrium will still be
linear by the time we get to the high concentrations of a bulk system. However, if it is
appropriate and useful we will write the Henry’s law as

𝑌𝑌 = 𝐻𝐻𝐻𝐻 (202)

We have dropped the subscript ‘e’ to remind ourselves that this is Henry’s law for a bulk
system.

4.13.3 Material Balance on Distributed Component

G´, Y1 = YT L´, XT

G´, YB L´, XN = XB

Figure 36 Bulk System, Operating Line

The component balance round the whole column on the distributed (soluble) component is:

𝐿𝐿′ 𝑋𝑋𝑇𝑇 + 𝐺𝐺 ′ 𝑌𝑌𝐵𝐵 = 𝐿𝐿′ 𝑋𝑋𝐵𝐵 + 𝐺𝐺 ′ 𝑌𝑌𝑇𝑇 (203)

and the operating line is

𝐿𝐿′ 𝐿𝐿′
𝑌𝑌𝑖𝑖+1 = 𝑋𝑋 + �𝑌𝑌𝑇𝑇 −
′ 𝑖𝑖
𝑋𝑋𝑇𝑇 � (204)
𝐺𝐺 𝐺𝐺 ′

Now, because L´ and G´ are the flow rates of the inert component that do not change phase,
they are both assumed to be constants and all the theory that we have developed for dilute
systems in our original nomenclature, applied to bulk absorption in our new nomenclature.

Absorption and Stripping 81 Separation Processes

 CPE230 Dr Alasdair Campbell

For instance, in Figure 37 we show the stage to stage construction for both stripping and
absorption in bulk systems
Y
Y
Yy
Y 4
yYout Top
T
YB 3

YT
1

X yYinB Bottom X
xout xin
XT XB X X B XT Xx
Figure 37 Absorption and Stripping of a Bulk Component

4.13.4 Linear Equilibrium

All the previous theory applies, but in the new notation. For instance, the Kremser Equation
for the number of stages for absorption by an initially pure liquid becomes

−1 𝑌𝑌 −1
log��1−𝐴𝐴′ �� 𝐵𝐵 �+𝐴𝐴′ �
𝑌𝑌𝑇𝑇
𝑁𝑁 = (205)
log 𝐴𝐴′

where

𝐿𝐿′
𝐴𝐴′ ≝ (206)
𝐺𝐺 ′ 𝐻𝐻

For stripping we get

−1 𝑋𝑋 −1
log��1−𝑆𝑆 ′ �� 𝑇𝑇 �+𝑆𝑆 ′ �
𝑋𝑋𝐵𝐵
𝑁𝑁 = (207)
log 𝑆𝑆 ′

where

𝐺𝐺 ′ 𝐻𝐻
𝑆𝑆′ ≝ (208)
𝐿𝐿′

All the theory we have developed for the maximum and minimum flow rates still applies and
all the restrictions to that theory also still apply.

Absorption and Stripping 82 Separation Processes

CPE230 Dr Alasdair Campbell

Example 14

A moist air stream is to be dried by contacting it in a counter current column with concentrated
sulphuric acid. The air is fed to the column at 1 kmol/s, and has an initial moisture content of
2.7 mol%. The aqueous sulphuric acid supplied to the column is 50 mol% H2SO4. The moisture
in the air is to be reduced to 0.79 mol%. Assuming that the ratio of solute (water) free air and
sulphuric acid is 1.5 times the minimum, find the water content of the exiting sulphuric acid
stream, as well as the number of stages required. In addition, calculate the ratios of the molar
inflow:outflow for both the gas and liquid streams.
LT GT L´ G´
xT = 0.5 yT = 0.0079 XT = 0.5 / (1 – 0.5) = 1 YT = 0.0079 / (1 – 0.0079) =
0.008

YB = 0.027 / (1 – 0.027) =
xB yB = 0.027 XB 0.0277
LB GB = 1 kmol/s L´ G´ = 1× (1 – 0.027)
= 0.973 kmol/s dry

In the gas stream, there is 1 kmol/s of wet air entering the column. The proportion of this which
is dry air is 1 – yB. The water free gas flow rate is therefore

G´ = G (1 – yB) = 1 × (1 – 0.027) = kmol/s dry air

Recall our definitions of molar ratio:

𝑦𝑦 𝑥𝑥
𝑌𝑌 = , 𝑋𝑋 =
1−𝑦𝑦 1−𝑥𝑥

We can use these equations to calculate the molar ratios at the inlet and outlet.

Now we must examine the equilibrium diagram. Because it has convex curvature, there exists
the possibility of a tangential pinch.

Absorption and Stripping 83 Separation Processes

CPE230 Dr Alasdair Campbell

0.03

0.025
moles dry air

0.02
moles H2 O

0.015
𝑌𝑌 =

0.01

0.005

0
0 5 10 15 20
moles H2 O
𝑋𝑋 =
moles H2 SO4

The gradient of the operating line giving the tangential pinch gives us:

𝐿𝐿′ 0.0277 − 0.008

� � = =
𝐺𝐺′ 𝑚𝑚𝑚𝑚𝑚𝑚 14.2 − 1

Hence the actual value used in the column is:

𝐿𝐿′
=
𝐺𝐺′

and L´ is

𝐿𝐿′ = 0.00225𝐺𝐺 ′ =

Plotting the operating line gives us X at the bottom of the column i.e.

𝑋𝑋𝐵𝐵 =

𝑋𝑋𝐵𝐵
𝑥𝑥𝐵𝐵 = =
1 + 𝑋𝑋𝐵𝐵

We know that the total liquid flow rate is given by: water free flow + water

𝐿𝐿 = 𝐿𝐿′ + 𝐿𝐿′𝑋𝑋 = 𝐿𝐿′(1 + 𝑋𝑋)

hence

Absorption and Stripping 84 Separation Processes

CPE230 Dr Alasdair Campbell

𝐿𝐿𝐵𝐵 𝐿𝐿′(1+𝑋𝑋𝐵𝐵 ) (1 + 𝑋𝑋𝐵𝐵 )

= = =
𝐿𝐿𝑇𝑇 𝐿𝐿′(1+𝑋𝑋𝑇𝑇 ) (1 + 𝑋𝑋𝑇𝑇 )

Similarly,

𝐺𝐺𝐵𝐵 𝐺𝐺′(1 + 𝑌𝑌𝐵𝐵 ) (1 + 𝑌𝑌𝐵𝐵 )

= = =
𝐺𝐺𝑇𝑇 𝐺𝐺′(1 + 𝑌𝑌𝑇𝑇 ) (1 + 𝑌𝑌𝑇𝑇 )

Because the mole fractions in the gas are much less than 1, the overall gas flow rate doesn’t
change very much (could consider it dilute). In contrast, L changes by a factor of 5 going from
the top of the column to the bottom.

McCabe-Thiele gives us 3 stages for the column.

4.14 What Do You Not Know Yet?

In absorption and stripping, we have not said anything about what we do if there is significant
mass transfer between the phases of all the components present. This is important at high
temperature where the liquid phase solvent may be evaporating as well as absorbing some or
all of the components from the gas phase. This situation is still not fully resolved in the research
literature.

Gas absorption is often accompanied by chemical reaction and is also, therefore, not
isothermal – the use of amine systems for CO2 capture is a classic example.

We have said nothing about column diameter, tray spacing, pressure drop, prediction of tray
efficiency, process control etc. We have said almost nothing about multi-component systems!

Lot’s more fun things to look forward to in future modules!

Absorption and Stripping 85 Separation Processes