Mass Transfer Operations Notas
Mass Transfer Operations Notas
Mass Transfer Operations Notas
Introduction:
What is chemical engineering?
Chemical engineering is the field of engineering that deals with industrial processes in which
raw materials are converted or separated into final useful products.
Who is the chemical Engineer?
The function of a qualified chemical engineer is to apply chemistry of a particular industrial
process through the use of coordinated scientific and engineering principles. Also develop the
laboratory results of chemists into economical chemical plants, he must develop, design and
engineer both the complete industrial process and the equipment used in it. Then his duties can
be stated as:
1- He must choose the raw material.
2- He must design and operate the plant efficiently, safely, and economically.
3- He must know that his products meet the requirements set by the customers.
4- When science does not give him a complete answer, he must use experience & judgment
to combine all sources of information to reach particular solution to processing problems.
Definition of Unit Operation:-
The physical operations necessary for manufacturing chemicals deals mainly with the
transfer and change of material and energy principles by physical means but also by
chemical – physical means such as mass transfer, heat transfer and momentum transfer (fluid
dynamics). There are two types of unit operation classifications:
The first one depends on the process type as:
1- Processes without chemical reaction, as distillation, crystallization, absorption,
extraction, and filtration.
2- Processes with chemical reactions as nitration, oxidation, and sulfonation.
While the second classification depends on the type of transfer occur within the process such as:
1- Mass transfer operations
Transfer of component from one phase to another such processes is:
Gas absorption, Distillation, Extraction, Crystallization, Drying, Evaporation, Leaching,
Stripping, Mechanical separation for example (Filtration, settling, size reduction, and
sedimentation).
2- Heat transfer operations
Accumulation and transfer of heat and energy such as:
Heat exchanger, Evaporation, Drying, Distillation.
3- Momentum transfer operations
Flow and transportation of fluid and solid such as:
Fluid flow, Mixing, Handling of solid, Fluid transportation.
1
Chapter Two
Mass Transfer & Diffusion
Mass transfer is the net movement of a component in a mixture from one location to
another location where the component exists at a different concentration. Often, the
transfer takes place between two phases across an interface. Thus, the absorption by a
liquid of a solute from a gas involves mass transfer of the solute through the gas to the
gas-liquid interface, across the interface, and into the liquid. Mass transfer models are
used to describe processes such as the passage of a species through a gas to the outer
surface of a porous adsorbent particle and into the pores of the adsorbent, where the
species is adsorbed on the porous surface. Mass transfer is also the selective
permeation through a nonporous polymeric material of a component of a gas mixture.
Mass transfer is not the flow of a fluid through a pipe. However, mass transfer might
be superimposed on that flow. Mass transfer is not the flow of solids on a conveyor
belt.
Mass transfer occurs by two basic mechanisms:
(1) Molecular diffusion by random and spontaneous microscopic movement of
individual molecules in a gas, liquid, or solid as a result of thermal motion; and
(2) Eddy (turbulent) diffusion by random macroscopic fluid motion.
Molecular and/or eddy diffusion frequently involves the movement of different
species in opposing directions. When a net flow occurs in one of these directions, the
total rate of mass transfer of individual species is increased or decreased by this bulk
flow or convection effect, which is a third mechanism of mass transfer. As will be
shown later, molecular diffusion is extremely slow, whereas eddy diffusion, when it
occurs, is orders of magnitude more rapid. Therefore, if large-scale separation
processes are to be conducted in equipment of a reasonable size, fluids must be
agitated, interfacial areas maximized, and distances in the direction of diffusion
minimized. In a binary mixture, molecular diffusion occurs because of one or more
different potentials or driving forces, including differences (gradients) of concentration
(ordinary diffusion), pressure (pressure diffusion), temperature (thermal diffusion),
and external force fields (forced diffusion) that act unequally on the different chemical
species present. Pressure diffusion requires a large pressure gradient, which is
achieved for gas mixtures with a centrifuge. Thermal diffusion columns or cascades
can be employed to separate liquid and gas mixtures by establishing a temperature
gradient across the mixture. More widely applied is forced diffusion in an electrical
field, to cause ions of different charges to move in different directions at different
speeds. In this chapter, only molecular diffusion caused by concentration gradients is
considered, because this is the most common type of molecular diffusion in
2
commercial separation processes. Furthermore, emphasis is on binary systems, for
which molecular diffusion theory is simple and applications are relatively
straightforward. Multi-component molecular diffusion, which is important in many
applications, is much more complex than diffusion in binary systems, and is a more
appropriate topic for advanced study using a specialized text.
Molecular diffusion occurs in solids and in fluids that are stagnant or in laminar or
turbulent motion. Eddy diffusion occurs in fluids in turbulent motion. When both
molecular diffusion and eddy diffusion occur, they take place in parallel and are
additive. Furthermore, they take place because of the same concentration difference
(gradient). When mass transfer occurs under turbulent flow conditions, but across an
interface or to a solid surface, conditions may be laminar or nearly stagnant near the
interface or solid surface. Thus, even though eddy diffusion may be the dominant
mechanism in the bulk of the fluid, the overall rate of mass transfer is controlled by
molecular diffusion because the eddy diffusion mechanism is damped or even
eliminated as the interface or solid surface is approached. Mass transfer of one or more
species results in a total net rate of bulk flow or flux in one direction relative to a fixed
plane or stationary coordinate system. When a net flux occurs, it carries all species
present. Thus, the molar flux of an individual species is the sum of all three
mechanisms. If Ni is the molar flux of species i with mole fraction Xi, and N is the
total molar flux, with both fluxes in moles per unit time per unit area in a direction
perpendicular to a stationary plane across which mass transfer occurs, then
Where:
Xi.N is the bulk-flow flux.
Each term in (1) is positive or negative depending on the direction of the flux relative
to the direction selected as positive. When the molecular and eddy diffusion fluxes are
in one direction and N is in the opposite direction, even though a concentration
difference or gradient of i exists, the net mass transfer flux, Ni, of i can be zero.
STEADY-STATE ORDINARY MOLECULAR DIFFUSION
Suppose a cylindrical glass vessel is partly filled with water containing a soluble red
dye. Clear water is carefully added on top so that the dyed solution on the bottom is
undisturbed. At first, a sharp boundary exists between the two layers, but after a time
the upper layer becomes colored, while the layer below becomes less colored. The
upper layer is more colored near the original interface between the two layers and less
colored in the region near the top of the upper layer. During this color change, the
motion of each dye molecule is random, undergoing collisions mainly with water
3
molecules and sometimes with other dye molecules, moving first in one direction and
then in another, with no one direction preferred. This type of motion is sometimes
referred to as a random-walk process, which yields a mean-square distance of travel
for a given interval of time, but not a direction of travel. Thus, at a given horizontal
plane through the solution in the cylinder, it is not possible to determine whether, in a
given time interval, a given molecule will cross the plane or not. However, on the
average, a fraction of the molecules in the solution below the plane will cross over into
the region above and the same fraction will cross over in the opposite direction.
Therefore, if the concentration of dye molecules in the lower region is greater than in
the upper region, a net rate of mass transfer of dye molecules will take place from the
lower to the upper region. After a long period of time, the concentration of dye will be
uniform throughout the solution. Based on these observations, it is clear that:
1. Mass transfer by ordinary molecular diffusion occurs because of a concentration
difference or gradient; that is, a species diffuses in the direction of decreasing
concentration.
2. The mass transfer rate is proportional to the area normal to the direction of mass
transfer and not to the volume of the mixture. Thus, the rate can be expressed as a flux.
3. Mass transfer stops when the concentration is uniform.
When solving mass transfer problems involving net movement of the mixture, it is not
convenient to use fluxes and flow rates based on VM as the frame of reference. Rather,
5
it is preferred to use mass transfer fluxes referred to stationary coordinates with the
observer fixed in space. Thus, from (Eq. 9), the total species velocity is:
Vi = VM + Vi D …10
Combining (Eq. 11) with (Eq. 5), (Eq. 6), and (Eq. 7), then
nA dX
NA = = X A . N - C . D AB ( A ) ..12
A dZ
Or can be written as:
dX A
N A = - C . D AB ( ) + X A (N A + N B ) ..12A
dZ
And
nB dX
NB = = X B . N - C . D Ba ( B ) ..13
A dZ
where in (Eq. 12) and (Eq. 13), ni is the molar flow rate in moles per unit time, A is the
mass transfer area, the first terms on the right-hand sides are the fluxes resulting from
bulk flow, and the second terms on the right-hand sides are the ordinary molecular
diffusion fluxes. Two limiting cases are important:
1. Equimolar counter diffusion (EMD)
2. Uni-molecular diffusion (UMD)
6
This idealization is closely approached in distillation. From (Eq. 12) and (Eq.13), we
see that in the absence of fluxes other than molecular diffusion, and
dX A dC A
N A = J A = - C . D AB ( ) ..16 or N A = - D AB ( ) ..16A
dZ dZ
And
dX B
N B = J B = - C . D Ba ( ) ..17
dZ
If the total concentration, pressure, and temperature are constant and the mole fractions
are maintained constant (but different) at two sides of a stagnant film between Z1 and
Z2 then (Eq. 16) and (Eq. 17) can be integrated from Zl to any Z between Z1 and Z2 to
give
C . D AB
JA = (X A1 - X A ) ...18
Z - Z1
And
C . D BA
JB = (X B1 - X B ) ...19
Z - Z1
Thus, in the steady state, the mole fractions are linear in distance, as shown in Fig.1a.
Furthermore, because c is constant through the film, where
C = CA+CB …20
7
By differentiation,
dC = 0 = dCA + dCB …21
Thus,
dCA = - dCB …22
From (Eqs. 3, 4, 15, and 22),
D AB D BA
= ...23
dZ dZ
Therefore, DAB = DBA.
This equality of diffusion coefficients is always true in a binary system of constant
molar density.
NOTE:
From the ideal gas law
PV = nRT
n P
=
V RT
P
C= Where C is the total mole concentration
RT
Using Dalton's law, then
n A PA
= = CA
V RT
Then
D AB dPA
JA = - ( ) Another formula of Fick's law for equi-molar gas diffusion
RT dZ
And
D AB PA2 − PA2
N A = −( )( )
RT Z 2 − Z1
Example 1
Two bulbs are connected by a straight tube, 0.001 m in diameter and 0.15 m in length.
Initially the bulb at end 1 contains N2 and the bulb at end 2 contains H2. The pressure
and temperature are maintained constant at 25°C and 1 atm. At a certain time after
allowing diffusion to occur between the two bulbs, the nitrogen content of the gas at
end 1 of the tube is 80 mol% and at end 2 is 25 mol%. If the binary diffusion
coefficient is 0.784 cm2/s, determine:
(a) The rates and directions of mass transfer of hydrogen and nitrogen in mol/s
(b) The species velocities relative to stationary coordinates, in cm/s
8
Solution:
(a) Because the gas system is closed and at constant pressure and temperature, mass
transfer in the connecting tube is equimolar counter diffusion by molecular diffusion.
The area for mass transfer through the tube, in cm2, is
A = Π (0.1)2/4 = 7.85 * 10-3 cm2.
The total gas concentration is
P 1 mol
C= = = 4.09 *10 -5
R . T (82.06)(298) cm 3
Take the reference plane at end 1 of the connecting tube. Applying (Eq. 18) to N2 over
the length of the tube,
C . D N 2H 2
n N2 = [ (X N 2 ) 1 - (X N 2 )] A
Z 2 - Z1
(b)For equimolar counter diffusion, the molar average velocity of the mixture, VM, is
Zero
Therefore, from (Eq. 9), species velocities are equal to species diffusion velocities.
Thus,
J N2 n N2 9.23 * 10 -9
VN 2 = (VN 2 ) D = = =
C N2 A C X N2 [(7.85 * 10 -5 )(4.09 * 10 −5 )( X N 2 )]
0.0287
VN 2 = in the positive Z direction
( X N2 )
Similarly
0.0287
VH 2 = in the negative Z direction
( X H2 )
9
z (cm) X N2 X H2 VN 2 (cm/s) VH 2 (cm/s)
o (end 1) 0.800 0.200 0.0351 -0.1435
5 0.617 0.383 0.0465 -0.0749
10 0.433 0.567 0.0663 -0.0506
15 (end 2) 0.250 0.750 0.1148 -0.0383
Note that species velocities vary across the length of the connecting tube, but at any
location, Z, VM = 0. For example, at Z = 10 cm, from (Eq. 8)
VM = (0.433)(0.0663) + (0.567)(-0.0506) = 0
Unimolecular Diffusion
In unimolecular diffusion (UMD), mass transfer of component A occurs through
stagnant component B. Thus,
NB = 0 …24
And
N = NA …25
Therefore, from (Eq. 12),
dX
N A = X A . N - C . D AB ( A ) ..26
dZ
This can be rearranged to a Fick's law form,
CD AB dX A
NA =- ( ) ..27
(1 - X A ) dZ
The factor (1 - XA) accounts for the bulk flow effect. For a mixture dilute in A, the bulk flow effect
is negligible or small. In mixtures more concentrated in A, the bulk flow effect can be appreciable.
For example, in an equimolar mixture of A and B, (1 - XA) = 0.5 and the molar mass transfer flux
of A is twice the ordinary molecular diffusion flux.
For the stagnant component, B, (Eq.13) becomes
dX B
0 = X B N A − CD AB ...28
dZ
dX B
Or X B N A = CD AB ...29
dZ
Thus, the bulk flow flux of B is equal but opposite to its diffusion flux.
At quasi-steady-state conditions that is, with no accumulation, and with constant molar density,
(Eq. 27) becomes in integral form:
ZA
C.D AB X A dX A
∫ dZ = − ∫ ...30
Z A1 N A XA (1 − X A )
1
Which upon integration yields
C.D AB 1− X A
NA = ln( ) ...31
Z − Z1 1 − X A1
10
Rearrangement to give the mole-fraction variation as a function of Z yield
⎡ N ( Z − Z1 ) ⎤
X A = 1 − (1 − X A1 ) exp ⎢ A ⎥ ...32
⎣ C . D AB ⎦
Thus, as shown in Fig. 1b, the mole fractions are nonlinear in distance.
An alternative and more useful form of (Eq. 31) can be derived from the definition of
the log mean. When Z= Z2, (Eq. 31) becomes
C.D AB 1 − X A2
NA = ln( ) ...33
Z 2 − Z1 1 − X A1
The log mean (LM) of (1 - XA) at the two ends of the stagnant layer is
(1 − X A ) − (1 − X A ) XA −XA
X BLM = (1 − X A ) LM = 2 1 = 1 2 ...34
ln ⎡(1 − X A ) /(1 − X A )⎤ ln ⎡(1 − X A ) /(1 − X A )⎤
⎢⎣ 2 1 ⎥⎦ ⎢⎣ 2 1 ⎥⎦
C.D AB ( X A1 − X A2 ) C.D AB (− ∆X A )
NA = . = ...35
Z 2 − Z1 (1 − X A ) LM (1 − X A ) LM (∆Z )
Or
C.D AB ( X A1 − X A2 ) C.D AB (−∆X A )
NA = . = ...35 A
Z 2 − Z1 X BLM X BLM (∆Z )
NA * A = NA mol/s
dV dZ
NA = CA = C A .A
dt dt
But
11
ρA
CA =
Mwt A
Then:
ρ A ⎛ dZ ⎞
NA = ⎜ ⎟
Mwt A ⎝ dt ⎠
Mwt
dZ = N A dt
ρA
Mwt ⎛⎜ C.D AB (X A1 − X A 2 ⎞
⎟dt
dZ = …36
ρ A ⎜⎝ (1 - X A ) LM Z ⎟
⎠
Rearrange the above equation
ρA ⎛ (1 - X A ) LM Z ⎞
dt = ⎜ ⎟dZ …37
⎜
Mwt ⎝ C.D AB (X A1 − X A 2 ) ⎟⎠
But, as shown before in the equimolar condition
P P
C= and C A = A then
RT RT
ρA ⎛ R T (PBLM ) Z ⎞
dt = ⎜ ⎟dZ …38
⎜
Mwt ⎝ PT .D AB (PA1 − PA 2 ) ⎟⎠
Knowing that:
(1 − PA ) − (1 − PA ) PA − PA
PBLM = (1 − PA ) LM = 2 1 = 1 2
ln ⎡(1 − PA ) /(1 − PA )⎤ ln ⎡(1 − PA ) /(1 − PA )⎤
⎢⎣ 2 1 ⎥⎦ ⎢⎣ 2 1 ⎥⎦
Finally by integrating the above equation for the following boundary condition:
t1 = 0 at Z1=Z0 and t2 = tf at Z2 = Zf we can determine the time required to drop the
level of liquid to a certain height:
ρA ⎛ R T (PBLM ) ⎞ Zf
t= ⎜ ⎟ ∫ ZdZ …39
Mwt ⎜⎝ PT .D AB (PA1 − PA 2 ) ⎟⎠ Z
0
Example 2:
As shown in Figure 2, an open beaker, 6 cm in height, is filled with liquid benzene at
25°C to within 0.5 cm of the top. A gentle breeze of dry air at 25°C and 1 atm is blown
by a fan across the mouth of the beaker so that evaporated benzene is carried away by
convection after it transfers through a stagnant air layer in the beaker. The vapor
pressure of benzene at 25°C is 0.131 atm. The mutual diffusion coefficient for benzene
in air at 25°C and 1 atm is 0.0905 cm2/s. Compute:
(a) The initial rate of evaporation of benzene as a molar flux in mol/cm2.s
(b) The initial mole fraction profiles in the stagnant layer
12
(c) The initial fractions of the mass transfer fluxes due to molecular diffusion.
(d) The initial diffusion velocities, and the species velocities (relative to stationary coordinates) in
the stagnant layer.
(e) The time in hours for the benzene level in the beaker to drop 2 cm from the initial level, if the
specific gravity of liquid benzene is 0.874. Neglect the accumulation of benzene and air in the
stagnant layer as it increases in height.
Note that the molecular diffusion fluxes are equal but opposite, and the bulk flow flux
of B is equal but opposite to its molecular diffusion flux, so that its molar flux, NB, is
zero.
14
From (Eq. 9), the diffusion velocities are given by
J J
Vi d = i = i (3)
Ci X i C
From (Eq. 10), the species velocities relative to stationary coordinates are:
Vi = Vi d + VM (4)
Using the above equations (2 to 4), we obtain:
Note that VB is zero everywhere, because its molecular diffusion velocity is negated by
the molar mean velocity.
(e) The mass transfer flux for benzene evaporation can be equated to the rate of change
of liquid benzene. Letting z = distance down from the mouth of the beaker and using
(Eq. 35) with ∆Z = Z
C.D AB (−∆X A ) ρ L ⎛ dZ ⎞
NA = = ⎜ ⎟ (5)
(1 − X A ) LM ( Z ) M L ⎝ dt ⎠
t=
0.874(0.933)
((2.5) 2 − (0.5) 2 ) = 64590 s = 17.94 hr
78.11(4.09 *10 - 5 )(0.0905)(0.131)
15
Diffusion through a varying cross – sectional area
mol
In previous cases NA ( ) was constant since the cases were at steady state and the
2
m .S
cross – sectional area was constant too.
Now, if the cross – sectional area (A) varies with Z (the direction of diffusion) such as
evaporation of water drop or in the case of naphthalene sphere sublimation, or even in
the case of diffusion in conical container, then
N mol
N A = A will not be constant since A is not constant ( N A mole flux is constant
A S
at steady state).
First, we will consider the case of diffusion from a sphere material (A) in a large
volume of gas (B):
Assumptions:
1- B is non – diffusing material.
2- At r = r2 PA = PA 2
16
1
If r2 >> r1 then =0 and PA 2 = 0 (dilute solution) also PBLM = PT , then
r2
N A ⎡ 1 ⎤ D AB
4π ⎢⎣ r1 ⎥⎦
=
RT
( )PA1 …43
Assumptions:-
B is non –diffusing material (i.e. NB = 0)
Then equation 27 may be applied
CD AB dX A
NA =- ( ) ..27
(1 - X A ) dZ
And for diffusion of gases the above equation could be written as follow:
17
D AB PT dP
NA =- ( A) ...46
RT(PT − PA ) dZ
Here NA is not constant with Z then by integrating the above equation we find
Z2 P
dZ − D AB PT A 2 dPA
NA ∫ = ∫ ...47
Z1 A RT (P
PA1 T − PA)
To over come the above integration we must find the relationship between A & Z,
which depends on the shape. For the above shape (conical):
π d2
A=
4
Therefore we must find a relationship between d & Z
d − d1
K= 2
2
K M
tan θ = =
L L−Z
K (L − Z )
M =
L
Then the diameter at height Z from the top is
2K (L − Z )
d = d1 + 2m = d1 + 2 M = d1 +
L
d1, K, and L are known. So the above equation 47 will be written as:
Z2 P
dZ − D AB PT A 2 dPA
NA ∫ = ∫ ...47A
π
Z1 ( d1 +
2K (L − Z ) RT PA1 (PT − PA)
4 L
18
Diffusivity of Gases or Vapors
Many attempts have been made to express the diffusivity in terms of other physical
properties and the following empirical equation by Gilliland gives a satisfactory
agreement with experimental data:
4.3 *10 − 4 T 1.5
1 1
+
Mwt A Mwt B
D AB =
( )
...48
2
PT V A1 / 3 + V B1 / 3
Where:
m2
D AB = Diffusivity in ( ), T = temperature in (K), P = absolute total pressure in (Pa)
s
MwtA, and MwtB = molecular weight of component A & B.
m3
VA, VB = molar volume of component A & B ( ).
kmol
From equation (48) we can notice that:
T1.5
Dα
P
And:
1.5
D1 ⎛ T1 ⎞ ⎛ P2 ⎞
=⎜ ⎟ ⎜ ⎟ ...49
D2 ⎜⎝ T2 ⎟⎠ ⎜⎝ P1 ⎟⎠
Another equation is used to calculate the diffusivity coefficient, it is semi empirical
equation proposed by Fuller et. al., that is:
0.00143 (T 1.75 )
D AB = ...50
2
1 ⎛ 1 1 ⎞
PT (M AB ) 2 ⎜ (∑ v A ) + (∑ v B ) 3 ⎟⎟
⎜ 3
⎝ ⎠
Where:
cm 2
D AB = Diffusivity in ( ), T = temperature in (K), P = absolute total pressure in
s
(atm)
2
M AB = ...51
⎛ 1 ⎞+⎛ 1 ⎞
⎜ Mwt ⎟ ⎜ Mwt ⎟
⎝ A⎠ ⎝ B⎠
MwtA, and MwtB = molecular weight of component A & B.
∑ v = summation of the atomic and structural diffusion volumes from references such
as the following table:
19
Table (1) Diffusion Volumes from Fuller, Ensley, and Giddings [J. Phys. Chem.,
73, 3679-3685 (1969)] for Estimating Binary Gas Diffusivity by the Method of
Fuller et al.
20
Maxwell's Law of diffusion For Binary System
Maxwell's postulated that the pressure gradient ( d PA ) in the direction of diffusion for
a constituent of two components gaseous mixture was proportional to:
A) The relative velocity of the molecules in the direction of diffusion.
B) The product of the molar concentration of the component.
Thus;
− dPA
= F .C A .C B (U A − U B ) . ..53
dz
U A & U B = mean molecule of A & B respectively
CA & CB = molar concentration of the component A & B respectively.
F = Coefficient
But
N N
UA = A & UB = B
CA CB
Also
PA = CA RT
Then
− dC A F
= (N A . CB − N B . CA ) . ..54
dz RT
Now, applying the two cases that have been considered before, i.e. equimolar
diffusion, and diffusion through stagnant layer, then we can reach to the final equation
to calculate the rate of mass transfer as shown below:
A) For equimolar diffusion
NA = - NB
− dC A F .N A
= (C B + C A ) . ..55
dz RT
RT dC A
NA = − ( ) . ..56
F .CT dz
Comparing with Fick's law, we find that
RT
D AB =
FCT
Or
RT RT
FAB = General form F =
D AB CT DCT
B) For diffusion through stagnant layer
NB = 0
21
Then
− dC A F
= (N A . CB ) . ..57
dz RT
And
RT dC A
NA = − ( ) . ..58
F .C B dz
C
Multiplying the right side of the equation by ( T ) and by comparison we find that
CT
D.CT dC A
NA = − ( ) . ..59
CB dz
This finally will result:
C .D 1 − X A2
N A = T AB ln( ) ...60
Z 2 − Z1 1 − X A1
Maxwell's Law for Multi-Component Mass Transfer
Consider the transfer of component A through a stationary gas consisting of
component B, C, D …Suppose that the total partial pressure gradient can be regarded
as being made up of series of terms each represent the contribution of the individual
component gases, the form of the binary system equation (Eq. 54) can be written as
follow:
− dC A FAB .N A .C B F .N .C F .N .C
= + AC A C + AD A D + Λ . ..61
dz RT RT RT
As found before for F
RT
F=
DCT
So, for diffusion of A through multi-component B, C, and D
RT RT RT
FAB = , FAB = , FAD = respectively
D AB CT D AC CT D AD CT
Therefore:
− dC A N A C B CC C
= ( + + D +Λ . ..62
dz CT D AB D AC D AD
− CT dC A
NA = ...63
⎛ CB CC CD ⎞ dZ
⎜⎜ + + + Λ ⎟⎟
⎝ D AB D AC D AD ⎠
⎛C − CA ⎞
Multiplying the above equation by ⎜⎜ T ⎟⎟ , then
⎝ CT − C A ⎠
22
1 −C dC
N = T A ...64
A ⎛⎛ C ⎞ ⎛ C ⎞ ⎛ ⎞ ⎞ CT − C A dZ
⎜⎜ 1
. B ⎟+⎜ 1 . C ⎟ + ⎜ 1 . CD ⎟⎟
⎜⎜ ⎜ D C −C ⎟ ⎜ D C −C ⎟ ⎜ D C − C ⎟ ⎟⎟
⎝ ⎝ AB T A ⎠ ⎝ AC T A ⎠ ⎝ AD T A ⎠⎠
Define
Cj
y 'j = then
CT − C i
CB CC CD
y ′B = , yC′ = and y ′D = Therefore
CT − C A CT − C A CT − C A
1 − CT dC A
NA = ...65
⎛ ⎛ y ′B ⎞ ⎛ y ′B ⎞ ⎛ yC ′ ⎞ ⎞ CT − C A dZ
⎜ ⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ ⎟⎟
⎜ D
⎝ ⎝ AB ⎠ ⎝ D AC ⎠ ⎝ D AD ⎠⎠
Finally
− CT dC A
N A = D AM ...66
CT − C A dZ
Where:
DAM is defined as the diffusivity coefficient of component A through a mixture of B,
C, D, ….
And
1
D AM =
⎛ ⎛ y ′B ⎞ ⎛ y C
′ ⎞ ⎛ y ′D ⎞⎞
⎜ ⎜⎜ ⎟ + ⎜ ⎟⎟ + ⎜⎜ ⎟⎟ ⎟⎟
⎜ D ⎟ ⎜D
⎝ ⎝ AB ⎠ ⎝ AC ⎠ ⎝ D AD ⎠⎠
23
Example 3:
A sphere of naphthalene having a radius of 2mm is suspended in a large volume of
still air at 318 K and 101.325 kPa. The surface temperature of the naphthalene can be
assumed to be at 318 K and its vapor pressure at this temperature is 0.555 mmHg. The
diffusivity (DAB) of naphthalene in air at 318 K & 1 atm is 6.92*10-6 m2/s, calculate:
1- The rate of evaporation of naphthalene from the surface.
2- Then find the partial pressure of naphthalene at a distance 20 mm from the surface
of the naphthalene sphere. Assume steady – state diffusion.
Solution:
1- To calculate the rate of evaporation of naphthalene use equation 42:
N A ⎡ 1 1 ⎤ D AB PT
⎢ − ⎥=
4π ⎣ r1 r2 ⎦ RT PBLM
(
PA1 − PA 2 )
PA1 = 0.555 mmHg ≡ 74 Pa
For large values of r2 that is r2 » r1 then:
1 1 1
〉〉 and ≈ 0 also PT = PBLM because PA 2 = 0 and PA1 « PT, therefore apply
r1 r2 r2
equation 43:
N A ⎡ 1 ⎤ D AB
⎢ ⎥
4π ⎣ r1 ⎦
=
RT
( )
PA1 and then
4π * 0.002 * 6.92 * 10 − 6 * 74
NA =
4π .r1.D AB
RT
( )
PA1 =
8.314 * 318
= 4.8654 * 10 − 9 mol / sec
The rate of evaporation from the surface is N A r
1
N A 4.8654 * 10 − 9
= 96.8 * 10 − 5
kmol
NA r = =
1 A 4π * (0.002) 2 m 2 . sec
2- To find the partial pressure of naphthalene at a distance 20 mm from the surface
apply the following equation:
N A ⎡ 1 1 ⎤ D AB PT ⎛⎜ PT − PA1 ⎞⎟
− =
4π ⎢⎣ r1 r2 ⎥⎦
ln which is the original form of equation 42, for
RT ⎜ PT − PA ⎟
⎝ 2 ⎠
the following conditions:
At r = r1 = 2 mm then PA1 = 74 Pa
And r = r2 = 22 mm then PA 2 is unkonwn and by applying the above equation
we can find PA 2 knowing that all other parameters in this equation is known.
24
Example 4:
An open tank is filled with 2ft of the top with pure methanol. The tank is tapered, as
shown below. The air within the tank is stationary but circulation of air immediately
above the tank is adequate to assure a negligible concentration of methanol at this
point. The tank and air space are at 77oF and 1 atm and the diffusivity of methanol in
air in such conditions is 0.62 ft2/s. Calculate the rate of loss of methanol from the tank
at steady – state. Given that the partial pressure of methanol at 77oF is 135 mm Hg.
Solution:
Since air does not diffuse in methanol, so
NB = 0
∴ the case is diffusing through stagnant layer with
variable area, apply equation 47
Z2 P
dZ − D AB PT A 2 dPA
NA ∫ = ∫
Z A
1
RT P (PT − PA) A1
Then we have to find the area (A) in terms of (Z)
1 m
tan θ = =
2 L−Z
L−Z
m=
2
d = 2m + 4
d =L−Z +4=6−Z
π π
∴A= (d 2 ) = (6 − Z ) 2
4 4
Then
P
4 * N A Z2 dZ − D AB * PT A2 dPA
∫ = ∫
π Z1 ( 6 − Z ) 2 R * T PA1 ( PT − PA )
For the following boundary conditions :
Z1 = 0 then PA1 = 135 mm Hg = 0.178 atm check d = 6ft
Z 2 = 2 then PA 2 = 0 atm check d = 4ft
Now we have to calculate all the parameters in the equation, so we can find N A ,
therefore:
25
PB1 = 760 − 135 = 625 mmHg = 0.822 atm
PB 2 = 760 - 0 = 760 mm Hg = 1 atm
PB2 − PB2 1 − 0.822
PBLM = = = 0.908 atm
PB2 1
ln ln
PB1 0.822
atm ft 3
Use R = 0.7302
lbmol * R
ft 2 − 4 ft
2
D AB = 0.62 = 1.722 *10
hr sec
Example 5:
A small diameter tube closed at one end was filled with acetone to within 18 mm of
the top and maintained at 290 K and 99.75 kPa with a gentle stream of air blowing
across the top. After 15 ksec, the liquid level had fallen to 27.5 mm. The vapor
pressure of acetone at that temperature is 21.95 kN/m2. Calculate the diffusivity of
acetone in air, given the following data:
Mwt of acetone = 58
The density of acetone (ρ) = 790 kg/m3
Solution:
Since evaporation is occurred through constant area, then apply equation 39 and by
integrating this equation we find:
ρA ⎛ ⎞ ⎛ Z2 Z 2 ⎞
t= ⎜ R T (PBLM ) ⎟⎜ f − 0 ⎟
Mwt ⎜⎝ PT .D AB (PA1 − PA 2 ) ⎟⎠ ⎜ 2 2 ⎟
⎝ ⎠
To apply this equation we must first calculate each parameter in the equation as:
PT = 99.75 kPa
T = 290 K
kPa * m 3
R = 8.314
kmol * K
26
PB2 − PB2 (99.75 − 0) − (99.75 − 21.95)
PBLM = = = 88.321 kPa
PB2 99.75
ln ln
PB1 77.8
Example 6:
An open conical vessel is filled with water up to 10 cm from its top, as shown in the
figure below. Calculate the time required to drop the level by 13 cm, given that the
diffusivity of water in air at 25oC & 1 atm is 0.256 cm2/s, and the vapor pressure of
water at 25oC is 0.0313 atm.
Solution:
To calculate the time required, we must first find the rate of diffusion using equation
47 that is
Z2 P
dZ − D AB PT A 2 dPA
NA ∫ = ∫
Z A
1
RT P (PT − PA)
A1
Then we must find the relation between the
diameter (d) & the height (Z)
From the figure:
27
1 r
tan θ = =
L L−Z
1 r
=
3 3−Z
3−Z 2( 3 − Z )
∴r = →d =
3 3
π π * 4( 3 − Z ) 2 π ( 3 − Z )2
A= d2 = =
4 4*3 3
Then substitute A in the above integration and solve for N A we find that
0.1
3* NA ⎛ 1 1 ⎞
⎜⎜
π ⎝ 3−Z
− ⎟⎟ =
D AB * PT
PA1 − PA2 ( )
PA2
3 − Z ⎠0 R * T * PBLM PA1
To apply this equation we must first calculate each parameter in the equation as:
PT = 1 atm
T = 298 K
3
kPa * m
R = 8.314
kmol * K
PB2 − PB2 (1 − 0) − (1 − 0.0313)
PBLM = = = 0.9843 atm
PB2 1
ln ln
PB1 0.9687
∫( )
2
ρ *π 0.23
N A *t = 3 − Z dZ
3 * Mwt 0.1
t=
1000 * π − 1 ⎡
*
3 *18 * N A 3 ⎢⎣
3
3 − 0.23 − ( ) ( )
3 − 0.1 ⎤ = sec
3
⎥⎦
28
Example 7:
Normal butanol (A) is diffusing through air (B) at 1 atm absolute pressure. Estimate
the diffusivity (DAB) for the following conditions and compare it with experimental
data:
A- For 0oC & 1atm.
B- For 26oC & 1atm.
C- For 0oC & 2 atm (absolute).
Solution:
A- Fuller equation is given as (Eq. 50)
0.00143 (T 1.75 )
D AB =
1 ⎛ 1 1 ⎞2
PT (M AB ) 2 ⎜⎜ (∑ v A ) 3 + (∑ v B ) 3 ⎟⎟
⎝ ⎠
So we have to estimate the parameters of the equations:
T = 0+273=273 K
PT = 1 atm
M AB Can be calculated from:
2 2
M AB = = = 41.67
⎛ 1 ⎞+⎛ 1 ⎞ ⎛ 1 1 ⎞
⎜ Mwt ⎟ ⎜ Mwt ⎟ ⎜ + ⎟
⎝ A⎠ ⎝ B⎠ ⎝ 74 29 ⎠
∑ V A = (4 *15.9) + (10 * 2.31) + (1 * 6.11) = 92.81 (The values of diffusion volumes
for each atom are taken from table (1).
∑ V B = 19.7
Then
0.00143 * (273)1.75 cm 2
D AB = 2
= 0.077737639
1* (41.67 ) 2 ⎛⎜ (92.81) 3 + (19.7) 3 ⎞⎟
1 1 1 s
⎝ ⎠
B- To estimate the value of the diffusivity at 26oC & 1atm use equation 52
1.75
D1 ⎛ T1 ⎞ ⎛ P2 ⎞
=⎜ ⎟ ⎜⎜ ⎟⎟
D2 ⎜⎝ T2 ⎟⎠ ⎝ P1 ⎠
But there is no change in the pressure (both are 1 atm) then:
1.75
D1 ⎛ T1 ⎞
=⎜ ⎟
D2 ⎜⎝ T2 ⎟⎠
1.75
⎛T ⎞ ⎛ 26 + 273 ⎞
1.75
cm 2
D2 = ⎜⎜ 2 ⎟⎟ * D1 = ⎜ ⎟ * 0.077737639 = 0.091153069
⎝ T1 ⎠ ⎝ 0 + 273 ⎠ s
C- To estimate the value of the diffusivity at 0oC & 2atm also use equation 52
1.75
D1 ⎛ T1 ⎞ ⎛ P2 ⎞
=⎜ ⎟ ⎜⎜ ⎟⎟
D2 ⎜⎝ T2 ⎟⎠ ⎝ P1 ⎠
But as we can see that the temperature is the same of part A, then:
29
D1 ⎛ P2 ⎞
=⎜ ⎟
D2 ⎜⎝ P1 ⎟⎠
And
⎛P ⎞ ⎛1⎞ cm 2
D2 = ⎜⎜ 1 ⎟⎟ * D1 = ⎜ ⎟ * 0.077737639 = 0.03886882
⎝ P2 ⎠ ⎝2⎠ s
Example 8:
Oxygen (A) is diffusing through non-diffusing gas mixture of methane (B) and
hydrogen (C) in the volume ratio of 2:1. The total pressure is 101.3 kPa and the
temperature is 0oC. The partial pressure of oxygen at two planes (2 mm) is 13 kN/m2
and 6.5 kN/m2. The diffusivity of Oxygen in Hydrogen ( DO2 H 2 ) = 6.99*10-5 m2/s
and the diffusivity of Oxygen in Methane ( DO2 CH 4 ) = 1.86*10-5 m2/s were measured
at 1 atm and 0oC. Calculate the rate of diffusion of oxygen in kmol/s through each
square meter of the two planes.
Solution:
Apply equation 66
− CT dC A
N A = D AM
CT − C A dZ
Rewrite the above equation for the gas phase as:
− D AM * PT dPA
NA =
RT * ( PT − PA ) dZ
And by integration the above equation we find:
D AM * PT ( PA1 − PA2 )
NA =
RT * PMLM ( Z 2 − Z1 )
Note:
PMLM means the average log mean of all partial pressures in the system accept A, that is
( PT − PA ) − ( PT − PA )
PMLM = 2 1
⎛ PT − PA ⎞
ln⎜⎜ 2⎟
PT − PA ⎟
⎝ 1 ⎠
PT = 101.3 kPa, PA = 13 kPa, and PA = 6.5 kPa. Then:
1 2
(101.3 − 6.5) − (101.3 − 13)
PMLM = = 90.2 kPa
⎛ 101.3 − 6.5 ⎞
ln⎜ ⎟
⎝ 101.3 − 13 ⎠
Now calculate DAM using the following equation:
30
1
D AM =
⎛ ⎛ y ′B ⎞ ⎛ y C
′ ⎞ ⎛ y ′D ⎞⎞
⎜ ⎜⎜ ⎟ + ⎜ ⎟⎟ + ⎜⎜ ⎟⎟ ⎟⎟
⎜ D ⎟ ⎜D
⎝ ⎝ AB ⎠ ⎝ AC ⎠ ⎝ D AD ⎠⎠
2 1
y ′B = and yC′ = ( volume ratio = mole ratio for gases)
3 3
(VT = VA + VB + VC )
VB VB
( y ′B = = )
VT − VA VB + VC
2
1 −5 m
D AM = = 2.46*10
⎛ ⎛ 0.6667 ⎞ ⎛ 0.3333 ⎞ ⎞ s
⎜⎜ ⎟+⎜ ⎟⎟
⎜⎜ −5 ⎟ ⎜ −5 ⎟ ⎟
⎝ ⎝ 1.86*10 ⎠ ⎝ 6.99*10 ⎠ ⎠
32
X B2 − X B1
X BLM = ...73
⎛ XB ⎞
ln⎜ 2 ⎟
⎜ XB ⎟
⎝ 1 ⎠
Note:
X A1 + X B1 = X A 2 + X B2 = 1
For dilute solutions X BLM ≈ 1.0
Then
⎛ C A − C A1 ⎞
N A = - D L ⎜⎜ 2 ⎟⎟ ...74
⎝ Z 2 − Z1 ⎠
Diffusivities in Liquids
Diffusion coefficient in liquids at 293 K is given in table 10.7 in volume 1 of chemical engineering
BY Coulson, J.M.; Richardson, J.F., fifth edition (page 506). In the absence of data, estimation of
diffusivities can be made using Wilke & Chang equation:
7.4 *10 -8 (φ B Mwt B )1 / 2 T
D AB = 75
µ B υ A0.6
Where:
DAB = diffusivity of solute A in very dilute solution in solvent B, (cm2/s)
MwtB = molecular weight of solvent B.
T = temperature (K).
gm
µB = viscosity of solvent B, (cP or )
cm .s
υA = Solute molar volume at its normal boiling point ( cm3/mol). For water as solute = 0.0756
m3/kmol
φB = Association factor for the solvent.
= 2.6 for Water as solvent.
= 1.9 for Methanol as solvent.
= 1.5 for Ethanol as solvent.
= 1.0 for unassociated solvents as Benzene, ethyl ether.
For diffusion of A through multi-component stagnant layer mixture, the following equation may be
used:
⎛C −C ⎞
− C ⎜ A A ⎟
N = D` T ⎜ 2 1 ⎟ ...76
A C
RM ⎜⎜ Z − Z ⎟
⎝ 2 1 ⎟⎠
Or
⎛X −X ⎞
− Cav ⎜ A A ⎟
N = D` ⎜ 2 1 ⎟ ...77
A X
RM ⎜⎜ Z − Z ⎟⎟
⎝ 2 1 ⎠
Where:
1
D` =
⎛⎛ X B
′ ⎞ ⎛ XC ′ ⎞ ⎛ XD ′ ⎞⎞
⎜ ⎜⎜ ⎟ + ⎜ ⎟ + ⎜ ⎟⎟ ⎟⎟
⎜ D ⎟ ⎜D ⎟ ⎜D
⎝ ⎝ AB ⎠ ⎝ AC ⎠ ⎝ AD ⎠ ⎠
XRM = remaining mole fraction log mean. (Mole fraction of all components except A).
33
Molecular Diffusion in Solid Phase
Diffusion in solids takes place by different mechanisms depending on the
diffusing atom, molecule, or ion; the nature of the solid structure, whether it be porous
or nonporous, crystalline, or amorphous; and the type of solid material, whether it be
metallic, ceramic, polymeric, biological, or cellular. Diffusion in solid phase can be c
classified as:
1- Diffusion which follows Fick's law and does not depend on the structure of solid.
2- Diffusion in which the structure of solid are important.
Where:
C A 2 & C A1 = concentrations at opposite side of the sphere.
34
In the case that gas diffuses through a solid, then the solubility of the gas in the solid is
directly proportional to the partial pressure of solute.
The solubility of a solute gas (A) in a solid is expressed as (S) in cm3 solute (at STP,
e.g. 0oC, and 1 atm) per cm3 solid per (atm, partial pressure of solute A), then:
cm 3 ( STP ) of A
S= ...83
3
cm solid . atm
To convert the solubility (S) of A to concentration (CA):
1
CA = S * * PA ...84
22.414
Where:
mol A cm 3 ( STP ) of A cm 3 A
CA = ,S = , PA = atm, and 22.414 =
3 3 mol
cm atm . cm ( solid )
In many cases the experimental data for diffusion of gases in solid are given as
permeability (PM) in cm3 of solute gas (A) at STP {0oC, and 1 atm} per diffusing per
second per cm2 cross-sectional area through of solid of 1 cm thick under a pressure
difference of 1 atm.
⎛ C A2 − C A1 ⎞
N A = − D AB ⎜⎜ ⎟ ...85
Z − Z ⎟
⎝ 2 1 ⎠
Applying (Eq. 84), then
S .PA1 S .PA2
C A1 = & C A2 = , then:
22.414 22.414
D AB .S ⎛ PA2 − PA1 ⎞ Pm ( PA1 − PA2 )
NA = − ⎜ ⎟ = ...86
22.414 ⎜⎝ Z 2 − Z1 ⎟⎠ 22.414( Z 2 − Z1 )
Where:
cm 3 ( STP ) of A
Pm = D AB .S = ...87
atm
cm 2 .S .
1cm
If series of solids are presented in series 1, 2, 3, .., etc of thickness of each L1, L2,
L3,…, etc respectively, then:
⎛ ⎞
⎜ ⎟
( PA1 − PA2 ) ⎜ 1 ⎟
NA = ⎟ ...88
22.414 ⎜ L1 L2 L3
⎜ + + + ...... ⎟
⎜ Pm P P ⎟
⎝ 1 m2 m3 ⎠
Where:
( PA1 − PA2 ) is the overall partial pressure difference.
35
Diffusion in which the structure of solid are important:
Porous Solids
When solids are porous, predictions of the diffusivity of gaseous and liquid solute
species in the pores can be made. This type of diffusion is also of great importance in
the analysis and design of reactors using porous solid catalysts. It is sufficient to
mention here that any of the following four mass transfer mechanisms or combinations
thereof may take place:
1. Ordinary molecular diffusion through pores, which present tortuous paths and
hinder the movement of large molecules when their diameter is more than 10% of the
pore diameter.
2. Knudsen diffusion, which involves collisions of diffusing gaseous molecules with
the pore walls when the pore diameter and pressure are such that the molecular
mean free path is large compared to the pore diameter.
3. Surface diffusion involving the jumping of molecules, adsorbed on the pore walls,
from one adsorption site to another based on a surface concentration-driving force.
4. Bulk flow through or into the pores.
Crystalline Solids
Diffusion through nonporous crystalline solids depends markedly on the crystal lattice
structure and the diffusing entity. For the cubic lattice (simple, body-centered, and
face-centered), the diffusivity is the same in all directions (isotropic). In the six other
lattice structures (including hexagonal and tetragonal), the diffusivity can be different
in different directions (anisotropic). Many metals, including Ag, AI, Au, Cu, Ni, Pb,
and Pt, crystallize into the face-centered cubic lattice structure. Others, including Be,
Mg, Ti, and Zn, form anisotropic hexagonal structures. The mechanisms of diffusion
in crystalline solids include:
1- Direct exchange of lattice position by two atoms or ions, probably by a ring rotation
involving three or more atoms or ions
2- Migration by small solutes through inter-lattice spaces called interstitial sites
3- Migration to a vacant site in the lattice
4-Migration along lattice imperfections (dislocations), or grain boundaries (crystal
interfaces)
Diffusion coefficients associated with the first three mechanisms can vary widely and
are almost always at least one order of magnitude smaller than diffusion coefficients in
low-viscosity liquids. As might be expected, diffusion by the fourth mechanism can be
faster than by the other three mechanisms. Typical experimental diffusivity values,
taken mainly from Barrier, are given in Table 2. The diffusivities cover gaseous, ionic,
and metallic solutes. The values cover an enormous 26-fold range. Temperature effects
can be extremely large.
Metals
Important practical applications exist for diffusion of light gases through metals. To
diffuse through a metal, a gas must first dissolve in the metal. As discussed by Barrier,
all light gases do not dissolve in all metals. For example, hydrogen dissolves in such
metals as Cu, AI, Ti, Ta, Cr, W, Fe, Ni, Pt, and Pd, but not in Au, Zn, Sb, and Rh.
Nitrogen dissolves in Zr, but not in Cu, Ag, or Au. The noble gases do not dissolve in
any of the common metals. When H2, N2, and O2 dissolve in metals, they dissociate
and may react to form hydrides, nitrides, and oxides, respectively. More complex
molecules such as ammonia, carbon dioxide, carbon monoxide, and sulfur dioxide also
dissociate. The following example illustrates how pressurized hydrogen gas can slowly
leak through the wall of a small, thin pressure vessel.
37
Table 2 Diffusivity of Solutes in Crystalline Metals and Salts
38
Unlike metals, where hydrogen usually diffuses as the atom; hydrogen apparently
diffuses as a molecule in glass. For both hydrogen and helium, diffusivities increase
rapidly with increasing temperature. At ambient temperature the diffusivities are three
orders of magnitude lower than in liquids. At elevated temperatures the diffusivities
approach those observed in liquids. Solubilities vary only slowly with temperature.
Hydrogen is orders of magnitude less soluble in glass than helium. For hydrogen, the
diffusivity is somewhat lower than in metals. Diffusivities for oxygen are also
included in Table 3 from studies by Williams and Sucov. At 1000°C, the two values
differ widely because, as discussed by Kingery, Bowen, and Uhlmann, in the former
case, transport occurs by molecular diffusion; while in the latter case, transport is by
slower network diffusion as oxygen jumps from one position in the silicate network to
another. The activation energy for the latter is much larger than for the former (71,000
cal/mol versus 27,000 cal/mol). The choice of glass can be very critical in high-
vacuum operations because of the wide range of diffusivity.
Ceramics
Diffusion rates of light gases and elements in crystalline ceramics are very important
because diffusion must precede chemical reactions and causes changes in the
microstructure.
Therefore, diffusion in ceramics has been the subject of numerous studies, many of
which are summarized in Figure 3, taken from Kingery et aI., where diffusivity is
plotted as a function of the inverse of temperature in the high-temperature range. In
this form, the slopes of the curves are proportional to the activation energy for
diffusion, E, where
D = Do exp (- E/RT) …92
39
An insert at the middle-right region of Figure 3 relates the slopes of the curves to
activation energy. The diffusivity curves cover a ninefold range from 10-6 to 10-15
cm2/s, with the largest values corresponding to the diffusion of potassium in [β-Al2O3
and one of the smallest values for carbon in graphite. In general, the lower the
diffusivity, the higher is the activation energy. As discussed in detail by Kingery et al.,
diffusion in crystalline oxides depends not only on temperature but also on whether the
oxide is stoichiometric or not (e.g., FeO and Fe0.95O) and on impurities. Diffusion
through vacant sites of non-stoichiometric oxides is often classified as metal-deficient
or oxygen-deficient. Impurities can hinder diffusion by filling vacant lattice or
interstitial sites.
40
Polymers
Thin, dense, nonporous polymer membranes are widely used to separate gas and liquid
mixtures. Diffusion of gas and liquid species through polymers is highly dependent on
the type of polymer, whether it be crystalline or amorphous and, if the latter, glassy or
rubbery. Commercial crystalline polymers are about 20% amorphous. It is mainly
through the amorphous regions that diffusion occurs. As with the transport of gases
through metals, transport of gaseous species through polymer membranes is usually
characterized by the solution-diffusion mechanism of (Eq. 90). Fick's first law, in the
following integrated forms, is then applied to compute the mass transfer flux
Gas species:
H i .Di Pmi
Ni = ( Pi1 − Pi 2 ) = ( Pi − Pi 2 ) ...93
Z 2 − Z1 Z 2 − Z1 1
Where: Pi is the partial pressure of the gas species at a polymer surface.
Liquid species:
K .D
N i = i i (Ci1 − Ci 2 ) ...94
Z 2 − Z1
Where: Ki, the equilibrium partition coefficient, is equal to the ratio of the
concentration in the polymer to the concentration, Ci, in the liquid adjacent to the
polymer surface. The product Ki Di is the liquid permeability.
Values of diffusivity for light gases in four polymers; given in Table 4, range from
1.3*10-9 to 1.6*10-6 cm2/s, which are orders of magnitude less than for diffusion of the
same species in a gas. Diffusivities of liquids in rubbery polymers have been studied
extensively as a means of determining visco-elastic parameters. In Table 5,
diffusivities are given for different solutes in seven different rubber polymers at near-
ambient conditions. The values cover a sixfold range, with the largest diffusivity being
that for n-hexadecane in polydimethylsiloxane. The smallest diffusivities correspond
to the case where the temperature is approaching the glass transition temperature,
where the polymer becomes glassy in structure. This more rigid structure hinders
diffusion. In general, as would be expected, smaller molecules have higher
diffusivities. A more detailed study of the diffusivity of n-hexadecane in random
styrene/butadiene copolymers at 25°C by Rhee and Ferry shows a large effect on
diffusivity of fractional free volume in the polymer. Diffusion and permeability in
crystalline polymers depend on the degree of crystallinity. Polymers that are 100%
crystalline permit little or no diffusion of gases and liquids. For example, the
diffusivity of methane at 25°C in polyoxyethylene oxyisophthaloyl decreases from
0.30*10-9 to 0.13*10-9 cm2/s when the degree of crystallinity increases from 0 (totally
amorphous) to 40%. A measure of crystallinity is the polymer density. The diffusivity
41
of methane at 25°C in polyethylene decreases from 0.193*10-6 to 0.057*10-6 cm2/s
when the specific gravity increases from 0.914 (low density) to 0.964 (high density). A
plasticizer can cause the diffusivity to increase. For example, when polyvinylchloride
is plasticized with 40% tricresyl triphosphate, the diffusivity of CO at 27°C increases
from 0.23*10-8 to 2.9*10-8 cm2/s.
42
Cellular Solids and Wood
Cellular solids consist of solid struts or plates that form edges and faces of cells, which
are compartments or enclosed spaces. Cellular solids such as wood, cork, sponge, and
coral exist in nature. Synthetic cellular structures include honeycombs, and foams
(some with open cells) made from polymers, metals, ceramics, and glass. The word
cellulose means "full of little cells."
A widely used cellular solid is wood, whose annual world production of the order of
1012kg is comparable to the production of iron and steel. Chemically, wood consists
of lignin, cellulose, hemi cellulose, and minor amounts of organic chemicals and
elements. The latter are extractable, and the former three, which are all polymers, give
wood its structure. Green wood also contains up to 25 wt% moisture in the cell walls
and cell cavities. Adsorption or desorption of moisture in wood causes anisotropic
swelling and shrinkage. The structure of wood, which often consists of (1) highly
elongated hexagonal or rectangular cells, called tracheids in softwood (coniferous
species, e.g., spruce, pine, and fir) and fibers in hardwood (deciduous or broad-leaf
species, e.g., oak, birch, and walnut); (2) radial arrays of rectangular-like cells, called
rays, which are narrow and short in softwoods but wide and long in hardwoods; and
(3) enlarged cells with large pore spaces and thin walls, called sap channels because
they conduct fluids up the tree. The sap channels are less than 3 vol. % of softwood,
but as much as 55 vol. % of hardwood.
Because the structure of wood is directional, many of its properties are anisotropic. For
example, stiffness and strength are 2 to 20 times greater in the axial direction of the
tracheids or fibers than in the radial and tangential directions of the trunk from which
43
the wood is cut. This anisotropy extends to permeability and diffusivity of wood
penetrates, such as moisture and preservatives. According to Stamm, the permeability
of wood to liquids in the axial direction can be up to 10times greater than in the
transverse direction.
Movement of liquids and gases through wood and wood products takes time during
drying and treatment with preservatives, fire retardants, and other chemicals. This
movement takes place by capillarity, pressure permeability, and diffusion.
Nevertheless, wood is not highly 3.3 One-Dimensional Steady-State and Unsteady-
State Molecular Diffusion 117 permeable because the cell voids are largely discrete
and lack direct interconnections. Instead, communication among cells is through
circular openings spanned by thin membranes with sub micrometer-sized pores, called
pits, and to a smaller extent, across the cell walls. Rays give wood some permeability
in the radial direction. Sap channels do not contribute to permeability. All three
mechanisms of movement of gases and liquids in wood are considered by Stamm.
The simplest form of diffusion is that of a water-soluble solute through wood saturated
with water, such that no dimensional changes occur. For the diffusion of urea,
glycerin, and lactic acid into hardwood, Stamm lists diffusivities in the axial direction
that are about 50% of ordinary liquid diffusivities. In the radial direction, diffusivities
are about 10% of the values in the axial direction. For example, at 26.7°C the
diffusivity of zinc sulfate in water is 5*10-6 cm2/s. If loblolly pines sapwood is
impregnated with zinc sulfate in the radial direction, the diffusivity is found to be
0.18*10-6 cm2/s.
The diffusion of water in wood is much more complex. Moisture content determines
the degree of swelling or shrinkage. Water is held in the wood in different ways: It
may be physically adsorbed on cell walls in monomolecular layers, condensed in
preexisting or transient cell capillaries, or absorbed in cell walls to form a solid
solution.
Because of the practical importance of the lumber drying rates, most diffusion
coefficients are measured under drying conditions in the radial direction across the
fibers. The results depend on temperature and swollen-volume specific gravity.
44
Example 9
Calculate the rate of diffusion of Acetic acid (A) across a film of non-diffusing water
(B) solution of 1 mm thick at 17 oC. The concentrations of acid on the opposite sides
of the film are 9 wt% and 3 wt% respectively. Given that the diffusivity of Acetic acid
in the solution is 0.95*10-9 m2/s at 17 oC and the specific gravity of pure Acetic acid is
1.049. The molecular weight of Acetic acid is 60 and for water is 18.
Solution
Use equation 72
⎛ C av ⎞⎛ X − X A1 ⎞
N A = - DL ⎜ ⎟⎜ A 2 ⎟
⎟
⎜ ⎟⎜ Z 2 − Z1
⎝ X BLM ⎠⎝ ⎠
DL = 0.95*10-9 m2/s
Z2 - Z1 = 1 mm = 0.001 m
Now convert the wt fractions to mole fractions as:
9
At Z1, X A1 = 60 = 0.0288
9 + 91
60 18
3
At Z2, XA2 = 60 = 0.0092
3 + 97
60 18
45
ρ1 = X 1 * ρ H 2 O + X 2 * ρ Aceticacid
gm kgm
ρ1 = (0.91 * 1) + (0.09 * 1.049) = 1.00441 = 1004.41
cm 3 m3
Av. Mwt = mol. fraction 1 * Mwt1 + mol. fraction 2 * Mwt 2
Av. Mwt point 1 = (0.0288 * 60) + (0.9712 *18)
Av. Mwt point 1 = 19.2
ρ 2 = X 1 * ρ H 2 O + X 2 * ρ Aceticacid
gm kgm
ρ 2 = (0.97 *1) + (0.03 * 1.049) = 1.00147 = 1001.47
cm 3 m3
Av. Mwt point 2 = (0.0092 * 60) + (0.9908 *18)
Av. Mwt point 1 = 18.3
⎛ 1004.41 ⎞ ⎛ 1001.47 ⎞
⎜ ⎟+⎜ ⎟
C av = ⎝ 19 .2 ⎠ ⎝ 18.3 ⎠
= 53.3
kmol
2 m3
Example 10
A sintered solid of silica (2mm) thick is porous with void fraction of 0.3 and tortuosity
of 4. the pores are filled with water at 298 K. at one face the concentration of KCl is
held at 0.1 mol / liter and fresh water flow rapidly by other face. Neglecting any other
resistance but that in porous solid, calculate the diffusion of KCl at steady state. Given
the diffusivity DAB is equal to 1.87*10-9 m2/s.
Solution
Since neglecting any other resistance but that in porous solid, then ordinary diffusion
is occurred and then the diffusion is calculated by equation no. 90
⎛ C A2 − C A1 ⎞
N A = − Deff ⎜⎜ ⎟
⎟
⎝ Z 2 − Z1 ⎠
And Deff is calculated by equation 89
D AB ε 1.98 *10 - 9 * 0.3 −10 m
2
Deff = = = 1.4025 * 10
τ 4 s
⎛ 0 − 0.1 ⎞
N A = −1.4025 *10 −10 ⎜ ⎟ = 7.01 * 10
− 9 kmol KCl
⎝ 0.002 ⎠ m2 * s
46
Unsteady – State Mass transfer
All mass transfer processes will have an initial period of time with unsteady – state
conditions where the concentration at certain point varies with time until steady – state
is reached.
Consider an element of dimensions dx, dy, dz;
X
NA X N A X + ∆X
The molar flow rate of species A by diffusion at the plane Z=Z is given by Fick's law:
⎛ ∂C ⎞
N A Z = - D AB * A⎜ A ⎟ ...95
⎝ ∂Z ⎠ Z
The molar flow rate of species A by diffusion at the plane Z=Z+∆Z is:
⎛ ∂C ⎞
N A Z = - D AB * A⎜ A ⎟ ...96
⎝ ∂Z ⎠ Z + ∆Z
The accumulation of species A in the control volume is:
∂C A
A ∆Z ...97
∂t
Since
Rate in – Rate out = accumulation
⎛ ∂C ⎞ ⎛ ∂C ⎞ ∂C
- D AB * A⎜ A ⎟ + D AB * A⎜ A ⎟ = A A ∆Z ...98
⎝ ∂Z ⎠ Z ⎝ ∂Z ⎠ Z + ∆Z ∂t
Rearranging and simplifying
⎛ (∂C A / ∂Z ) Z + ∆Z − (∂C A / ∂Z ) Z ⎞ ∂C A
D AB ⎜ ⎟= ...99
⎝ ∆ Z ⎠ ∂t
In the lime, as ∆Z→0
∂C A ∂ 2C A
= D AB ...100
∂t ∂Z 2
The above equation is Fick's second law for one–dimensional diffusion. The more
general form, for three dimensions rectangular coordinates, is
47
∂C A ⎛ ∂ 2C ∂ 2 C A ∂ 2 C A ⎞⎟
= D AB ⎜ A + + ...101
∂t ⎜ ∂x 2 ∂y 2 2 ⎟
∂z ⎠
⎝
For one-dimensional diffusion in the redial direction only, for cylindrical and spherical
coordinates, Fick's second law becomes, respectively,
∂C A D AB ∂ ⎛ ∂C A ⎞
= ⎜r ⎟ ...102
∂t r ∂r ⎝ ∂r ⎠
And
∂C A D AB ∂ ⎛ 2 ∂C A ⎞
= ⎜r ⎟ ...103
∂t r 2 ∂r ⎝ ∂r ⎠
Analytical solutions to these partial differential equations are available for a variety of
boundary conditions. Only two of the most common useful solutions will be
considered.
Semi-infinite Medium
The boundary conditions for this case to solve (Equ.100) are:
∂C A ∂ 2C A
= D AB
∂t ∂Z 2
Boundary conditions:
At t = 0 0< Z < ∞ CA = CA0
t>0 Z =0 CA = CAi (CAi : initial concentration)
t>0 Z=∞ CA = CA0
To solve the above partial differential equation, either the method of combination of
variables or the Laplace method is applicable. The result, in terms of the fractional
accomplished concentration change (θ), is
C A − C A0 ⎛ Z ⎞
θ= = erfc⎜⎜ ⎟ ...104
C As − C A0 ⎟
⎝ 2 D AB t ⎠
Equation 104 is used to compute the concentration in the semi-infinite medium, as a
function of time and distance from the surface, assuming no bulk flow. Thus, it applies
most rigorously to diffusion in solids, and also to stagnant liquid and gases when the
medium is dilute in the diffusing solute.
The instantaneous rate of mass transfer across the surface of the medium at Z = 0 can
be obtained by taking the derivative of (Equ. 104) with respect to distance and
substituting it into Fick's first law applied at the surface of the medium. Then
⎛ ∂C A ⎞ ⎛ C As − C A0 ⎞ ⎛ Z 2 ⎞
n A = − D AB A⎜ ⎟ = D AB A⎜⎜ ⎜
⎟exp − ⎟ ...105
⎝ ∂Z ⎠Z =0 π ⎟ ⎜ 4 D t ⎟
⎝ D AB ⎠t ⎝ AB ⎠
Z =0
Thus
48
n A Z=0 =
D AB
πt
(
A C As − C A0 ) ...106
We can determine the total number of moles of solute, NA, transferred into the semi-
infinite medium by integrating equation 106 with respect to time:
( )
t D AB t dt D AB t
N A = ∫ n A z = 0 dt = A C As − C A0 ∫ = 2 A(C As − C A0 ) ...107
0 π 0 t π
49
Convective Mass Transfer for Binary gas Mixture
In previous sections we have considered molecular diffusion in stagnant fluid (laminar
flow) where the rate of diffusion is slow. To increase the rate of mass transfer, the
fluid velocity is increased until turbulent mass transfer occurs. As given before in
equation (1):
Ni = Xi.N + molecular diffusion flux of i + eddy diffusion flux of i ...1
The eddy diffusion flux term is given by:
dC A
J *A Z = - E d ...112
dZ
Then
dX
N A = - C * (D AB + E D ) * ( A ) + X A (N A + N B ) ...113
dZ
This is the general equation used to calculate the mass transfer.
1- For equi - molar mass transfer
A) For gases:
− ( D AB + E D ) dPA PA
NA = * + * (N A + N B ) ...114
R *T dZ PT
For equi – molar transfer: (NA = - NB)
Then equation 114 will be:
− ( D AB + E D ) dPA
NA = * ...115
R *T dZ
By integrating equation 115 from PA1 at Z1 to PA2 and Z 2 then:
( D AB + E D ) ( PA1 − PA2 )
NA = * ...116
R *T (Z 2 − Z 2 )
Because the film thickness can not be measured or is not known, then the value (Z2-Z1)
is not known (the distance of the path), also the amount of ED can not be measured,
then, the term (individual mass transfer coefficient) is used as shown below:
'
N A = KG (PA1 - PA 2 ) ...117
Where:
' ( D AB + E D )
KG = ...118
R * T ( Z 2 − Z1 )
Also for gases another form of equation 117 can be written as:
N A = K 'y (y A1 - y A 2 ) ...119
A) For Liquids:
Similar to what done for gases, the rate of mass transfer is:
N A = K L' (C A1 - C A 2 ) = K x' (X A1 - X A 2 ) ...120
50
The general form for the mass transfer is:
N A = K C' (C A1 - C A 2 ) ...121
All these individual mass transfer coefficients are related to each other.
For gases:
N A = K C' (C A1 - C A 2 ) = K G
'
(PA1 - PA 2 ) = K 'y (y A1 - y A 2 ) ...122
So
'
'
KG K 'y
KC = =
RT CT
Home work: find the relations for the liquid phase.
2- For uni - molecular mass transfer
A) For gases:
− ( D AB + E D ) dPA PA
NA = * + * (N A + N B ) ...114
R *T dZ PT
For uni – molar transfer: (NB = 0)
Then equation 114 will be:
− ( D AB + E D ) dPA PA
NA = * + * (N A ) ...123
R *T dZ PT
By integrating equation 123 from PA1 at Z1 to PA2 and Z 2 then
( DAB + ED ) * PT ( PA − PA )
NA = * 1 2
...124
R * T * PBLM (Z 2 − Z 2 )
Similarly as done in section 1:
N A = K G (PA1 - PA 2 ) = K y (y A1 - y A 2 ) ...125
B) For Liquids:
Similar to what done for gases, the rate of mass transfer is:
N A = K L (C A1 - C A 2 ) = K x (X A1 - X A 2 ) ...126
51
Methods to Determine the Mass transfer Coefficient
The mass transfer coefficient can be evaluated using empirical correlations involving
dimensionless numbers. These empirical equations were obtained from experimental
data using various types of fluids, different velocities, and different geometrics. The
most important groups (dimensionless numbers) are:
1- Reynolds Number (Re No.):
Re. No indicate the degree of turbulence.
ρud
Re No =
µ
Where:
ρ = density of flowing mixture fluid (solute A and solvent B).
µ = viscosity of flowing mixture fluid (solute A and solvent B).
d = diameter of the pipe in which the fluid is flow.
Dp May be used in Reynolds number equation instead of d when flow across a sphere,
also L may be used if the flow was above a flat plate.
U = mean average velocity in the pipe.
If the flow was across a packed bed, U will be substituted by:
U
U=
ε
Where
U is the superficial velocity of empty cross- section of packed bed column.
ε = void fraction.
2- Schmidt Number (Sc No.):
µ
Sc No =
ρ D AB
3- Sherwoood Number (Sh No.):
K C' d
Sh No =
D AB
4- Stanton Number (St No.):
Sh K C' K C' ρ
Sh No = = =
Re * Sc u Gm
5- J-factor
In general j-factors are uniquely determined by the geometric configuration and the
Reynolds number.
52
Now after introducing the dimensionless groups, now we will present the correlations
used to calculate the mass transfer coefficient.
1- Using the analogy or similarity of momentum, heat, and mass
transfer using Chilton – Colburn Analogy:
f K C' * ρ
jM = jH = jD = = ( Sc) 2 / 3 ...128
2 Gm
2- For flow inside pipe
For both gases and liquids, where Re > 2100
K C' d
Sh No = = 0.023(Re) 0.833 ( Sc) 0.333 ...129
D AB
Note:
Sc = 0.5 – 3 for gases
Sc > 100 for liquids
Or
jM = jH = j D = 0.023(Re) − 0.2 ...130
For 10000 < Re < 1000000
Sh No =
K C' d
D AB
(
= 2 + 0.552 Re 0.53 * Sc 0.333 )
...139
For liquids:
Sh No =
K C' d
D AB
(
= 2 + 0.95 Re 0.5 * Sc 0.333 ) ...140
For Re = 2 - 2000
And also for liquids
Sh No =
K C' d
D AB
(
= 2 + 0.347 Re 0.62 * Sc 0.333 ) ...141
7- For fluidized beds of packed with sphere, for both gases and liquids:
0.86
j D = 0.01 + ...143
0.58
Re − 0.483
In the previous sections, diffusion and mass transfer within solids and fluids were
considered, where the interface was a smooth solid surface. Of greater interest in
separation processes is mass transfer across an interface between a gas and a liquid or
between two liquid phases. Such interfaces exist in absorption, distillation, extraction,
and stripping. At fluid-fluid interfaces, turbulence may persist to the interface. The
following theoretical models have been developed to describe mass transfer from a
fluid to such an interface.
Film Theory
A simple theoretical model for turbulent mass transfer to or from a fluid-phase
boundary was suggested in 1904 by Nernst, who postulated that the entire resistance to
mass transfer in a given turbulent phase is in a thin, stagnant region of that phase at the
interface, called a film. This film is similar to the laminar sub – layer that forms when
a fluid flows in the turbulent regime parallel to a flat plate. This is shown
schematically in Figure 4. For the case of a gas-liquid interface, where the gas is pure
component A, which diffuses into nonvolatile liquid B. Thus, a process of absorption
of A into liquid B takes place, without desorption of B into gaseous A. Because the
gas is pure A at total pressure P = PA, there is no resistance to mass transfer in the gas
phase. At the gas-liquid interface, equilibrium is assumed so the concentration of A,
CAi is related to the partial pressure of A, PA, by some form of Henry's law, for
55
example, CAi= HAPA. In the thin, stagnant liquid film of thickness δ, molecular
diffusion only occurs with a driving force of (C Ai − C AB ) . Since the film is assumed
to be very thin, the entire diffusing A passes through the film and into the bulk liquid.
If, in addition, bulk flow of A is neglected, the concentration gradient is linear as in
Figure 4. Accordingly, Fick's first law, (Eq. 3), for the diffusion flux integrates to:
JA =
D AB
δ
(C Ai − C Ab ) = C * δD AB (X Ai − X Ab ) ...144
Figure 4 Film theory for mass transfer from a fluid – fluid interface into a liquid.
If the liquid phase is dilute in A, the bulk-flow effect can be neglected and (Eq. 144)
applies to the total flux:
D
(
N A = AB C Ai − C Ab =
δ
) C * D AB
δ
(
X Ai − X Ab ) ...145
If the bulk-flow effect is not negligible, then, from (Eq. 31),
C.D AB 1 − X Ab C.D AB
NA = ln( )= (X A i - X A b ) ...146
Z 2 − Z1 1 − X Ai δ (1 - X A ) LM
Where:
In practice, the ratios DAB/δ in (Eq. 145) and DAB/δ*(1- XA)LM in (Eq. 146) are
replaced by mass transfer coefficients and K c' and K c , respectively, because the film
thickness, δ, which depends on the flow conditions, is not known.
56
The film theory, which is easy to understand and apply, is often criticized because it
appears to predict that the rate of mass transfer is directly proportional to the
molecular diffusivity. This dependency is at odds with experimental data, which
indicate a dependency of Dn, where n ranges from about 0.5 to 0.75. However, if
DAB/δ is replaced with K c' , which is then estimated from the Chilton-Colburn analogy,
2/3
(Eq. 128), we obtain K c' proportional to D AB , which is in better agreement with
experimental data. In effect, δ depends on DAB (Sc No.). Regardless of whether the
criticism of the film theory is valid, the theory has been and continues to be widely
used in the design of mass transfer separation equipment.
Penetration Theory
A more realistic physical model of mass transfer from a fluid-fluid interface into a
bulk liquid stream is provided by the penetration theory of Higbie, shown
schematically in Fig. 5. The stagnant-film concept is replaced by Boussinesq eddies
that, during a cycle, (1) move from the bulk to the interface; (2) stay at the interface
for a short, fixed period of time during which they remain static so that molecular
diffusion takes place in a direction normal to the interface; and (3) leave the interface
to mix with the bulk stream. When an eddy moves to the interface, it replaces another
static eddy. Thus, the eddies are intermittently static and moving. Turbulence extends
to the interface. In the penetration theory, unsteady-state diffusion takes place at the
interface during the time the eddy is static. This process is governed by Fick's second
law, (Eq. 100), with boundary conditions
At t = 0 0 ≤ Z ≤ ∞ CA = CAb
t>0 Z =0 CA = CAi (CAi : initial concentration)
t>0 Z=∞ CA = CAb
Figure 5 Penetration theory for mass transfer from a fluid – fluid interface into a
liquid.
57
Thus, the solution can be written by a rearrangement of (Eq. 104):
C − CA ⎛ Z ⎞
θ = Ai = erf ⎜⎜ ⎟ ...147
C Ai − C Ab ⎟
⎝ 2 D AB c ⎠
t
Where tc = "contact time" of the static eddy at the interface during one cycle. The
corresponding average mass transfer flux of A in the absence of bulk flow is given by
the following form of (Eq. 107):
NA = 2
D AB
(
π * tc
)
C A i − C Ab ...148
Or (
N A = K c C A i − C Ab ) ...149
Thus, the penetration theory gives
D AB
Kc = 2 * ...150
π * tc
This predicts that Kc is proportional to the square root of the molecular diffusivity,
which is at the lower limit of experimental data. The penetration theory is most useful
when mass transfer involves bubbles or droplets, or flow over random packing. For
bubbles, the contact time, to of the liquid surrounding the bubble is taken as the ratio
of bubble diameter to bubble rise velocity. For example, an air bubble of 0.4 cm
diameter rises through water at a velocity of about 20 cm/S. Thus, the estimated
contact time, tc, is 0.4/20 = 0.02 S. For a liquid spray, where no circulation of liquid
occurs inside the droplets, the contact time is the total time for the droplets to fall
through the gas. For a packed tower, where the liquid flows as a film over particles of
random packing, mixing can be assumed to occur each time the liquid film passes
from one piece of packing to another. Resulting contact times are of the order of about
1 S. In the absence of any method of estimating the contact time, the liquid-phase mass
transfer coefficient is sometimes correlated by an empirical expression consistent with
the 0.5 exponent on DAB, given by (Eq. 150)with the contact time replaced by a
function of geometry and the liquid velocity, density, and viscosity.
We can express the rate of mass transfer in terms of mass transfer coefficients that can
be determined from any suitable theory, with the concentration gradients visualized
more realistically as in Figure 6b. In addition, we can use any number of different
59
mass transfer coefficients, depending on the selection of the driving force for mass
transfer.
Figure 6 Concentration gradients for two – resistance theory: (a) film theory; (b) more
realistic gradients.
For gas phase, under dilute or equimolar counter diffusion (EMD) conditions, we write
the mass transfer rate in terms of partial pressure:
(
N A = K g' PAb − PAi ) ...152
Or for stagnant layer mass transfer (non – volatile liquid), the rate of mass transfer can
be written as:
(
N A = K g PAb − PAi ) ...153
Where ( ' ) refers to the equimolar counter diffusion case.
The above equations can be written in terms of mole fractions as:
K g'
NA =
PT
( ) ( )
y Ab − y Ai = K 'y y Ab − y Ai ...154
And
NA =
Kg
PT
( ) (
y Ab − y Ai = K y y Ab − y Ai) ...155
60
NA =
K l'
CT
( ) (
x Ai − x Ab = K x' x Ai − x Ab ) ...158 for equi – molar mass transfer
And
K
( ) (
N A = l x Ai − x Ab = K x x Ai − x Ab
CT
) ...159 for uni - molar mass transfer
Assume a column where a gas and a liquid are contacted. At any point (P) in the
column, the gas phase has a mole fraction of A (certain composition of A) that is yA.
And the liquid has a mole fraction of A (certain composition of A) that is xA. Then at
that point an equilibrium between the gas phase and the liquid phase exist at y Ai and
x Ai at point (m)
1
(y) m ole fraction ofgas phase
0.8
0.6
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
(x) mole fraction of liquid phase
In the above figure plot the equilibrium data and also and point (p). Draw a straight
line from point (P) to intersect the equilibrium curve at point (1), let it be q.
The slop of this straight line (pq) is:
y A − y Ai K x'
Slop = =−
x A − x Ai K 'y
Then, to estimate the interface composition, we must know the mass transfer
coefficient for both gas phase and liquid phase also the equilibrium data must be
known.
The equilibrium data are presented by a curve for concentrated solution, and by a
straight line for dilute solution.
62
Case 2:
For uni-molecular diffusion (mass transfer of A through stagnant layer of B)
Let yA = yAb
and xA = xAb
(
N A = K y y A − y Ai = K x ) (x Ai − x A ) ...168
Or
K 'y
NA =
y Ai
(
y A − y Ai = ) K x'
x Ai
(x Ai − x A ) ...168A
LM LM
Then
K (y A − y Ai )
− x =
Ky (x A − x Ai ) ...169
But the slop of the drawn between points (P) & (m) is
y A − y Ai K x'
Slop = =−
x A − x Ai K 'y
Therefore substitute Ky and Kx by there equivalents as
K 'y
Ky =
y Ai
LM
And These relations must be found from the previous section
K x'
Kx =
x Ai
LM
63
(1 − x A ) − (1 − x Ai )
x Ai =
LM (1 − x A )
ln
(1 − x Ai )
And
(1 − y Ai ) − (1 − y A )
y Ai
LM
=
(1 − y Ai )
ln
(1 − y A )
From the slop equation we can conclude that there is a difficulty in calculating the
interface composition ( y Ai and x Ai ), because they are already exist in the left hand
side of the slop equation. This problem will be solved by trail and error by following
these steps:
1- Assume a value for ( y Ai and x Ai ) and let it be equal to (1).
LM LM
2- Calculate the slop.
3- Plot the line (pq).
4- From the intersection point read xAi and yAi .
5- Calculate ( y Ai and x Ai ), then recalculate the slop of the line (pq) and let it be
LM LM
(slop)2 if the value of the two slops are equal then the assumed values of
( yA and x Ai ) is correct. If not, then use the value of (slop)2 to estimate a new
iLM LM
values of xAi and yAi by repeating steps (3-5) until you will reach not more the 10%
change in the value of the slop.
Note:
For dilute solutions ( y Ai and x Ai ) are equal to (1)
LM LM
1
(y) m ole fraction ofgas phase
0.8
0.6
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
(x) mole fraction of liquid phase
65
Now, the rate of mass transfer which will be calculated based on overall driving force
can be written as follow:
A) For EMD
'
N A = K oy ( y A − y *A ) ...170
B) For UMD
N A = K oy ( y A − y *A ) ...171
Where
'
K oy and Koy are the overall mass transfer coefficients with units (kmol/m2. s.mol
fract), which are based on the overall driving force in the gas phase.
y*A : is the value of the mole fraction of A in the gas phase that would be in
equilibrium with xA.
The relationship between the overall mass transfer coefficient and the
individual mass transfer coefficient
A) Case 1 EMD
We can write the overall driving force as follow by adding and subtracting y Ai , that is:
y A − y *A = ( y A − y Ai ) + ( y Ai − y *A )
But
y Ai − y *A
m2 =
x A − x Ai
Then
y A − y *A = ( y A − y Ai ) + m2 ( x A − x Ai )
Substitute each driving force by its equivalent in the EMD case, that is:
66
NA NA m2 * N A
= +
'
K oy K 'y K x'
Finally
1 1 m
= + 2 ...174
' ' '
K oy K y K x
Equation 174 can be explained as follow:
The total resistance for mass transfer is equal to the summation of individual resistance
for each phase 9the gas phase & the liquid phase).
The same procedure could be made to find the relationship between the over all mass
transfer coefficient for the liquid phase and the individual mass transfer coefficients,
the relation is:
1 1 1
= + ...175
' ' '
K ox m3 K y K x
For the dilute solutions:
m1 = m2 = m3 = m
Then:
1 1 m
= + ...176
'
K oy K 'y K x'
1 1 1
= + ...177
'
K ox mK 'y K x'
B) Case 1 UMD
In this case only solute A is transferred through the interface and no B is transferred.
As shown in above:
K 'y
Ky =
y Ai
LM
And
K x'
Kx =
x Ai
LM
Then
K 'y
NA =
y Ai
(y A − y Ai ) = x K x'
(x Ai − x A ) ...168A
LM
AiLM
For the over all driving force, and by using the over all mass transfer coefficient, the
mass transfer rate is calculated by:
67
( ) (x*A − x A )
' '
K oy * K ox
NA = yA − yA = ...178
y *ALM x *A LM
Where:
*
x ALM =
(
(1 − x A ) − 1 − x*A )
(1 − x A )
ln
(
1 − x*A )
And
( )
1 − y*A − (1 − y A )
( )
*
y ALM =
1 − y*A
ln
(1 − y A )
Using the same procedure done in the EMD, we can find the relationship between the
overall mass transfer coefficient and the individual mass transfer coefficients, as
shown:
For the gas phase:
y *ALM y Ai m2 x Ai
= LM
+ LM
...179
' ' '
K oy Ky Kx
For the liquid phase:
x *ALm y Ai x Ai
= LM
+ LM
...180
'
K ox m3 K 'y K x'
Again for dilute solution when:
m1 = m2 = m3 = m
Then
x*ALM = y*ALM = x Ai = y Ai = 1
LM LM
Question:
Is there an overall mass transfer coefficient base on partial pressure for
the gas phase, and other one based on concentration for the liquid phase?
Answer:
Yes, these are presented as Kog, and Kol.
Question:
Find the relationships between these overall mass transfer coefficients
and the individual mass transfer coefficient?
68
Example 11
A large volume of pure gas (B) at 2 atm absolute pressure is flowing over a surface
from which pure (A) is evaporating. The liquid (A) completely wets the surface which
is a blotting paper. Hence the partial pressure of (A) at the surface is the vapor
pressure of (A) at 298 K which is (0.2 atm). The K`y has been estimated to be
(6.78*10-5 kmol / s.m2. mol fraction). Calculate the vaporizing rate (NA), and the
individual mass transfer coefficient Kg.
Solution:
Since the case is mass transfer through a stagnant layer (non-diffusing layer of B).
Then the mass transfer rate is calculated by
N A = K y (y A1 - y A 2 )
PA1 0.2
y A1 = = = 0.1 & yA2 = 0 (large volume of B)
PT 2
K 'y
Θ Ky =
y B LM
y B2 − y B1 (1 − 0) − (1 − 0.1)
y B LM = = = 0.95
⎛ y B2 ⎞ ⎛ 1 ⎞
ln⎜ ⎟ ln⎜ ⎟
⎜ yB ⎟ ⎝ 0.9 ⎠
⎝ 1 ⎠
K 'y 6.78 * 10 − 5
= 7.138 *10 − 5
kmol
Ky = =
y B LM 0.95 m 2 . S. mol fractiom
kmol
N A = K y (y A1 - y A 2 ) = 7.138 * 10 - 5 * (0.1 - 0) = 7.138 *10 - 6
m2S
K 'y = K g * PB Lm = K g * y B LM * PT
K 'y 6.78 * 10 − 5
= 3.76 *10 − 5
kmol
Kg = =
y B LM * PT 0.95 * 2 m 2 * S * atm
69
Example 12
A large volume of pure water at 26 oC is flowing parallel to a flat plate of solid
benzoic acid. The length of the plate in the direction of flow is L = 0.244 m. The water
velocity is 0.061 m/s. the solubility of Benzoic acid in water at that temperature =
0.0295 kmol/m3. The diffusivity of Benzoic acid in water (DAB = 1.24*10-9 m2/s).
Calculate the mass transfer coefficient (Kc) and the rate of mass transfer flux NA.
Assume dilute solution, given the following data:
µ = 8.71*10-4 kg/m.s
ρ = 996 kg/m3
Solution
Calculate Re No.
ρ u L 996 * 0.061 * 0.244
Re No. = = = 17000
µ −4
8.71 * 10
For liquid flow parallel a flat plate use the following equation
j M = j H = j D = 0.99(Re) − 0.5
LUρ
For 600 < Re < 50000 where Re =
µ
j D = 0.99(Re) − 0.5 = 0.99 * (17000) - 0.5 = 0.00758
Since
K C' * ρ 2 / 3 KC
'
jD = ( Sc) = ( Sc) 2 / 3
Gm u
µ 8.71 * 10 - 4
Sc = = = 705
ρD AB 996 * 1.24 *10 -9
j * u 0.00758 * 0.061 m
K C' = D = = 5.83 * 10 - 6
( Sc) 2 / 3 (705) 2/3 s
Since the benzoic acid (A) diffuses through stagnant layer of large volume of Water
(B), NB = 0
Because we are dealing with dilute solution, therefore XBLM = 1
K C' 5.83 *10 − 6
KC = = = 5.83 * 10 − 6
X B LM 1
kmol
N A = K C (C A1 - C A 2 ) = 5.83 * 10 - 6 * (0.0295 - 0) = 1.719 * 10 - 7
m2S
70
Example 13
A wetted wall column of inside diameter (2 in) contains air and CO2 flowing at 3 ft/s.
at a certain point in the column, the CO2 concentration in the air is 0.1 mol fraction, at
the same point in the column, the concentration of CO2 in the water at the water – air
interface is 0.005 mole fraction. The column operates at 10 atm and 25 oC. Calculate
the mass transfer coefficient and the mass flux at the point of consideration, given the
following data:
The diffusivity of CO2 in air at 25 oC and 1 atm = 0.164 cm2/s
The density of air at STP = 0.0808 lb/ft3
The viscosity of air at 25 oC and 10 atm = 0.018 cP
Hennery constant = 1640 atm / mol fraction
Solution
Since we are dealing with a case of fluid flow inside a pipe, then to calculate the mass
transfer coefficient use the following equation:
K C' d
Sh No = = 0.023(Re) 0.833 ( Sc) 0.333
D AB
But we are dealing with mass transfer through stagnant layer (transfer of CO2 from air
to water only), then
PT
K C = K C'
PB LM
Or
K C * PB LM * d
Sh No = = 0.023(Re) 0.833 ( Sc) 0.333
PT * D AB
And
P * D AB
KC = T * 0.023(Re) 0.833 ( Sc) 0.333
PB LM * d
To calculate the dimensionless groups we must first correct the physical properties
from their conditions to the operation condition (10 atm and 25 oC), also we must
convert all the units of all the quantities to the SI system.
ft 1m m
u=3 * = 0.915
s 3.28 ft s
2.54cm 1m
d = 2in = 2in * * = 0.0508m
in 100cm
µ = 0.018cP = 0.018 * 10 − 3
kg
m*s
71
p * mwt
ρ=
R *T
Then
⎛ P2 ⎞ 10
(ρ 2 )298K ,10atm ⎜⎜⎝ T2 ⎟⎟⎠ 298
= =
(ρ1 )STP ⎛ P1 ⎞ 1
⎜⎜ ⎟⎟
⎝ T1 ⎠ 273
3
lb 1kg ⎛ 3.28 ft ⎞
lb kg
ρ1 = 0.0808 = 0.0808 * *⎜ ⎟ = 1.296
ft 3 ft 3 2.2lb ⎝ 1m ⎠ m3
kg 10 * 273 kg
∴ ρ 2 = 1.296 * = 11.973
m3 298 m3
(D AB )2 ⎛ P ⎞ ⎛T ⎞
1.5
= ⎜⎜ 1 ⎟⎟ * ⎜⎜ 2 ⎟⎟
(D AB )1 ⎝ P2 ⎠ ⎝ T1 ⎠
2
cm 2 cm 2 ⎛ 1m ⎞ −5 m
2
( D AB )1 = 0.164 = 0.164 *⎜ ⎟ = 1.64 *10
s s ⎝ 100cm ⎠ s
1.5 2
− 5 ⎛ 1 ⎞ ⎛ 298 ⎞ −6 m
(D AB ) = 1.64 *10 *⎜ ⎟*⎜ ⎟ = 1.64 * 10
⎝ 10 ⎠ ⎝ 298 ⎠ s
PB2 − PB1
PB Lm =
⎛ PB2 ⎞
ln⎜ ⎟
⎜ PB ⎟
⎝ 1 ⎠
PA1 = y1 * PT = 0.1 *10 = 1atm
PB1 = 10 − 1 = 9atm
PA2 can be estimated from Henery law
PA = x A * H
PA2 = 0.005 *1640 = 8.2atm
PB2 = 10 − 8.2 = 1.8atm
1.8 − 9
∴ PB LM = = 4.473612
1.8
ln( )
9
Now we can calculate the mass transfer coefficient
72
P * D AB
KC = T * 0.023(Re) 0.833 ( Sc) 0.333
PB LM * d
0.833 0.333
P * D AB ⎛ ρ *u *d ⎞ ⎛ µ ⎞
KC = T * 0.023 * ⎜ ⎟ ⎜⎜ ⎟⎟
PB LM * d ⎝ µ ⎠ ⎝ ρ * D AB ⎠
0.333
−6
* 0.023 ⎛ 11.973 * 0.915 * 0.0508 ⎞
0.833
⎛ 0.018 * 10 − 3 ⎞
*⎜ ⎟
10 * 1.64 * 10
KC = *⎜
⎜ ⎟⎟
⎝ −3
⎠ ⎜ 11.973 * 1.64 * 10 − 6 ⎟
4.473612 * 0.0508 0.018 * 10 ⎝ ⎠
m
K C = 0.008868
s
N A = K g ( PA1 − PA2 )
K m 1
K g = C = 0.008868 *
R *T s atm * l m3
1000mol
0.082 * * * 298 K
mol * K 1000l kmol
K g = 3.62642 * 10 − 4
kmol
m 2 * s * atm
N A = 3.62642 * 10 − 4 * (8.2 − 1) = 2.611 *10 − 3
kmol
m2 * s
73
Example 14
A solute (A) is being diffused from a gas mixture of A & B in a wetted wall column
with a liquid flowing as a film downwards along the wall. At a certain point in the
column the gas bulk contains A of yA=0.38 (mole fraction) and at the liquid bulk the
concentration of A is xA= 0.1 mole fraction.
The column is operated at 1 atm and 25 oC and the equilibrium data in such conditions
are as follow:
K x' = 1.967 * 10 − 3
kmol A
s * m 2 * mole fraction
Calculate the interfacial composition and the flux NA.
Solution
Since we are dealing with a mass transfer operation through stagnant layer, therefore
we can not use the slop as:
K x' (y A − y Ai)
slop = − =
K 'y (x A − x Ai)
But we must use the form:
K x'
x Ai
Slop = − LM
K 'y
y Ai
LM
Then:
1- Draw the equilibrium curve.
2- Plot point P (yA = 0.38, xA = 0.1).
3- Assume y Ai and x Ai =1.
LM LM
4- Draw a line from point P with a slop of:
K x' − 1.967 *10 − 3
slop = − = = -1.342662116
' −3
Ky 1.465 * 10
74
5- Extent the line until it will intercept the equilibrium curve at point (m).
6- Read (xAi)1 and (yAi)1
(xAi)1 =0.246 and (yAi)1 = 0.18
7- Calculate y Ai And x Ai .
(1 − x A ) − (1 − x Ai ) (1 − 0.1) − (1 − 0.246)
LM LM
x Ai = = = 0.824847594
LM (1 − x A ) (1 − 0.1)
ln
(1 − x Ai ) ln
(1 − 0.246)
And
(1 − y Ai ) − (1 − y A ) (1 − 0.38) − (1 − 0.18)
y Ai
LM
=
(1 − y Ai ) = ln (1 − 0.38) = 0.715346711
ln
(1 − y A ) (1 − 0.18)
8- Recalculate the slop using the equation:
K x'
x Ai − 1.967 * 10 − 3 0.824847594
Slop = − LM
= = −1.164419871
−3
K 'y 1.465 * 10
y Ai 0.715346711
LM
Since (slop)1≠ (slop)2
x Ai = = = 0.818459811
LM (1 − x A ) (1 − 0.1)
ln
(1 − x Ai ) ln
(1 − 0.258)
And
(1 − y Ai ) − (1 − y A ) (1 − 0.38) − (1 − 0.199)
y Ai
LM
=
(1 − y Ai ) = ln (1 − 0.38) = 0.70664075
ln
(1 − y A ) (1 − 0.199)
12- Recalculate the slop using the equation:
75
K x'
x Ai − 1.967 * 10 − 3
Slop = − LM
= 0.81845811 = −1.159225845
K 'y 1.465 * 10 − 3
y Ai 0.70664075
LM
Since (slop)2 ≠ (slop)3
13-Draw a second line of slop = - 1.159225845
14- Read (xAi)1 and (yAi)1
(xAi)1 =0.257 and (yAi)1 = 0.199
15- Calculate y Ai And x Ai .
(1 − x A ) − (1 − x Ai ) (1 − 0.1) − (1 − 0.257)
LM LM
x Ai = = = 0.819
LM (1 − x A ) (1 − 0.1)
ln
(1 − x Ai ) ln
(1 − 0.257)
And
(1 − y Ai ) − (1 − y A ) (1 − 0.38) − (1 − 0.199)
y Ai
LM
=
(1 − y Ai ) = ln (1 − 0.38) = 0.70664
ln
(1 − y A ) (1 − 0.199)
16- Recalculate the slop using the equation:
K x'
x Ai − 1.967 * 10 − 3
Slop = − LM
= 0.819 = −1.15846
K 'y 1.465 *10 − 3
y Ai 0.70664
LM
Since (slop)3 ≈ (slop)4
Then we have reached the correct interface composition which is equal to:
(xAi)1 =0.257 and (yAi)1 = 0.199
And
x Ai = 0.819
LM
y Ai = 0.70664
LM
K 'y 1.465 *10 − 3
(
N A = K y * y A − y Ai = ) y Ai
* ( y A − y Ai ) =
0.70664
* (0.38 − 0.2)
LM
N A = 3.7317 *10 − 4
kmol
m2 * s
76
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4
77
Example 15
Using the same data as in example 14, calculate the overall mass transfer coefficient
Koy, the flux (NA) and the percent resistance in the gas and liquid films of the total
resistance. Base your calculation on the case that A is the only solute transferred.
Solution
1 1 m2
= +
'
K oy K 'y K x'
Knowing that
'
K oy K 'y K x'
K oy = , Ky = and K x =
* y Ai x Ai
y ALM LM LM
So we have to calculate m2, y Ai , x Ai and y*ALM .
LM LM
Then we have to repeat the solution done in previous example to find y Ai , x Ai
LM LM
We can make use of the results in example 14, then:
x Ai = 0.819
LM
y Ai = 0.70664
LM
From the plot we find x*A and y*A
x *A = 0.35
y *A = 0.052
Using the following relation to calculate m2
y Ai − y *A
m2 =
x Ai − x A
0.2 − 0.052
m2 = = 0.9427
0.257 − 0.1
( )
1 − y *A − (1 − y A ) (1 − 0.052 ) − (1 − 0.38)
( )
*
y ALM = = = 0.773
1− yA*
ln
(1 − 0. 052 )
ln
(1 − y A ) (1 − 0.38)
Now
K 'y 1.465 * 10 − 3
Ky = = = 2.0732 * 10 − 3
y Ai 0.70664
LM
78
K x' 1.967 *10 − 3
Kx = = = 2.4017 * 10 − 3
x Ai 0.819
LM
1 1 m2
= +
'
K oy K 'y K x'
1 1 0.9427
= + = 1161.852
' −3 −3
K oy 1.465 * 10 1.967 *10
'
K oy = 8.607 * 10 − 4
8.607 * 10 − 4
'
K oy
= 1.1134 *10 − 3
kmol
K oy = =
y *ALM 0.773 m 2 * s * mole fraction
N A = K oy * ( y A − y *A ) = 1.1134 * 10 − 3 (0.38 − 0.052)
N A = 3.6521 *10 − 4
kmol
m2 * s
1
In this example the resistance of the gas film is represented by: , the resistance of
Ky
m2 1
the liquid film is represented by: and the overall resistance is .
Kx K oy
1 y Ai 0.70664
= LM
= = 482.348
Ky ' −3
Ky 1.465 * 10
1 y *ALm 0.773
= = 898.1062
K oy ' − 4
K oy 8.607 *10
gas resistance 482.348
The percent gas resistance = * 100 = *100 = 53.707%
overall resistance 898.106
liquid resistance 392.512
The percent liquid resistance = * 100 = * 100 = 43.704%
overall resistance 898.1062
As you can see that the summation of percentages does not equal too 100% and that
due to errors in calculating the interface compositions.
79
Summary
1. Mass transfer is the net movement of a component in a mixture from one region to
another region of different concentration, often between two phases across an
interface. Mass transfer occurs by molecular diffusion, eddy diffusion, and bulk flow.
Molecular diffusion occurs because of a number of driving forces, including
concentration (the most important), pressure, temperature, and external force fields.
2. Fick's first law for steady-state conditions states that the mass transfer flux by
ordinary molecular diffusion is equal to the product of the diffusion coefficient
(diffusivity) and the negative of the concentration gradient.
3. Two limiting cases of mass transfer are equimolar counter diffusion (EMD) and
unimolecular diffusion (UMD). The former is also a good approximation for dilute
conditions. The latter must include the bulk-flow effect.
4. When experimental data are not available, diffusivities in gas and liquid mixtures
can be estimated. Diffusivities in solids, including porous solids, crystalline solids,
metals, glass, ceramics, polymers, and cellular solids are best measured. For some
solids-for example, wood-diffusivity is an anisotropic property.
5. Diffusivity values vary by orders of magnitude. Typical values are 0.10, 1 * 10-5,
and 1* 10-9 cm2/s for ordinary molecular diffusion of a solute in a gas, liquid, and
solid, respectively.
6. Fick's second law for unsteady-state diffusion is readily applied to semi-infinite and
finite stagnant media, including certain anisotropic materials.
80
9. A number of models have been developed for mass transfer across a two-fluid
interface and into a liquid. These include the film theory, penetration theory, surface
renewal theory, and the film-penetration theory. These theories predict mass transfer
coefficients that are proportional to the diffusivity raised to an exponent that varies
from 0.5 to 1.0.Most experimental data provide exponents ranging from 0.5 to 0.75
81
Chapter Three
Absorption and Stripping
Introduction:
In absorption (also called gas absorption, gas scrubbing, and gas washing), a gas
mixture is contacted with a liquid (the absorbent or solvent) to selectively dissolve one
or more components by mass transfer from the gas to the liquid. The components
transferred to the liquid are referred to as solute or absorbate.
Absorption is used to separate gas mixture; remove impurities, contaminants,
pollutants, or catalyst poisons from gas; or recovery valuable chemicals. Thus, the
species of interest in the gas mixture may be all components, only the component(s)
not transferred, or only the component(s) transferred.
The opposite of absorption is stripping (also called de-sorption), wherein a liquid
mixture is contacted with gas to selectively remove components by mass transfer from
the liquid to the gas phase.
The absorption process involves molecular and turbulent diffusion or mass transfer of
solute [A] through a stagnant layer gas [B] then through non – diffusing liquid [C].
There are two types of absorption processes:
1- Physical process (e.g. absorption of acetone from acetone – air mixture by water.
2- Chemical process, sometimes called chemi-sorption (e.g. absorption of nitrogen
oxides by water to produce nitric acid.
Equipment:
Absorption and stripping are conducted in tray towers (plate column), packed column,
spray tower, bubble column, and centrifugal contactors. The first two types of these
equipment will be considered in our course for this year.
1- Tray tower:
A tray tower is a vertical, cylindrical pressure vessel in which vapor and liquid, which
flow counter currently, are contacted on a series of metal trays or plates. Liquid flows
across any tray over an outlet weir, and into a down comer, which takes the liquid by
gravity to the tray below. The gas flows upward through opening in each tray,
bubbling through the liquid on the other tray. A schematic diagram for the flow
patterns inside the tray column is shown below.
In Chemical Engineering vol. 2 by J. M. Coulson & J. F. Richardson a review for the
types of trays used in plate towers are given in page 573.
82
Liquid in
2- Packed tower:
The packed column is a vertical, cylindrical pressure vessel containing one or more
section of packing material over who's the liquid flows down wards by gravity as a
film or as droplets between packing elements. Vapor flows upwards through the
wetted packing contacting the liquid. The sections of packing are contained between a
lower gas – injection support plate, which holds the packing, and an upper grid or mish
hold – down plate, which prevent packing movement. A liquid distributor, placed
above the hold – down plate, ensures uniform distribution of liquid as it enters the
packing section.
General Design Consideration
Design or analysis of an absorber (or stripper) requires consideration of a number of
factors, including:
1- Entering gas (liquid) flow rate, composition, temperature, and pressure.
2- Design degree of recovery (R) of one or more solutes.
3- Choice absorbent (stripping) agent.
4- Operating pressure and temperature and allowable pressure drop.
5- Minimum absorbent (stripping) agent flow rate and actual absorbent (stripping)
agent flow rate as a multiple of the minimum rate needed to make the separation.
6- Number of equilibrium stages.
7- Heat effects and need for cooling (heating).
8- Type of absorber (stripper) equipment.
9- Height of absorber (stripper) column.
10- Diameter of absorber (stripper) column.
The ideal absorbent should have
a- High solubility for the solute(s) to minimize the need for absorbent.
b- A low volatility to reduce the loss of absorbent and facilitate separation of
absorbent from solute(s).
83
c- Be stable to maximize absorbent life and reduce absorbent makeup requirement.
d- Be non – corrosive to permit use of common material of construction.
e- Have a low viscosity to provide low pressure drop and high mass and heat
transfer rates.
f- Be non – foaming when contacted with gas so as to make it unnecessary.
g- Be non – toxic and non – flammable to facilitate its safe use.
h- Be available, if possible.
The most widely absorbent used are water, hydrocarbon oils, and aqueous solutions of
acids and bases. While the most common stripping agents used are water vapor, air,
inert gases, and hydrocarbon gases.
Equilibrium Relations Between Phases
The equilibrium relationship in gas – liquid system is given in terms of Henry's law
constants such as:
A- for dilute solutions:
P = H*x
A A
Or y = H` * x
A A
Where:
atm
H: is the Henry's law constant in
mole fraction
mole fraction of solute in gas
H`: is the modified Henry's law constant in ( ), that means
mole fraction of solute in liquid
( H` = H )
P
T
B- for concentrated solutions:
The (yA–xA) data, may give the equilibrium relationship, which is obtained from K-
value.
In the following section, the physical process type of absorption will be considered.
1- Tray column
The absorption process can be carried out in countercurrent flow process, which may
be carried out in a single stage unit or multistage unit.
A- Single stage:
Consider the countercurrent flow process that operates under isobaric, isothermal,
continuous steady-state flow condition. The liquid phase enters from the top while the
gas enters from the bottom, with known amounts and composition for both.
Phase equilibrium is assumed to be achieved, that is the two-exit streams (liquid &
gas) leave in equilibrium with each other. Consider the following figure:
84
LT GT
xT yT
Single stage
LB GB
xB yB
Where:
L: moles of liquid phase/ unit time
G: moles of gas phase/ unit time
We will denote the streams in the top of the equipment by (T), and those in the bottom
of the equipment by (B).
Then, the overall material balance equation for the above consideration can be written
as:
L +G =L +G
T B B T
(Mole balance can be used because there is no chemical reaction)
[A] M.B.
L *x + G * y = L * x + G * y
T T B B B B T T
Where:
x: is the mole fraction of solute A in the liquid phase.
y: is the mole fraction of solute A in the gas phase.
Since only component A is redistributed between the two phases then, a balance
component for A can be written as follow:
x y x y
L` ( T ) + G ` ( B ) = L` ( B ) + G ` ( T )
1- x 1- y 1- x 1- y
T B B T
The above equation is called the operating equation where:
L`: Moles of inert liquid (solute–free absorbent) / unit time
G `: Mole of inert gas {solute–free gas (or career gas)} / unit time
The operating equation can also be written as:
L` * X + G ` * Y = L` * X + G ` * Y
T B B T
Where:
x moles of solute A in the liquid phase
X= =
1 - x moles non - A components (inerts) in the liquid phase
85
y moles of solute A in the gas phase
Y= =
1 - y moles non - A components (inerts) in the gas phase
That is, X is the mole ratio of (A) in the liquid phase to the moles of inert liquid phase,
and Y is the mole ratio of (A) in the gas phase to the moles of inert gas phase.
Both L`and G `are constants and usually known and xB, yT are in equilibrium. So the
use of Henry's law for dilute solution will solve the operating equation.
Note:
If yB> 0.1 or xT > 0.1 then the solution is considered as concentrated solution.
If yB< 0.1 & xT < 0.1 then the solution is considered as dilute solution.
Example 1:
In a single stage absorption process, a gas mixture at 1 atm pressure containing air and
H2S is contacted with pure water at 30oC. The exit gas and liquid streams reach
equilibrium. The inlet gas flow rate is (200 kmole / hr), with 0.2 mole fraction of H2S,
while the liquid flow rate entering is 600 kmole/hr. Calculate the amount and
composition of the two outlet phases. Assume water does not vaporize to the gas
phase. Given:
H = 0.0609*104 atm/mole fraction (at 1 atm and 30oC).
Solution:
Since the inlet liquid is pure water, thus
xT = 0
And, L` = LT = 600 kmole/hr
G` = GB (1-yB)
= 200 * (1-0.2) = 160 kmole/hr
Applying the operating equation of the single stage absorption
x y x y
L` ( T ) + G ` ( B ) = L` ( B ) + G ` ( T )
1- x 1- y 1- x 1- y
T B B T
0 0.2 x y
600 * ( ) + 160 * ( ) = 600 * ( B ) + 160 * ( T )
1- 0 1 - 0.2 1- x 1- y
B T
x y
40 = 600 * ( B ) + 160 * ( T )
1- x 1- y
B T
Since xB is in equilibrium with yT and Henry's constant is given, then
H
y = *x = 609 * x
T P B B
total
Substitute in the above equation then
x 609 * x
40 = 600 * ( B ) + 160 * ( B )
1- x 1 - 609 * x
B B
86
Solve the equation for xB
xB = 0.000327116 and then, yT = 0.199213892
To calculate the total flow rates:
L` 600
L = = = 600.1963341 kmole/hr
B 1- x 1 - 0.000327117
B
G` 160
G = = = 199.8036659 kmole/hr
T 1- y 1 - 0.199213892
T
1
2
3
n-1
n
n+1
LB xB GB yB
For application to an absorber, make an overall material balance for the section
denoted:
O.M.B.
L +G = Ln + G
T n +1 T
Solute (A) M.B.
L *x + G *y = Ln * x n + G * y
T T n +1 n + 1 T T
Then
G *y = Ln * x n + G * y - L * x
n +1 n +1 T T T T
87
Where:
y n + 1 : The composition of solute A in the gas phase that leaves the plate (n + 1)
x n : The composition of solute A in the liquid phase that enters the plate (n+1)
Since the solute A is the only component that transferred between the two phases, then
the above equation can be written as:
y x y x
G ` ( n + 1 ) = L` ( n ) + G ` ( T ) - L` ( T )
1- y 1- x 1- y 1- x
n +1 n T T
ِAnd
y L ` xn y L ` xT
( n +1 ) = ( ) + ( T )- ( ) ...1
1- y ` 1 - x 1 - y G ` 1 - x
n +1 G n T T
Or
L` L`
Y = X n + (Y - X ) ...2
n +1 ` T ` T
G G
L` L`
This is the equation of straight line, with a slop of `
and (Y -
T `
X ) is its
T
G G
intercept with Y- axis. It is called the operating line equation.
When using the normal axis (y, x) the operating line equation will appear as a curve
and not as a straight line unless we are dealing with dilute solutions, otherwise we
have to convert the equilibrium data to (Y, X) and use the axis as (Y, X) axis to draw
the operating line and the equilibrium data.
When dealing with dilute solutions, we can use the ordinary axis and the operating line
equation will appear as a straight line also the equilibrium data as
y
n + 1 and x n will nearly equal to y and x respectively.
n+1 n
1- y 1- x
n +1 n
88
5- Continue drawing the vertical lines and horizontal lines as in step 4 (shown in the
fig.).
6- Count the triangles constructed, this number represents the number of theoretical
plates.
7- To calculate the actual number of plates, use the following equation:
The theoretical number of plates
The actual number of plates =
η
Where:
η = the efficiency of the plate
Note:
YT = (1-R) * YB
89
Minimum absorbent flow rate:
In the design of absorbers GT, GB, yB, yT, and xT are set by process requirements, but
the quantity of entering liquid LT is subjected to choice in which finally xB depend on.
In the following figure, there are four operating lines for different flow rates, where
each operating line passes through the terminal point, P (yT, xT), at the top of the
L`
column and corresponding to a different slop ` .
G
L` L`
( )
operating
= 1.5 * ( )
min
G ` G `
Then to find xB (which is normally will not be known), we must follow these steps:
1- Draw the equilibrium data.
2- Set the point P (yT, xT) on the figure.
L`
3- Find ( )
min
from the figure.
G `
90
4- Find the slop of line 4 (PM) xB either from the figure or using the following
equation:
L y -y
( ) = B T
G min x * - x
B T
y
Where x*B = B or yB = m*xB
m
Or this value can be withdrawn from the figure.
L`
5- Calculate ( )
operating
by using suitable factor (1.1 – 1.5) or as defined in the
G `
problem. Then calculate the actual value (operating) of LT.
6- Substitute the operating line slop in the equation of the operating line to find xB.
y L ` xB y L ` xT
( B )= ( ) + ( T )- ( )
1- y ` 1 - x 1 - y ` 1 - x
B G B T G T
7- Continue to find the number of theoretical plates.
ﻓﺄن اﻗﻞ ﻣﻴ ﻞ ﻳﻤﻜ ﻦ ﺣ ﺴﺎﺑﻪ ﻋ ﻦ ﻃﺮﻳ ﻖ،ﻓﻲ ﺣﺎﻟﺔ آﻮن ﻣﻌﻠﻮﻣﺎت اﻟﺘﻮازن ﺗﻌﻄﻲ ﺷﻜﻞ ﻣﻨﺤﻨﻲ آﻤﺎ ﻣﻮﺿﺢ ﻓﻲ اﻟﺸﻜﻞ
.( ﻟﻤﻨﺤﻨﻲ اﻟﺘﻮازن و آﻤﺎ ﻣﻮﺿﺢ ﻓﻲ اﻟﺸﻜﻞ ادﻧﺎﻩP) رﺳﻢ ﻣﻤﺎس ﻣﻦ اﻟﻨﻘﻄﺔ
91
For stripping:
The process can be shown in the figure:
LT xT GT y T
N
n+1
n
n-1
3
2
1
ﻧﻔﺲ اﻟﺘﻄﺒﻴﻖ اﻟﺴﺎﺑﻖ و ﺑﻨﻔﺲ اﻟﺘﺮﺗﻴﺐ اﻟﺬي اﺳﺘﺨﺪم ﻓﻲ ﺷﺮح ﺣﺴﺎﺑﺎت اﻻﻣﺘﺼﺎص ﻳﻄﺒﻖ هﻨﺎ اﻳﻀﺎ" ﻣﻊ اﻻﺧﺬ ﺑﻨﻈﺮ
اﻻﻋﺘﺒﺎر ان ﻣﻨﺤﻨﻲ اﻟﺘﻮازن هﻨﺎ ﻋﻨﺪ رﺳﻤﻪ ﺳ ﻴﻜﻮن اﻋﻠ ﻰ ﺧ ﻂ اﻟﺘ ﺸﻐﻴﻞ اي ﻋﻜ ﺲ ﺣﺎﻟ ﺔ اﻻﻣﺘ ﺼﺎص .و هﻨ ﺎ اﻳ ﻀﺎ"
ﻳﺠﺐ اﻻﻧﺘﺒﺎﻩ هﻨﺎ ﻋﻠﻰ اﻧﻨﺎ ﻧﺘﻌﺎﻣﻞ ﻣﻊ اﻗﻞ آﻤﻴﺔ ﻣﻦ اﻟﻄﻮر اﻟﻐﺎزي اﻟﺪاﺧﻞ اي ) (Gminو ﺑﺘﻘﻠﻴ ﻞ )` (Gﺳ ﻴﺰﻳﺪ اﻟﻤﻴ ﻞ
اﻟﻰ ان ﻳﻘﻄﻊ ﻣﻨﺤﻨﻲ اﻟﺘﻮازن و آﻤﺎ ﻣﻮﺿﺢ ﻓﻲ اﻻﺷﻜﺎل ادﻧﺎﻩ.
92
Algebraic method for determining the number of equilibrium stages:
Graphical methods for determining equilibrium stages have great educational value
because of fairly complex multistage problem can be readily followed and understood.
Furthermore, one can quickly gain visual insight into the phenomena involved.
However, the application of graphical method can become very tedious when (1) the
problem speciation fixes the number of stages rather than the percent recovery of
solute, (2) when more than one solute is being absorbed or stripped, (3) when the best
operating conditions of temperature and pressure are to be determined so that the
location of the equilibrium curve is unknown, and/or (4) if very low or very high
concentrations force the graphical construction to the corners of the diagram so that
multiple y – x diagrams of varying sizes and dimensions are needed. Then the
application of an algebraic method may be preferred.
When the flow rate of L and G are constant and the process is countercurrent and the
equilibrium line is a straight line and can be presented by the relation y = m* x, also if
the operating line is a straight line (i.e. dilute solution), a simplified analytical equation
found by Kremser et al. can be used to find the number of theoretical plates as follow:
1- For absorption:
y −y N+1 - A
B T =A
y - y* A N+1 - 1
B T
Then
y - m*x
log[ B T ( 1- 1 ) + 1 ]
y - m*x A A
N= T T
log A
Where
N: Theoretical number of plates.
y* = m * x (Equilibrium relation for dilute solution)
T T
A: absorption factor which is the ratio of slop of operating line ( L G ) to the slop of
equilibrium line (m), then:
L
A=
m*G
For high degree of absorption (high % recovery):
1- Use large number of plates.
2- Use high absorption factor (A), since m is constant by the system, then to get
high value of A means ( L G ) must be larger than m.
The most economic value of A is about (1.3) {suggested by Calburn}.
93
2- For stripping:
( ) N+1 - ( )
1 1
x −x
T B = A A
y 1 N+1
x -( B) ( ) -1
T m A
And
y
x - ( B m)
log[ T ( 1 - A ) + A]
y
x - ( B m)
N= B
1
log ( )
A
Where:
y
B = x*
m B
اﻻﺷﺘﻘﺎﻗﺎت ﻣﻄﻠﻮﺑﺔ ﻟﻠﻤﻌﺎدﻻت اﻟﻤﺴﺘﺨﺪﻣﺔ ﻓﻲ ﺣﺴﺎب ﻋﺪد اﻟﺼﻮاﻧﻲ و ﻳﻤﻜﻦ اﻻﻃﻼع ﻋﻠﻴﻪ ﺑﻤﺮاﺟﻌﺔ آﺘ ﺎب اﻟﻬﻨﺪﺳ ﺔ
Chemical engineering (vol. 2) by J. M.Coulson &J. F. . 573 اﻟﻜﻴﻤﻴﺎوﻳﺔ )اﻟﺠﺰء اﻟﺜﺎﻧﻲ( اﻟﺼﻔﺤﺔ
Richardson
Example 2:
When molasses is fermented to produce a liquor containing ethyl alcohol, a CO2- rich
vapor containing a small amount of ethyl alcohol is evolved. The alcohol can be
recovered by absorption with water in a sieve-tray tower. For the following conditions,
determine the number of equilibrium stages required for countercurrent flow of liquid
and gas, assuming isothermal, isobaric conditions in the tower and neglecting mass
transfer of all components except ethyl alcohol.
Entering gas:
180 kmole/hr; 98 mol% CO2, 2 mol% ethyl alcohol; 30 oC, 110 kPa
Entering liquid absorbent:
100 % water; 30 oC, 110 kPa
Required recovery (absorption) of ethyl alcohol (R): 97%
Equilibrium relationship: y = 0.57*x
Solution:
YT = (1- R) * YB
y 0.02
Y = B = = 0.02041
B 1- y 1 - 0.02
B
yT = 0.03 * 0.02041 = 0.00061 (mole fraction of ethyl alcohol in the top)
x* = yB/m = 0.02/0.57 = 0.03509
B
94
L y -y
( ) = B T
G min x * - x
B T
L 0.02 - 0.00061
( ) = = 0.5529
G min 0.03509 − 0
Take the operating factor 1.5
Then:
L`
( ) = 1.5 * 0.5529 = 0.8293
` operating
G
L` y -y
( ) = B T
operating x - x
G` B T
L` 0.02 - 0.00061 0.0194
( ) = =
G ` operating x - 0 x
B B
So xB = 0.0194/ 0.8293
xB = 0.0234 (the mole fraction of the liquid exits from the bottom)
Now
Draw the equilibrium data using the given equation (the result curve is a straight line)
Draw the operating line using the two points (yB, xB) and (yT, xT) { because we are
dealing with dilute solution yB < 0.1}
From the figure calculate the number of plates = 7
95
equilibrium line
operating line
0.03
0.025
0.02
0.015
y
0.01
0.005
0
0 0.01 0.01 0.02 0.02 0.03 0.03
x
96
2- Packed column:
Absorption and stripping are frequently conducted in packed columns, particularly
when:
(1) the required column diameter is less than 2 ft.
(2) the pressure drop must be low, as for a vacuum service.
(3) corrosion consideration favor the use of ceramic or polymeric material.
(4) low liquid holdup is desirable.
Structured packing is often favored over random packing for revamps to overcome
capacity limitation of trayed towers.
The countercurrent packed towers operates in a different manner from plated towers.
In packed towers the fluids are in contact continuously in their path through the tower,
while in plated towers the fluids are contacted occasionally. So, packed columns are
continuous differential contacting devices that do not have the physically
distinguishable stages found in tray towers. Thus, packed columns are best analyzed
by mass transfer considerations rather than by the equilibrium – stage concept.
Nevertheless, in practice packed – towers performance is often analyzed on the basis
of equivalent equilibrium stages using a packed height equivalent to a theoretical
(equilibrium) plates (stage), called the HETP or HETS and defined by the equation:
packed height l
HETP = = T
number of equivalent equilibrium stages Nt
The most important reason to use packed towers in absorption is to provide a large
contact area as possible as can. There are many types of packing such as, Rasching
rings, Lessing rings, Bert Saddles, Pall ring, and many others, most of these types of
packing are made of cheep inert materials such as glass, ceramic, graphite or, plastic,
but some times it may be made of stainless –steal.
A schematic diagram of the packed tower is shown in below.
97
Calculations of the height of packing: xT yT
A- for dilute mixtures LT GT
Consider the following figure, then
O. M. B. on the denoted section
L + ( G + dG ) = ( L + dL ) + G
∴ dG = dL
N
[A] M. B.: N = A
A A
d (G* y) = d (L*x)
LB GB
xB yB
Θ d (G * y) = N
A
And
Θ N A = k y ( y − y i )dA
or
N A = k oy ( y − y * )dA
Then
d (Gy ) = k y ( y − y i )dA
or
d (Gy ) = k oy ( y − y * )dA
Where:
dA: is the interfacial area associated with differential tower length, it is very difficult
to measure in packed towers so, for packed towers
dA = a*s*dZ
Where:
a: Specific interfacial area per unit volume of packing (m2/m3)
s: empty tower cross-sectional area (m2)
dZ: differential height in (m)
Then:
d (Gy ) = k y a ( y − y i ) sdz = k oy a ( y − y * ) sdz
So
Z yB yB
d (Gy ) d (Gy )
∫0 dZ = y∫ k y a( y − yi )s = ∫k
yT oy a ( y − y ) s
*
T
In absorption of dilute solution the flow rate does not effected by the small change in
concentration, so
G +G
G = Gm = T B (i.e. constant through out the column)
2
98
Gm y B dy Gm y B dy
∴z = =
k y as y∫ y − yi k oy as y∫ y − y *
T T
The equation is based on gas phase and is used in calculating the height of the packing
(Z) are applied only for dilute or weak solution which leads to the assumption that Lm
& Gm are constant through out the solution. Don't forget that the solutions can be
considered dilute if the mole fraction of the solute in the inlet streams (gas & liquid)
are less than 0.1 (i.e. ≤ 10%). So the operating line can be considered as a straight line
and also the equilibrium curve.
Therefore the integration terms:
y −y
yB
dy
∫
yT y− y *
= B *T
( y − y )m
Where:
( y B − y B* ) − ( yT − yT* )
( y − y )m =
*
( y − y B* )
ln B
( yT − yT* )
Also
y B − yT
yB
dy
∫ y− y = اﻻﺷﺘﻘﺎق ﻣﻄﻠﻮب ﻓﻲ آﻴﻔﻴﺔ اﻟﻮﺻﻮل اﻟﻰ اﻟﻨﺎﺗﺞ ﻟﻜﻼ اﻟﻤﻌﺎدﻟﺘﻴﻦ
yT i ( y − yi ) m
Where:
( y B − y Bi ) − ( yT − yTi )
( y − yi ) m =
( y − y Bi )
ln B
( yT − yTi )
Now let us name the following terms
1- Gm is called the height of transfer unit based on gas film (HTU)y with the units of
k y as
(m)
2- Gm is called the height of transfer unit based on overall gas phase (HTU)oy with the
k oy as
units of (m)
yB
3- ∫ dy is called the number of transfer units based on gas film (NTU)y (without
y− y
yT i
units).
yB
4- ∫ dy * is called the number of transfer units based on overall gas phase (NTU)oy
y− y
yT
(without units).
Then the height of packing is calculated as follow:
yB
Gm dy
Z = (NTU)y * (HTU)y =
k y as ∫ y− y
yT i
Or
99
yB
Gm dy
Z = (NTU)oy * (HTU)oy =
koy as ∫ y− y
yT
*
All the above derivations can be done on the basis of liquid phase which will lead to:
x x
L B dx L B
dx
z= m ∫ = m ∫ *
k x as xT xi − x k ox as xT x − x
Where:
1- Lm is called the height of transfer unit based on liquid film (HTU)x with the units
k x as
of (m)
2- Lm is called the height of transfer unit based on overall liquid phase (HTU)ox with
k ox as
the units of (m)
xB
3- ∫ dx is called the number of transfer units based on liquid film (NTU)x (without
x −x
xT i
units).
xB
4- ∫ *dx is called the number of transfer units based on overall liquid phase (NTU)ox
x −x
xT
That is:
Z = (NTU)ox * (HTU)ox= (NTU)x * (HTU)x
The terms kya, kxa are called the volumetric film coefficient and their units are:
kmole
Sec * molefraction * m 3 packing
Where kOya, kOxa are called the volumetric overall mass transfer coefficient and their
units:
kmole
Sec * molefraction * m 3 packing
Therefore, depending on the liquid film (overall mass transfer coefficient) the
derivation will be as:
d (lx) = k x a( x − xi ) sdz = k ox a( x − x * ) sdz
So
Z xB B x
d (lx) d (lx)
∫ dZ =
0
∫x k x a( x − xi )s x∫ k ox a( x − x * )s
=
T T
In absorption of dilute solution the flow rate does not effected by the small change in
concentration, so
L +L
L = Lm = B T (i.e. constant through out the column)
2
xB xB
Lm dx L dx
∴z =
k x as ∫
xT
= m
x − xi kox as ∫ x−x
xT
*
(without units).
Also
1- Lm is called the height of transfer unit based on liquid film (HTU)L with the units
k L as
of (m)
2- Lm is called the height of transfer unit based on overall liquid phase (HTU)OL with
k OL as
the units of (m)
cAB
dc A
3- ∫ is called the number of transfer units based on liquid film (NTU)L (without
c AT
c Ai − c A
units).
cAB
dc A
4- ∫ is called the number of transfer units based on overall liquid phase
cA − cA
*
c AT
(NTU)OL
Home work:
1-what are the units of kG, kOG, kL, and kOL?
2- Solve the integration to find the number of transfer unit?
The following terms, (HTU)G, (HTU)y, (HTU)L, and (HTU)x, , are called the
individual height of transfer units. While the terms (HTU)OG, (HTU)oy, (HTU)OL, and
(HTU)ox are the overall height of transfer units.
101
The relation between the individual and overall height of transfer units:
1- For gas film
From previous lectures
1 1 m1
= +
k Oy k y k x
Then finally
1
(HTU) oy = (HTU) y + * (HTU) x
A
102
Then finally
(HTU) ox = A * (HTU) y + (HTU) x
Home work: find the relation between the individual height of transfer units based on
(kG, kL) and the overall height of transfer units which are based on the (kOG, kOL).
In absorption, the equation which relates the individual mass transfer coefficient to the
overall mass transfer coefficient, as given in the above, is:
1 1 m1
= + (For gas phase)
k Oy ky kx
1 1 1
= + (For liquid phase)
k Ox m2 * k y k x
Or
1 1 m1
=
+ (For gas phase)
k OG k G k L
1 1 1
= + (For liquid phase)
k OL m2 * k G k L
And as mentioned before, that for dilute solutions: m=m1=m2
We have to two cases:
A) Very soluble gas
If m is very small, the equilibrium line is almost horizontal, that is a small change in
the value of yA in the gas phase will result a large change in the value of xA in the
liquid phase. So the solute is then called very soluble in liquid, then the term m → 0
kx
(because m is very small, or we can say that the resistance in the liquid phase is
negligible). Then:
1 1
≈
k oy ky
Also in such case, it is called (gas controlling process)
Then, the point m will move down very close to point E, that means:
y − y* ≈ y − y
i
B) Very insoluble gas
If m is very large, the equilibrium line is almost vertical, that is a large change in the
value of yA in the gas phase will result a small change in the value of xA in the liquid
phase. So the solute is then called very insoluble in liquid, then the term 1 → 0
m*k y
103
(because m is very large, or we can say that the resistance in the gas phase is
negligible). Then:
1 1
≈
k ox kx
Also in such case, it is called (liquid controlling process)
Then, the point m will move up very close to point D, that means:
x* − x ≈ x − x
i
` ` k `x a/(1 − x )
( k y a, and k x a ) then the slop = − B } to intersect the equilibrium line at m .
1
k `y a/(1 − y )
B
Read (yiB, xiB) as shown in the figure.
k a
5- From point P2 (xT, yT) draw a line of slop = − x {and if the data is given as
k ya
` ` k `x a/(1 − x )
( k y a, and k x a ) then the slop = − T } to intersect the equilibrium line at m .
2
`
k y a/(1 − y )
T
Read (yiT, xiT) as shown in the figure.
Note:
If the overall mass transfer coefficient is being used then y*B , y*T , x*B , and x*T must be
determined as shown in the figure instead of steps 4 &5.
6- Calculate (y-yi)m and (xi-x)m or (y-y*)m and (x*-x)m depending on the given data
using the equations given in previous sections.
7- Calculate (HTU), (NTU) depending on the given data, finally calculate Z.
104
yB P1
yiB
yT P2
yiT
105
Example 3:
When molasses is fermented to produce a liquor containing ethyl alcohol, a CO2- rich
vapor containing a small amount of ethyl alcohol is evolved. The alcohol can be
recovered by absorption with water in a tower packed with 1.5-in metal Pall rings. For
the following conditions, determine the height of packing required for countercurrent
flow of liquid and gas, assuming isothermal, isobaric conditions in the tower and
neglecting mass transfer of all components except ethyl alcohol. Given that
(HTU)oy=2.0 ft.
Entering gas:
180 kmole/hr; 98 mol% CO2, 2 mol% ethyl alcohol; 30 oC, 110 kPa
Entering liquid absorbent:
100 % water; 30 oC, 110 kPa
Required recovery (absorption) of ethyl alcohol (R): 97%
Equilibrium relationship: y = 0.57*x
Solution:
YT = (1- R) * YB
y 0.02
Y = B = = 0.02041
B 1- y 1 - 0.02
B
yT = 0.03 * 0.02041 = 0.00061 (mole fraction of ethyl alcohol in the top)
x* = yB/m = 0.02/0.57 = 0.03509
B
L y -y
( ) = B T
G min x * - x
B T
L 0.02 - 0.00061
( ) = = 0.5529
G min 0.03509 − 0
Take the operating factor 1.5
Then:
L`
( ) = 1.5 * 0.5529 = 0.8293
operating
G`
L` y -y
( ) = B T
operating x - x
G` B T
L ` 0.02 - 0.00061 0.0194
( ) = =
operating
G` x -0
B
x
B
So xB = 0.0194/ 0.8293
xB = 0.0234 (the mole fraction of the liquid exits from the bottom)
Now:
y B − yT
yB
(NTU)oy= ∫ dy =
y− y
yT i ( y − yi ) m
106
(y B − y *B ) − (y T − y *T ) (0.02 - 0.57 * 0.0234) - (0.00061 - 0.57 * 0)
(y − y ) m =
*
= = 0.002531459
(y B − y *B ) ln(
0.02 - 0.57 * 0.0234
ln
(y T − y *T ) 0.00061 - 0.57 * 0
0.02 - 0.00061
∴ (NTU) oy = = 7.66
0.002531459
Z = (HTU)oy * (NTU)oy = 2 * 7.66 = 15.32 ft
Example 4:
Experimental data have been obtained for air containing 1.6 vol% SO2 being scrubbed
with pure water in a packed column of 1.5 m2 in cross-sectional area and 3.5 m in
packed height. Entering gas and liquid flow rates are 0.062 and 2.2 kmole/sec,
respectively. If the outlet mole fraction of SO2 in the gas is 0.004 and column
temperature is near ambient, calculate from the data:
(a) The (NTU)oy for absorption of SO2
(b) The (HTU)oy in meters
(c) The volumetric overall mass transfer coefficient, koya for SO2 in (kmole/m3 s
mole fraction)
Solution
(a) First we must find xB
[SO2] O.M. B.
L *x + G * y = L *x + G * y
T T B B B B T T
And finally
L` * X + G ` * Y = L` * X + G ` * Y
T B B T
L`= LT (because xT = 0)
L` = 2.2 kmole/s
G` = GB *(1- yB) = 0.062 * (1- 0.016) = 0.061008 kmole/s
Then
0.016 0.004
(L * 0 ) + [ 0.061008 * ( )] - [0.061008 * ( )]
X =
T 1 - 0.016 1 - 0.004 = 0 + 0.000992 - 0.000245012 X
B B
2.2 2.2
= 0.00034
Then
xB = 0.00034 (dilute solution XB = xB)
y B − yT
yB
(NTU)oy= ∫ dy * =
y− y yT ( y − y* )m
(y B − y *B ) − (y T − y *T ) (0.016 - 40 * 0.00034 ) - (0.004 - 40 * 0)
(y − y * ) m = = = 0.003132184
(y B − y *B ) ln(
0.016 - 40 * 0.00034
)
ln
(y T − y *T ) 0.004 - 40 * 0
0.016 - 0.004
∴ (NTU) oy = = 3.8312
0.003132184
(b) Z = (HTU)oy * (NTU)oy
107
(HTU)oy = Z / (NTU)oy= 3.5/3.8312 = 0.9136 m
Gm
(c) (HTU) oy =
k oy as
Gm 0.062 kmole
∴ k oy a = = = 0.045245
(HTU) oy s 0.9136 *1.5 m 3 Sec mole fraction
Example 5:
Acetone is being absorbed by water in a packed tower having a cross-sectional area of
0.186 m2 at 293 k and 101.3 kPa. The inlet air contains 2.6 mol% acetone and the out
let 0.5 mol%. The gas flow rate is 13.65 kmole air inlet/ hr. The pure water inlet flow
rate is 45.36 kmole/ hr. The film coefficients for the given conditions in the tower are:
komle
k `y a = 3.78x10 - 2
Sec . m 3packing . mole fraction
komle
k `x a = 6.16x10 - 2
Sec . m 3packing . mole fraction
The equilibrium data is given by the equation: (y = 1.186 x)
Calculate the height of packing using:
(a) k`ya
(b) k `x a
Also calculate ( k `oy a ) and the height of packing based on this coefficient.
Solution:
(a) Z = (HTU)y * (NTU)y
Where:
y B − yT
yB
(NTU)y = ∫ dy =
y− y yT i ( y − yi ) m
( y B − y Bi ) − ( yT − yTi )
And ( y − yi ) m =
( y − y Bi )
ln B
( yT − yTi )
G
Also (HTU)y = ` m
k y as
Then we have to find xB so then we can find each of xiB, yiB, xiT, and yiT
[A] M. B.
L` * x + G ` * y = L` * x + G ` * y (Dilute solution)
T B B T
L`= LT = 45.31 kmole/hr (xT = 0)
G` = 13.65 kmole/hr (given in data)
Then:
(0 * 45.36) + (0.026 *13.65) − (0.005 *13.65)
x = = 0.00632
B 45.36
108
Now to find the interface composition:
1- Plot the equilibrium data and the operating line as shown in the figure.
2- Find the slop of line P1M1
- k `x a - 6.16x10 - 2
(1 − x B ) (1 - 0.00632)
slop = = = - 1.5974
k `y a 3.78x10 - 2
(1 − y B ) (1 − 0.026)
3- Draw the line P1M1 of slop -1.5974, then read xiB, yiB
xiB = 0.013 , yiB = 0.0154
equilibrium line
0.06
0.05
y (mole fraction)
0.04
0.03
0.02
0.01
0
0 0.005 0.01 0.015 0.02
x (mole fraction)
5- Draw the line P1M1 of slop -1.6215, then read xiB, yiB
xiT = 0.00186 , yiT = 0.002
6- Calculate (NTU)y
y B − yT
yB
(NTU)y= ∫ dy =
y− y yT i ( y − yi ) m
109
( y B − y Bi ) − ( yT − yTi ) (.026 - 0.0154) - (0.005 - 0.002)
( y − yi ) m = = = .00682
( y B − y Bi ) 0.026 - 0.0154)
ln ln(
( yT − yTi ) 0.005 − 0.002
110
0.012604
(HTU)x = = 1.1 m
− 2
6.16x10 * 0.186
Z = 1.1 * 1.6764 = 1.844 m
G ` * dy
dZ =
s * k y a(1 − y ) 2 ( y − yi )
k `y a
Θ k ya = ( yim ≠ 1 concentrated system )
y im
Then
y
G` B dy
Z= ∫
s y k` a
y
T
(1 − y ) 2 ( y − y )
y i
im
ﻻ ﻳﺨﺮج ﺧﺎرج اﻟﺘﻜﺎﻣﻞ ﻻﻧﻪ آﻤﺎ ذآﺮﻧﺎ ﺳﺎﺑﻘﺎ ﺑﺎﻧﻪ ﻳﺘﻐﻴﺮ ﻣﻊ ﺗﻐﻴﻴﺮ اﻟﺘﺮآﻴﺰ k `y a او اﻟﻤﻘﺪار k y a اﻟﻤﻘﺪار هﻨﺎ
As done above the following can be driven:
111
y
G` B dy
Z= ∫
s y k` a
T oy
(1 − y ) 2 ( y − y* )
y *
m
Or
x
L` B dx
Z= ∫
s x k` a
T x (1 − x) 2 ( x − x)
x i
im
Or
x
L` B dx
Z= ∫
s x k` a
T ox (1 − x) 2 ( x* − x)
x*
m
All the above design equations cannot be solved analytically and must be solved by
graphical integration as the following steps:
1- Draw the operating curve by using the following equation:
n
y
+ 1 L ` xn y
T L ` xT
( )= ( ) + ( )- ( )
1- y G ` 1- x 1- y G ` 1- x
n +1 n T T
By assuming xn and then find yn+1, plot (yn+1 vs. xn).
2- Draw the equilibrium data (given.
3- The values of ( k `x a ) & ( k `y a ) are obtained from empirical equations (will be given
to you), where the film coefficient are functions of ( G ny ) ( Kgm total gas ) & ( Lnx )
2 m * Sec
( Kgm total liquid ) where (n & m) in the range of (0.2 – 0.8).
m 2 * Sec
Gy, Lx will be calculated for different values of y & x using that were found by the
operating equation. So, for different values of y, x in the tower convert the value of G
& L to Gy, Lx as follow:
y
G y = G ` * Mwt of inert gas (carrier) + G `( ) * Mwt absorbed gas
1− y
(mass flow rate of carrier gas) + mass flow rate of absorbed gas at any section in the
column
112
x
L x = L` * Mwt of inert liquid + L` ( ) * Mwt absorbed gas
1− x
mass flow rate of carrier liquid + mass flow rate of absorbed gas at any section in the
column
Then the value of ( k `x a ) & ( k `y a ) are calculated as mentioned before using the
empirical equation for each value of Gy & Lx which means calculate ( k `x a ) & ( k `y a )
for each y &x.
4- Starting with tower bottom composition P1 (yB, xB), the interface composition (yiB,
xiB) is determined by plotting line P1M1 with a slop calculated by:
k `x a
x
Slop = - im
k `y a
y
im
Where:
(1 - x ) − (1 − x )
x = B iB
im 1− x
ln( B)
1− x
iB
(1 - y ) − (1 − y )
y = B iB
im 1− y
ln( B)
1− y
iB
Starting with xim = 1 - xB & yim = 1 - yB then, the slop of the line P1M1 is calculated by
trail and error.
5- The second point taken is P2 (yT, xT) and determine the slop of line P2M2 using the
same procedure done in step 4 by using xim = 1- xT & yim = 1- yT as the starting point.
Repeating this technique for several intermediate points in the tower to find the
interface composition for each point.
6- Set up the following table from the collected data:
x y Gy Lx k `x a k `y a xi yi xim yim (1-y)2 (y-yi) F(y)
xB y B
x1 y 1
x2 y 2
x3 y 3
xT yT
113
Where:
1
f(y) =
k `y a
(1 − y)2 (y − y )
y i
im
7-
8- Plot f(y) vs. (y) and then calculate the area under the curve starting from yB to yT.
To calculate the area under the curve. There are two methods:
(a) By counting the squares formed and multiply the number of squares by the area of
each square as shown below.
114
Absorption with chemical reaction (chemisorption)
Some processes involve chemical reaction of gas (A) with one component in the liquid
phase (B). Absorption with chemical reaction enlarges the liquid capacity and increase
the absorption rate. In the following some absorption processes involve chemical
reaction:
Solute gas (A) Liquid (B) reagent
CO2 Carbonates, hydroxides, Ethanol amines
NO2 H2 O
CO Cuprous, ammonium chloride
SO2 Ca(OH)2, KOH
Cl2 H2O, FeCl2
H2 S Ethanol amines, Fe(OH)2
SO3 H2SO4
NO FeSO4, Ca(OH)=, H2SO4
C2H4 KOH
115
( ﻟﻼﻧﺘ ﺸﺎر ﻓ ﻲ ﻃﺒﻘ ﺔA) ( ﻳﻨﺘﻬﻲ ﻓﻲ ﻃﺒﻘﺔ اﻟﺴﺎﺋﻞ ﻣﻤﺎ ﻳﺴﺘﺪﻋﻲ اﻟﻐﺎزB) اﻟﺘﻔﺎﻋﻞ ﻳﺤﺼﻞ ﻣﺒﺎﺷﺮة ﻣﻤﺎ ﻳﻌﻨﻲ انΘ
اﻟﺴﺎﺋﻞ و ﺗﺘﻜﻮن ﻟﺬﻟﻚ ﻃﺒﻘﺔ اﺧﺮى )ﻣﻨﻄﻘﺔ اﻟﺘﻔﺎﻋﻞ( ﻓﻲ ﻃﺒﻘﺔ اﻟﺴﺎﺋﻞ ة اﻟﺘﻲ ﻳﺤ ﺪث ﻓﻴﻬ ﺎ اﻟﺘﻔﺎﻋ ﻞ و هﻨ ﺎ ﻳﻜ ﻮن ﺗﺮآﻴ ﺰ
( ﺳﻴﻜﻮن ﻓﻲ اﻟﻄﺒﻘﺔ اﻟﺘﻲ ﻳﺤﺪث ﻓﻴﻬﺎ اﻟﺘﻔﺎﻋ ﻞ ﺛ ﻢ ﻳﺒ ﺪأ ﺑﺎﻟﻨﻘ ﺼﺎن ﻟﺬوﺑﺎﻧ ﻪAB) ان اﻋﻠﻰ ﺗﺮآﻴﺰ ﻟﻠﻤﺎدة. ﺻﻔﺮB وA
ﻓ ﻲ ﻃﺒﻘ ﺔ اﻟ ﺴﺎﺋﻞA أن ﻣﻮﻗﻊ اﻟﻄﺒﻘﺔ اﻟﺘﻲ ﻳﺤﺪث ﻓﻴﻬﺎ اﻟﺘﻔﺎﻋﻞ ﻳﺘﺤ ﺪد ﻋﻨ ﺪﻣﺎ ﻳﺘ ﺴﺎوى ﻣﻌ ﺪل اﻧﺘ ﺸﺎر.ﻓﻲ ﻃﺒﻘﺔ اﻟﺴﺎﺋﻞ
. AB ﻓﻲ ﻃﺒﻘﺔ اﻟﺴﺎﺋﻞ اﻟﺬي ﻳﺘﻜﻮن ﻓﻴﻪB ﻣﻊ ﻣﻌﺪل اﻧﺘﺸﺎر
Based on this model, Van Krevelen & Hoftijzer derived the chemical acceleration
factor or (Enhancement factor E):
H a [1 − (E − 1)Z]0.5
E= .....1
tanh[H a {1 − (E − 1)Z}0.5 ]
Where:
C m C n −1D
2
( K
Ha: Hatta number = n + 1 AB B Ai A
.....2
k
L
m
D C
n A Ai
Z= ...3 (Z is a parameter)
D C
B B
Where:
m
: Stochiometric constant ratio, (m) moles of B react with (n) moles of A.
n
KAB: Chemical reaction rate constant (m3/s * kmole)
CB: Concentration of B in the liquid bulk (kmole/m3)
CAi: Concentration of A at the interface (kmole/m3)
DA: Diffusivity of A in liquid (m2/s)
DA: Diffusivity of B in liquid (m2/s)
kL: Liquid phase mass transfer coefficient by physical absorption (m/s)
This relationship is shown in Log-Log plotting presented in vol. 2 of Chemical
engineering by J. M.Coulson &J. F. Richardson (fig. 12.12) page 549.
116
Let us consider the reaction model equation of the form:
dC
A = K Cn Cm
dt AB A B
y y
G` * ( ) + L * x B = L * x B + G` * ( 1 ) ...1
1-y 0 1- y
1
y L y L
( )= *x +( 1 - *x ) ...2
1-y G ` B 1 - y G ` B 0
1
(Equation of operating line assuming L is constant because when comparing the moles
of the reactant component to the total moles we will found that it is small ( the moles
of B is large), beside that the liquid phase will gain the moles of the product which
will replace the losses in A.
The height of packing is calculated by the equation:
y
G` n y dy
im
Z= ∫ ...3
k as y (1 − y ) 2 ( y − y )
y 1 i
117
Note:
The equilibrium curve is y = f (xA)
While the operating curve is y = g (xB)
Example 6:
A tower packed with 25.4 mm (1") ceramic Rasching rings is to be designed to absorb
SO2 from air using pure water at 293 K and 101.3 kPa absolute pressure. The entering
gas contains 20 mol% SO2 and that leaving 2 mol%. The inert air flow at 6.53x10-4
kmole/ air/s and the inert water flow is 0.042 kmole water/s. The tower cross-sectional
area is 0.0929 m2. For dilute SO2, the film mass transfer coefficients at 293 K are
given by the following equations:
k `y a = 0.0594 * G 0.7 0.25 3
y * L x kmole/s m mole fraction
k `x a = 0.152 * L0.82 3
x kmole/s m mole fraction
Where:
Gy is: kg total gas flow rate /m2 s
Lx is: kg total liquid flow rate /m2 s
Calculate the tower height, given the equilibrium data for SO2/air-water system at 293
K and 101.3 kPa as follow:
xA mole fraction Y*A mole fraction
SO2 in liquid SO2 in gas
0.0000562 0.000658
0.0001403 0.000158
0.0002800 0.004210
0.0004220 0.007630
0.0005640 0.011200
0.0008420 0.018550
0.0019650 0.051300
0.004200 0.12100
0.006980 0.21200
Solution;
y
G` B dy
Z= ∫
s y k` a
y
T
(1 − y ) 2 ( y − y )
y i
im
Step one:
From the given data, plot the equilibrium data.
G`=6.53x10-4 kmole/s
L` = 0.0042 kmole/s
yB = 0.2
118
yT = 0.02
xT = 0
Make material balance to calculate xB
y ` x x y
B = L *( B − T ) + T
1 − y B G ` 1 − x B 1 − xT 1 - yT
0.2 0.042 x 0 0.02
= *( B − )+
1 − 0.2 0.000653 1 − x 1 − 0 1 - 0.02
B
Then: xB = 0.00355
Step two:
Now the operating equation can be written as:
y
n + 1 = 0.042 * ( x n − 0 ) + 0.02
1− y 0.000653 1 − x n 1 − 0 1 - 0.02
n +1
Now plot the operating line by assuming yn+1 to find xn starting from yT to yB, the
following data are obtained:
Xn Yn+1
0 0.02
0.000332 0.04
0.000855 0.07
0.00207 0.13
0.002631 0.16
0.00356 0.2
Step three:
By using the equations of mass flow calculate Gy, Lx for each y, x as shown
y
G y = G ` * 29 + G ` * * 64
1− y
For example, use y = 0.2, whereas x = 0.00356
Therefore:
0.2 2
G y = [(0.000653 * 29) + 0.000653 * * 64]/0.0929 = 0.3164 kg / m sec
1 - 0.2
x
L x = L` *18 + L` * * 64
1− x
0.00356
L x = (0.042 *18) + 0.042 * * 64]/0.0929 = 8.241 kg/m 2 sec
1 - 0.00356
Step four:
Now we have to calculate k`ya and k`x a for each value of Gy, Lx estimated in above
using the following equations:
k `y a = 0.0594 * G 0.7 0.25 3
y * L x kmole/s m mole fraction
k `y a = 0.0594 * (0.3164) 0.7 * (8.241) 0.25 = 0.04496 kmole/s m 3 mole fraction
119
k `x a = 0.152 * L0.82 3
x kmole/s m mole fraction
k `x a = 0.152 * (8.241) 0.82 = 0.857 kmole/s m 3 mole fraction
Next we have to calculate the interfacial composition yi, xi for each y, x estimated
above using the technique explained in the first chapter using trial and error.
The first point to be considered is yT = 0.02, xT = 0:
The first trial
Assume xim = 1- xT = 1- 0.00355 = 0.99645
And yim = 1- yB = 1- 0.2 = 0.8
k `x a
x 0.857
Slop = - im = - 0.99645 = - 15.3
k `y a 0.04496
0.8
y
im
Plotting this on y – x diagram we get yi = 0.1688, and xi = 0.00566
Now, calculate the new yim, xim which will equal to 0.816, 0.996 respectively, calculate
the value of the new slop:
k `x a
x 0.857
Slop = - im = - 0.995 = - 15.6
k `y a 0.04496
0.816
y
im
As you can see the new value of the slop is different from the first one, so, repeat the
trials until you find a better results of yi, xi.
Repeat steps (3 to 5) for each value of (y, x) found from the operating curve equation.
Step six:
Calculate the height of the tower:
y
G` B dy
Z= ∫
s y k` a
y
T
(1 − y ) 2 ( y − y )
y i
im
All the results of calculations are represented in the following table, and by using
graphical integration between y= 0.2, &y= 0.02 to get height of the tower required,
and that was equal to 1.588.
120
X G` y im
y Gy Lx kxa kya yi xi (1-y)2 yim y-yi *
k y as (1 − y ) 2 ( y − y i )
0.02 0 0.213 8.138 0.848 0.03398 0.009 0.00046 0.96 0.985 0.011 19.295
0.04 0.000335 0.2226 8.147 0.849 0.03504 0.0235 0.00103 0.9216 0.968 0.0165 12.77
0.07 0.000855 0.2378 8.162 0.85 0.03674 0.0476 0.00185 0.8649 0.941 0.0224 9.292
0.13 0.00201 0.2712 8.196 0.853 0.04032 0.1015 0.00355 0.7569 0.885 0.0285 7.131
0.16 0.002631 0.2895 8.214 0.855 0.04224 0.13 0.0045 0.7056 0.855 0.03 6.718
0.2 0.00355 0.3164 8.241 0.857 0.04496 0.1685 0.00565 0.64 0.816 0.0315 6.327
121
Example 7:
A tower packed with 25.4 mm (1") ceramic Rasching rings is to be designed to absorb
SO2 from air using pure water at 293 K and 101.3 kPa absolute pressure. The entering
gas contains 20 mol% SO2 and that leaving 2 mol%. The inert air flow at 6.53x10-4
kmole/ air/s and the inert water flow is 0.042 kmole water/s. The tower cross-sectional
area is 0.0929 m2 for dilute SO2 the overall mass transfer coefficient:
k`OG a =16 kmole/hr m3 packing 105 pa
Calculate the tower height on the basis of the overall driving force that is [(NTU)oy,
(HTU)oy], given the equilibrium data for SO2/air-water system at 293 K and 101.3 kPa
as follow:
xA mole fraction Y*A mole fraction
SO2 in liquid SO2 in gas
0.0000562 0.000658
0.0001403 0.000158
0.0002800 0.004210
0.0004220 0.007630
0.0005640 0.011200
0.0008420 0.018550
0.0019650 0.051300
0.0042000 0.121000
0.0069800 0.212000
Solution
G` B
y
y *m dy
Z= ∫
k oy s y T (1 − y ) 2 ( y − y * )
`
Step one: From the given data, plot the equilibrium data.
G`=6.53x10-4 kmole/s = 2.3508 kmole/hr
L` = 0.0042 kmole/s
yB = 0.2
yT = 0.02
xT = 0
Make material balance to calculate xB
yB ` x x y
L
= *( B − T )+ T
1 − y B G ` 1 − x B 1 − xT 1 - yT
0.2 0.042 x 0 0.02
= *( B − )+
1 − 0.2 0.000653 1 − x B 1 − 0 1 - 0.02
Then: xB = 0.00355
Step two:
122
Assume xn, and find yn* for each one assumed using the equilibrium curve (you use
the equilibrium data given instead of assuming).
Step three: for each xn assumed calculate y using the operating equation curve:
y 0.042 x 0 0.02
= *( n − )+
1 − y 0.000653 1 − x n 1 − 0 1 - 0.02
Step four:
Now, calculate: (1 − y) 2 , ( y − y * ), y*m and then establish the following table:
G `y*
f ( y) = m
x Y* y (1-y)2 Y*m y-y*
k s(1 − y) ( y − y * )
` 2
oy
0.0000562 0.000658 0.023459 0.953631 0.987897 0.952973 1.71921642
0.0001403 0.000158 0.028592 0.943634 0.985557 0.943476 1.75076248
0.0002800 0.004210 0.037001 0.927368 0.979303 0.923158 1.80912837
0.0004220 0.007630 0.045403 0.911256 0.973362 0.903626 1.86949741
0.0005640 0.011200 0.053661 0.895557 0.967414 0.884357 1.93184301
0.0008420 0.018550 0.06943 0.865961 0.955784 0.847411 2.05990726
0.0019650 0.051300 0.128193 0.760047 0.909712 0.708747 2.67086378
0.0042000 0.121000 0.225818 0.599358 0.825482 0.478358 4.55352709
0.0069800 0.212000 0.320886 0.461196 0.732208 0.249196 10.0760174
Step four:
Plot f(y) vs. y and then calculate the area under the curve.
Z = Area under the curve = m
123
SUMMARY
1. A liquid can be used to selectively absorb one or more components from a gas
mixture. A gas can be used to selectively desorb or strip one or more components from
a liquid mixture.
2. The fraction of a component that can be absorbed or stripped in a countercurrent
depends on the number of equilibrium stages and the absorption factor.
3. Absorption and stripping are most commonly conducted in trayed towers equipped
with sieve or valve trays, or in towers packed with random or structured packing.
4. Absorbers are most effectively operated at high pressure and low temperature. The
reverse is true for stripping. However, high costs of gas compression, refrigeration.
and vacuum often preclude operation at the most thermodynamically favorable
conditions.
5. For a given gas flow rate and composition, a desired degree of absorption of one or
more components, a choice of absorbent, and an operating temperature and pressure,
there is a minimum absorbent flow rate, that corresponds to the use of an infinite
number of equilibrium stages. For the use of a finite and reasonable number of stages,
an absorbent rate of 1.5 times the minimum is typical. A similar criterion, holds for a
stripper.
6. The number of equilibrium stages required for a selected absorbent or stripping
agent flow rate for the absorption or stripping of a dilute solution can be determined
from the equilibrium line, and an operating line, using graphical. algebraic, or
numerical methods. Graphical methods, offer considerable visual insight into stage-by-
stage changes in compositions of the gas and liquid streams.
7. Packed column height can be estimated using the HTU/NTU concepts, with the
latter having a more fundamental theoretical basis in the two-film theory of mass
transfer.
8. One significant advantage of a packed column is its relatively low pressure drop per
unit of packed height, as compared to a trayed tow
124
Chapter Four
Distillation
In distillation (fractionation), a feed mixture of two or more component is separated
into two or more products, including, and often limited to an overhead distillate and a
bottoms, whose compositions differ from that of the feed. Most often, the feed is a
liquid or vapor – liquid mixture. The bottoms products is almost always a liquid, but
the distillate may be a liquid or a vapor or both. The separation requires that (1) a
second phase be formed so that both liquid and vapor phases are present and can
contact each other on each stage within a separation column, (2) the components have
different volatilities so that they will partition between the two phases to different
extents, and (3) the two phases can be separated by gravity or other mechanical means.
Distillation differs from absorption and stripping in that the second phase is created by
thermal means (vaporization and condensation) rather than by the introduction of a
second phase that usually contains an additional component or component not present
in the feed mixture. The word distillation is derived from Latin word destillare, which
means dripping or trickling down. By at least the sixteenth century, it was known that
the extent of separation could be improved by providing multiple vapor-liquid contacts
(stages) in a so-called Rectificatorium. The term rectification is derived from the Latin
words recte facere, meaning to improve. Modern distillation derives its ability to
produce almost pure products from the use of multistage contacting.
A- Distillation of binary mixtures
Binary Mixture: mixture contains only two components differ in boiling point.
Equipment and Design Consideration
Industrial distillation operations are most commonly conducted in trayed towers, but
packed columns are finding increasing use. Occasionally, distillation columns contain
both trays and packing. Types of trays and packing are identical to those used for
absorption and stripping.
A schematic diagram for the distillation column for binary mixture is presented below.
125
Factors that influence the design or analysis of a binary distillation operation include:
1. Feed flow rate, composition, temperature, pressure, and phase condition.
2. Desired degree of separation between two components.
3. Operating pressure (which must be below the critical pressure of the mixture).
4. Vapor pressure drop, particularly for vacuum operation.
5. Minimum reflux ratio and actual reflux ratio.
6. Minimum number of equilibrium stages and actual number of equilibrium stages (stage
efficiency).
7. Type of condenser (total, partial, or mixed).
8. Degree of sub-cooling, if any, of the liquid reflux.
9. Type of re-boiler (partial or total).
10. Type of contacting (trays or packing or both).
11. Height of the column.
12. Feed entry stage.
13. Diameter of the column.
14. Column internals.
Equilibrium Relationships
Experimental data have been published for several thousand binary and many multi-
component systems. The book by Chu et al (1956) and Hala et al (1968, 1973) cover
most of the published experimental data.
Otherwise the equilibrium relationship between yA and xA can be obtained by two
methods either in terms of the relative volatility (αAB), or using the vapor pressure.
A- The volatility (α) of any component can be defined as the ratio of the partial
pressure to the mole fraction in the liquid phase of that component, so
P P
α = A and αB = B
A x xB
A
Where:
α Po
α = A = A
AB α
B Po
B
126
y y
A A
K ( x ) ( x ) Po
α = A = A = A = A
AB K y 1 - y Po
B ( Bx ) ( A ) B
B 1- x
A
2- Another method to estimate the relative volatility is by making use of Dalton's law,
Raoult's law, as shown below:
PT = ∑ P and then P = y * PT (Dalton's law)
i A A
0
P = P *x (Raoult's law)
A A A
P0 * x P0 * x
y = A A and y = B B
A P B PT
T
y y
Po ( Ax ) ( Ax )
α = A = A = A
AB P o y 1 - y
B ( Bx ) ( A
B 1- x )
A
Now to make use of the relative volatility (αAB) value estimated by both methods:
P
A
α x P0
Θα = A = A = A
AB α P 0
PB
B B
x
B
Using Dalton's law to substitute Pi ( yi* PT), then
y *x y (1 − x )
α = A B= A A
AB y * x x (1 − y )
B A A A
Rewrite the above equation either in terms of yA
α *x
y = AB A ...1
A 1 + (α − 1)x
AB A
y
x = A ...2
A α − y (α - 1)
AB A AB
Equations (1) & (2) are used also to calculate the equilibrium composition of a binary
system from the vapor pressure or volatility.
B- Using the vapor pressure, where by making use of Dalton's law and Raoult's law as
follow:
P0 * x 0 *x
PB
Θy = A A and y = B
A PT B PT
And Θ y A + y B = 1
Then
127
P0 * x P0 * x
A A + B B =1
PT PT
P − P0
∴x = T B ...3
A
P0 − P0
A B
So, from equation (3) we can calculate the composition of A in liquid phase if the
vapor pressure of components A, B are known. Also the composition of A & B in
vapor phase could be calculated from the following equation:
P0 * x
y = A A ...4
A PT
Note: the vapor pressure can be estimated using Antioe equation:
B
ln P o = A − where (A,B and C are constants which can be found from references).
C +T
Let us first define some important terms which will be considered in this section:
More volatile component (M. V. C.): components of lower boiling point and it
concentrate in vapor phase.
Less volatile component (L. V. C.): components of higher boiling point and it
concentrate in liquid phase.
Boiling point curve:
For binary mixture the equilibrium data can be presented in T-x-y diagram as shown in
below:
128
1- If we start we start with a mixture (A+B), where A is the more volatile
component (M. V. C.) & component B is the L. V. C. and the composition of the
mixture is xA.
2- At a point R the mixture is all liquid at temperature T1, xA mole fraction at fixed
pressure.
3- Start heating the mixture, and when T2 is reached, the mixture will start boiling
at point N; where the composition of the first bubble yA is in equilibrium with the
liquid composition xA. So we will call T2 the bubble point of the mixture.
4- As we continue heating, the composition of the liquid xA will move to the left of
the figure, since A is redistributed between the two phase (the vapor will be rich
with A).
5- At temperature T3, all the mixture is vaporized and reaches the saturation
condition (saturated vapor), therefore we will call T3 the dew point of the
mixture.
6- At temperature T4, the mixture will be superheated vapor.
Example 1:
Vapor pressures of Benzene – Toluene mixture are given in the table below. Assuming
that this mixture follow Rault's law, calculate and plot the boiling point diagram and
the equilibrium composition curve. Note the total pressure is 1 atm.
P0 * x
y = A A
A PT
At T = 85 oC
760 - 345
∴x = = 0.78
A 877 - 345
877 * 0.78
y = = 0.9
A 760
Repeat for all the given temperatures establish the following table:
T 0C 80.1 85 90 100 110 110.6
xA 1 0.78 0.581 0.258 0.017 0
yA 1 0.9 0.777 0.456 0.039 0
95
90
85
80
75
0 0.2 0.4 0.6 0.8 1
x,y
130
y
k = i
i x
i
If ∑ yi = 1 , then the assumption is correct.
If ∑ yi 〉 1 that means that the assumed temperature is larger than the bubble point, and
then assumes smaller temperature for second trail.
If ∑ yi 〈1 that means that the assumed temperature is lower than the bubble point then
assumes larger temperature for second trail.
For dew point temperature calculation at fixed pressure, assume a temperature and
then from literatures find ki for each component at that temperature and pressure.
Calculate xi using the following equation:
y
k = i
i x
i
If ∑ xi = 1 , then the assumption is correct.
If ∑ xi 〉 1 that means that the assumed temperature is larger than the dew point, and
then assumes smaller temperature for second trail.
If ∑ xi 〈1 that means that the assumed temperature is lower than the bubble point, and
then assumes larger temperature for second trail.
131
For the binary system we consider the following types of distillation:
1- Batch distillation.
2- Flash distillation.
3- Multi-stage distillation.
1-Batch Distillation
In batch separation operations, a feed mixture is charged to the equipment and one or
more products are withdrawn. A familiar example is laboratory distillation, shown in
Figure 1, where a liquid mixture is charged to a still pot, retort, or flask and heated to
boiling. The vapor formed is continuously removed and condensed to produce a
distillate.
The composition of both the initial charge and distillate change with time; there is no
steady state. The still temperature increases and the relative amount of lower-boiling
components in the charge decreases as distillation proceeds.
Batch operations can be used to advantage under the following circumstances:
1. The capacity of a facility is too small to permit continuous operation at a practical
rate.
2. It is necessary, because of seasonal demands, to distill with one unit different feed
stocks to produce different products.
3. It is desired to produce several new products with one distillation unit for evaluation
by potential buyers.
4. Upstream process operations are batch wise and the composition of feed stocks for
distillation vary with time or from batch to batch.
5. The feed contains solids or materials that form solids, tars, or resin that plug or foul
a continuous distillation column.
A- DIFFERENTIAL DISTILLATION
The simplest case of batch distillation, as discussed by Lord Rayleigh, is differential
distillation, which involves use of the apparatus shown in Figure 1. There is no reflux;
at any instant, vapor leaving the still pot with composition yD is assumed to be in
equilibrium with perfectly mixed liquid in the still. For total condensation, yD=xD.
Thus, there is only a single equilibrium stage, the still pot. This apparatus is useful for
separating wide boiling mixtures. The following nomenclature is used for variables
that vary with time, t, assuming that all compositions refer to a particular species in the
muIti-component mixture.
D = instantaneous distillate rate, mol/h
y = YD = XD = instantaneous distillate composition, mole fraction
W = moles of liquid left in still
x = Xw = composition of liquid left in still, mole fraction
o = subscript referring to t = 0
132
Figure 1 Differential distillation.
For any component in the mixture:
Instantaneous rate of output = DyD
Instantaneous rate of depletion in the still=
d dx dw
− ( w * xw ) = − w w − xw
dt dt dt
The distillate rate and, therefore, the rate of depletion of the liquid in the still depend
on the rate of heat input to the still. By material balance at any instant:
d dx dw
( w * xw ) = w w + xw = -Dy D ...5
dt dt dt
Multiplying by dt:
wdxw + xwdw = yD(-Ddt) = yDdw …6
Since by total balance (- Ddt = dW) Separating variables and integrating from the
initial charge condition:
xW
dx W W
dw ⎛W ⎞
∫ = ∫ = ln ⎜⎜ ⎟⎟ ...7
y
xW0 D − xW W0 w ⎝ 0⎠
W
This is the well-known Rayleigh equation, which was first applied to the separation of
wide-boiling mixtures such as HCl-H20, H2SO4-H20, and NH3-H20. Without reflux, yD
and xw are in equilibrium and (equation 7) simplifies to:
x
dx ⎛W ⎞
∫ = ln ⎜⎜ ⎟⎟ ...8
x0 y − x ⎝ 0⎠
W
Equation (8) is easily integrated only when pressure is constant, temperature change in
the still pot is relatively small (close-boiling mixture), and K-values are composition
independent. Then y = K*x, where K is approximately constant, and (equation 8)
becomes
⎛W ⎞ 1 ⎛ x⎞
ln⎜⎜ ⎟⎟ = ln⎜⎜ ⎟⎟ ...9
⎝ 0⎠
W K − 1 ⎝ 0⎠
x
For a binary mixture, if the relative volatility a is assumed constant, substitution of
(equation 1) into (equation 8), followed by integration and simplification, gives
133
⎛W ⎞ 1 ⎡ ⎛ xo ⎞ ⎛ 1 − x ⎞⎤
ln⎜ 0 ⎟ = ⎢ ⎜ ⎟
ln + α ln ⎜⎜ ⎟⎟⎥ ...10
⎝ ⎠
W α AB − 1 ⎣ ⎝ ⎠x
AB
⎝ 1 − x0 ⎠⎦
If the equilibrium relationship y = f{x} is in graphical or tabular form, integration of
(equation 8) can be performed graphically or numerically. The final liquid remaining
in the still pot is often referred to as the residue.
The average composition of the distillate can be calculated from the following
equation:
(x D )avg = ( yD )avg = W0 *Wx0 −−WW * x ...11
0
The Rayleigh equation (equation 7) can be applied to any two components, A and B of
a multi – component mixture. Thus for a binary mixture if we let
M A = W * xW A ...12
Where: MA is the moles of A in the still at any time.
Then
dM A y D A
= ...13
dM B y D B
For constant αAB
y *x
D w
α AB = A B ...14
y *x
D W
B A
Then equation 13 becomes
dM A ⎛ xW ⎞
= α AB ⎜ A ⎟ ...15
dM B ⎜ xW ⎟
⎝ B⎠
Substituting equation 12 for both A and B into equation 15 gives
dM A ⎛ dM B ⎞
= α AB ⎜⎜ ⎟⎟ ...16
dM B ⎝ dM A⎠
Integration from the initial – charge condition gives
⎛M ⎞ ⎛M ⎞
ln⎜ A ⎟ = α AB * ln⎜ B ⎟ ...17
⎜MA ⎟ ⎜ MB ⎟
⎝ o ⎠ ⎝ o ⎠
This equation is useful for determining the effect of relative volatility on the degree of
separation that can be achieved by Rayleigh distillation.
134
Example 2
A batch still is loaded with 100 kmol of a liquid containing a binary mixture of 50
mol% benzene in toluene. As a function of time, make plots of
(a) Still temperature,
(b) Instantaneous vapor composition,
(c) Still-pot composition,
(d) Average total distillate composition.
Assume a constant boil-up rate of 10 kmol/h and a constant relative volatility of 2.41
at a pressure of 101.3 kPa (1 atm).
Equilibrium data is given as follow at 101.3 kPa:
Temp., oC 105.3 101.5 98.0 95.1 92.3 89.7 87.3 85.0 82.7 81.4
xBen 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.95
yBen 0.208 0.372 0.507 0.612 0.713 0.791 0.857 0.912 0.959 0.98
Solution:
Initially, Wo = 100 kmol, xo= 0.5.
Solving (equation 10) for W at values of x from 0.5 in increments of 0.05, and
determining corresponding values of time from t = (Wo - W)/10, the following table is
generated:
T, hr 2.12 3.75 5.04 6.08 6.94 7.66 8.28 8.83 9.35
W, kmole 78.85 62.51 49.59 39.16 30.59 23.38 17.19 11.69 6.52
x= xW 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05
135
Example 3:
Repeat Example 2, except instead of using a constant value of 2.41 for the relative
volatility use the vapor-liquid equilibrium data for benzene-toluene at 101.3 kPa, given
in the data, to solve the problem graphically or numerically with (equation 8) rather
than (equation 11).
1
Equation (8) can be solved by plotting versus x with a lower limit of x0= 0.5.
y−x
Using the equilibrium data given for y as a function of x, points for the plot in terms of
benzene are as follows:
The area under the plotted curve from xo= 0.5 to a given value of x is equated to
⎛W ⎞
ln⎜⎜ ⎟⎟ and W is computed for Wo = 100 kmol. In the region from x = 0.5 to 0.3, the
⎝ W0 ⎠
value of l/(y – x) changes only slightly. Therefore, a numerical integration by the
trapezoidal rule is readily made:
For x = 0.4:
⎛ W ⎞ 0.4 dx ⎡ 1 ⎤ ⎡ 4.695 + 4.717 ⎤
ln⎜⎜ ⎟⎟ = ∫ ≈ ∆x ⎢ ⎥ = (0.4 − 0.5) ⎢ ⎥ = - 0.4706
⎝ 0 ⎠ 0.5
W y − x ⎣ y − x ⎦ avg ⎣ 2 ⎦
W/Wo = 0.625,W = 0.625(100) = 62.5kmol
136
For x=0.3
⎛ W ⎞ 0.3 dx ⎡ 1 ⎤ ⎡ 4.695 + 4.717 + 4.717 + 4.831⎤
ln⎜⎜ ⎟⎟ = ∫ ≈ ∆x ⎢ ⎥ = ( 0. 3 − 0. 5) ⎢ ⎥ = - 0.948
⎝ W0 ⎠ 0.5 y − x ⎣ y − x ⎦ avg ⎣ 4 ⎦
These results show that a sharp separation between A and B for 50% vaporization of A
is only achieved if αAB= 100.
Furthermore, the purity achieved depends on the percent vaporization of A.
For αAB = 100, if 90% of A is vaporized, the mole fraction of B in the total distillate
increases from 0.0136 to 0.0247. For this reason, it is common to conduct a binary
137
batch distillation separation of M.V.C. (only for binary light key, LK) and L.V.C.
(only for binary heavy key, HK) in the following manner:
1. Produce a distillate LK cut until the limit of impurity of HK in the total distillate is
reached.
2. Continue the batch distillation to produce an intermediate cut of impure LK until the
limit of impurity of LK in the liquid in the still is reached.
3. Empty the HK-rich cut from the still.
4. Recycle the intermediate cut to the next still charge.
For desired purities of the LK cut and the HK cut, the fraction of intermediate cut
increases as the LK-HK relative volatility decreases.
When the column is operated at a constant reflux ratio R, the concentration of the
M.V.C. in the top product will continuously fall. Over a small interval of time (dt), the
top product composition will fall from xD to (xD – dxD). If in this time the amount of
product is dD, then the material balance for the M.V.C. will be as follow:
138
⎡ dx D ⎤
= dD ⎢ x D -
2 ⎥⎦
D
A ⎣
DA = xD * dD
XD * dD = - d(WxW)
But
dD = - dW
Then
- xd * dW = - (W * dxW) – (xW * dW)
WdxW = (xd – xW) *dW
xW
W1
dW 1 dx
= w
∫ ∫
W2 W xW x −x
2 d W
W x
W1 1 dx
ln = ∫ w
W2 x W x − x
2 d W
Or
W x
W 1 dx w
ln 1 = ∫ (yD=xD)
W y −x
2 x W2 d W
1
The right – hand side of this equation can be integrated by plotting vs. xw. This
yd − xw
will enable the ratio of the initial to final quantity in the still to be found for any
desired change in xw, and hence the amount of distillate D.
The heat to be supplied to provide the reflux will now be:
QR = λ * R * D
139
⎡x − x ⎤ a
D = S * ⎢ s1 s2 ⎥ = * S
b 1 ⎢ x −x ⎥ b 1
⎣ d s2 ⎦
Vn+1 D
yn+1 xD
L
xD
R x
y n +1 = * xn + D
R +1 R +1
If φ is the intercept on the y-axis of any operating line in U.O.L. equation:
Then:
xD
=φ
R +1
Or
xD
R= −1
φ
The above equations enable the final reflux ratio to be found for any desired end
concentration in the still, and also give the total quantity of distillate obtained.
When comparing the operation at constant reflux ratio with that at constant product
composition, we find that there is a difference in the total amount of steam used in the
distillation, for a given quantity Db of product that is the most important point to be
considered.
If the reflux ratio R is assumed to be adjusted continuously to keep the top product at
constant quality, then at any moment the reflux ratio is given by
dL b
R=
dD p
Then
Lp R2
∫ dL p = ∫ RdD p ...1
0 R1
To provide reflux dLb requires the removal of heat λdLb in the condenser; (λ is latent
heat per mole). Thus, the heat to be supplied in the re-boiler QR to provide this reflux
during the total distillation is given by:
Lp R2
Q = λ * ∫ dL p = λ * ∫ RdD p ...2
R 0 R1
Equations 1 and 2 can be integrated graphically if the relation between R and Db is
found.
140
Example 5:
A mixture of ethyl alcohol and water with 0.55 mole fraction of alcohol is distilled in batch column
to give a top product of 0.75 mole fraction of alcohol. The column has four ideal plates. The
distillation is stopped when the reflux ratio has to be increased beyond 4.
Find the amount of distillate obtained and the heat required per kmole of product, given the
following data:
x 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
y 0 0.42 0.52 0.58 0.61 0.65 0.7 0.75 0.82 0.9 1.0
λaver. = 4000 J/kmol
Solution:
xD
1- Find for R = 4 1
R +1
0.75
φ= = 0.15
5 0.8
2- Plot operating line for the points
Point 1: y = x = xD = 0.75 0.6
Point 2: y = φ = 0.15, x = 0
y
R φ xS2 Dp
0.5 0.5 0.56 -5.26316
0.85 0.405405 0.55 0
Ignored
1 0.375 0.5 20
1.5 0.3 0.37 47.36842
2 0.25 0.2 63.63636
3
4
0.1875
0.15
0.075
0.05
70.37037
71.42857
Note: you can start from R = 4 and then start to reduce its
value until reaching x = xS2
141
Now plot Dp vs. R and calculate the area under the curve obtained, where:
R2
Area = ∫ RdD p = 96 Kmole
R1
R2
QR = λ * ∫ RdD p
R1
QR = 96 * 4000 = 384000 kJ
For producing 71.4 Kmole
QR = 384000 / 71.4 = 5.32 MJ
142
2- FLASH OR EQUILIBRIUM DISTILLATION
The simplest separation process is one in which two phases in contact are brought to physical
equilibrium, followed by phase separation. If the separation factor between two species in the two
phases is very large, a single contacting stage may be sufficient to achieve a desired separation
between them; if not, multiple stages are required. For example, if a vapor phase is in equilibrium
with a liquid phase, the separation factor is the relative volatility, α, of a volatile component called
the light key, LK, with respect to a less-volatile component called the heavy key, HK, where, αLK,
HK =KLK/KHK. If the separation factor is 10000, an almost perfect separation is achieved in a single
stage. If the separation factor is only 1.1, an almost perfect separation requires hundreds of stages.
For single stage, a liquid mixture is partially vaporized, then pressure is reduced. The vapor and
liquid in contact reach equilibrium, then separation into vapor phase from top and liquid phase
from the bottom as shown in below:
143
Case 1:
Known product composition and known feed composition
Draw the operating line between the following tow points:
Point 1 is at x = xf which lies on the 45o line, while the second point is the product composition (y,
x) which lies on the equilibrium curve.
1- f v
Calculate the slop of the operating line, which equals to (- ) and then calculate fv.
fv
Case 2:
Known product amount and known feed composition
1- f v
Draw the operating line from point (y = x = xf) with a slop of (- ) , then from the intersect point
fv
of the operating line with the equilibrium curve; calculate the composition of the product.
0.8
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
144
Example 6:
100 kmole/hr of a liquid mixture containing 60 mol% n-butane and 40 mol% n-hexane is subjected
to an equilibrium flash distillation giving products at 250oF. 80 % of the hexane is to recovered in
the liquid. Calculate the vapor flow rate and the composition of the liquid and vapor, and the
required pressure.
Solution:
V ,y
F=100 kmole/hr
xfb = 0.6
xfh= 0.4 S ,x
Basis: 1 hr
Hexane in feed = 0.4 * 100 = 40 kmole
Hexane in liquid = 0.8 * Hexane in feed
Hexane in liquid = 0.8 * 40 = 32 kmole
So, Hexane in vapor = 40 – 32 = 8 kmole
O.M. B.
F=V+S
Hexane M. B.
F * xfh = V * yh + S * xh
y
Θk = i
i x
i
Then
y
Fh = Vh + S * h * (V)
k
h V
yh *V
∴ Fh = Vh + S * ( )
k h *V
S
∴ Fh = Vh + Vh * ( )
k h *V
Fh
Vh =
S
1+
k h *V
S F
Or = h −1 …1
k h * V Vh
Benzene M. B.
Repeat the same steps as above to derive the following equation:
S F
= b −1 …2
k b * V Vb
Dived equation 2 by equation one, then
F
( b ) −1
k V
h = b ...3
k F
b ( h ) −1
V
h
Where:
145
Fi = molar flow rate of component (i) in feed
Vi = Fi = molar flow rate of component (i) in vapor
In equation 3, Fh = 40, Vh = 8, and Fb = 60
But because the pressure of the process is unknown, then we have three variables, namely (kh, kb,
and Vb) in that equation, which can not be solved directly.
Then, this equation is solved by trial and error, by assuming a pressure, to find the k-values, which
must be done as in the bubble point temperature calculation (Σ yi = Σ (ki * xi) = 1), then evaluating
the correct values of ki, Vb can be calculated.
Assume PT = 10 atm, then from k-value charts for PT = 10 atm, and T = 250oF:
kh = 0.48 and kb = 1.87
Solving equation (3) for Vb, then:
Vb = 29.603 kmole/hr
V = Vh + Vb = 8 + 29.603 = 37.603 kmole/hr
Remember
Σ yi = Σ (ki * xi) = 1
Then we must test find xh, and xb
Hexane in S Benzene in S
x = and x =
h Total S b Total S
S = F – V = 100 – 37.603 = 62.397 kmole/hr
32
x = = 0.5128
h 62.397
x = 1 - 0.5128 = 0.4872
b
Σ yi = (0.48 * 0.5128) + (1.87 * 0.4872) = 1.1572
Which means that the assumption was not correct; we have to assume another pressure higher than
the first one and let it be 11.6 atm.
At PT = 11.6 atm and T = 250 oF
kh = 0.415 and kb = 1.615
Repeat the same calculation as in the first step, you will find that:
Vb = 29.6 kmole/hr, V = 37.6 kmole/hr, S = 62.4 kmole /hr, xh = 0.5128205, and xb = 0.4871795
Then:
Σ yi = (0.415 * 0.5128205) + (1.615 * 0.487195) = 1
So, the assumption was correct.
The final results are:
S = 62.4 kmole/hr
V = 37.6 kmole/hr
xb = 0.4872 yb = 1.615 * 0.487195 = 0.787
xh = 0.5128 yb = 0.415 * 0.512805 = 0.213
146
C- RECTIFICATION WITH REFLUX
It is also called fractionation (distillation) or multiple stages with reflux.
On each stage there are liquid and vapor flow counter-currently, mixed and equilibrated on each
stage. Therefore, the liquid and vapor leave the stage are in equilibrium.
The final vapor comes from overhead contains high concentration of the M. V. C., is condensed
and portion of it removed as top product, while the remaining liquid is returned to the top tray of
the column as reflux liquid.
The liquid leaving the bottom, which is less in the M. V. C. or in other words rich of the L. V. C.,
is withdrawn as bottom product, portion of this product is evaporated in the re-boiler and sent back
to the bottom stage (or tray).
The methods of calculation of the number of stages are:
1- McCabe – Thiele Method
In 1925, McCabe and Thiele published an approximate graphical method for combining the
equilibrium with the operating – line curves to estimate, for a given binary feed mixture and
column operating pressure, the number of equilibrium stages and the amount of reflux required for
a desired degree of separation of the feed.
The McCabe – Thiele method determine not only N, the number of equilibrium stages, but also
Nmin, Rmin, and the optimal stage for feed entry. Following the application of the McCabe – Thiele
method, energy balances are applied to estimate condenser and re-boiler heat duties.
Besides the equilibrium curve, the McCabe – Thiele method involve a 45o reference line, separate
operating lines for the upper rectifying (enriching) section of the column and the lower stripping
(exhausting) section of the column, and a fifth line (the q-line or feed line) for the phase or thermal
condition of the feed. The most important assumptions made to apply this method are:
1- The two components have equal and constant molar enthalpies of vaporization (latent heat).
2- Component sensible enthalpy changes (Cp ∆T) and heat of mixing are negligible compared to
latent heat changes.
3- The column is well insulated so that heat loss is negligible.
4- The pressure is uniform throughout the column (no pressure drop).
These assumptions are referred to as the McCabe – Thiele assumptions leading to the condition of
constant molar flow rate in both sections of the column (rectifying & stripping) or we can say
constant molar flow rate in the tower.
147
A- Equations for rectification (enriching) section:
O. M. B. for the entire column:
F=D+W
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
148
B- Equations for stripping section:
O. M. B. around the selected section (stripping section:
Vm+1 = Lm - W
[A] M. B. for the selected section:
Vm+1 * ym+1= Lm * xm - W * xw
0.8
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
150
1
0.8
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
Location of feed tray and the number of theoretical trays (or stages):
Using the equilibrium data, and the equations (1, 2, and 3), the number of theoretical plates and the
feed tray can be determined as follow:
1- Complete the material balance so the values of all streams and their composition are known.
2- Plot the equilibrium data.
3- Locate xD, xF, xw on the 45o line.
x
4- Plot the U. O. L. between xD on the 45o line & the point (y = D , x =0).
R +1
o q
5- Plot the q-line from the point of xf on the 45 line with slop of .
q -1
6- From the point of intersection of q-line with the U. O. L. draw a line between this point and
the point y=x=xw on the 45o line (L. O. L.).
7- Either from xD, or from xw step off the number of theoretical stages.
8- Locate the feed plate from the figure.
NOTE:
Theoretical number of stages = the number of stages from the curve – 1
0.8
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
151
Total and minimum reflux for McCabe – Thiele Method
A reflux ratio must be established before any quantitative fractionation design calculation can be
made. It was pointed out that a distillation column can only produce the desired products between
the limits of minimum reflux and total reflux.
1
0.9
0.8
0.7
0.6
0.5
y
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
x
This gives the minimum number of plates that can possibly used to obtain the desired separation.
This requires infinite size of condenser, re-boiler, and tower diameter for a given feed rate.
B- Minimum reflux:
From figures before, if the efflux ratio reduced, the slop of the U. O. L. is decreased and the
intersection of the u. O. L. with the q-line and with L. O. L. move far from the 45o line and closer
to the equilibrium curve so the number of stages (required to achieve the desired products mole
fraction xD, xw) increases.
152
When the two operating lines touch the equilibrium curve, a (Pinch point) at y`, and x` occurs.
Where the number of stages required becomes infinite, so the slop of the U. O. L. is as follow:
R x − y`
min = D
R +1 x − x`
min D
For vertical q-line (y`, and x`) are substituted by (yf, and xf) as shown in the figure and then the
above equation can be written as
−y x
D f
R =
min y −x
f f
And for horizontal q-line y`, and x`) are substituted by (yf, and xc) as shown in the figure and then
the above equation can be written as
x −x
D f
R = 1
min x −x
f c 0.9
0.8
Or can be determined graphically as 0.7
shown:
0.6
NOTE: 0.5
y
The number of plates and minimum reflux ratio can be found analytically using Underwood &
Fenske equations:
xA xB
log[( )D * ( )W ]
xB xA
N +1=
logα average
For small variation in α
α aaver. = (αD * αw)½
αD: Relative volatility of the overhead product
αw: Relative volatility of the bottom product
While the minimum reflux ratio equation is:
1 x D α(1 − x D )
R = { − }
min α −1 x 1− x
f f
153
So if the number of plates is plotted vs. reflux ratio, the resulting curve is shown as below:
154
Whereas for the condenser:
- V *λ
WC = n
C * (t − t )
P 2 1
WC: mass flow rate of cold water required in kg/hr.
Vn: vapor mole flow rate of overhead in kmole/hr.
λ: latent heat of vaporization of the condensate at the top of the column in kJ/kmole.
CP: heat capacity of water in (kJ/kg. oC).
(t2-t1): the temperature difference of outlet and inlet of cooling water (oC).
F VD
xf yD
m+1
yw
xw
W
155
1
0.9
0.8
0.7
0.6
0.5
y
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
x
2- Enriching tower:
This process is used for mixtures lean in M. V. C.
Feed is introduced from the bottom of the column as saturated vapor or superheated vapor, where
no re-boiler is used. The composition of bottom product is comparable to feed composition
(slightly less than xf).
O. M. B. on the top of the column
Vn+1 = Ln + D
[A] M. B. on the bottom of the column
yn+1 * Vn+1 = xn * Ln + xD * D
The operating line equation is:
L
y n +1 = n * x + D* xD
n
Vn +1 Vn +1
Vn = F if the feed is saturated vapor
And Vn = (1-q) * F if the feed is superheated vapor
O. M. B. on the column
F+S=D+W
[A] M. B.
F * xf = D * xD + W * xW
The enriching section is not affected, because no change done but for the stripping section, the
material balance will be:
O. M. B.
Lm + S = Vm+1 + W
[A] M. B.
Lm * xm = Vm+1 * ym+1 + W * xW
156
Lm W *xw
y m+1 = * xm −
Vm+1 Vm+1
If the steam was saturated steam, then S = Vm+1 and Lm+1 = W
That will cause to write the L. O. L. as
W W *xw
y m+1 = * xm −
S S
This is an equation of straight line with slop of (W/S), to draw this equation, use the following two
points:
The first one is at
Y = 0 then x = xW,
And the second one is
W*xw
At y = xm, then x m =
W −S
Using direct steam is the simplest heater construction and it is used when one component of the
feed mixture is water.
0.8
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
157
Vs+1 = Vn+1 = Ln + D
Vn+1 = Ln + D
[A] M. B.
ys+1 * Vs+1 = xs * Ls + xD * D + xZ * Z
Rewrite for y and substitute Vs+1 by Vn+1, so
L D * xD + Z* x
y S+1 = S *x + Z
V S Vn +1
n +1
L −Z D* xD + Z* x
y S+1 = n * xS + Z
L +D Ln + D
n
Ln
ΘR=
D
R − ZD x + ZD*x
∴ y = *x + D Z
S +1 R +1 S R +1
R − ZD L
The above equation is the operating line equation for the side stream of slop ( S ) and
R + 1 Vn + 1
+ ZD*x
Z ( D * x D + Z * x Z ). So to calculate the number of plates
x
intersect the y-axis at y = D
R +1 Vn +1
using the McCabe-Thiele method, follow the steps:
1- Complete the M. B., so the mole fraction of all streams must be known.
2- Plot the equilibrium data and locate xD, xZ, xf, xW on the 45oline.
x
3- Plot the U. O. L. using the points, (y = x = xD), (y = D , x =0).
R +1
4- Plot the side stream q-line which has a slop = ∞ (saturated liquid), from the point y =x =xZ.
5- Plot the side stream operating line between the points; the intersection of the U.O.L. & the side
x + ZD*xZ R − ZD
stream q-line and the (y = D , x =0), with slop .
R +1 R +1
6- Plot the q-line of the feed from the point y =x = xf as before (depending on the feed condition).
7- From the point of intersection of the side stream operating line and the q-line of the feed draw
the L.O.L. to intersect the 45o line at y =x = xw.
8- Step –off the stages as before.
Then the theoretical number of plates = N-1
As shown in below
158
1
0.8
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
159
Ln
ΘR=
D
F F *x
R + 1D x − 1 f1 D
∴y = xr + D
r+1 R +1 R +1
F
R + 1D
This is the operating line equation with slop of and intersects the y-axis at
R +1
F *x
x − 1 f1 D
D
R +1
The lower operating line is the same as before, that is:
Lm W
y m+1 = * xm − *xW
Vm+1 Vm+1
Lm
With slop and it intersects the 45o line (that is y = x) at x = xw. And at x = 0 it intersects the
V
m+1
W*xw
y-axis at y = −
Vm+1
So to calculate the number of plates using the McCabe-Thiele method, follow the steps:
1- Complete the M. B., so the mole fraction of all streams must be known.
2- Plot the equilibrium data and locate xD, xZ, xf, xW on the 45oline.
x
3- Plot the U. O. L. using the points, (y = x = xD), (y = D , x =0).
R +1
4- Plot the first feed q-line (depending on its entering conditions), from the point y =x =xF1.
5- Plot the second feed operating line between the points; the intersection of the U.O.L. & the first
F *x F
x − 1 f1 D R + 1
feed q-line and the (y = D , x =0), with slop D.
R +1 R +1
6- Plot the q-line of the second feed from the point y =x = xf2 as before (depending on the feed
condition).
7- From the point of intersection of the side stream operating line and the q-line of the second feed
draw the L.O.L. to intersect the 45o line at y =x = xw.
8- Step –off the stages as before.
Then the theoretical number of plates = N-1
As shown in below
160
1
0.8
0.6
y
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
x
161
Stage efficiency
Graphical and algebraic methods for determining stage requirements for distillation assume
equilibrium with respect to both heat and mass transfer at each stage. Thus, the number of
equilibrium stages (theoretical stages, ideal stages, or ideal plates) is determined or specified when
using those methods. The McCabe-Thiele method assumes that the two phases leaving each stage
are in thermodynamic equilibrium. In industrial countercurrent, multistage equipment, it is not
always practical to provide the combination of residence time and intimacy of contact required to
approach equilibrium closely. Hence, concentration changes for a given stage are usually less than
predicted by equilibrium.
The simplest approach for staged columns is to apply an overall stage (or column) efficiency,
defined by lewis as:
N
E = t
0 Na
Where E0 is the fractional overall stage efficiency, usually less than 1.0 (may vary from 0.3 to 1.0);
Nt is the calculated number of equilibrium (theoretical) stages; Na is the actual number of
contacting trays (stages, plates), usually greater than Nt required. Based on the results of extensive
researches conducted over a period more than 60 years, the overall stage efficiency has been found
to be a complex function of the:
1. Geometry and design of the contacting trays.
2. Flow rates and flow paths of vapor and liquid streams.
3. Compositions and properties of vapor and liquid streams.
There are several empirical correlations for estimating the overall stage efficiency.
Another method used to describe the performance of the distillation column is the stage efficiency,
which is frequently used to describe individual tray performance for individual components; one of
these methods is Murphree plate efficiency. This efficiency can be defined on the basis of either
phase and, or for a given component, is equal to the change in actual composition in the phase,
divided by the change predicted by equilibrium. This definition applied to the vapor phase can be
expressed as:
y −y
E MV = n n +1
*
yn − y
n +1
Where:
yn: average actual composition of vapor leaving tray n.
yn+1: average actual composition of vapor leaving tray n (leaving tray n+1).
y*n : Concentration of vapor in equilibrium with liquid composition x n leaving the tray to the
down comer.
This equation gives the efficiency in vapor terms, but if the concentrations in the streams are used
then the plate efficiency is given as:
x − xn
E ML = n + 1
x − x*n
n +1
Where:
xn: average actual composition of liquid leaving tray n.
xn+1: average actual composition of liquid leaving tray n+1 (entering tray n).
x *n : Composition of liquid in equilibrium with vapor composition yn leaving the tray to the down
comer.
162
The ratio E MV is shown graphically in below.
There are many empirical expressions for efficiency such as Drickamer & Bardford, andChu in
which physical properties of feed and relative flow rates of liquid and vapor inside the column, (see
vol. 2 pages 507-511 for more details).
Binary Mixture non-equi-molal flow conditions
In the previous methods the case of constant molar latent heat has been considered with no heat of
mixing, and hence constant molar rate of reflux in the column. These simplifying assumptions are
extremely useful in that they enable a simple geometrical method to be used for finding the change
in concentration on the plate and, whilst they are rarely entirely true in industrial conditions, they
often provide convenient start for design purposes.
For non-ideal system where the molar latent heat is no longer constant and there is a substantial
heat of mixing, the calculations become much more tedious. For binary mixture of this kind a
graphical model has been developed by Ruhemann, Ponchon, and Savarit, based on the use of an
enthalpy – composition chart.
A typical enthalpy – composition or (H-y-x chart) is indicated in the following diagram:
y, x
Mole (mass) fraction in vapor, liquid phases
163
This diagram is based on the following geometrical properties, as illustrate in the following figure:
Q
A +Q = C
A quantity of mixture in any physical state is known as "phase" and it is denoted by mass
composition and enthalpy.
The phase is indicated upon the diagram by a point which shows enthalpy and composition, but
which does not show the mass.
If m is the mass, x the composition and H is the enthalpy of the phase, then the addition of two
phases A and B to give phase C is governed by:
mA + mB = mC
mA * xA + mB * xB = mC * xC
mA * HA + mB * HB = mC * HC
Similarly, if an amount Q of heat is added to mass mA of a phase, the increase in enthalpy from HA
to HC will be given by
HA + (Q/mA) = HC
Thus, the addition of two phases A and B is shown on the diagram by point C on the straight line
joining the two phases, whilst the difference A-B is found by a point C on the extension of the line
AB.
In figure (1), a phase represented by C in the region between the dew-point and boiling-point
curves is considered, then this phase will be divided into two phases A and B at the ends of a tie
line through the point C, so that:
MA CB
=
MB CA
The H-y-x chart, therefore, enables the effect of adding two phases with or without the addition of
heat to be determined geometrically. The diagram may be drawn for unit mass or for one mole of
material, though as a constant molar reflux does not now apply, it is more convenient to use unit
mass as the basis. Thus, working with unit mass of product, the mass of the individual streams as
proportions of the product will be calculated,
Equilibrium Data:
It takes into account the latent heat of vaporization, sensible heat and heat of solution or mixing.
So, the saturated liquid line in H-y-x diagram is:
hL = HA + HB + ∆Hsol
164
hL = xA * CpA * (T-Tref) + (1- xA) * CpB * (T-Tref) + ∆Hsol
HV = hL * λmix
Where:
λmix = xA * λA + (1-xA) * λB
x A & yA
yA
Z
xA
The tie (ab) represents the enthalpy
For non – adiabatic mixing which is similar to that for adiabatic but with addition of Q (Q is heat
of mixing or heat losses or the net), so
S1 * H1 + S2 * H2 = S3 * H3 + Q
165
Example 7
Two pure liquids A & B are to be mixed adiabatically with each other and then mixed with another
mixture of A & B. The pure A & pure B streams are in the ratio of 4:1 (mass ratio). Initially pure A
has an enthalpy of 4 units/ kgm while pure B has initially 2 units/ kgm. Mixture 2 has a mass
fraction of B (0.75) with an enthalpy of 7 units/ kgm. Determine the composition of mixture 3 if its
enthalpy is 5 units/ kgm. What mass of mixture 2 is required with respect to 1 kgm of pure A used
in the mixture 1.
Mixture 2
A
Mixture 1 Mixture 3
B
Solution
Basis 1 kgm of A
Θ Mixture 1 composed of 4 kgm A and 1 kgm of B
That means that mass fraction of B = 0.2
So total mass of mixture 1 (S1) = 1.25 kgm
Enthalpy of mixture 1 (h1) = xA * hA + (1- xA) * hB
h1 = (0.8 * 4) + (1 – 0.8) * 2 = 3.6 units/ kgm
S1 Z 2 − Z 3
Θ = 8
S 2 Z 3 − Z1
From the figure:
Z3 = 0.42
1.25 0.75 − 0.42 6
∴ =
S 0.42 − 0.2
2
S2 = 0.833 4
0
0 0.2 0.4 0.6 0.8 1
Operating Data:
As done before we will divide the operating data into five parts, the first part is the equilibrium
data, the second part is similar to the 45o line, the third part is above the feed the forth part is below
the feed and finally the fifth part which concern the feed entrance.
Part one:
The equilibrium data is given as (hL vs. x), and (hV vs. y) which must be plot on the same diagram.
Some times the enthalpies of liquid and vapor must be calculated as mentioned in above.
Part Two:
166
Overall column M.B.
F=D+W …1
[M.V.C.] M. B.
F * xf = D * xD + W * xw …2
[M.V.C.] E. B.
F * hf + qr = qc + D * h D + W * h w …3
Then:
q q
F * h = D * (h + c ) + W * (h − r )
f D D W W
q
Let Qc = c (heat lost by condenser per kgm of distillate)
D
q
And Qr = r (heat gained by re-boiler per kgm of residue)
W
Then:
F * h = D * h` + W * h` ...4
f D W
Where:
h` = h + Q c
D D
h` = h − Qr
W W
Substitute equation 1 into 2 and 4, then
(D + W) * xf = D * xD + W * xw
D * (xD – xf) = W * (xW – xf) …5
(D + W) * h = D * h ` + W * h `
f D W
D * (h ` - h ) = W * (h - h ` ) ...6
D f f W
Dived equation 5 by 6
x −x x −x
D f = f W
h` − h h − h`
D f f W
This equation states that h ` , hf and h ` is on a straight line at the points D`, f, and W`
D W
Where:
D`: is the pole point of all operating lines above feed.
W`: is the pole point of all operating lines below feed.
f: is the feed point.
167
Part three:
Operating lines above the feed:
Total M. B.
Vn = Ln+1 + D …1
[M.V.C.] M. B.
Vn * yn = Ln+1 * xn+1 + D * xD …2
Enthalpy M. B.
Vn * hn = Ln+1 * hn+1 + D * (hD + Qc)
Or
Vn * hn = Ln+1 * hn+1 + D * ( h ` ) …3
D
Substitute equation (1) into (2) and (3)
(D + Ln+1) * yn = Ln+1 * x n+1 + D * xD
D * (xD – yn) = Ln+1 * (yn – x n+1) …4
(D + L n + 1 ) * h n = L n + 1 * h n + 1 + D * h `D
D * (h `D − h n ) = L n + 1 * (h n − h n + 1 ) ...5
Dived equation 4 by 5
x − yn yn − x
D = n +1
`
h − hn h − h
D
n n +1
Which is also as before an equation of straight line, which states that D`, yn, and xn+1 are points of
the line.
Now, part five which concerns the effect of the feed conditions, this is not similar to that in
McCabe-Thiele method; i.e. there is no q-line. The effect of feed conditions is directly presented
on the diagram by locating the feed enthalpy and its composition. The feed enthalpy is calculated
depending on its condition either if it is sub-cooled liquid, or saturated liquid, or two phase
mixture, or saturated vapor, or superheated vapor.
1- For sub-cooled liquid:
hf = hL – ∆hL
∆hL = xA * CpA * (Tbubble – Tentry) + (1- xA) * CpB (Tbubble – Tentry)
Where:
hL = xA * CpA * (T-Tref) + (1- xA) * CpB * (T-Tref) + ∆Hsol
This point is located at x = xf below the bubble curve by a value of ∆hL.
2- for saturated liquid:
This point is located directly on the bubble curve at x = xf.
3- For two phase mixture:
hf = (1 –X)* hL+ X * hV
Where: X is the quality of the feed.
4- For saturated vapor:
This point is located directly on the dew curve at x = xf.
5- For superheated vapor:
hf = hV + CPmixv * (Tentry - Tdew)
Where:
hV = hL * λmix
λmix = xA * λA + (1-xA) * λB
Reflux Ratio
Lr
The reflux R= can be determined as follow:
D
Vn = Lr + D
qC = Vn * (HD – hD)
Where: HD is the enthalpy of saturated vapor whose composition is y = xD
169
q
Θ h` = h + c
D D D
q
c = h` − h
D D D
V * (H − h )
∴ n D D = h` − h
D D D
Lr
Substitute Vn by and rearrange the above equation then:
D
(L + D (h ` − h )
r = D D
D (H − h )
D D
(h` − h )
R +1 = D D
(H − h )
D D
`
(h − h )
R= D D -1
(H − h )
D D
`
h −h −H +h
R= D D D D
(H − h )
D D
`
(h − H )
R= D D ًﻋﻼﻗﺔ ﻤﻬﻤﺔ ﺠﺩﺍ
(H − h )
D D
Now to calculate the number of theoretical plates required, follow the steps:
1- Complete the material balance and energy balance, so all the streams and their composition are
known in addition to that calculated the heat required in the re-boiler (not very necessary) and the
heat rejected in the condenser (also not necessary).
2- Plot the equilibrium data.
3- Draw vertical lines from xD, Xf, and xW.
4- Locate D`, f, and then plot the line joining these two points and then extent to intersect the
vertical line drawn from xW (then point w` is located). Point D` is located at x = xD and hD` which
is calculated from the equation of the reflux ratio, while f is located at x = xf and depending on the
feed condition.
5- From the equilibrium data find xn*, and draw the tie line.
6- Join the point D` with xn* it will intersect the dew curve. The intersection point is y1.
7- Repeat steps 5 & 6 until you accede xf.
8- From point W` plot a line to join W` with xn+1 and extent to intersect the dew point curve.
9- Read yk, and from the equilibrium data find xk* (draw the tie line).
10- Join the point W` with xk* and extent to intersect the dew curve. The intersection point is yk+1.
11- Repeat steps 9 & 10 until you accede xW. Step-off the number of plates (N) from the figure.
Then:
The theoretical number of plates = N -1.
Check the figure below (for saturated feed case)
170
Total reflux (minimum number of plates) & Minimum reflux ratio:
A- Total reflux
Lr
As R= is increased, the operating point D` must be located at higher position since
D
q
h ` = h + c increased (decreasing D) R = ∞ when D 0
D D D
At the same time the point W` become in lower position, since D`, f, W` lie on the same line. This
behavior was also noticed before on the y-x diagram when the operating lines become closer to the
45o line which yield the minimum number of plates, so
On the enthalpy diagram as R ∞ the point D` & W` goes to ∞, then all the operating lines will
become vertical and parallel. So to find the minimum number of plates, no product
(i.e. R ∞ ) draw a vertical line from the point (hD, xD) on the bubble curve until it intersects the
dew curve, then from equilibrium data draw the tie line between yD and x1 then draw a vertical line
from the bubble curve to intersect the dew curve and carry on repeating this procedure until you
reach xW the number of triangular form represent the minimum theoretical plates needed, as shown
in the figure.
171
B- Minimum reflux ratio (infinite number of plates)
As done before when using minimum reflux ratio the number of plates required to achieve the
required separation will approach to infinity. To calculate the minimum reflux ratio follow the
following steps:
1- Locate xD and draw a vertical line.
2- Locate xF and find yF from (y-x diagram) which is in equilibrium with xF and locate yF on the
dew curve.
3- Draw a line connecting (yF, xF) and extend it until intersects the vertical line drawn from xD.
4- The point of intersection of the above two lines will called Hm`, and this value will be used to
calculate the value of the minimum reflux ratio (Rmin) as shown in below:
H` − H
R min = m D
H −h
D D
172
Multi-component Distillation
A multi-component distillation problem contains many variables. The conditions of equilibrium are
more complex, thus yA depend not only on xA but also on the relative proportion of the other
components.
The simplest relationship can be used for equilibrium is:
y A = kA * x A
K-values varies with temperature and pressure, some of it are presented in figure 11.38 in
(chemical engineering vol. 2) and also in any thermodynamic text book.
Other methods used for many systems which are chemically similar where the relative volatilities
remain constant over a wide range of temperature and composition.
Light key and heavy key
Suppose four components namely (A, B, C, and D) are to be distilled where A is the M.V.C. and D
is the L.V.C., the final products are shown in below:
Feed Top product Bottom product
A A --
B B B
C C C
D -- D
Component B is the highest component appearing in the bottom product so it is called the "light
key component"; where the component C is the heaviest component appearing in the top
product, component C is the "heavy key component".
The mean purpose of fractionation is the separation B from C.
Equilibrium Data:
Let the four components (A, B, C, and D) are to be distilled where B is the heavy key component
(HK) then:
yA + yB +yC + yD = 1
And
y y
A + C + yD + yB = 1 ...1
yB yB yB yB yB
P
A
x
∴ α AB = A
P
B
x
B
But PA = yA * PT
Then
y
A
x
∴ α AB = A
y
B
x
B
Or
α *x
y = AB A ...2
A x
B
y
B
173
k
Also ∴ α AB = A
k
B
Hence:
y x
A =α * A
y AB x
B B
The same is done for each component, then substitute in equation 2 to get:
x x x x 1
α * A +α * B +α * C +α * D=
AB x BB x CB x DB x
B B B B yB
Or
n =i
∑ α*x
n =1 =
1
x y
B B
Or
B = n∑= iα * x
x
y
B n =1
Substitute in equation 2, then:
α *x
y = AB A ...3
A n =i
∑α*x
n =1
α *x
y = CB C ...4
C n =i
∑α*x
n =1
α *x
y = DB D ...5
D n =i
∑α*x
n =1
Equations 3, 4, and 5 are the equilibrium equations.
Calculations of bubble point and dew point
Vapor pressure of a liquid at certain temperature and total pressure can be found from Henry's law.
P = P0 * x Also P = y *P
A A A A A T
And P = H * x where: H is Henry's constant
A A
P H*x
y= A = A = k *x
P P A A
T T
So the partial pressure for component A is given as:
PA= PT * kA * xA
Then at bubble point the total vapor pressure = P * (Σ k * x) = PT
If any inert gas present, then:
Pinert = PT * [1- (Σ k * x)]
174
To calculate the bubble point of multi-component mixture (boiling point of a mixture):
At bubble point:
1- Total vapor pressure of the mixture = ambient pressure (PT).
2- Σ k * x = 1
So by trail and error the bubble point can be predicted following the steps:
A- Assume a probable value of bubble point, (PT is known).
B- At this temperature read k values from figures (monogram fig.) for each component.
C- Find (Σ k * x).
If the (Σ k * x) = 1 then the assumption is correct.
If the (Σ k * x) > 1 then the assumed temperature is too high.
If the (Σ k * x) < 1 then the assumed temperature is less than the bubble point.
y
For the calculation of dew point, follow the same steps as in above but for ∑ k =1.
Example 8
Determine the vapor pressure and the bubble point of the liquid mixture with the following
composition under total pressure of 2 bar and a temperature of 25oC:
Component n- C4H10 n- C5H12 n- C6H14 n- C7H16
Mol% 20 40 30 10
Solution
From monogram vol. 6 figure 8.3 read k-value for each component at 2 bar and 25oC as shown in
the table below:
Component Mole fraction k-value Xi * ki
n-C4 0.2 1.25 0.25
n-C5 0.4 0.35 0.14
n-C6 0.3 0.109 0.0327
n-C7 0.1 0.031 0.0031
Θ Σxi * ki = 0.426 ≠ 1
Therefore there is an inert in the vapor phase.
So, the inert gas pressure = PT(1 - Σxi * ki) = 2 * (1 – 0.426) = 1.148 bar
The vapor pressure mixture = 0.852 bar
Since at 25oC the vapor pressure of the liquid is less than the total pressure; therefore the bubble
point of the mixture at 2 bar is larger than 25oC. Assume it is equal to 50oC, then
Component Mole fraction k-value Xi * ki
n-C4 0.2 2.3 0.46
n-C5 0.4 0.8 0.32
n-C6 0.3 0.29 0.087
n-C7 0.1 0.09 0.009
Σxi * ki = 0.876 < 1
Assume T = 60oC
Component Mole fraction k-value Xi * ki
n-C4 0.2 2.8 0.56
n-C5 0.4 1.0 0.4
n-C6 0.3 0.38 0.114
n-C7 0.1 0.14 0.014
Σxi * ki = 1.088 > 1
175
Let T = 56oC
Component Mole fraction k-value Xi * ki
n-C4 0.2 2.59 0.518
n-C5 0.4 0.92 0.368
n-C6 0.3 0.34 0.102
n-C7 0.1 0.21 0.012
Σxi * ki = 1.0
So the bubble point of the mixture = 56oC
The vapor pressure of each component are roughly estimated by
PA = PT * (kA * xA), then
Component Mole fraction k-value Xi * ki PA (bar)
n-C4 0.2 2.59 0.518 1.036
n-C5 0.4 0.92 0.368 0.736
n-C6 0.3 0.34 0.102 0.204
n-C7 0.1 0.21 0.012 0.024
A- Flash distillation
For multi-component system the equilibrium relation will y = k * x as shown in above, so for
component i
y i = ki * x i
As derived before in binary system, the operating equation for multi-component mixture is written
as:
x 1 - f v yi
y = fi - ( )* ... *
i fv fv k
i
Or
k *x
y = i fi ... * *
i f * (k − 1) + 1
v i
For multi-component calculation to find yi (mole fraction of vapor leaving the flash drum:
1- Assume fv (always < 1), then calculate yi for each component.
2- Find Σ yi
If the (Σ yi) = 1 then the assumption is correct.
If the (Σ yi) ≠ 1 then assume another fv.
To calculate the heat duty done to heat the feed. Make energy balance as:
F * Hf + q = S * Hs + V * Hv
Example 9:
The following mixture is to be subjected to equilibrium flash vaporization at 150oC and 1930
kN/m2. Determine the bubble point and the dew point of this mixture and then find the amount and
composition of the liquid and vapor phases resulting from this separation process, for feed rate of
100 kmole/hr with the following composition:
2- Calculation of dew point temperature, we will assume the mixture is a vapor mixture at PT =
1930 kn/m2, assume T = 190oC
y fi
component Ki yfi ki
C3 3.85 0.28 0.073
n-C4 2.0 0.24 0.12
n-C5 1.25 0.24 0.192
n-C6 0.78 0.08 0.103
n-C8 0.3 0.16 0.533
177
yi
∑k = 1.02 > 1
i
3- To find the amounts and the composition of liquid and vapor at 150oC and 1930 kN/m2, we will
assume fV and then find the yi of each component using the following equation.
If Σ yi = 1 then the assumed fV is correct, if not assume a new one.
k *x
y = i fi
i f * (k − 1) + 1
v i
Then we must find the k-value for each component at T = 150oC and P = 1930 kN/m2
Feed components C3 n-C4 n-C5 n-C6 n-C8
k-value 3 1.42 0.85 0.48 0.17
178
1.42 * 0.24
y = = 0.2773
n -C
4 0.54475 * (1.42 − 1) + 1
0.85 * 0.24
y = = 0.2215
n -C
5 0.54475 * (0.85 − 1) + 1
0.48 * 0.08
y = = 0.0536
n -C
6 0.54475 * (0.48 − 1) + 1
0.17 * 0.16
y = = 0.0496
n -C
8 0.54475 * (0.17 − 1) + 1
Σ yi = 1.004 ≈ 1, then assumption is correct
V
fV =
F
F =100 kmole/he
Then
V = 54.4475 kmole/hr
L = 45.5525 kmole/hr
Then to find the composition of the liquid phase make a material balance for each component as
follow:
F * xfi = V * yi + L * xi
Then
F* x − V * y
x = Fi i
i L
179
V = 54.4475 kmol/hr
yC3 = 0.402
yn-C4 = 0.2773
yn-C5 = 0.2215
yn-C6 = 0.0536
yn-C8 = 0.0496
F = 100 komle/hr
xfC3 = 0.28
xfn-C4 = 0.24
xfn-C5 = 0.24
xfn-C6 = 0.08
xfn-C8 = 0.16
L = 45.5525 kmole/hr
xC3 = 0.1342
xn-C4 = 0.1954
xn-C5 = 0.2621
xn-C6 = 0.1116
xn-C8 = 0.292
B- fractionating column:
To calculate the number of trays needed in fractionating column, Lewis and Matheson method is
used, which is based on Lewis and sorel method when the composition of liquid on any phase is
known, then vapor composition in equilibrium can be calculated from vapor pressures or relative
volatilities. So the composition of each component above the feed plate entrance can be calculated
from the upper operating line equation and below the feed plate the lower operating line equation
must be used.
Keep in mind that for each component an upper operating equation and lower operating equation
must be found. In below the steps which must be followed to calculate the number of theoretical
plates.
1- Complete the material balance, so the mole flow rate of each component in the two products
must be known (i.e. D, and W is calculated and their composition also).
2- Calculate the reflux ratio.
3- Exam the feed condition in order to calculate (q) if it is necessary.
4- Find the U.O.L. & L.O.L. equations for each component.
5- Starting from the bottom, use the equilibrium equation and substitute xWi in the following
equation for each component to find ymi
α *x
ij i
y =
i n =i
∑α*x
n =1
180
6- Substitute ymi in the L.O.L equation and find x(m+1)i for each component.
7- Repeat steps 6 & 7 until you reach xmi > xfi then substitute ymi obtained from the equilibrium
equation in the U.O.L. equation for each component to find xmi and repeat the same
procedure as done in steps 6 &7 until you reach xni = xDi. Step off the number of plates
required (N).
Theoretical number of plates = N -1
Minimum reflux ratio
The method used for determine the minimum reflux ratio for binary mixture graphically can not be
used when we deal with multi-component system, as we can not draw an equilibrium curve for this
system. There are two methods for determining the minimum reflux:
1- Colburn's Method:-
The following equation was proposed by Colburn to calculate the minimum reflux ratio:
1 x x
R min = −1 [ DL - α LH * DH ]
α LH x nL x nH
Where:
αDL: The relative volatility of L.K. component to the H.K. component.
xDL: The mole composition of the L.K. component in the distillate.
xDH: The mole composition of the H.K. component in the distillate.
XnL: The mole composition of the L.K. component in the pinch point.
XnH: The mole composition of the H.K. component in the pinch point.
Where:
rf
x nL =
(1 + rf )(1 + ∑ α x fh )
x
x nH = nL
r
f
For rf: is the estimated ratio of the key components on the feed plate.
For a liquid feed at its bubble point, rf equals to the ratio of the key components in the feed.
x
r = nL
f x
nH
Otherwise rf is calculated as the ratio of the key component in liquid part of the feed.
xfh: is the mole fraction of each component in liquid portion of feed heavier than heavy key in feed.
α: relative volatility of the components relative to the H.K.
2- Underwood's Method
For conditions where the relative volatilities remain constant, Underwood has developed the
following equations from which Rmin may be calculated:
α A * x fA α B * x fB α C * x fC
+ + + ..... = 1 − q
αA − θ αB − θ αC − θ
And
α A * x DA α B * x DB α C * x DC
+ + + ..... = R min + 1
αA − θ αB − θ αC − θ
Where:
181
xfA, xfB, xfC, xDA, xDB, xDC …etc are the mole fraction of components A,B,C, …etc in the feed and
distillate, A being the light key and B is the heavy key.
αA, αB, αC, are the relative volatilities of components with respect to the heavy key.
q: is the heat required to vaporize one mole of feed to the molar latent heat of the feed.
θ: is the root of the first equation where:
αH < θ < αL
Example 10:
Suppose a mixture of hexane, Heptane and Octane to be separate to give products as shown in the
table. What will be the value of the minimum reflux ratio, if the feed is liquid at its boiling point?
Then find the minimum number of plates required. Investigate the change in N with R and find the
number of plates if R=10. (Plot N vs. R).
182
Solution
The L.K. is Heptane, H.K. is Octane
Then αHO = 2.7, αHeoO = 2.2, αOO = 1.0
Using Underwood's method:
α A * x fA α B * x fB α C * x fC
+ + + ..... = 1 − q
αA − θ αB − θ αC − θ
Where q = 1 (saturated liquid), then
2.7 * 0.4 2.22 * 0.35 1 * 0.25
+ + =0
2.7 − θ 2.22 − θ 1− θ
αH < θ < αL or 1< θ < 2.22
The above equation is solved by trail and error, so assume θ = 1.15
α * xf
∴∑ = −0.243
α−θ
Assume θ = 1.17, then
α * xf
∴∑ = −0.024 ≈ 0 This is acceptable value
α−θ
Now substitute the value of θ in the following equation
α A * x DA α B * x DB α C * x DC
+ + + ..... = R min + 1
αA − θ αB − θ αC − θ
2.7 * 0.534 2.22 * 0.453 1 * 0.013
+ + = R min + 1
2.7 − 1.17 2.22 − 1.17 1 − 1.17
Rmin = 0.827
The minimum number of plates Nmin can be calculated by:
log[(x L /x H ) D * (x H /x L ) B ]
N min + 1 =
log(α LH ) av
Here consider αLH is constant for F, D and W, then
log[(0.453/0.013) * (0.96/0.04)]
N min + 1 =
log(2.22) av
Nmin +1 = 8.5
Nmin = 7.5
Now to find the effect of R on the number of plates use the equation:
N − N min 1.805 R − Rmin
y= = 1 − exp[1.49 + 0.315 * x − 0.1 ] ..* Where x =
N+2 x R +1
Establish a table as shown in below by assuming R starting from Rmin to a value of 10 for each
value find x for each R and then substitute in equation (*) and find y then evaluate N.
Finally plot N vs. R as given in below:
Θ Rmin = 0.827 ≈ 0.83
And Nmin = 7.5
183
21
17
N 13
5
0 3 6 9 12
R
Example 11:
A mixture of Ortho, Meta and Para mono-nitro-toluene containing 60, 4 and 36 mol% respectively
of the three isomers is to be continuously distilled to give a top product of 98 mol% Ortho, the
bottom product is to contain 12.5 mol% Ortho. The mixture is to be distilled at a temperature of
410 K requiring a pressure in the boiler of about 6 kN/m2 (0.06 bar). If a reflux ratio of 5 is used,
how many theoretical plates will be required and what will be the approximate composition of the
top product stream? Given the following data:
αOP = 1.7
Both at a range of 380 to 415 K
αMP = 1.16
Solution:
As a first estimation, suppose the distillate to contain 0.6 mol% Meta and 1.4 mol% Para. Then by
material balance find the composition of the bottom.
Basis 100 kmole of feed
D is the top product with composition of xDO (mole fraction of Ortho).
W is the bottom product with composition of xWO (mole fraction of Ortho).
O. M. B.
F=D+W
[Ortho] O. M. B.
60 = D * xDO + W * xWO
60 = (100-W) * 0.98 + 0.125 * W
Then D = 55.56 kmole and W = 44.44 kmole
The M. B. will give the compositions and amounts of all streams as shown in the following table:
Feed Distilled Bottom
Component
kmole Mol% kmole Mol% kmole Mol%
Ortho (O) 60 60 54.44 98 5.56 12.5
Para (P) 36 36 0.79 0.6 35.21 79.2
Meta (M) 4 4 0.33 1.4 3.67 8.3
Now find the operating lines equations for each component:
184
Above the feed
Liquid flow rate Ln = R *D = 55.56 * 5 = 277.8 kmole
Vapor flow rate Vn = D * (R + 1) = 55.56 *(5 + 1) = 333.4 kmole
Assuming the feed enters as saturated liquid, then
Below the feed
Liquid flow rate LM = Ln + F = 277.8 + 100 = 377.8 kmole
Vapor flow rate Vm = Lm – W = 377.8 – 44.4 = 333.4 kmole
Therefore the lower operating lines for each component will be as:
L m+1 W
ym = * xm − *xW
Vm Vm+1
For Ortho
377.8 44.4
y Om = * x Om+1 − * 0.125
333.4 333.4
y Om = 1.133 * x Om+1 - 0.01666 ...1
For Meta
y Mm = 1.133 * x Mm+1 - 0.011 ...2
For Para
y Pm = 1.133 * x Pm +1 - 0.105 ...3
The upper operating lines are:
For Ortho
y On = 0.833 * x On +1 + 0.163 ...4
For Meta
y Mn = 0.833 * x Mn +1 + 0.001 ...5
For Para
y Pn = 0.833 * x Pn +1 + 0.002 ...6
Starting from the bottom, the composition of liquid leaving the still is in equilibrium with the vapor
entering the column; therefore the composition of this vapor ySi is found from the equation of
relative volatility:
α *x
ij i
y =
i n =i
∑α*x
n =1
For Ortho
α *x
y SO = OP SO
α *x + α *x + α *x
OP SO MP SM PP SP
1.7 * 0.125
y SO = = 0.193
1.7 * 0.125 + 1.16 * 0.08 + 1* 0.792
For Meta
α *x
y SM = MP SM
α *x + α *x + α *x
OP SO MP SM PP SP
1.16 * 0.083
y SP = = 0.0875
1.7 * 0.125 + 1.16 * 0.08 + 1* 0.792
185
For Para
α *x
y SP = PP SP
α *x +α *x +α *x
OP SO MP SM PP SP
1* 0.792
y SP = = 0.7195
1.7 * 0.125 + 1.16 * 0.08 + 1* 0.792
Use these values of ysi in the lower operating lines equations (equations 1, 2 and 3) to find the x1i
Then
x1O = 0.185
x1M = 0.087
x1P = 0.728
Now find y1i by reusing the equilibrium equations and then re-substitute in the L.O.L. equations to
find x2i, the following table is established
Component xS α * xS yS x1 α * x1 y1 x2 α * x2
Ortho o 0.125 0.2125 0.193 0.185 0.3145 0.275 0.257 0.4369
Meta M 0.083 0.0963 0.0875 0.087 0.1009 0.088 0.087 0.1018
Para P 0.792 0.792 0.719 0.728 0.728 0.637 0.656 0.656
Sum 1.0 1.1008 1.0 1.0 1.1434 1.0 1.0 1.1947
Component y2 x3 α * x3 y3 x4 α * x4 y4 x5
Ortho o 0.366 0.338 0.5746 0.46 0.421 0.7157 0.547 0.4975
Meta M 0.085 0.085 0.0988 0.079 0.079 0.09164 .07 0.0715
Para P 0.549 0.577 0.577 0.461 0.5 0.5 0.383 0.431
Sum 1.0 1.0 1.2502 1.0 1.0 1.30734 1.0 1.0
Component α * x5 y5 x6 α * x6 y6 x7
Ortho o 0.847575 0.622 0.5636 0.95812 0.6821 0.617
Meta M 0.08299 0.061 0.0635 0.07366 0.0524 0.056
Para P 0.431 0.317 0.3729 0.3729 0.2655 0.327
Sum 1.35969 1.0 1.0 1.40468 1.0 1.0
Here we have exceed the composition of the feed, then by using the same procedure replacing the
L.O.L. with the U.O.L. equations and continue the calculation as
Component x7 α * x7 y7 x8 α * x8 y8 x9 α * x9 y9
Ortho o 0.617 1.0489 0.728 0.678 1.1526 0.7771 0.737 1.2529 0.822
Meta M 0.056 0.06496 0.045 0.053 0.0615 0.0415 0.048 0.05563 0.037
Para P 0.327 0.327 0.227 0.269 0.269 0.1814 0.215 0.215 0.141
Sum 1.0 1.44086 1.0 1.0 1.4831 1.0 1.0 1.52358 1.0
Component x10 α * x10 y10 x11 α * x11 y11 x12 α * x12 y12
Ortho o 0.791 1.3447 0.862 0.839 1.4263 0.875 0.879 1.4943 0.922
Meta M 0.043 0.05 0.032 0.037 0.04292 0.027 0.031 0.03596 0.022
Para P 0.166 0.166 0.106 0.124 0.124 0.078 0.09 0.09 0.056
sum 1.0 1.5607 1.0 1.0 1.59322 1.0 1.0 1.62026 1.0
186
Component x13 α * x13 y13 x14 α * x14 y14 x15 α * x15 y15
Ortho o 0.911 1.5487 0.943 0.936 1.5912 0.959 0.955 1.6235 0.972
Meta M 0.025 0.029 0.018 0.02 0.0232 0.014 0.015 0.01755 0.01
Para P 0.064 0.064 0.039 0.044 0.044 0.027 0.03 0.03 0.018
sum 1.0 1.6417 1.0 1.0 1.6584 1.0 1.0 1.67105 1.0
xBenzene 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
yBenzene 0 0.2 0.38 0.51 0.63 0.71 0.78 0.85 0.91 0.96 1.0
Solution:
Basis 100 kmole of feed
A total material balance gives:
100 = D + W
[Benzene M. B.]
100 * 0.4 = 0.9 * D + 0.1 * W
Thus:
40 = 0.9 * (100- W) + 0.1 * W
Whence:
W = 62.5 kmole and D = 37.5 kmole
Ln
R=
D
Ln = 3 * D = 112.5
And
Vn = Ln + D = 150
187
Thus, the upper operating line equation is:
112.5 37.5
yn = * x n +1 + * 0.9
150 150
Or
y n = 0.75 * x n +1 + 0.225
Since the feed is all at its boiling point, it will all run down as increased reflux to the plate below.
Thus
Lm = Ln + F = 112.5 +100 = 212.5
Vm = Lm – W
Vm = 212.5 – 62.5 = 150 = Vn
Thus, the lower operating line will be
212.5 62.5
∴ ym = * x m+1 − * 0.1
150 150
Or
y m = 1.415 * x m+1 - 0.042
Continue as described in the lectures.
Example 13
A liquid mixture of Benzene – Toluene is to be distilled in a fractionating column. The feed of 100
kmole/hr is liquid containing 45 mol% Benzene enters the column at a temperature 39 K below its
bubble point.
The top product contains 95 mol% Benzene and the bottom product contain 10 mol% Benzene.
The reflux ratio is 4:1.
The specific heat of the feed is 159 kJ/(kmol. K). the molar latent heat of vaporization of both
Benzene and Toluene may be assumed constant and equal 32100 kJ/ kmol. Equilibrium data for
Benzene – Toluene mixture is:
xBenzene 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
yBenzene 0 0.2 0.38 0.51 0.63 0.71 0.78 0.85 0.91 0.96 1.0
Calculate:
1- The number of theoretical plates required.
2- The vapor flow rates inside the column.
3- The liquid flow rates inside the column.
4- The heat transfer rate in the re-boiler in (kJ/hr).
188
Steam distillation
Where the material to be distilled has a high boiling point, and particularly where decomposition
might occur if direct distillation were employed, the process of steam distillation can be used.
Steam is passed directly into the liquid in the still; the solubility of steam in the liquid must be very
low. Steam distillation is perhaps the most common example of differential distillation.
Two cases are possible. The steam may be superheated and provide sufficient heat to vaporize the
material concerned, without itself condensing. Alternatively, some of the steam may condense,
producing a liquid water phase. In either case, assuming the gas laws to apply, the composition of
the vapor produced can be obtained from the following relationship:
m A m B PA y A PA
/ = =
M A M B PB y B P − PA
=
Where the subscript A refers to the component being recovered and B to steam, and:
m = mass
M = molecular weight
PA, PB = partial pressure of A, B
P = total pressure
The second type of problem occurs where the relative volatility of 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
189
large cross-section containing many trays. An example of the second type of problem is the
separation of n-heptane from methyl cyclohexane. Here the relative volatility is only 1.08 and a
large number of plates are 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
separation relatively easy. Benedict and Rubin have defined these two processes in the following
way. 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 concentration. 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
concentration on all the plates.
The third component added to the binary mixture is sometimes known as the entrainer or the
solvent.
190