B49CE Student Guide v1 PDF

CONTENTS 1
B49CE
Multiphase Thermodynamics
Heriot-Watt University
Edinburgh EH14 4AS, United Kingdom
Produced by Heriot-Watt University, 2015
HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

CONTENTS 2
Copyright 2015 Heriot-Watt University
All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system or transmitted in any form or by any means without
express permission from the publisher.
This material is prepared to support the degree programmes in Chemical

and Petroleum Engineering.
Distributed by Heriot-Watt University

Multiphase Thermodynamics

CONTENTS 3
Acknowledgements
Thanks are due to the members of Heriot-Watt, School of Engineering and Physical
Science who planned and generated this material.
We would like to acknowledge the assistance and contributions from colleagues across
the University and students in preparing this and support material.

CONTENTS 4
Contents
TOPIC 1: PURE SPECIES PHASE EQUILIBRIUM

1.1 INTRODUCTION .................................................................................................................. 3
1.2 CRITERION FOR EQUILIBRIUM ............................................................................................. 4
1.2.1 Pure Species Phase Equilibrium ......................................................................... 4
1.2.2 Pure Species Chemical Potential ........................................................................ 8
1.2.3 Pure Species Gibbs Energy Relation................................................................... 9
1.2.4 Pressure Dependence of Pure Species Chemical Potential .............................. 10
1.3 PURE SPECIES FUGACITY ................................................................................................ 12
1.3.1 Fugacity as a Criterion for Pure Species Phase Equilibrium .............................. 14
1.3.2 Fugacity using Residual Volume ....................................................................... 15
1.3.3 Example 1.1 ...................................................................................................... 16
1.3.4 Fugacity using Compressibility Factors ............................................................. 17
1.3.5 Compressibility Factor Charts............................................................................ 17
1.3.6 Example 1.2 ...................................................................................................... 20
1.3.7 Fugacity using Charts ........................................................................................ 21
1.3.8 Example 1.3 ...................................................................................................... 22
1.3.9 Fugacity of Pure Saturated Liquids .................................................................... 23
1.3.10 Fugacity of Compressed Liquid States ........................................................... 25
1.3.11 Example 1.4 .................................................................................................... 27
1.3.12 Fugacity Coefficient using Cubic EOS ............................................................. 28
1.4 TUTORIAL 1..................................................................................................................... 29
1.5 BIBLIOGRAPHY ................................................................................................................ 31
APPENDIX A.......................................................................................................................... 32
Simple Gas Z-Factor Chart......................................................................................... 32
Non-Simple Gas Z-Factor Chart ................................................................................. 33
APPENDIX B.......................................................................................................................... 35
Simple Gas Fugacity Coefficient Chart High Pressure............................................. 35
Non-Simple Gas Fugacity Coefficient Chart High Pressure ..................................... 36
Simple Gas Fugacity Coefficient Chart Low Pressure ............................................. 37
Non-Simple Gas Fugacity Coefficient Chart Low Pressure ...................................... 38
APPENDIX C ......................................................................................................................... 39
Fugacity Coefficient using Cubic EOS ........................................................................ 39

CONTENTS 5
TOPIC 2: VLE FOR NON-POLAR MIXTURES

2.1 INTRODUCTION .................................................................................................................. 3
2.1.1 The Apportioning Rule, the Recombination Rule and Gibbs-Duhem.................... 3
2.2 CRITERIA FOR PHASE EQUILIBRIUM .................................................................................... 5
2.2.1 Gibbs Free Energy of a Mixture ........................................................................... 5
2.2.2 Daltons Law........................................................................................................ 7
2.2.3 Fugacity Coefficient of a Component in a Mixture................................................ 7
2.2.4 Ideal Solutions and Lewis-Randall Rule .............................................................. 8
2.2.5 Criterion for Multicomponent Phase Equilibrium ................................................ 10
2.2.6 Fugacity as a Criterion for Mixture Phase Equilibrium........................................ 12
2.2.7 Finding Fugacity Coefficient of a Component in a Mixture ................................. 13
2.3 VLE FOR NON-POLAR SYSTEMS ...................................................................................... 15
2.3.1 Rigorous Phi-Phi Approach to VLE .................................................................... 15
2.3.2 Liquid and Vapour as Ideal Solutions ................................................................ 17
2.3.3 Raoults Law: Liquid as an Ideal Solution and Vapour as an Ideal Gas ............. 18
2.3.4 Phi-Phi VLE Summary ....................................................................................... 19
2.3.5 Finding K-Values using Charts .......................................................................... 20
2.3.6 Example 2.1 ...................................................................................................... 22
2.3.7 Vapour Pressure Correlation ............................................................................. 25
2.3.8 Temperature-Composition Diagram for Binary System ...................................... 27
2.3.9 Significance of K-Values.................................................................................... 28
2.4 PHASE EQUILIBRIUM CALCULATIONS ................................................................................ 29
2.4.1 Bubble Point Calculations .................................................................................. 29
2.4.2 Dew Point Calculations...................................................................................... 31
2.4.3 Example 2.2 ...................................................................................................... 32
2.4.4 Example 2.3 ...................................................................................................... 34
2.4.5 Flash Calculations ............................................................................................. 36
2.4.6 Example 2.4 ...................................................................................................... 42
2.5 TUTORIAL 2..................................................................................................................... 44
2.6 BIBLIOGRAPHY ................................................................................................................ 51
APPENDIX A.......................................................................................................................... 52
De Priester Charts see next page............................................................................ 52
A1 High Temperature Chart: ...................................................................................... 53
A2 Low Temperature Chart: ....................................................................................... 54
A3 Antoine Coefficient Table ...................................................................................... 55
APPENDIX B.......................................................................................................................... 56
Rigorous Phi-Phi VLE Worksheet ............................................................................... 56
APPENDIX C ......................................................................................................................... 61
Review of Partial Molar Properties ............................................................................. 61
C-1 Partial Molar Property Overview .......................................................................... 62
C-2 Definition of Partial Molar Property ...................................................................... 64
C-3 Partial Molar Properties from Correlations ........................................................... 66
C-3 Partial Molar Properties from Diagrams ............................................................... 68
C-4 Recombination Rule ............................................................................................ 71
C-5 Gibbs-Duhem Equation....................................................................................... 72
C-6 Partial Molar Property Relations .......................................................................... 73

CONTENTS 6
TOPIC 3A: VLE FOR POLAR LIQUID MIXTURES

3A.1 INTRODUCTION ............................................................................................................... 3
3A.2 MODELS OF IDEALITY AND STANDARD STATES .................................................................. 4
3.2.1 Lewis-Randall Model of Ideality ........................................................................... 4
3.2.2 Henrys Law Model of Ideality .............................................................................. 5
3A.3 REAL MIXTURES ............................................................................................................. 7
3.3.1 Excess Properties ............................................................................................... 7
3.3.2 Activity Coefficients ............................................................................................. 8
3A.4 THE GAMMA-PHI APPROACH TO VLE ............................................................................. 12
3.4.1 Rigorous Gamma-Phi Approach to VLE ............................................................ 13
3.4.2 Vapour Phase as an Ideal Solution.................................................................... 14
3.4.3 Vapour Phase as an Ideal Gas .......................................................................... 15
3.4.4 Both Phases as Ideal Solutions ......................................................................... 15
3.4.5 Raoults Law...................................................................................................... 16
3A.5 PHASE EQUILIBRIUM DIAGRAMS ..................................................................................... 17
3.5.1 P-x-y Diagrams.................................................................................................. 18
3.5.2 T-x-y Diagrams .................................................................................................. 19
3.5.3 Example 3A.1 .................................................................................................... 22
3.5.4 The x-y Diagram ................................................................................................ 24
3.5.5 Example 3A.2 .................................................................................................... 26
3A.6 MODELS OF SOLUTION BEHAVIOUR ................................................................................ 28
3.6.1 The One Parameter Model ................................................................................ 28
3.6.2 Margules Two Parameter Model ....................................................................... 31
3.6.3 Van Laars Two Parameter Model ..................................................................... 32
3.6.4 Wilsons Local Composition Model .................................................................... 35
3.6.5 Example 3A.3 .................................................................................................... 36
3A.7 TUTORIAL 3A ................................................................................................................ 37
3A.8 BIBLIOGRAPHY .............................................................................................................. 40
APPENDIX A.......................................................................................................................... 41
Real Mixtures ............................................................................................................. 41
A.1 Property Changes on Mixing ................................................................................ 42
A.2 Property Changes on Mixing for Ideal Solutions ................................................... 43
A.3 Activity of a Species in a Mixture.......................................................................... 45
A.4 Excess Properties ................................................................................................ 47
APPENDIX B.......................................................................................................................... 49
B.1 Positive & Negative Deviation from Ideality .......................................................... 49
B.2 Formation of Azeotropes ...................................................................................... 52
APPENDIX C ......................................................................................................................... 54
Gamma-Phi VLE Worksheet for Azeotrope ................................................................ 54

CONTENTS 7
TOPIC 3B: VAPOUR-LIQUID EQUILIBRIUM, LIQUID-LIQUID EQUILIBRIUM

AND MISCIBILITY
3B.1 INTRODUCTION ............................................................................................................... 3
3B.2 PURE SPECIES ............................................................................................................... 4
3.2.1 P-T Diagrams ...................................................................................................... 4
3.2.2 P-v Diagrams ...................................................................................................... 6
3B.3 BINARY SYSTEMS AND MISCIBLE VLE .............................................................................. 7
3.3.1 P-T Diagrams ...................................................................................................... 8
3.3.2 P-x-y Diagrams.................................................................................................. 13
3.3.3 T-x-y Diagrams .................................................................................................. 14
3.3.4 x-y Diagrams ..................................................................................................... 15
3.3.5 Azeotropes ........................................................................................................ 15
3B.4 PARTIALLY MISCIBLE SYSTEMS ...................................................................................... 20
3.4.1 Liquid-Liquid Solubility Diagrams ....................................................................... 20
3.4.2 VLLE and P-x-y Diagram ................................................................................... 22
3.4.3 VLLE and T-x-y Diagram ................................................................................... 24
3.4.4 VLLE at Higher Temperature ............................................................................. 26
3.4.5 Quantitative Treatment of VLLE ........................................................................ 28
3.4.6 Example 3B.1 .................................................................................................... 34
3B.5 IMMISCIBLE SYSTEMS .................................................................................................... 36
3.5.1 Quantitative Treatment ...................................................................................... 37
3.5.2 Example 3B.2 .................................................................................................... 39
3B.6 TUTORIAL 3B ................................................................................................................ 40
3B.7 BIBLIOGRAPHY .............................................................................................................. 41
APPENDIX A.......................................................................................................................... 42
Rigorous Approach to LLE ......................................................................................... 42
A.1 Liquid-Liquid Equilibrium ...................................................................................... 43
APPENDIX B.......................................................................................................................... 46
Heating or Cooling a VLLE System. ........................................................................... 46
B.1 Isobaric Cooling from A ..................................................................................... 47
B.2 Isobaric Cooling from B ..................................................................................... 48
B.3 Isobaric Cooling from C ..................................................................................... 49
B.4 Isobaric Vaporisation from X .............................................................................. 50
TOPIC 4: CHEMICAL EQUILIBRIUM

4.1 INTRODUCTION .................................................................................................................. 3
4.2 REACTION STOICHIOMETRY ............................................................................................... 5
4.2.1 Extent of Reaction ............................................................................................... 5
4.2.2 Example 4.1 ........................................................................................................ 7
4.3 EQUILIBRIUM CRITERION .................................................................................................... 9
4.3.1 Equilibrium Constant ......................................................................................... 11
4.3.2 Approximate Temperature Dependence of K ..................................................... 14
4.3.4 Accurate Temperature Dependence of K........................................................... 17
4.3.5 Example 4.2 ...................................................................................................... 20
4.4 GAS PHASE CHEMICAL EQUILIBRIUM ................................................................................ 22

CONTENTS 8
4.4.1 Rigorous Non-Ideal Mixture ............................................................................... 24

4.4.2 Ideal Solution Mixture ........................................................................................ 24
4.4.3 Ideal Gas Mixture .............................................................................................. 25
4.4.4 Effect of Pressure on Chemical Equilibrium ....................................................... 25
4.4.6 Calculation of Equilibrium Conversion ............................................................... 26
4.4.7 Example 4.3 ...................................................................................................... 29
4.5 LIQUID PHASE CHEMICAL EQUILIBRIUM ............................................................................. 31
4.6 HETEROGENEOUS REACTIONS ......................................................................................... 36
4.6.1 Heterogeneous Reactions with Solids ............................................................... 37
4.7 TUTORIAL 4..................................................................................................................... 38
4.8 BIBLIOGRAPHY ................................................................................................................ 44
TOPIC 5: VOLUMETRIC BEHAVIOUR OF REAL FLUIDS AND SOLIDS

5.1 INTRODUCTION .................................................................................................................. 4
5.1.1 P-T Diagram ........................................................................................................ 4
5.1.2 P-v Diagrams ...................................................................................................... 6
5.1.3 Relation between P-T & P-v Diagrams ................................................................ 8
5.2 VOLUME EXPANSIVITY AND ISOTHERMAL COMPRESSIBILITY ................................................. 9
5.2.1 Liquids and Solids ............................................................................................. 10
5.2.2 Gases................................................................................................................ 11
5.2.3 Worked Example 5.1 ......................................................................................... 12
5.2.4 Worked Example 5.2 ......................................................................................... 13
5.2.5 Reversible Volume Work ................................................................................... 14
5.2.6 Worked Example 5.3 ......................................................................................... 15
5.3 COMPRESSIBILITY FACTOR .............................................................................................. 16
5.3.1 Two-Parameter Generalised Correlations.......................................................... 17
5.3.2 Three-Parameter Generalised Correlations ....................................................... 18
5.4 VIRIAL EOS .................................................................................................................... 21
5.4.1 Truncated Virial EOS ......................................................................................... 23
5.4.2 Worked Example 5.4 ......................................................................................... 24
5.4.3 Worked Example 5.5 ......................................................................................... 25
5.4.4 Worked Example 5.6 ......................................................................................... 26
5.4.5 Worked Example 5.7 ......................................................................................... 27
5.5 SECOND VIRIAL COEFFICIENT B ..................................................................................... 28
5.5.1 Experimental Determination .............................................................................. 29
5.5.2 Generalised Correlations ................................................................................... 30
5.5.3 Worked Example 5.8 ......................................................................................... 31
5.5.4 Worked Example 5.9 ......................................................................................... 32
5.5.5 Worked Example 5.10 ....................................................................................... 33
5.5.6 Worked Example 5.11 ....................................................................................... 34
5.6 CUBIC EOS .................................................................................................................... 35
5.6.1 Manual Iteration Schemes ................................................................................. 35
5.6.2 Generalised Correlations for Cubic Parameters................................................. 37
5.6.3 Worked Example 5.12 ....................................................................................... 38

CONTENTS 9
5.6.4 RKS EOS in Cubic Form ................................................................................... 40

5.6.5 Worked Example 5.13 ....................................................................................... 41
5.7 BIBLIOGRAPHY ................................................................................................................ 43

CONTENTS 10

TOPIC 1. Pure Species Phase Equilibrium 1
Topic 1
Pure Species Phase Equilibrium
Contents
TOPIC 1: PURE SPECIES PHASE EQUILIBRIUM
1.1 INTRODUCTION .................................................................................................................. 3
1.2 CRITERION FOR EQUILIBRIUM ............................................................................................. 4
1.2.1 Pure Species Phase Equilibrium ......................................................................... 4
1.2.2 Pure Species Chemical Potential ........................................................................ 8
1.2.3 Pure Species Gibbs Energy Relation................................................................... 9
1.2.4 Pressure Dependence of Pure Species Chemical Potential .............................. 10
1.3 PURE SPECIES FUGACITY ................................................................................................ 12
1.3.1 Fugacity as a Criterion for Pure Species Phase Equilibrium .............................. 14
1.3.2 Fugacity using Residual Volume ....................................................................... 15
1.3.3 Example 1.1 ...................................................................................................... 16
1.3.4 Fugacity using Compressibility Factors ............................................................. 17
1.3.5 Compressibility Factor Charts............................................................................ 17
1.3.6 Example 1.2 ...................................................................................................... 20
1.3.7 Fugacity using Charts ........................................................................................ 21
1.3.8 Example 1.3 ...................................................................................................... 22
1.3.9 Fugacity of Pure Saturated Liquids .................................................................... 23
1.3.10 Fugacity of Compressed Liquid States ........................................................... 25
1.3.11 Example 1.4 .................................................................................................... 28
1.3.12 Fugacity Coefficient using Cubic EOS ............................................................. 29
1.4 TUTORIAL 1..................................................................................................................... 30
1.5 BIBLIOGRAPHY ................................................................................................................ 32
APPENDIX A.......................................................................................................................... 33
Simple Gas Z-Factor Chart......................................................................................... 33
Non-Simple Gas Z-Factor Chart ................................................................................. 34
APPENDIX B.......................................................................................................................... 36
Simple Gas Fugacity Coefficient Chart High Pressure............................................. 36
Non-Simple Gas Fugacity Coefficient Chart High Pressure ..................................... 37
Simple Gas Fugacity Coefficient Chart Low Pressure ............................................. 38
Non-Simple Gas Fugacity Coefficient Chart Low Pressure ...................................... 39
APPENDIX C ......................................................................................................................... 40
Fugacity Coefficient using Cubic EOS ........................................................................ 40

Prerequisite knowledge
Process Industries C.
Process Engineering B.
Kinetics and Thermodynamics
Learning objectives
After studying this Pure Species Phase Equilibrium topic you should be able
to:
Write down phase equilibrium criterion for a pure species.
Identify and differentiate between the potentials for heat transfer, work
transfer and interphase mass transfer.
Define (for a pure species) the chemical potential, the fugacity and the
fugacity coefficient and write down relationships between them.
Show how fugacity may be determined from measurable P v T
data.
Explain the background to Corresponding States Principle and
compressibility factor. Outline and distinguish between the two-
parameter and the three-parameter Law of Corresponding States.
Review and solve problems using generalised compressibility factor
charts.
Demonstrate how generalised compressibility charts may be used to
determine the fugacity coefficient and fugacity knowing only reduced
temperature, reduced pressure and acentric factor.
Show how the fugacity coefficient and the fugacity may be calculated
for both a saturated and a subcooled liquid.

1.1 Introduction
In previous courses (Process Engineering B) a general chemical reaction was

written as follows:
aA + bB cC + dD
At chemical equilibrium the concentration of each species [ A], [B ], [C ], [D ]

taking part in the reaction was related to the equilibrium constant K through
the expression
K=
[C ]c [D]d
[A]a [B]b
This provides a criterion for chemical equilibrium but it only applies (in this
form) to an ideal system. Later it will be shown how this may be adapted and
applied to non-ideal systems.
What is also important is the degree to which a given species will transfer from
one phase to another. This type of problem is known as phase equilibrium.
Important operations, such as distillation, absorption and extraction, are
concerned with the transfer of species between different phases.
The rate at which a species is transferred from one phase to another depends
on how far the system is from equilibrium. This degree of departure from
equilibrium is usually referred to as the driving force or simply the potential.
Far from equilibrium the driving force (potential) is high and the rate of
transfer of species between phases is high.
Close to equilibrium the driving force (potential) is low and the rate of
transfer of species between phases is correspondingly low.
As equilibrium is approached, the rate of transfer of species between
phases becomes slower as the potential is reduced.
Phase equilibrium implies zero driving force (zero potential), zero
transfer of a species between phases and no change in composition of
the different phases.
Thus, equilibrium is approached asymptotically since it is a theoretical
entity that requires the potential to be zero.
In science equilibrium means that there are no changes. However, in

engineering phase equilibrium is assumed when no further changes in phase
composition are measurable.
Liquid-vapour equilibrium systems are the most common, but other phase
equilibrium situations are possible, such as liquid-solid, solid-solid, vapour-
solid and even vapour-liquid-solid and vapour-liquid-liquid systems.

1.2 Criterion for Equilibrium
For mechanical equilibrium there can be no pressure gradient (this is the

potential for work transfer). For thermal equilibrium there can be no
temperature gradient (this is the potential for heat transfer). But, for phase
equilibrium:
What is the potential that drives interphase mass transfer?
What is the criterion for phase equilibrium?
How can this criterion be used to solve phase equilibrium problems,
especially for non-ideal systems?
The potential that drives interphase mass transfer is not immediately

obvious. However, it is only when this potential approaches zero that phase
equilibrium can be assumed.
There are two distinct problems that need to be solved in order to understand
phase behaviour:
1. Why do different phases exist?
2. How do we deal with mixtures and changes in composition?
Although this subject deals principally with mixtures, it is better to start with a
discussion of phase equilibrium involving a single species. A common example
would be a steam-water system (but equally an ice-water, or an ice-steam
system are also possible, as indeed is an ice-steam-water system).
1.2.1 Pure Species Phase Equilibrium
Consider a closed system with two phases but only one component (this
could be a steam-water system).
Take it that both phases within this closed system are at constant temperature
and pressure, T & P , and that any heat or work, Q or W , that is transferred
between the system/surroundings takes place reversibly.
Initially, the system is not in phase equilibrium and net mass transfer of the
component is free to take place in either direction. The total entropy change of
the surroundings dS SURR , units of (kJ/K), is given by
dQ SURR
dS SURR =
TSURR
But since heat transfer (either to or from the surroundings) can only come from
the system, it follows that the heat exchanged between the system dQSYS and
surroundings dQSURR must be equal and opposite.

Which may be written as follows:
dQSURR = dQSYS
Also, since the heat transfer is reversible, it follows that heat can only be
transferred through an infinitesimal temperature difference. Thus, the system
temperature T must be equal to surroundings temperature TSURR .
Therefore, the total entropy change of the surrounding dS SURR becomes
dQSYS
dS SURR =
T
The second law is given by (Process Engineering B)
dS SYS + dS SURR 0
Substitute the entropy change of the surroundings into the second law to get
dQSYS
dS SYS 0
T
or
dQSYS TdS
Notice that the surroundings have now been eliminated. The non-flow energy
equation (Process Engineering B) is given by
dQSYS = dU + PdV
In which U is the total system internal energy (kJ) and V is the total system
volume (m3). Here, internal energy and volume are both extensive properties.
Substitute the last equation into the previous one and re-arrange
dU + PdV TdS 0
If T & P remain constant this may be written
dU T , P + d (PV )T , P d (TS )T , P 0
Which is identical to
d (U + PV TS )T , P 0 ..(1.1)

The Gibb's free energy (Process Engineering B) is defined as
G = H TS
In which H is the total system enthalpy (kJ) and G is the total system Gibbs
free energy (kJ). Enthalpy (Process Engineering B) is defined as
H = U + PV , so that the above may also be written as
G = U + PV TS
Substitute this last expression into equation (1.1) leads to
dGT , P 0 ....(1.2)
Equation (1.2) is used here to describe the situation where there are two
phases of a pure substance, together forming a closed system, but initially not
in phase equilibrium:
It shows that (for a spontaneous process) the total Gibbs energy of a
system at constant temperature and pressure always gets smaller, or
stays the same. It can never get larger.
The inequality dG < 0 applies to irreversible processes when the
system is not in equilibrium. This is the direction for spontaneous
change.
The equality dG = 0 applies to reversible processes and only occurs
when the system is at equilibrium.
Thus, the system always moves in a direction which minimises its Gibbs
energy and the point of equilibrium must, therefore, occur at the minimum
Gibbs energy. Mathematically equilibrium occurs whenever equation (1.3)
below is satisfied
dG = 0 ...(1.3)
For the case of a pure species distributed between two phases, call them the
-phase and the -phase (for instance, this could be a steam ( ) water ( )
system), the total Gibbs energy G of both phases together is given by
G = n g + n g (1.4)
In which g and g are the molar Gibbs energies (kJ/mol) of each phase
separately (intensive properties) and n & n are the number of moles of
each phase. Substitute this into equation (1.2) gives
d (n g + n g ) 0

After expanding out the differential this becomes
g dn + n dg + g dn + n dg 0
At this stage it is worth remembering that the temperature and pressure are
both constant. Under these conditions, what will happen to dg and dg ?
From a previous course (Process Engineering B) the fundamental property

relation was derived
dg = vdP sdT ..(1.5)
In which the volume v is the molar volume (m3/mol) and the entropy s is the
molar entropy (kJ/(mol K)) both intensive properties.
If T & P are constant, then equation (1.5) shows that there can be no change
in molar Gibbs energy of either phase and the phase equilibrium expression
reduces to
g dn + g dn 0
But, since there are only two phases present dn = dn and this becomes
g dn g dn 0
For a pure species which initially is not in equilibrium but is distributing

between two phases, this becomes
(g g )dn 0 ..(1.6)
If the system consists of two phases & and if g > g (while not in
equilibrium initially), then it follows that:
The first term g dn must be larger than the second term g dn .
In order to satisfy the inequality, the change in the number of moles in
the - phase must be less than zero, i.e. dn < 0 .
This means that the pure species must be transferred from the -
phase to the -phase.
Equation (1.4) shows that the total Gibbs energy of the system must
decrease as the system approaches equilibrium.
The transfer will continue until equation (1.6) is satisfied. This
corresponds to the point of equilibrium and minimum total Gibbs
energy.

The opposite will happen if g > g . However, if the molar Gibbs energies of
both phases are equal g = g , then equation (1.6) shows that the equality is
satisfied. This is the reversible case and the system must be in equilibrium.
Thus, the potential for interphase mass transfer is the molar Gibbs free
energy and when the condition below is satisfied, according to equation (1.6),
net interphase mass transfer must cease
g = g ....(1.7)
Only when this potential (driving force) is zero g g = 0 , will net mass
transfer stop and only then can the two phases co-exist in equilibrium.
Water and steam at 101.325 kPa have exactly the same molar Gibbs energy,
but only at 100oC. This means that pure water and steam can only co-exist in
equilibrium at 100oC when the pressure is 101.325 kPa.
This potential is the so called chemical potential and, for a pure species, it
is denoted by .
1.2.2 Pure Species Chemical Potential
In a previous course (Process Engineering B) it was shown that, for a pure

species, the chemical potential is equal to the molar Gibbs energy. The
chemical potential, like all driving forces, must be intensive (just like T & P ).
Thus, for a pure species
= g .(1.8)
In order to make the chemical potential and Gibbs free energy compatible, the
intensive versions of the system properties are used, that is g (kJ/mol), h
(kJ/mol) and s (kJ/(mol K)), which leads to
= g = h Ts
In summary, if any two phases & (involving a pure substance) are to co-
exist in equilibrium, then the chemical potential (molar Gibbs free energy)
must be the same in each phase
= ...(1.9)a
g = g (1.9)b

If the pressure is specified, then this condition is satisfied at some fixed

temperature and vice versa. Only one of these two properties can be varied
(the other is automatically fixed by the above phase constraint). Hence, the
system only has one degree of freedom.
Thus, for any phase change (such as any point on the freezing,
sublimation or vaporisation curves) the chemical potential change and
the Gibbs free energy change g between the co-exiting phases must be
= 0 (1.10)a
g = 0 (1.10)b
Comparing Gibbs free energy and entropy:
Situation Gibbs Free Energy Entropy

Equilibrium State: (G )T , P =0 (S )TOTAL = 0 (Reversible)
Non-Equilibrium State: (G )T , P <0 (S )TOTAL > 0 (Irreversible)
Any Change: (G )T , P 0 (S )T , P 0 (Any process)
The advantage of using the Gibbs free energy is that only system properties
are needed. With entropy, the entropy change of the surroundings is also
needed.
1.2.3 Pure Species Gibbs Energy Relation
Equation (1.5) was taken directly from Process Engineering B, but, due to its
importance, a quick review of its derivation follows. The total Gibbs energy is
defined as
G = H TS
dG = dH TdS SdT
The total enthalpy is defined as
H = U + PV
dH = dU + PdV + VdP
Substitute dH into the previous equation gives
dG = dU + PdV + VdP TdS SdT

But, the first law for a reversible process may be written as
dU = dQ + dW = TdS PdV
Substitute this last equation into the previous one to get
dG = TdS PdV + PdV + VdP TdS SdT
After cancellation this leads directly to equation (1.5) and shows that Gibbs
free energy fundamentally depends on T & P as follows:
dG = VdP SdT
Given that dG is an exact differential, it is possible to write the extensive form

of the relationship between the isothermal pressure-dependence of total Gibbs
free energy and total system volume, since SdT cancels
G
= V ...(1.11)
P T
Equation (1.5) may also be written in terms of molar properties as follows:
dg = vdP sdT .(1.12)
And, this in turn leads the intensive version of equation (1.11)
g
= v .(1.13)
P T
1.2.4 Pressure Dependence of Pure Species Chemical Potential
According to equation (1.8) the molar Gibbs free energy is the chemical
potential, so that equation (1.13) may also be written as

= v .(1.14)
P T
In which it is clear that the isothermal pressure-dependence of the pure

species chemical potential is equal to the molar volume (m3/mol). Consider the
isothermal variation of molar Gibbs energy g (chemical potential) with
pressure, as given by equation (1.12)
d = vdP ...(1.15)

Integrate the above expression, at constant temperature, between some very

low pressure state P * and any higher pressure state P
d = vdP
* P*
P
* = vdP
P*
It is clear that as P * 0 , v and the integral on the right tends to infinity.

Thus, if is finite at pressure P , then the low pressure value for chemical
potential * , as P * 0 .
The variation in chemical potential versus pressure is also highly non-linear.

Both these features make the chemical potential badly behaved
(mathematically speaking) at low pressure.
The value of properties as pressure approaches zero is very important as

gases become ideal at this low-pressure limit. Because of the very non-linear
behaviour and the large negative values (as pressure approaches zero)
chemical potential is difficult to use some other variable would be better.
A good idea would be to seek an auxiliary property, one that depends directly
on but, because of the way the auxiliary property is defined, behaves
differently at low pressure.
A place to start looking for such an auxiliary property is to examine the

isothermal variation of for an ideal-gas. If equation (1.15) is true for any
substance in any state, then it must also be true for an ideal gas
dg ig = v ig dP
The superscript ig is to denote that the substance being considered is an

ideal gas. If the fluid is an ideal gas, then the ideal gas Equation of state, EOS,
may be substituted into the above to give
RT
d ig = dP ..(1.16)
P
Or, after integration
ig = RT ln P + ...(1.17)

1.3 Pure Species Fugacity
The previous equation may be applied to a real fluid, so long as the real
pressure P is replaced by some new property which allows the equation to
hold for real fluids (not just ideal gases).
This new property is given the symbol f , has the same units as pressure
(kPa) and is called fugacity. Thus, for a real fluid the chemical potential is
given by
= RT ln f + ..(1.18)
The fugacity f may be considered an auxiliary property of the chemical

potential . Subtract equation (1.17) from equation (1.18), and cancel out the
temperature-dependent constant of integration to get
f
ig = RT ln
P
( )
The term on the left ig is the isothermal change in chemical potential,
between real fluid and ideal gas, and this is related to the ratio f / P . This
expression may also written in differential form as follows:
( )
d ig = RT d ln ...(1.19)
Where, the ratio f / P is called the fugacity coefficient defined as
f
= ...(1.20)
P
Integrating equation (1.19) along an isothermal path from some low pressure
state to a higher pressure leads to
( ) = RT ln + ..(1.21)
ig
( )
At P = 0 , corresponding to the ideal gas state, ig = 0 , the constant of
integration will be zero. This means that at P = 0 the fugacity coefficient
must be = 1 and f = P .
Thus, equation (1.21) becomes
( ) = RT ln ..(1.22)
ig

The term on the left is the difference between the chemical potential of a real
fluid and that of an ideal gas at the same temperature and pressure.
As the pressure P 0 , the chemical potential must approach its ideal gas
( )
value ig 0 , in which case 1 and f P .
This completes the second part of the definition where the fugacity approaches
the system pressure as the pressure approaches zero. Thus, for any ideal gas
(in the limit of zero pressure)
( ) = 0 ..(1.23)
ig
ig = 1 (1.24)
f ig = P ..(1.25)
Summarising: f is an auxiliary property of . Changes in are tied to

changes in f through equation (1.18). However, because = 0 , it follows
that f P as P 0 .
Fugacity fulfils the same function as chemical potential, but is better

behaved. At low pressure f P , whereas at low pressure the chemical
potential tends to minus infinity.
The fugacity coefficient was defined by equation (1.20) as follows:
f
=
P
However, due to equation (1.25) the fugacity coefficient of a pure species

could equally well be defined by
f
= ig
..(1.26)
f
Notice, is a correction factor that corrects the fugacity f between the ideal
gas and real fluid state at the same temperature and pressure.
In exactly the same way as the compressibility factor Z , introduced in Y2

(Process Industries C), corrects the molar volume v between the ideal gas
state and the real fluid state at the same temperature and pressure.
Thus Z corrects the molar volume and corrects the fugacity between the
ideal gas reference and the real fluid state.

1.3.1 Fugacity as a Criterion for Pure Species Phase Equilibrium
For a pure species the criterion for phase equilibrium (in terms of chemical
potential) between solid S , liquid L and vapour V phases is expressed by an
equation, similar to equation (1.9a), as follows:
S = L = V ...(1.27)
For instance, this could be applied at the Triple Point for any pure substance.
Simpler expressions (but now for only two co-existing phases) could be used
along the vaporisation, freezing and sublimation curves.
Re-arrange equation (1.22) to get
= ig + RT ln
And then substitute this into equation (1.27) as follows:
( ig
) ( ) (
+ RT ln S = ig + RT ln L = ig + RT ln V )
However, since the temperature and pressure is the same for all phases this
becomes
ln S = ln L = ln V
S = L = V .(1.28)
f S
= f L
= f V ..(1.29)
Thus, for a pure species in phase equilibrium, equation (1.27) states that the
chemical potential must be the same in all phases.
Again, for a pure species in phase equilibrium, equation (1.29) states that the
fugacity must be the same in all phases. This is the criterion for equilibrium
that is used in practice because it is easier to work in term of fugacity.
When trying to find numerical values for the fugacity f it is better to first find
the fugacity coefficient (it is dimensionless). Once this value has been
found then the fugacity f may be recovered from (1.26).
This is exactly analogous to molar volume v . It is better to find the

compressibility factor Z first (it is dimensionless) and then determine the
molar volume v from the definition of Z .

1.3.2 Fugacity using Residual Volume
Equation (1.15) shows how the chemical potential of a real fluid varies with
pressure at constant temperature in accordance with
d = vdP
Likewise, for an ideal gas, it is possible to write
d ig = v ig dP
Subtracting the last equation from the previous one leads to
( ) ( )
d ig = v v ig dP
Integrate the previous equation from P = 0 , where ig = 0 to any other

higher pressure, where ig 0 , leads to
ig = (v v ig )dP .(1.30)
P
The term on the right v v ig is known as the residual volume (m3/mol) and in
many texts is given the symbol v R . It is just the difference between the real
fluid molar volume and the ideal gas molar volume at the same T & P .
The residual molar volume is sometimes also expressed by the symbol

leading to the integral on the right which was used in past papers
P
ig = dP
0
However, substitute (1.30) into equation (1.22) and set v v ig = v R = (which

is the more usual notation) to get
P
1
RT 0
ln = v R dP .(1.31)
Given P v T data, it is possible (using the above expression) to find a

numerical value for and then, through equation (1.26), a numerical value
may be found for the real fluid fugacity f .

1.3.3 Example 1.1
Estimate the fugacity of gaseous propane at 2.4 MPa. and 344 K using the
following P v T data for propane at 160o F (complete this example by hand
in the workbook).
Molar Mass of propane is 44.097 kg/kmol
P v v ig vR P v R P
(MN/m2) (m3/kg) (m3/kg) (m3/kg) (kN/m2) (kJ/kg)
0 --- - - - -
0.1724 0.3691
0.3447 0.1810
0.6895 0.08652
1.0342 0.05512
1.2066 0.04598
1.3790 0.03903
2.4000 0.01783
(v )
1 R
P
RT
Estimate of fugacity coefficient at 24 bar and 344 K =
Estimate of fugacity at 24 bar and 344 K f = (kPa)

1.3.4 Fugacity using Compressibility Factors
Equation (1.31) may also be written as follows
(v v )dP ..(1.32)
P
1
ln = ig
RT 0
P
v v ig
ln = dP
0 RT RT
Pv Pv ig dP
P
ln =
0
RT RT P
But the compressibility factor Z is defined as Z = Pv / RT (remember v is

the molar volume given by v = V / n) in which case the equation above may
also be written as follows:
P
ln = (Z 1)
dP
..(1.33)
0
P
1.3.5 Compressibility Factor Charts
If experimental P v T data is available, then the fugacity may be estimated

either using either equation (1.31), in terms of residual volume v R , or equation
(1.33), in terms of compressibility factor Z .
If no experimental data is available, compressibility factors may be obtained

from Z -factor charts or tables. This approach was previously discussed in
Broadly there are two types of Z -factor correlations that have been
developed:
1. Two-Parameter correlations.
2. Three-Parameter correlations.
A two-parameter Z -factor correlation takes the general form
Z = Z (Tr , Pr )

The reduced temperature Tr (dimensionless) and the reduced pressure Pr

(also dimensionless) are the two parameters and these have already
previously (Process Industries C) been defined as follows:
T
Tr = ..................................................................(1.34)
TC
P
Pr = ..(1.35)
PC
Where,
T is absolute temperature of the system (K).
TC is absolute critical temperature of the fluid (K).
P is absolute pressure of the system (kPa).
PC is absolute critical pressure of the fluid (kPa).
Two-parameter correlations are based on the two-parameter Law of

Corresponding States (also known as the Corresponding States Principle
CSP):
All fluids at same Tr and Pr deviate from ideality to the same extent
and, therefore, have the same compressibility factor Z .
The above statement holds for simple gases with symmetrical charge
distribution (such as noble gases Ar, Kr and Xe).
For other non-simple gases, a three-parameter correlation is needed. A

three-parameter generalised correlation relates some property to reduced
temperature Tr , reduced pressure Pr and a third parameter called the
acentric factor .
Thus, a three-parameter generalised correction for the compressibility factor

takes the general form
Z = Z (Tr , Pr , )
Three-parameter correlations are based on the three-parameter Law of

Corresponding States:
All fluids with the same acentric factor and at the same Tr and Pr
deviate from ideality to the same extent and, therefore, have the same
compressibility factor Z

The acentric factor is easily determined by a single vapour pressure

measurement P sat at a specified reduced temperature of Tr = 0.7 :
= log10 ( Pr sat ) T =0.7 1 .(1.36)

r
And,
P sat
Prsat = ....(1.37)
PC
For simple gases (Ar, Kr, Xe) the acentric factor is zero. The three-parameter
correlations are more accurate than two-parameter correlations and have a
good accuracy over a full range of states (particularly when using tables).
Three-parameter graphical correlations use two charts: the Z 0 component

chart (sometimes written as Z ( 0 ) ); the Z 1 component chart (sometimes written
as Z (1) ). These two component parts of Z are combined in a simple way:
Z = Z 0 + Z 1 ...(1.38)
Charts for the Z 0 and Z 1 components are presented in Appendix A:

Z 0 is the major component of Z due to simple gas behaviour.
Z 1 is the minor adjustment to Z due to non-simple behaviour.
The superscripts are used to identify the component part of the compressibility
factor; note well, they are NOT powers of Z .

1.3.6 Example 1.2
Estimate the fugacity of gaseous propane at 2.4 MPa. and 344 K. Estimate the
compressibility factor Z (at each Pr ) from the charts in Appendix A. Complete
this example by hand in this workbook using the following critical and other
properties for propane:
Molar Mass 44.097 kg/kmol

Critical Temperature TC 369.8 K
Critical Pressure PC 42.48 bar
Acentric Factor 0.152
P Tr Pr Z0 Z 1 Z
(MN/m2)
0 0 1 0 1
0.1724
0.3447
0.6895
1.0342
1.2066
1.3790
2.4000
P Z 1 P Z 1
P
(MN/m2)
P
(kN/m2) P
(kN/m2)-1
0 0 -
0.1724
0.3447
0.6895
1.0342
1.2066
1.3790
2.4000
Z 1
P
P
Estimate of fugacity coefficient at 24 bar and 344 K =
Estimate of fugacity at 24 bar and 344 K f = (kPa)

1.3.7 Fugacity using Charts
Equation (1.33) may also be written in terms of reduced temperature and

reduced pressure as follows:
P = Pr PC dP = dPr PC
T = Tr TC dT = dTr TC
Pr
dPr
ln = (Z 1) P
0 r
..(1.39)
Tr
Substitute equation (1.38) into the above to get
( )
Pr Pr
dP dPr
ln = Z 1 r
0
+ Z1 (1.40)
0
Pr 0
Pr
Tr Tr
Recognise that the two integrals on the RHS are simply component parts of
ln and, labelling them ln 0 and ln 1 , leads to
ln = ln 0 + ln 1
( )( )
= 0 1 ...(1.41)
The integrals in equation (1.40) have been evaluated as functions of reduced

temperature and reduced pressure and are also available in the form of
charts/tables.
See Appendix B to view these three-parameter fugacity coefficient charts. In

addition, tabulations are also available (Smith, Van Ness, Abbott, 2005).

1.3.8 Example 1.3
Estimate the fugacity of gaseous propane at 2.4 MPa. and 344 K using the
charts supplied in Appendix B. Complete this example by hand in this
workbook using the following critical and other properties for propane:
Molar Mass 44.097 kg/kmol

Critical Temperature TC 369.8 K
Critical Pressure PC 42.48 bar
Acentric Factor 0.152
Tr =
Pr =
0 =
1 =
Fugacity coefficient at 24 bar and 344 K =
Fugacity at 24 bar and 344 K f = (kPa)

1.3.9 Fugacity of Pure Saturated Liquids
P v and P T diagrams may be linked to each other. Notice how the

saturated vapour curve on the P T diagram (right hand) corresponds to the
two-phase envelope on P v diagram (left hand). The P v diagram only
shows the vapour/liquid region, the solid region is excluded.
The saturated vapour pressure curve terminates at the upper end at the critical
point, labelled C . At the lower end it terminates at the Triple Point, labelled
TP .
CONSTANT T P
P C
C
4 4
3 2 LIQUID
P sat 2&3
1 1
VAPOUR
A B SOLID TP
v T
The fugacity is determined is determined at constant temperature. Points 1-2-

3-4 shows one such isotherm:
If P < P sat , then fluid is a single-phase vapour see state 1.
If P > P sat , then fluid is a single-phase liquid see state 4.
If P = P sat fluid state is anywhere within the two-phase envelope, i.e.
anywhere between state 2 and state 3.
State 2 corresponds to a saturated vapour state (dryness fraction
one) where no liquid is present.
State 3 corresponds to a saturated liquid state (dryness fraction zero)
where no vapour is present.
Between state 2 and state 3 the dryness fraction varies continuously
from one to zero.
Within two phase envelope 2 3 the state is continually changing but both
T & P are constant. From point 3 to point 4 notice the steepness of the
isotherm. This is the subcooled liquid region and liquids are nearly
incompressible.

As he pressure is increased at constant temperature T (from zero pressure

below point 1), the fugacity increases. When the vapour becomes saturated
at point 2, the fugacity of the saturated vapour may be found from P v T
data using the usual integral:
fV
P2
(Z 1) P
dP
ln V =
P State 2 0
Further slight compression leads to a very small quantity of liquid being

formed. The two phases split into a small amount of saturated liquid (state 3)
and a much larger quantity of saturated vapour (state 2).
These two phases are at the same temperature and pressure and are in
equilibrium. Equation (1.5) shows that dg = 0 (along any tie-line). Integrating
between saturated liquid and saturated vapour states leads to
g 2V g 3L = 0
g 2V = g 3L .....(1.42)
There is always only two equilibrium phases co-existing (just different amounts
of each) and the molar Gibbs energy is the same for each of the two phases.
In view of equation (1.8) it follows immediately that
2V 3L = 0
2V = 3L ....(1.43)
Integrating the relationship d = RTd ln f (the differential form of equation

(1.18)) between the same two ends stated leads to
f 2V f 3L = 0
f 2V = f 3L ....(1.44)
Finally dividing equation (1.44) by the pressure (which is the same for both
saturated and saturated vapour states) leads to
V2 3L = 0
V2 = 3L ....(1.45)

At any point along a horizontal tie-line (linking saturated liquid, point 3, and
vapour, point 2, states) T & P remains constant. This means that there is no
change in Gibbs energy, chemical potential, fugacity or fugacity coefficient
along a tie-line.
Thus, once the fugacity coefficient of the saturated vapour state is known (by
any of the methods discussed previously), it follows that the fugacity coefficient
of the saturated liquid state must have the same numerical value.
When a phase change occurs the fugacity is calculated in three separate steps
as follows:
1. The fugacity of the saturated vapour is found, at the target saturation
temperature, by integrating volumetric data from zero pressure to the
vapour pressure.
2. The fugacity of the saturated liquid in equilibrium, making use of
equation (1.44), must be the same.
3. The fugacity of the compressed (or subcooled) liquid can then be
estimated using the method given in the section below.
Charts may be extended to cover the liquid region but, because liquids are
harder to compress (solids are nearly incompressible), properties such as
molar volume and fugacity, for liquids, are only weak functions of pressure.
The analytical method outlined below is a better approach.
1.3.10 Fugacity of Compressed Liquid States
Equation (1.15) was derived before
d = vdP
And from the differential form of equation (1.18)
d = RTd ln f
Equating these two leads to
v
d ln f = dP
RT
Integrating (along an isotherm) from P sat to any higher pressure P (now in

the compressed liquid region)
f P
1
d ln f =
sat RT v L dP
f P sat

The left side of this equation then becomes
P
f 1
ln
f sat
=
RT v L dP
P sat
Notice how the lower limit of integration is P sat and not zero pressure, so that
there is no difficulty in terms of the right hand integral tending to infinity (liquid
molar volumes are finite at the saturation pressure).
Also, liquid molar volume v L is a weak function of P and, as a first-order

approximation they may be regarded as constant (this is often taken to be the
saturated liquid volume v Lsat ).
P
f v Lsat
ln
f sat

RT dP
P sat
or
ln
f
=
(
v Lsat P P sat )
.(1.46)
sat
f RT
Referring to the previous diagram the integration is performed between points

3 and 4 along the isotherm in the subcooled liquid region. Applying these
subscripts to equation (1.46) leads to
f 4 v Lsat (P4 P3 )
ln = ....(1.47)
f3 RT
Remember f 3 will be known from equation (1.44). This means that, if a

numerical value can be placed on the term on the right, then the fugacity f 4
(of the subcooled liquid) may be estimated using equation (1.47).
The above derivation assumes liquids at low/moderate pressure, where the

pressure, P << PC - otherwise the assumption of near constant v Lsat starts to
break down. Equation (1.47) is often written as either
fL v Lsat P P sat
= exp
( )

f sat RT
Or,
f L v sat P P sat
= sat P sat exp L
( )
.(1.48)
RT

L sat
In which f is the fugacity of the compressed liquid state at T & P and f
is the fugacity of the saturated liquid at T & P sat .
The exponential term on the right hand side of equation (1.48) is known as the
Poynting correction factor. It corrects the fugacity between the saturation
state f sat and the compressed liquid state f L .
Very often the Poynting correction factor adds little to the overall accuracy and
it is often taken to be unity (the dimensionless tem within the square brackets
is usually a very small number).
Setting the Poynting correction factor to unity is equivalent to ignoring the

pressure correction (between saturation and compressed liquid states). In this
case the liquid fugacity becomes f L = sat P sat = f sat .
We will make use of this simplification in Topic 3A, VLE for Polar Liquid
Mixtures.

1.3.11 Example 1.4
Estimate the fugacity of liquid propane at 4.8 MPa. and 344 K using the same
gaseous P v T data for propane at 160oF (344 K) as provided in Example
(1.1). The vapour pressure of propane at 344 K is 2.56 MPa.
Molar Mass of propane is 44.097 kg/kmol
P v v ig vR P v R P
(MN/m2) (m3/kg) (m3/kg) (m3/kg) (kN/m2) (kJ/kg)
0 --- - - - -
0.1724 0.3691
0.3447 0.181
0.6895 0.08652
1.0342 0.05512
1.2066 0.04598
1.3790 0.03903
2.4000 0.01783
2.5600
(v )
1 R
P
RT
Fugacity coefficient (saturated vapour) at 25.6 bar and 344 K =
Fugacity (saturated vapour) at 25.6 bar and 344 K f =
The specific volume and pressure data for the liquid (past the point of
saturation) is given below:
P v
10 3
(MN/m2) (m3/kg)
2.551 2.446
2.758 2.446
3.447 2.404
4.800 2.338
Fugacity (subcooled liquid) at 48 bar and 344 K f =
Fugacity coefficient (subcooled liquid) at 48 bar and 344 K =

1.3.12 Fugacity Coefficient using Cubic EOS
Please refer to Appendix C to see how the Peng-Robinson Equation of State

EOS may be used to find the fugacity coefficient of a pure species.
Also outlined in Appendix C is how to estimate the pure species vapour

pressure using an EOS. This approach is based on the fact that two co-
existing phases must have the same fugacity coefficient.
The material in Appendix C may be very useful for Design Projects or

Experimental Projects. Also, these methods are widely used in Process
Simulators to estimate thermodynamic properties of fluids.

1.4 Tutorial 1
1. Estimate the fugacity coefficient, and hence the fugacity, of ethylene at

50.5 bar and 24.95oC using the following data:
Molar mass of ethylene is 28.05 g/mol.
Table of P v T data for ethylene at 24.95oC
Pressure Specific Volume

(bar) ( 10 4 m 3 /mol )
1.0 245.00
5.1 47.80
10.1 23.20
15.2 15.00
20.2 10.80
25.3 8.34
30.3 6.66
35.4 5.44
40.4 4.52
45.5 3.78
50.5 3.17
Solutions:
= 0.73854
f = 37.30 bar
2. Estimate the fugacity of gaseous ethane, temperature constant at

275K, using the following methods:
a) At 24.6 bar assuming ideal gas behaviour.
b) At 24.6 bar using the Z factor method.
c) At 24.6 bar using the generalised fugacity coefficient charts.
Critical properties for ethane are given below:
PC = 48.72 bar
TC = 32.17 o C
= 0.100

Use the pressure increments and thermodynamic data provided:
Pressure
(bar)
0.0
1.0
2.0
4.0
8.0
12.0
14.8
19.7
24.6
Solutions:
a) f = 24.6 bar
b) f = 18.43 bar
c) f = 18.97 bar
3. Solid carbon dioxide at -59.11oC has a vapour pressure of 4.09918 bar

and specific volume of 0.000657 m3/kg. The molar mass of carbon
dioxide is 44.01 kg/kmol
Calculate the fugacity of solid carbon dioxide at 10 bar and -59.11oC

using the following the P v T data for gaseous carbon dioxide.
Pressure Specific Volume

3
(bar) (m / kg )
0.01020 33.36600
0.03061 13.02200
0.07483 5.32870
0.13605 2.92280
0.37415 1.05870
0.95236 0.41189
2.44897 0.15600
4.09918 0.09120
Solution:
f = 2.83 bar

1.5 Bibliography
J.R.Howell, R.O. Buckius. Fundamentals of Engineering Thermodynamics.

McGraw-Hill, SI ed., 1987.
M.J. Moran, H.N. Shapiro. Fundamentals of Engineering Thermodynamics.

Wiley 2nd ed., 1993.
R.E. Sonntag, C. Borgnakke. Fundamentals of Thermodynamics. Wiley, 5th

ed., 1998.
S.I. Sandler. Chemical and Engineering Thermodynamics. Wiley, 3rd ed., 1999.
Smith, J.M., van Ness, H.C., Abbott, M.M., Introduction to Chemical

Engineering Thermodynamics. McGraw-Hill. 7th ed.,2005.

Appendix A
Simple Gas Z-Factor Chart
This chart is compiled using the Benedict-Rubin-Webb (BWR) Equation of

State EOS as modified and improved by Lee-Kesler.

Non-Simple Gas Z-Factor Chart
This chart is also compiled using the Benedict-Rubin-Webb (BWR) Equation of

State EOS as modified and improved by Lee-Kesler.

Reproduced and scanned from Smith, Van Ness, Abbott, 5th. Ed.

Appendix B
Simple Gas Fugacity Coefficient Chart High Pressure

Non-Simple Gas Fugacity Coefficient Chart High Pressure

Simple Gas Fugacity Coefficient Chart Low Pressure

Non-Simple Gas Fugacity Coefficient Chart Low Pressure

Appendix C
Fugacity Coefficient using Cubic EOS
This Appendix will be useful since many process simulators find pure species
fugacity coefficients this way.
The PR EOS can also be used to find pure component fugacity coefficients
as follows:
1. Find pure component aT & b parameter from the three equations

given below; all that is needed are the critical properties of the pure
species (critical temperature, critical pressure and acentric factor).
RTC
b = 0.0778 .(C.1)
PC
2
R 2TC
aT = 0.45724 a (T ) ...(C.2)
PC
[ (
(T ) = 1 + (0.37464 + 1.5422 0.26992 2 ) 1 Tr 0.5 )]
2
..(C.3)
2. Convert these dimensional aT & b parameters into dimensionless

A & B parameters:
aT P
A= .................(C.4)
R 2T 2
bP
B= ...(C.5)
RT
3. Find the three roots of the PR EOS written below in standard cubic
form (notice that all terms in this equation are dimensionless):
Z 3 + ( B 1) Z 2 + ( A 2 B 3B 2 ) Z + ( AB + B 2 + B 3 ) = 0
..(C.6)

4. Select one of the three roots (see point 5 below) and substitute this
value of Z into expression below to find the pure species :
ln = (Z 1) ln(Z B)
aT ( )
Z + 1+ 2 B
ln
( )
2 2 bRT Z + 1 2 B
..(C.7)
5. The largest root should be selected when P < P sat . Equation (C.7)
then returns the vapour fugacity coefficient.
The smallest root should be selected when P > P sat . Equation (C.7)
then returns the liquid fugacity coefficient.
These equations are normally only used to find the fugacity coefficient of non-
polar/hydrocarbon liquids. For better liquid phase property prediction, special
adjustments are made to the some of the equations listed above.
For liquid and vapour phases to co-exist in equilibrium, equation (1.29) must
be satisfied as follows:
f L
= fV
Since both phases are at the same T & P this is the same as
L = V (C.8)
If the temperature T is fixed and the pressure P is varied and if both the
largest and the smallest roots are substituted into equation (C.7), then both
liquid and vapour phase fugacity coefficients L & V will be returned at the
prevailing temperature T .
These results could be plotted as two versus P curves one

corresponding to the pure vapour phase, the other to the pure liquid phase.
The point where the two curves intersect corresponds to equation (C.8).
Referring to point 5 above, when the two phases have the same fugacity
coefficient, the smallest root corresponds to the saturated liquid Z L , the
largest root to the saturated vapour Z V and P = P sat .
Hence it is possible, using this approach, to calculate the vapour pressure

Pi sat of a pure species at any temperature T - the temperature must lie
between the Triple Point TP and the Critical Point C of the species.

The vapour part of the versus P curve is virtual when P > P sat . The liquid
part of the versus P curve is virtual when P < P sat . Where the two curves
meet is the only point where the two phases can co-exist in equilibrium
together; this point corresponds to P = P sat .
For water at 100oC this point will be 101.325 kPa (the vapour pressure of
water at 100oC). Since this vapour pressure corresponds to 1 atm, it means
that the NBP of water is 100oC.



TOPIC 2: VLE for Non-Polar Mixtures 1
Topic 2
VLE for Non-Polar Mixtures
Contents
TOPIC 2: VLE FOR NON-POLAR MIXTURES
2.1 INTRODUCTION .................................................................................................................. 3
2.1.1 The Apportioning Rule, the Recombination Rule and Gibbs-Duhem.................... 3
2.2 CRITERIA FOR PHASE EQUILIBRIUM .................................................................................... 5
2.2.1 Gibbs Free Energy of a Mixture ........................................................................... 5
2.2.2 Daltons Law........................................................................................................ 7
2.2.3 Fugacity Coefficient of a Component in a Mixture................................................ 7
2.2.4 Ideal Solutions and Lewis-Randall Rule .............................................................. 8
2.2.5 Criterion for Multicomponent Phase Equilibrium ................................................ 10
2.2.6 Fugacity as a Criterion for Mixture Phase Equilibrium........................................ 12
2.2.7 Finding Fugacity Coefficient of a Component in a Mixture ................................. 13
2.3 VLE FOR NON-POLAR SYSTEMS ...................................................................................... 15
2.3.1 Rigorous Phi-Phi Approach to VLE .................................................................... 15
2.3.2 Liquid and Vapour as Ideal Solutions ................................................................ 17
2.3.3 Raoults Law: Liquid as an Ideal Solution and Vapour as an Ideal
Gas........................................................................................................................ 18
2.3.4 Phi-Phi VLE Summary ....................................................................................... 19
2.3.5 Finding K-Values using Charts .......................................................................... 20
2.3.6 Example 2.1 ...................................................................................................... 22
2.3.7 Vapour Pressure Correlation ............................................................................. 25
2.3.8 Temperature-Composition Diagram for Binary System ...................................... 27
2.3.9 Significance of K-Values.................................................................................... 28
2.4 PHASE EQUILIBRIUM CALCULATIONS ................................................................................ 29
2.4.1 Bubble Point Calculations .................................................................................. 29
2.4.2 Dew Point Calculations...................................................................................... 31
2.4.3 Example 2.2 ...................................................................................................... 32
2.4.4 Example 2.3 ...................................................................................................... 34
2.4.5 Flash Calculations ............................................................................................. 36
2.4.6 Example 2.4 ...................................................................................................... 42
2.5 TUTORIAL 2..................................................................................................................... 44
2.6 BIBLIOGRAPHY ................................................................................................................ 51
APPENDIX A.......................................................................................................................... 52
De Priester Charts see next page............................................................................ 52
A1 High Temperature Chart: ...................................................................................... 53
A2 Low Temperature Chart: ....................................................................................... 54
A3 Antoine Coefficient Table ...................................................................................... 55
APPENDIX B.......................................................................................................................... 56
Rigorous Phi-Phi VLE Worksheet ............................................................................... 56

APPENDIX C ......................................................................................................................... 61
Review of Partial Molar Properties ............................................................................. 61
C-1 Partial Molar Property Overview .......................................................................... 62
C-2 Definition of Partial Molar Property ...................................................................... 64
C-3 Partial Molar Properties from Correlations ........................................................... 66
C-3 Partial Molar Properties from Diagrams ............................................................... 68
C-4 Recombination Rule ............................................................................................ 71
C-5 Gibbs-Duhem Equation....................................................................................... 72
C-6 Partial Molar Property Relations .......................................................................... 75
Learning objectives
After studying this VLE for Non-Polar Mixtures topic you should be able to:
Write down definition for a partial molar property
Identify and differentiate nomenclature for mixture property, pure
component properties and partial molar properties.
Review partial molar properties and their importance in terms of
developing phase equilibrium strategies for non-polar systems.
Define the apportioning rule, the recombination rule and the Gibbs-
Duhem rule for partial molar properties.
For any pure component property relation, find the corresponding
partial molar property analogue.
Define the Lewis-Randall Rule and Daltons Law.
Derive the Gibbs free energy in terms of the intensive properties of the
system and then identify the chemical potential.
Define the fugacity coefficient of a component in a mixture and
demonstrate ways of obtaining numerical values.
Derive the criteria for multicomponent phase equilibrium in terms of
chemical potential and the fugacity of a component in a mixture.
Discuss the concept of ideal solutions and derive partial molar
properties of components in an ideal solution.
Define K value and demonstrate how these quantities may be used
to solve VLE problems using a variety of simplifying assumptions.
Demonstrate the use of de Priester charts.
Derive expressions for Raoults Law and Raoults Law K values.
Define and explain the Antoine vapour pressure correlation.
Explain and execute bubble/dew point and flash VLE calculations.
Outline the rigorous phi-phi approach to VLE and how problems may
be solved using an EOS and a maths package.

2.1 Introduction
Understanding phase equilibrium (involving multi-component mixtures) is a key

skill that must be mastered by Chemical Engineers. This topic shows how to
derive and solve phase equilibrium expressions to cover such systems.
The property that a component seems to have in a mixture is called a partial

molar property. A detailed review of partial molar properties is given in
Appendix C, what follows here are the basic partial molar property relations.
2.1.1 The Apportioning Rule, the Recombination Rule and Gibbs-Duhem
As shown in a previous module (Kinetics and Thermodynamics) any solution

property m may be apportioned between the various components in the
solution. For component i the apportioning rule is given by
(nm )
mi = ..(2.1)
ni T , P ,n j
Where,
mi is the partial molar property of species i in the mixture at T & P .
m is the overall solution property at the same T & P .
mi is the pure-component property of i at the same T & P .
n is the total number of mole of the mixture.
In equation (2.1) the partial molar property mi of species i is defined as the

partial derivative of the total solution property M = nm with respect to the
mole number of species i :
T & P must be held constant.
Constant n j means that the mole number of all other species except
the i th (the one that is being varied) must be held constant.
The partial molar property mi is generic, it could be either vi , u i , hi , s i or g i .

Other properties, such as chemical potential, fugacity and fugacity coefficient,
may also be apportioned between the components in the solution and have
their own respective partial molar analogues.
If properties can be apportioned, then it must be possible to mathematically

recombine them. Thus, it should be possible to recover some mixture molar
property m from it corresponding partial molar properties mi .

The recombination rule for partial molar properties is given by
n
m = xi mi .(2.2)
i =1
The partial molar properties are functions not only of T & P but also of
composition. If the partial molar properties can be recombined using equation
(2.2), then they cannot vary independently of each other.
The property relation that expresses this dependency is called the Gibbs-
Duhem equation. In its more general form (see Appendix C) it is given by
m m
P dP + T dT xi dmi = 0 (2.3)
T ,x P, x i
However, if T & P are held constant, then equation (2.3) reduces to the most
widely used form of the Gibbs-Duhem equation
x dm
i
i i = 0 ....(2.4)
This can also be expressed at constant T & P , for molar volume, as follows:
n dv = dV
i
i i
i
i =0
Hence, any change in the mole number of a species will change its extensive
partial molar volume so the above expression must sum to zero at equilibrium.
If T and/or P are changed, then this relationship may be used to find
changes in phase equilibrium composition.
In Appendix C it is also shown that for every pure component property relation,
such as, for instance the two well-known expressions below
g i = hi Ts i and, hi = u i + Pvi
There exits an identical relation between their corresponding partial molar

properties. Thus, it is possible to immediately write
g i = hi Ts i and, hi = u i + Pvi

2.2 Criteria for Phase Equilibrium
The Gibbs free energy is of fundamental importance in both phase and

chemical equilibrium and it needs to be examined closely
Any extensive property M may be expressed in terms of T & P and the mole
number of each species n1 , n 2 ......ni as follows:
M = f (T , P, n1 , n 2 ......ni )
In the same way, the total Gibbs free energy of a mixture G may also be
expressed in terms of T & P and n1 , n 2 ......ni as follows:
G = f (T , P, n1 , n2 ......ni )
2.2.1 Gibbs Free Energy of a Mixture
For a mixture the total differential dG is given by
G G G
dG = dP + dT + dni
P T ,n T P ,n i ni T , P , n
j
But, since the total Gibbs free energy is G = ng , then this may be written as
(ng ) (ng ) (ng )

d (ng ) = dP + dT + dni ..(2.5)
j
The first two partial derivatives in the above equation have been shown to be
nv and - ns (Process Engineering B) and this leads to
(ng )
d (ng ) = (nv )dP (ns )dT + dni ...(2.6)
i ni T , P , n
j
Compare the last term with the general definition of a partial molar property
from the apportioning rule, as expressed by equation (2.1) as follows;
(nm )
mi =
ni T , P ,n j

The last term in equation (2.6) is, therefore, g i
(ng )
gi = ..(2.7)
ni T , P , n j
Substituting equation (2.7) into equation (2.6) leads to
d (ng ) = (nv )dP (ns )dT + g i dni ..(2.8)

i
Divide through by n , the total number of moles of all species in the mixture
dg = vdP sdT + g i dxi ........(2.9)

i
Thus, the Gibbs free energy is fundamentally a function of the intensive

variables of the system, namely T , P & xi . Hence, Chemical Engineers can
alter phase equilibrium by changing any of these properties T , P & xi .
In the previous topic the relationship ( g i = i ) was shown to hold for pure
components. Thus, it follows that the partial molar analogue must be
g i = i ...(2.10)
In view of equation (2.7), the chemical potential of a component in a mixture

i must also be given by
(ng )
i = ..(2.11)
ni T , P , n
j
So that equation (2.9) may also be written as
dg = vdP sdT + i dx .......(2.12)

i
If i is the chemical potential of the pure species, then i is the chemical

potential of the species in a mixture at the same T & P . The meaning of the
circumflex over the chemical potential will become apparent later.
At constant T & P it will be shown that equation (2.8) leads to a general

criterion for Vapour Liquid Equilibrium (VLE). However, it is better to first
review some fugacity and fugacity coefficient relations.

2.2.2 Daltons Law
From earlier modules (Process Industries) the partial pressure of a species in

an ideal gas mixture p i is the pressure the species would exert if ni moles of
the species occupied the entire mixture volume V at the same temperature T
as the mixture.
ni RT
pi =
V
While, the total pressure P is related to the total number of moles n of all
species in accordance with the ideal gas EOS below
nRT
P=
V
Dividing the former expression by the latter leads to Daltons Law
p i ni
= = yi
P n
p i = Py i .....(2.13)
Equation (2.13) is known as Daltons Law. In many texts the partial pressure is
denoted by p i and the total pressure by P . However, the difference between
uppercase P and lowercase p i can lead to confusion.
Therefore, in this text the partial pressure will be referred to as y i P and

uppercase P will be reserved to denote the system (or total) pressure.
2.2.3 Fugacity Coefficient of a Component in a Mixture
The following expressions were derived in Topic 1:
fi
i =
f i ig
For an ideal gas igi = 1 and since at low pressure f P
f i ig = P

It follows that the fugacity coefficient of a component in a mixture

i and the
fugacity of a component in an ideal gas fi ig are given by their corresponding
partial molar analogues:

= f i (2.14)
i
fi ig
fi ig = y i P ..(2.15)
Substitute equation (2.15) into equation (2.14) to get an alternate definition for
i as follows:

i = f i ..(2.16)

Py i
Notice that, as P 0 the fugacity of a pure species approaches the system

pressure f i P , while the fugacity of a component in a mixture approaches
its partial pressure fi y i P
2.2.4 Ideal Solutions and Lewis-Randall Rule
Although often used for liquid phase mixtures, the ideal solution concept may
also be applied to gas phase mixtures. Pure-components, when mixed
together, form an ideal-solution whenever there is no heat or volume change
on mixing effects.
This means that the molar volume that a component appears to have in an
ideal solution is the same as its pure component molar volume
viid = vi = vi ...(2.17)
The superscript "id " is used to differentiate an ideal solution from an ideal
gas. These are different models and this should be clearly understood.
For a gas to behave as ideal gas the volume of the molecules, and the forces
of attraction between molecules, must both be negligible. On the other hand,
for a solution to behave as an ideal solution it is only necessary that:
1. Like and unlike molecules have similar size and shape.
2. Like and unlike molecules have similar forces of attraction.

Examples of ideal solutions are optical isomers and adjacent members of

homologous series, i.e. an n-hexane/n-heptane mixture.
Not many liquid mixtures behave as ideal solutions principally because, in the
liquid phase, the molecules are closer together and any difference in size,
shape and forces of attraction tends to be significant.
On the other hand, most gas mixtures at low to moderate pressure (although
not ideal gases) may behave as ideal solutions. This is because the molecules
are much further apart and any differences between like and unlike species
are less pronounced.
Notwithstanding, real solutions (even liquids) can behave as ideal solutions in

the limit of very high concentration. The question is: for an ideal solution, how
does the fugacity of a component in a mixture i behave?
In Topic 1 it was shown that the pure component fugacity coefficient may be
calculated by solving the integral on the right of the expression below:
(v v )dP
P
1
ln = ig
RT 0
It follows that the partial molar analogue of this expression is given by
(v )
P
i = 1
ln i viig dP ....(2.18)
RT 0
But for an ideal solution viid = vi . Substitute this into equation (2.18) to get the
fugacity coefficient of a component in an ideal solution
idi
(v )
P
1
ln
idi = i viig dP
RT 0
In addition, the partial molar volume of an ideal gas, at the same T & P as the
mixture, must be equal to its pure component value viig = viig . Thus, the
previous expression becomes
(v )
P
id = 1
ln viig dP .(2.19)
i i
RT 0

Compare this to the pure component fugacity coefficient i that was derived
earlier in Topic 1
(v )
P
1
ln i = i viig dP
RT 0
It is clear that the fugacity coefficient of a component in an ideal solution

idi
must be equal to its pure component value i

id = ....(2.20)
i i
Now, substitute the definition of

idi into the left and the definition of i into
the right and then re-arrange
fi id f
= i
yi P P
fi id = y i f i ....(2.21)
This important equation is known as the Lewis-Randall Rule and shows that
the fugacity of a component fi id depends on the pure component fugacity f i
and the composition y i .
Even for a non-ideal liquid, the Lewis-Randall Rule applies to a component in

the mixture, as the solution approaches purity in that component. For liquids it
is customary to denote compositions by xi otherwise the Lewis-Randall Rule
is the same.
2.2.5 Criterion for Multicomponent Phase Equilibrium
It has been shown that thermal equilibrium occurs when the potential for heat
transfer is zero, i.e. the temperature (the thermal potential) is the same either
side of the system boundary at this point heat transfer stops.
It has also been shown that mechanical equilibrium occurs when the potential
for work transfer is zero, i.e. when the pressure (the mechanical potential) is
the same either side of the system boundary at this point work transfer
stops.

It has also been shown that (for a pure species) phase equilibrium occurs
when the potential for mass transfer between the phases is zero, i.e. when the
chemical potential (the mass transfer potential) is the same in either phase:
At this point net interphase mass transfer stops and both phases co-
exist together in equilibrium.
If the pressure is fixed, this can only occur at one temperature (for
water at 1 atm pressure this temperature corresponds to the NBP of
100oC).
Hence, as phase equilibrium is approached, interface mass transfer

approaches zero. The criterion for pure species phase equilibrium was
developed in Topic 1. The derivation for multicomponent phase equilibrium
follows the same steps up to the point where it was shown that, for phase
equilibrium, dG = 0 .
The multicomponent system under consideration consists of phases and N

components (these are all contained within a closed system). In addition, all
the phases must be at a uniform T & P .
Following identical steps, as outlined in Topic 1, mixture phase equilibrium

between these phases and N components occurs whenever
dG = 0
After making the substitution g i = i into equation (2.8), the change in total
Gibbs energy G within any phase, denoted as the -phase, is given by

(
d (ng ) = (nv ) dP (ns ) dT + i dni )
i
Likewise for any other phase, denoted as the -phase, this becomes

(
d (ng ) = (nv ) dP (ns ) dT + i dni )
i
For the moment restrict the case to only two phases, and . In a VLE
system, for instance, could be vapour and could be liquid. This result will
be extended to all phases later).
Adding the above two equations together, at constant T & P , gives the total
change in Gibbs energy dG for both the -phase and the -phase
dG = d (ng ) + d (ng )


Thus, the total Gibbs free energy of both phases at constant T & P is given by
( )
dG = i dni + i dni ( )
i i
However, (for the moment) there are only two phases within this closed
system. Thus, any change in the number of moles of any species i in the -
phase must produce an equal and opposite change in the number of moles of
species i in the -phase.
dni = dni ...(2.22)
And if both phases are in equilibrium it is known that dG = 0 . Thus, for

equilibrium between these two phases
( )dn = 0
i
i i i
However, dni is arbitrary, so that the only way the above summation can
always be zero is when the chemical potential (of each species) in both
phases is the same at equilibrium.
i = i (i = 1,2......N ) (2.23)
This reasoning holds for any other pair of phases (taking care to keep one
phase out of each pair of phases the same).
This leads to the following general criterion for phase equilibrium between all
phases (the criterion must hold for each component in the mixture):
i = i = i = ........ i . (i = 1,2......N ) ...(2.24)
When this condition is satisfied net mass transfer of all species, between all
phases, ceases and the phases are all in equilibrium.
2.2.6 Fugacity as a Criterion for Mixture Phase Equilibrium
It has been shown that all relations between pure component properties have
their partial molar property analogues. Thus, the chemical potential of a
component in a mixture i must be given by
i = RT ln fi + ...(2.25)

In which is the temperature dependent constant of integration. Substitute

equation (2.25) into the general criterion for equilibrium, equation (2.24). After
cancelling, since all phases are at the same T & P , this becomes
fi = fi = ............ = fi (i = 1,2......N ) ..(2.26)
Equation (2.24) is the general criterion for phase equilibrium in terms of the
chemical potential. However it is difficult to apply since i is mathematically
badly behaved at low pressure.
Equation (2.26) is an equally general criteria for multicomponent phase

equilibrium, but now in terms of the fugacity of a component in a mixture fi :
fi behaves in a similar way as i but is easier to deal with.

is related to f through its definition - see equation (2.14) or
i i
equation (2.16)
i is related to measurable P v T data see below.
2.2.7 Finding Fugacity Coefficient of a Component in a Mixture
The link between i and measurable P v T data was found earlier to be

given by equation (2.18). This equation is shown again below:
(v )
P
i = 1
ln i viig dP
RT 0
Solving multicomponent phase equilibrium problems in principle (although not

in practice) is quite straightforward:
Solve equation (2.18), or similar, to find the fugacity coefficient of a
component in the mixture i for both phases.
Use the definition of
i , equation (2.14) or (2.16), to find the fugacity
of all the components in the mixture fi .
Adjust the intensive properties ( T , P, xi or y i ) until the condition for
phase equilibrium, equation (2.26), is satisfied.
In principle this is easy, but in practice

i are complex functions of the
intensive properties of the mixture ( T , P, xi y i ) and this complicates the
problem.

There are two general approaches:

1. Make simplifying assumptions that lead either to simpler algebraic
equations (with less complex functional dependency), or to charts
which are straightforward to use.
2. Retain the rigorous approach and use a Mathematics package to help
solve all the equations needed to satisfy the phase equilibrium
condition complex functional dependency is then not an issue.
The use of the circumflex above the fugacity coefficient, as in

i , is now
apparent. Compare the expression from Topic 1
P
1
ln =
RT 0
v R dP
With equation (2.18), it partial molar analogue
P
i = 1 viR dP
RT 0
ln
From the first equation we see that solution property m ln while it partial
molar analogue is mi ln
. In other words the partial molar property
i
includes the logarithm.
The circumflex denotes that properties such as are indeed properties of a

i
component in a mixture but not, by themselves, partial molar properties.
The same argument applies to fi . It is a property of a component in a mixture

but not, by itself, a partial molar property. In fact in this case the partial molar
property is ln( fi / xi ).
The argument for the circumflex above i is a little different. It has been
shown that i = g i , but whereas g i may appear as either intensive or
extensive form, the same is not true of i which, being a potential, may only
be intensive.
Section 2.3 outlines how to solve these phase equilibrium relationships for
non-polar systems (often hydrocarbon mixtures).

2.3 VLE for Non-Polar Systems
VLE problems in the hydrocarbon industry have typically been solved using so
called K values. Other methods are now possible but so many K value
charts have been generated that the method is still used.
The original charts have largely been superseded by EOS and computer
simulation packages. However, K value charts are still helpful in terms of
visualising and solving VLE problems.
The K i value of any species in the mixture is defined as follows:
yi
Ki = .....(2.27)
xi
2.3.1 Rigorous Phi-Phi Approach to VLE
The most useful criterion for phase equilibrium is given by equation (2.26). For
purely VLE problems equation this becomes
fiV = fi L (i = 1,2......N ) ...(2.28)
Where the vapour and liquid phases are indicated by superscripts applied to
each fugacity. The vapour phase fugacity coefficient
V is defined as
i
V
Vi = f i

fi ig
But for an ideal gas fi ig = y i P which leads to
fiV =
Vi y i P ....(2.29)
With the phi-phi approach to VLE, the same expression is used to find the
fugacity coefficient of a component in the liquid phase
iL mixture.
fi L =
L x P ..(2.30)
i i
Notice that the ideal-gas is used as reference state for both vapour & liquid
phases. All that changes, for the liquid phase, is that xi is substituted for y i .

Substitute equations (2.29) and (2.30) into equation (2.28) then, after
cancellation, the general phi-phi VLE relationship becomes
iL xi ...... (i = 1,2,3.....N ) ....(2.31)

Vi y i =

Or, in terms of K value notation

iL
Ki = ........................................(2.32)

Vi
Equation (2.31) is suitable for non-polar mixtures and has the distinct
advantage of being predictive (not based on empirical correlations). A single
EOS is used to predict both liquid and vapour
i values.
Many equations of state have been used historically the Benedict-Webb-

Rubin BWR equation was popular and many modifications to it were
developed to improve its accuracy.
More recently cubic equations of state have become more popular. The
Redlich-Kwong-Soave RKS and the Peng Robinson PR EOS are now very
accurate, even quite close to the critical point.
VLE problems (for non-polar liquids) are usually solved using equation (2.31).
However, this may be recast as follows:
(v )
P
iL = 1
ln i
L
viig dP
RT 0
(v )
P
V = 1
ln V
viig dP
i i
RT 0
Subtracting the second equation from the first and combining the integral
terms leads to

(v )
iL 1
P
ln = i
L
viV dP .

iV
RT 0
In which case
(v )
P
1
ln K i = i
L
viV dP .(2.33)
RT 0

The rigorous phi-phi approach to VLE is better suited to computer solution. A

MathCad worksheet is supplied in Appendix B as an example. The answer is
verified in a standard textbook (Smith, Van Ness and Abbott, 2005).
Either equation (2.32) or equation (2.33) may be used to find the species
K i value and hence solve VLE problems. Equation (2.32) is the most direct.
However, due to the complex functional dependency of

iL = f (T , P, xi ) and
Vi = f (T , P, y i ) two special cases, more amenable to hand calculation, will

be examined.
2.3.2 Liquid and Vapour as Ideal Solutions
If both the liquid phase and the vapour phase can be assumed to be ideal
solutions, then equation (2.20) applies
id = . In this case equation (2.31)
i i
reduces to
Vi y i = iL xi ...... (i = 1,2,3.....N ) ......(2.34)
Or, in terms of K values
iL
K i = V ..........................................(2.35)
i
Notice that these K i values do not depend on composition. An alternate way

of writing these expressions are as follows:
f iV y i = f i L x i ...... (i = 1,2,3.....N )
fiL
K i = V ..(2.36)
fi
Equations (2.34) and (2.35) are preferred because equations of state are set
up to calculate dimensionless fugacity coefficients.

2.3.3 Raoults Law: Liquid as an Ideal Solution and Vapour as an Ideal

Gas
For VLE the criterion for equilibrium is given by equation (2.28)
fiV = fi L (i = 1,2......N )
The left hand side of this expression is given by equation (2.29)
fiV =
Vy P
i i
But if the vapour phase is an ideal gas, then all

V = 1 and each f V is equal
i i
to the partial pressure of the component in the vapour phase mixture
fiV = y i P ......(2.37)
The liquid phase fugacity coefficient

iL was shown to be given by equation
(2.30) as follows:
fi L =
iL y i P
id = and the above expression

But if the liquid phase is an ideal solution i i
becomes
fi L = iL xi P
fiL
fi =
L
xi P
P
fi L = x i f i L ..(2.38)
Of course equation (2.38) is just the Lewis-Randall rule. Now substitute

equation (2.37) and (2.38) into (2.28) to get
y i P = x i f i L (i = 1,2......N ) ..(2.39)
At very low pressure the pure component fugacity approached the system
pressure f i P .

Thus, the fugacity of a pure liquid f i L (at low pressure) must approach the
pressure exerted by the pure liquid at the same temperature as the mixture.
Thus, for an ideal solution at very low pressure the liquid phase fugacity is
simply f i L = Pi sat and equation (2.39) becomes
y i P = xi Pi sat . (i = 1,2......N ) .(2.40)
This is known as Raoults Law and is the simplest VLE model, often called the
ideal VLE model. It assumes an ideal gas, the vapour phase, in equilibrium
with an ideal solution, the liquid phase.
Raoults Law should be used with caution, since many systems depart from
Raoults Law and it cannot be used to model azeotropic behaviour as shown in
Topics 3A and 3B. Notwithstanding, for non-polar systems Raoults Law is
often applicable.
2.3.4 Phi-Phi VLE Summary
1. Completely rigorous VLE:
iL xi ...... (i = 1,2,3.....N )
Vi y i =
and
iL
Ki = V

i
2. Both liquid and vapour phases are ideal solutions:
Vi y i = iL xi ...... (i = 1,2,3.....N )
and
iL
Ki = V
i
3. Ideal solution (liquid) in equilibrium with an ideal gas (vapour) which is

equivalent to Raoults Law:
y i P = xi Pi sat . (i = 1,2......N )

From the last expression, the Raoults Law K values are given by
Pi sat
Ki = ..(2.41)
P
It gets progressively easier to solve VLE problems (and find the K values)
as one moves from case 1 through to case 3.
There is also a 4th case (to be considered in Topic 3). This is the case where
the vapour phase is modelled as an ideal gas mixture (low pressure), but the
liquid phase is modelled as a non-ideal solution.
The VLE expressions based on these assumptions is known as modified

Raoults Law and will be discussed in the next topic. Many polar systems fit
this model of solution behaviour.
2.3.5 Finding K-Values using Charts
The properties that are needed to solve multicomponent VLE problems

(namely the component K i values) are usually obtained from an Equation of
State EOS.
In the past, the Benedict-Webb-Rubin BWR equation of state has proved

very accurate for hydrocarbon mixtures the equation is given below:
RT C 1 1 aa 1+ / v2
P= + B0 RT A0 02 2 + (bRT a ) 3 + + c exp 2
v T v v v v
6 3 2
vT
The molar average boiling point was used to take into account the effect of
composition. With this information K i -values were calculated (using the BWR
EOS) for 12 light hydrocarbons, for pressures up to 250 bar.
The results were presented as 324 charts known as Kellogg charts with K i -
values plotted a function of T , P and molar average boiling point. However,
the charts are laborious to use, especially when frequent estimates of K i -
values are needed. As a result, simpler methods have been developed.
De Priester prepared a set of nomographs (from the Kellogg charts) and these
are often used in place of Kellogg charts. In the de Priester nomographs, the
composition dependence of the K i -values has been completely removed
using an averaging process see Appendix A for a copy of these charts.

The results are less accurate than using Kellogg charts, but are adequate for
first-order estimates; they are also helpful when trying to visualise how K i -
values change with T & P .
There are two de Priester charts see Appendix A:

1. Low Temperature Chart -70oC to 20oC 1 bar to 50 bar
2. High Temperature Chart 10oC to 200oC 1 bar to 50 bar
As mentioned above, many EOS have been used to solve VLE problems.
Currently attention has focussed on using cubic EOS. Cubic EOS were
reviewed and discussed in an earlier course (Process Industries C).
It was also explained (in Process Industries C) how cubic EOS may be used to
find not only vapour but also liquid properties; in the past they were used
mainly for vapour phase properties, but recent refinements have led to
improved liquid phase property estimation.
See Appendix B for all the relevant equations and a worked example that
shows how to solve a rigorous VLE problem using the rigorous phi-phi method.

2.3.6 Example 2.1
Determine the equilibrium constant K i for some substance i in a multi-

component VLE system. Assume that both the liquid and vapour phases
behave as an ideal solution.
For the component in question:
Pr = 0.35 , Tr = 1.0 and Z C = 0.27
Use the generalised fugacity coefficient chart provided over the page. When
using this chart pure i , at these conditions, will be found to be a vapour. This
means that the liquid phase fugacity state is fictitious and, hence, the liquid
phase fugacity coefficient iL must be found by extrapolation from the real
liquid states to this fictitious liquid state.



2.3.7 Vapour Pressure Correlation
If a VLE system is ideal it will conform to Raoults Law and the P x y

diagram, for a binary system, will look like the one shown below:
Constant T The system pressure P is

plotted on the y axis.
sat
Subcooled X P1
Liquid P x1 Species 1 is the MVC,
P y1 while 2 is the LVC.
sat L-V
P X
2
The P x1 line (bubble
P
Superheated point curve) is the
Vapour saturated liquid line.
0 The P y1 line (dew point

0 1.0
curve) is saturated vapour
x1 & y1
line.
If the system is an ideal VLE system and if the temperature is below the
critical temperature of both components, then the system pressure will vary
from P2sat (LVC) to P1sat (MVC) and the P x1 line will be straight.
To understand why an ideal VLE system behaves in this way write down
Raoults Law twice; first for species 1 and then for species 2:
y1 P = x1 P1sat .(2.42)
y 2 P = x 2 P2sat ....(2.43)
The term on the left of equation (2.42) is the partial pressure of pure
component 1, while the term on the left of equation (2.43) is the partial
pressure of pure component 2.
Further, y1 + y 2 = 1 (just as x1 + x 2 = 1 ) and the partial pressure of 1 plus

the partial pressure of 2 is just the total system pressure P which leads to
P = y1 P + y 2 P = x1 P1sat + x 2 P2sat
P = x1 P1sat + (1 x1 )P2sat

After some re-arrangement this becomes
( )
P = P2sat + x1 P1sat P2sat ....(2.44)
If P is plotted versus x1 , then equation (2.44) is the equation of the P x1

line (saturated liquid line). Clearly, equation (2.44) takes the form y = mx + c .
Thus, is follows that the P x1 line must be straight.
However, this result is only true for ideal systems. For non-ideal systems the
P x1 line is not straight, but the system pressure P must still lie between
P1sat & P2sat (assuming there are no supercritical components present).
Raoults Law K values are given by
y i = K i xi
Pi sat
Ki =
P
Therefore, some sort of correlation is needed to predict the vapour pressure. A

good correlation for the vapour pressure P sat is the Antoine equation.
B
ln P sat = A (2.45)
T +C
In which A, B & C are species specific constants which are widely available in
standard reference books see Appendix A3 (Smith, Van Ness and Abbott,
2005). Take note that the Antoine equation is sometimes expressed as
B
log10 P sat = A
T +C
Clearly the numerical value of the constants A, B & C differ (even for the
same species) depending on the units used for T & P sat whether or not
natural logarithms are used.
For a given species, it should be noted Pi sat is a function of temperature T

only. Note that the vapour pressure Pi sat only has physical meaning on the
vaporisation curve between the triple point and critical point temperatures.

2.3.8 Temperature-Composition Diagram for Binary System
A P x y diagram was shown previously; in this case the temperature was

held constant. A T x y diagram (for the same ideal system) is shown
below; in this case the pressure is fixed:
Constant P
Again 1 is MVC and 2

Superheated is LVC.
T2sat Dew Point Vapour Region
Curve
The T x1 curve (bubble

point curve) is saturated
T x1 T y1
L-V Two-phase liquid line.
Region
T
T y1 curve (dew point
Subcooled
Liquid Region Bubble Point T
1
sat
curve) is saturated
Curve
0
vapour line
0 1.0
x1 & y1
Notice that, even though system is still ideal, neither the T x1 curve nor the
T y1 curve is a straight line. This is because the vapour pressure of each
component varies with temperature:
The saturated liquid curve, and the subcooled liquid region, now lies below
the two-phase region. The saturated vapour curve, and the superheated
vapour region, now lies above the two-phase region.
The intercepts T2sat & T1sat are the boiling points of pure 2 and pure 1
respectively (at the fixed pressure P ). The Antoine equation may also be used
to find these two pure component saturation temperatures.
To do this the Antoine equation must be solved in reverse, i.e. what

temperature Ti sat is needed to maintain the vapour pressure equal to the
applied pressure Pi sat = P ? The answer to this question is the pure
component saturation temperature Ti sat .
Distillation columns operate with a more or less constant pressure, but with
significant temperature variations along the column. For a binary system, this
type of equipment is best represented by a T x y diagram.

2.3.9 Significance of K-Values
Raoults Law for an N component VLE system is given by
y i P = xi Pi sat . (i = 1,2......N )
The Raoults Law K i values are then
Pi sat
Ki = . (i = 1,2......N )
P
These Raoults Law K i values depend only on T & P but will vary from
pure species to pure species at the same T & P . Thus, for any pure species:
It is clear that as the pressure decreases the K i values will increase.
And as the temperature increases the K i values will increase.
These findings may be confirmed by running a ruler horizontally up a
de Priester chart (as the ruler goes up, the temperature rises and
pressure drops). As this is done, the K i values for all species will
increase.
Now, keeping T & P constant, read from right to left and note that the de
Priester K i values of each pure species shows an increasing trend. Going
from right to left corresponds to going from heavier to lighter species.
Heavier species have a higher molar mass (molecular weight) than lighter
species. Species with a lower molar mass have a higher value of Pi sat at the
same temperature T .
The K i value is best thought of as the volatility of a component in a

mixture. An Increasing K i value corresponds to a greater volatility of the
component. Higher volatility is associated with:
1. A decrease in pressure P (releasing more vapour).
2. An increase in temperature T (greater Pi sat for all species).
3. A decrease in molar mass (lighter species have greater Pi sat than
heavier species, at the same T ).
Distillation works by separating components that have a difference in volatility;

the larger the difference in volatility, the easier the separation. As will be seen
later azeotropes occur when both species have the same volatility. Distillation
alone cannot separate an azeotropic mixture.

2.4 Phase Equilibrium Calculations
For bubble point or dew point calculations usually either T or P will be fixed
and either xi or y i will be fixed; the other unknowns may then be found from
the phase equilibrium relationships.
Thus, there are usually four possible combinations (although others are
possible) depending on the unknown variables. For distillation it is often more
realistic to fix the pressure and then calculate the unknown temperature.
2.4.1 Bubble Point Calculations
If the pressure P and the liquid phase compositions xi are fixed, then this is
known as a bubble point calculation.
The objective is to find the temperature T of the mixture, followed by the

composition of the first bubble of vapour y i in equilibrium with this liquid.
The calculations are iterative and may be laid out in the same way whether
Raoults Law or a de Priester chart is used to find the K i values. In either
case proceed as follows:
1. Fix P & x i .
2. Set each Pi sat = P and use the Antoine equation in reverse to calculate
Ti sat these are the boiling points of each component at pressure P :
Bi
Ti sat = C i ..(2.46)
Ai ln Pi sat
3. Guess an initial trial temperature T( 0 ) using the approximation:
T( 0 ) = xi Ti sat (2.47)
i
4. Take a ruler to a de Priester chart, set the left hand side to the specified
pressure P and the right hand end of the ruler to the guessed
temperature T( 0 ) - now read off all K i -values from the chart.

If Raoults Law is used, then find all K i -values using the expression
below:
Pi sat
Ki = . (i = 1,2......N )
P
To find these K i -values, Pi sat will needed for each component (use the
Antoine equation at temperature T( 0 ) ).
5. Calculate all y i using phase equilibrium relationship below this holds

whether a de Priester chart or Raoults Law is used:
y i = K i xi ..(2.48)
6. Sum up all the vapour phase mole fractions and check whether or not
they sum to one:
y = K x
i
i
i
i i = 1.0 (2.49)
Check is y
i
i 1 0.05 ....(2.50)
7. In all likelihood equation (2.50) will not be satisfied. Return to step 3 and
adjust the guessed temperature T( 0 ) to a new value T(1) . It will be fairly
obvious whether to increase or decrease the temperature.
If equation (2.50) sums to less than one, then K i values need to be

increased. This means a higher volatility is required; at a fixed pressure,
this means a greater temperature is needed.
8. With this new temperature T(1) repeat steps (4) through (6) until
equation (2.50) is satisfied within the stated tolerance.
This scheme was originally developed for de Priester charts and is purely trial-
and-error. The charts themselves cover only a limited range of hydrocarbon
components and the nomographs are often difficult to read accurately.
When using Raoults Law more efficient schemes have been developed. They
are usually two-stage schemes: the first stage involves iterating on
temperature T (until it converges within some tolerance); the second-stage
involves finding the compositions y i (Smith, Van Ness and Abbot, 2005).

2.4.2 Dew Point Calculations
A dew point calculation involves fixing the pressure P and the vapour phase
compositions y i . The objective is then to find the temperature T and the
composition of the first drop of liquid xi in equilibrium with this vapour.
Apart from the fact that P and y i are now fixed, step 1, step 2, step 3 and
step 4 are the same. In step 5 equation (2.48) is reversed since y i is fixed and
the purpose of a dew point calculation is to find xi as follows:
yi
xi = ..(2.51)
Ki
The equations in step 6 are also reversed (what is needed is x i ):
yi
x = K
i
= 1.0 ..(2.52)
i i i
Check is x
i
i 1 0.05 ...(2.53)
Repeat steps 4 through 6 until the liquid phase mole fractions all sum to one
within some tolerance thus, T and xi have been estimated. Again, if
Raoults Law is used a more efficient two-stage iteration scheme is available.
In the two cases described above the pressure P was fixed: then either xi
was fixed (bubble point route to estimate T & y i ); or y i was fixed (dew point
route to estimate T & x i ).
Another two cases are possible where the temperature T is fixed together
with one set of phase compositions. Very similar bubble point and dew point
calculations are then possible; these are easier since Pi sat do not change.
It is often better to spend time writing a worksheet (for a maths package) using
the rigorous phi-phi approach see Appendix B (this is a Mathcad worksheet,
although any package may be used).
For a binary, it is also possible to fix T & P and then calculate x i & y i , but
then, to be two-phase, T must lie between T1sat & T2sat at prevailing P .

2.4.3 Example 2.2
Find the bubble point temperature of the liquid mixture defined below:
Liquid
Component Antoine A Antoine B Antoine C
Mol Fraction
Ethane 0.15 15.6637 1511.42 -17.16
Propane 0.20 15.7260 1872.46 -25.16
i-butane 0.60 15.5381 2032.73 -33.15
n-butane 0.05 15.6782 2154.90 -34.42
At P = 770 kPa or P = 5775.5 mmHg .
Note the data for the Antoine constants have P & T units of (mmHg) and (K)
respectively.


2.4.4 Example 2.3
Find the dew point temperature for the same mixture in Example 2 at the same
pressure.
Liquid
Component Antoine A Antoine B Antoine C
Mol Fraction
Ethane 0.15 15.6637 1511.42 -17.16
Propane 0.20 15.7260 1872.46 -25.16
i-butane 0.60 15.5381 2032.73 -33.15
n-butane 0.05 15.6782 2154.90 -34.42
At P = 770 kPa or P = 5775.5 mmHg .
Note the data for the Antoine constants have P & T units of (mmHg) and (K)
respectively.


2.4.5 Flash Calculations
Consider a subcooled liquid mixture with an overall composition z1 and then

consider that this mixture is gradually expanded at constant temperature. At
some pressure, the liquid becomes saturated at its bubble point:
Constant T
z1
From 12 pressure
1 P x1 X P
1
sat decreases, but fluid
remains entirely liquid.
P 2
P y1
sat
At 2 liquid becomes
P X
2
saturated. The liquid
P
composition is unchanged
at x1 = z1 .
0
x1 y1 Read off to left to find P
0 1.0
then right and down (as
x1 & y1
shown) to find y1 .
Notice that T & x1 are fixed and P & y1 are read-off from the diagram. At
point 2 there is actually no vapour present; however, moving fractionally away
from 2 produces the first bubble of vapour with composition y1 .
A dew point problem works in exactly the opposite sense. A superheated

vapour is compressed isothermally until it intersects the saturated vapour
curve. Now T & y1 are fixed and P & x1 are read-off the diagram.
On the saturated vapour curve itself there is actually no liquid present, it is only
upon moving fractionally away from this point that the first drop of liquid will be
formed this liquid has composition x1 .
In the case of flash calculations the temperature and/or pressure is altered so

that inside the flash drum a two-phase mixture can be separated. The flash
drum constitutes a single equilibrium stage.
The mixture is separated inside the drum with saturated vapour V leaving
from the top and saturated liquid L leaving from the bottom of the drum. With
flash calculations the unknowns are both the vapour flowrate and composition
V & y1 and the liquid flowrate and composition L & x1 .

As well as more unknowns, flash calculations require more equations. Not only
are phase equilibrium relations needed (as in dew point and bubble point
calculations), but also flash calculations require mass balances.
A degree of freedom analysis will show that any flash calculation requires two
variables to be specified; in this case they will be taken to be T & P . For
bubble/dew point calculations the number of variables that need to be
specified is found from the Gibbs phase rule.
Two important types of flash are isothermal and adiabatic flash. In the case
of an isothermal flash, shown below, both T & P are fixed:
Vapour
V, yi
T, P
Feed
Steam
F, zi, T,
PF
If F = 1 , then
Fz i = ni : moles T
entering in the
Condensate
feed.
Liquid
L, xi
The vapour offtake valve controls the pressure within the flash drum while the
steam valve controls the temperature within the drum. Flashing may also
involve cooling of the single phase mixture.
In the case of an adiabatic flash there is no heating/cooling system and no

heat is transferred. Thus, the amounts of V and L leaving, the stream
compositions and the final temperature T all depend on the feed enthalpy hF .
Isothermal P T flash calculations are easier because the heat balance is

decoupled from the mass balances and equilibrium relations. In the case of
adiabatic flash calculations all these relations must be solved simultaneously.

Only a P T flash will be considered here. Be aware that a necessary

condition for an isothermal P T flash is that the pressure P must lie
between PBP (bubble-point pressure) and PDP (dew-point pressure).
Consider that a subcooled liquid at temperature T , pressure PF , with overall

composition z i enters the flash drum at a feed rate F kmol/hr.
It is normal to initialise the calculation so that F = 1 kmol/h. The results may

then be scaled up to the actual feed rate F kmol/hr at the end of the
calculation (in this case Fz i = ni or moles entering in the feed).
A typical P T flash may be visualised as follows:
Constant T
Subcooled liquid enters
the drum at T & PF
PF 1 P x1 X P1
sat
and composition z1 or
PBP P y1 n1 moles of species 1
P L
2 V
if F = 1 .
P2sat X PDP
P Pressure then drops to
P , while temperature
is kept constant at T .
0
0 x1 z1 y1 1.0 Liquid flashes to V
x1 & y1 kmol/h of vapour and
In a flash drum there is sufficient residence time for the phases to separate
and exit as follows: saturated liquid stream with flowrate L and composition
x1 ; saturated vapour stream with flowrate V and composition y1 .
State 1 corresponds to the inlet stream, with overall composition z i (or n1

moles of species 1 if F = 1 ). State 2 corresponds to the mixing point on
the horizontal tie-line.
The horizontal tie-line shows that both streams leaving the drum V & L are at
the same T & P . This must be the case, given that the two exit streams are in
phase equilibrium with each other.

Exactly how point 2 bisects the horizontal tie-line fixes the outlet V / L ratio
of the two phases leaving the drum, the exact relationship is known as the
Lever Rule:
As P PBP (bubble-point) then V 0 & L 1 .
As P PDP (dew point) then V 1 & L 0 .
Thus, bubble point and dew point calculations may be thought of as two
special cases of a more general flash calculation:
1. A bubble point calculation can be thought of as a limiting flash
calculation where the liquid exits the drum with the same flow and
composition as the feed entering:
There is actually no vapour leaving the drum; however, any
vapour that tends to leave the saturated liquid will have
composition equal to the bubble point composition.
If P > PBP only subcooled liquid leaves the drum and flash
vaporisation of the liquid into a two-phase mixture cannot occur.
2. A dew point calculation can be thought of as a limiting flash calculation
where the vapour exits the drum with the same flow and composition
as the feed entering:
There is actually no liquid leaving the drum; however, any liquid
that tends to condense from the saturated vapour will have
composition equal to the dew point composition.
If P < PDP only superheated vapour leaves the drum and flash
condensation of vapour into a two phase mixture cannot occur.
With the feed flow initialised to one ( F = 1 kmol/h) there is a single overall
mass balance given by
1 = V + L ......(2.54)
If there are N components present, then there are N 1 independent

component mass balances
z i = Vy i + L xi ...(2.55a)
Or,
n1 = Vy i L xi ....(2.55b)
There are N independent equilibrium relations given by
y i = K i xi .. (i = 1,2......N )

Where, each Raoults Law K values may be found from equation (2.41)
Pi sat
Ki = .. (i = 1,2......N )
P
Substitute equation (2.54) into equation (2.55a) and then eliminate y i using
the expression y i = K i xi to get
z i = (1 L )K i xi + L xi
Or,
ni = (1 L )K i xi + L xi
Rearranging this expression in favour of xi leads to the final result below:
zi
xi = . (i = 1,2......N ) .(2.56a)
( )
1 L K i + L

Or,
ni
xi = . (i = 1,2......N ) ....(2.56b)
(1 L )K i + L

A P T flash calculation can be carried out by iterating on equation (2.56a) or

(2.56b). Note that four different variants of this expression are possible,
depending on which variables are eliminated.
A P T flash calculation entails the following trial-and-error approach:

1. Fix F , z i (or ni ), T , PF and initialise F = 1 kmol/h.
2. Ensure flash pressure P lies between bubble point pressure PBP and
the dew point pressure PDP .
3. Knowing T & P calculate all K i from equation (2.41), or look up the
K values on a de Priester chart.
4. Guess a starting value for L . A value of 0.5 is sensible, unless the
pressure is much nearer the bubble point (then set it between 0.5 and
1.0) or the dew point (then set it between 0 and 0.5).
5. Compute all xi from equation (2.56a) or (2.56b).
6. Calculate x
i
i and check if x
i
i 1 0.05 .

7. If x
i
i 1 > 0.05 then go back to step (4) and adjust value of L and
continue iterating until the tolerance condition is satisfied.

8. Once tolerance condition is satisfied the calculation has ended.
Use the converged value for L to calculate fractional vapour flow from
equation (2.54)
V = 1 L
Calculate all the vapour phase compositions y i from the phase equilibrium
relationship:
y i = K i xi
Remember, the K i -values do not change throughout the procedure because

T & P fixed. Finally scale-up L and V based on the actual feed flow F
kmol/h.
If needed the energy balance may be solved for Q (kW). All the enthalpies
are fixed if the system T & P and the stream compositions are known.
Q = VhV + L hL FhF ..(2.57)
This energy balance is derived from the steady flow energy equation. For a
flash drum the work term is zero, the kinetic and potential energy change
terms are zero, there are two exit streams and one inlet stream.
An adiabatic flash is more complicated because now the temperature floats,

the left side of equation (2.57) is set to zero (adiabatic case). In this case the
energy balance (now an enthalpy balance), the mass balances and all the
phase equilibrium relations must be satisfied simultaneously.

2.4.6 Example 2.4
A feed stream enters a flash drum at 80oF and 150 psia. The overall molar
composition of this feed stream z i is shown below:
Component zi
CH 4 0.05
C2H8 0.10
C3H8 0.30
i-C 4 H 10 0.55
Calculate the fractional flowrates and compositions of both the exit liquid and
vapour streams leaving the drum.
Conversions:
o
C= ( o
F 32 ) 5
9
1 psi = 6894.8 Pa 6.9 kPa


2.5 Tutorial 2
1. A mixture of hydrocarbons of the following composition is flashed in a

separator at 40oC and 1,000 kPa. Determine what fraction leaves the
separator as liquid and the composition of both product streams.
(Use L = 0.5 for the first iteration).
Component Mole Fraction

Methane 0.10
Ethane 0.25
Propane 0.30
n-Butane 0.10
i-Butane 0.20
n-Pentane 0.05
(Answer L= 0.085, compositions:
Component xi yi
Methane 0.006 0.108
Ethane 0.067 0.268
Propane 0.194 0.310
n-Butane 0.201 0.090
i-Butane 0.307 0.190
n-Pentane 0.225 0.034
2. The product from the top of a distillation column passes into a

condenser and is partially condensed at 14 bara and 40oC. It is then
passed into a separator where both phase are separated. The mixture
has a composition as given below.

Ethane 0.20
Ethene 0.17
Propane 0.23
n-Butane 0.21
n-Heptane 0.10
n-Octane 0.09
Determine the fraction of the mixture which is condensed and the

composition of the two phases (Use L = 0.5 for the first iteration).

(Answers L= 0.600, compositions:
Component xi yi
Methane 0.111 0.333
Ethane 0.071 0.320
Propane 0.221 0.243
n-Butane 0.284 0.099
i-Butane 0.164 0.003
n-Pentane 0.149 0.001
3. Vapour leaving the top plate of a distillation column at a pressure of

6.5 bara has the following composition and is fed to a total condenser.

Ethane 0.12
Propane 0.26
i-Butane 0.62
a) Determine the operating temperature of the condenser (assume no

pressure drop across the condenser and use 40oC as a first
iteration).
b) The temperature of the reflux drum, which receives the condensed

stream, is 5oC higher than that of the condenser. Estimate what
percentage of the liquid stream will re-evaporate (assume no
pressure drop through the reflux drum and use L = 0.5 for the first
iteration).
(Answers a) 12oC and b) 15%)

4. The overhead vapour stream from a fractionating column passing a

condenser has the following composition:

Ethane 0.15
Propane 0.20
i-Butane 0.60
n-Butane 0.05
It is desired to operate so that 75% of the stream is liquefied in the

condenser. If the material leaves the condenser at 25oC, what is the
pressure required and what is the composition of the vapour product?
(HINT: take a first iteration of 1,000 kPa).
(Answer P = 720 kPa, composition:
Component yi
Ethane 0.349
Propane 0.255
i-Butane 0.372
n-Butane 0.022
5. If the overhead vapour stream described in the previous question is in

equilibrium with the liquid leaving the top bubble cap plate in the
column, calculate the range of operating temperatures within which the
top plate can be operated. Assume that the pressure drop between the
top plate of the column and the condenser is 50 kPa (HINT: take 40oC
as the first iteration).
(Answer 17oC to 40oC)

6. A hydrocarbon mixture of the composition below is to be separated

using a distillation column operating at 15 bara:

Methane 0.17
Ethane 0.24
Propane 0.28
n-Butane 0.31
It is required that the feed enters the column as a two-phase mixture.

What is the possible range of feed temperatures (HINT: take -40oC and
50oC as first iterations).
(Answer 50oC to 54oC)
7. A hydrocarbon stream, of the composition below, is to be partially

condensed by spraying liquid propane into it.

Methane 0.350
Ethane 0.275
Propane 0.375
If the molar ratio of liquid propane to the vapour stream is 1:4 and the
process is carried out at 30 bara and 5oC, calculate the composition
and quantities of both phases leaving the process (Use L = 0.5 for the
first iteration).
(Answers: L = 3.55 moles, V= 1.45 moles(based on 1 mole of liquid

entering, compositions
Component xi yi
Ethane 0.133 0.638

Propane 0.227 0.204
i-Butane 0.639 0.160

8. A sample of a vapour mixture has been partially analysed and has the
following composition:

Methane 0.08
Ethane 0.17
Propane 0.25
Unknown X 0.50
It is suspected that the unknown component is either i-butane or n-

butane. As a test the vapour is cooled at 1.4 bara and it begins to
condense at -14oC. By the time -26oC is reached, approximately half of
the mixture is condensed. Find Unknown X (HINT: the component may
be a mixture).
(Answer: The unknown component X is likely a mixture of i-butane

and n-butane in a 4:1 ratio. The dew point and liquid fraction
calculations tend point in this direction)
9. The vapour product from the top of a distillation column has the
following composition:

Ethane 0.11
Propane 0.66
n-Butane 0.23
The mixture passes to a partial condenser operating at 10 bara. The

vapour is taken off as a product and the liquid is refluxed to the column.
Estimate the operating temperature of the condenser (HINT: try 40oC) if
the reflux ratio is 5:1 (Note the reflux ratio is the ratio of liquid returned
to the column to the product removed from the condenser). Also
evaluate the composition of the product and reflux streams.
What would be the effect on the K value for propane if the ethane were
replaced by methane?

(Answer: 25oC, compositions
Component xi yi
Ethane 0.079 0.261

Propane 0.660 0.660
n-Butane 0.260 0.078
10.
a) Explain the circumstance under which the Lewis-Randall rule and
Raoults law are valid.
b) 100 kmols of an equimolar mixture of octane and water flows into a

flash separator maintained at 10.3 bara and 205oC. Evaluate the
amounts of vapour and liquid formed and the composition of both
phases. Use Raoults law and the vapour pressure given below in
your calculation (Use L = 0.5 as the first iteration).
B
lnP sat = A +
T
(cont)
In which P sat = vapour pressure (N/m2) and T = temperature (K.).
Component A B
Octane 22.66 -4468.6
Water 24.99 -5068.6
c) Can you see any reasons why your answer should be inaccurate or
invalid?

(Answers: a)The Lewis-Randall rule applies to.Raoults law

applies to..b) 50 kmol of liquid and 50 kmol of vapour formed,
compositions
Component xi yi
Octane 0.631 0.370

Water 0.368 0.632
(Answer: c) Validity? Accuracy?)

2.6 Bibliography
Gmehling, J, Onken, U, Arlt, W, Vapour-Liquid Equilibrium Data Collection,

Chemistry Data Series, DECHEMA, Frankfurt, 1977-1990.



ed., 1998.


Appendix A
De Priester Charts see next page.
Scanned from Equilibrium Stage Separation Operations in Chemical

Engineering, Henley & Seader, Wiley, p277

A1 High Temperature Chart:

A2 Low Temperature Chart:

A3 Antoine Coefficient Table

Appendix B
Rigorous Phi-Phi VLE Worksheet
This material will be useful for later Design Projects and Experimental
Projects; see next page for further details.

Modified versions of the PR EOS are now very accurate for vapour and liquid
phase property estimation, especially non-polar/hydrocarbon mixtures. The PR
has largely replaced the much more cumbersome BWR EOS and is easy to
program with the help of a mathematics package.
In the past cubic EOS were not accurate enough to model the behaviour of
polar liquid mixtures. As a result, empirical correlations were developed. These
correlations will be discussed in Topic 3A.
However, very recent developments, using cubic EOS in conjunction with

advanced mixing rules, has led to big improvements in their prediction of
polar liquid properties.
As this approach is refined, and gains wider acceptance, empirical correlations

should be needed less and less. However, there is likely to be an on-going
need for two different methods: cubic EOS for non-polar liquid mixtures (Topic
2); empirical correlations for polar liquid mixtures (Topic 3A).
The PR approach begins by estimating individual pure-component aTi and bi

parameters. These hark back to the a & b parameters originally developed by
VdW over a hundred years ago.
All that is needed are the critical pressure, the critical temperature and the
acentric factor (for each pure species). From these values pure-component aTi
and bi parameters are found from the following three equations:
2
R 2TCi
aTi = 0.45724 a (T ) i ...(B.1)
PCi
RTC i
bi = 0.0778 .....(B.2)
PCi
[ ( )(
(T ) i = 1 + 0.37464 + 1.5422 i 0.26992 i 2 1 Tr 0.5
i
)]
2
.(B.3)
The bi parameter is a species specific constant, while the aTi parameter is not
only species specific (acentric factor) but also depends on temperature (thus
the subscript T ).

The alpha function (T ) i was developed to give better accuracy when

predicting vapour pressure data. The next step is to find the dimensionless Ai
and Bi parameters (for each pure species) as follows:
aTi P
Ai = .....(B.4)
R 2T 2
bi P
Bi = ..(B.5)
RT
Once the pure component parameters have been found, mixture parameters
are next (the subscript mix is not really required but is used in order to
emphasise that these are not pure component values).
Use standard mixing rules, as shown below, to calculate mixture aTmix and
bmix parameters. These formulae are just a way of weighting individual aTi and
bi parameters to mixture parameters.
aTmix = y i y j (aT i aT j ) (1 k ij )..(B.6)

0.5
i j
bmix = y i bi ..(B.7)
i
k ij is known as the Binary Interaction Parameter BIP. It is used to correct the

property prediction so that it better matches experimental results. If BIPs are
unknown then set them to zero, in which case equation (B.6) reduces to
2
0.5
aTmix = y i aT i ..............(B.8)
i
Mixture Amix and Bmix parameters are now calculated from expressions below:
aTmix P
Amix = ....(B.9)
R 2T 2
bmix P
Bmix = ..(B.10)
RT

Solve the cubic EOS below for Z = Z mix using the two Amix and Bmix
parameters just found (the mix subscript is now dropped)
Z 3 + ( B 1) Z 2 + ( A 2 B 3B 2 ) Z + ( AB + B 2 + B 3 ) = 0 ..(B.11)
The largest root is the vapour root. Label this root the vapour mixture
compressibility factor Z V . Now solve the equation below for the fugacity
coefficient of the component in the vapour phase mixture:
2 y j Aij
ln i =
Bi
( )
Z 1 ln(Z B mix )
Amix j

B
i ln V
( )
Z V + 1 + 2 B mix
( )
V V V
B mix 2 2 B mix Amix B mix Z + 1 2 B mix

(B.12)
Notice that
Vi depends not only on T & P but also on the vapour phase
composition y i (this is the complex functional dependency discussed
previously).
Equations (B.6) through (B.11) are now recalculated using liquid phase
compositions xi in place of vapour phase composition y i . Three roots are
again obtained; now choose the smallest root since this will be the liquid root.
Label this root the liquid mixture compressibility factor Z L (again drop the
subscript mix) and then use this value in conjunction with equation (B.13),
shown below, to calculate a value for
iL :
2 x j Aij
ln
= B
( )
Z 1 ln(Z B mix )
A j

B
i ln L
( )
Z L + 1 + 2 B mix
( )
L i L L mix
i
B mix 2 2 B mix Amix B mix Z + 1 2 B mix

(B.13)
Notice that
L
i depends not only on T & P but also on the liquid phase
composition xi - mirroring, once more, the complex functional dependency
discussed previously.

The rigorous approach to VLE leads to the following equilibrium relation

(developed earlier) for species any species i in the mixture:

Vy =
i i
Lx
i i
And this in turn finally leads to the K i - value for this species:
iL
Ki = V

i
A full VLE calculation is more complicated since these calculations must be

repeated for all species and the results constrained so that unknown mole
fractions (either liquid or vapour) sum to one.
It is easy to see that hand calculations would involve complex iteration

schemes, since one set of compositions and either the temperature or the
pressure would initially be unknown.
However, all the above equations are readily programed into a maths package
by setting out the known variables and then declaring which parameters are
functions of the unknown variables.
The package then finds a solution (in terms of the unknown variables) that
meets the imposed phase equilibrium constraints. These worksheets may be
shared and adapted to rigorously solve a wide variety of VLE problems.
Below is a fully worked out methane/butane bubble point problem using the
rigorous phi-phi approach to VLE:
The liquid mixture composition and temperature are specified.
The vapour phase composition and pressure are calculated by the
worksheet.
At this temperature the methane is supercritical; that is the system
temperature is above the critical temperature of methane.
Thus, methane cannot exist as a pure liquid at this temperature.
This means that high concentrations of methane cannot co-exist in VLE
beyond some cut-off mixture concentration.
Once the calculation has converged, there is a final run through to find the
numerical values of all parameters as they might appear to verify the answer.
The worked example is rather long to be included in these lecture notes but
may be obtained from the course lecturer, see file below.
Ph-Phi Worked
Example.pdf

Appendix C
Review of Partial Molar Properties
This material is supplied as a helpful background to partial molar properties,

the various rules associated with partial molar properties and the notation that
is widely used in an engineering context.
Partial molar properties were introduced in Y2, Introduction to Kinetics and

Thermodynamics and this was (quite rightly) slanted in favour of a Chemistry
background. The theory was generally applied only to ideal systems.
The material in this appendix is slanted in favour of Chemical Engineering.

Students would do well to review the material and pay attention to both
notation and symbols which does vary between disciplines.
The apportioning rule, the recombination rule, the graphical interpretation

and the limiting values of partial molar properties are all particularly important;
of particular importance is the relationship between partial molar properties
and systems at equilibrium.
The solution of both phase and chemical equilibrium problems, for non-ideal
systems, is not possible without a thorough understanding of partial molar
properties, their dependencies and behaviour.
The correction of fugacity of a component in a mixture (closely linked to a

partial molar property) between real mixtures and ideal mixtures is of key
importance.
Also Appendix C, section C-6, is important because it shows that all relations
between pure component properties have their corresponding partial molar
analogues.
How these correction factors are calculated and applied tends to be more
widely used in engineering and industry.
See next page for more details.

C-1 Partial Molar Property Overview
If a mixture property m were simply the mole fraction weighted pure

component values ( m1 & m 2 ), then we could write
m = x1 m1 + x 2 m 2 >> m = x1 m1 + (1 x1 )m 2
m = m 2 + x1 (m1 m 2 ) (C.1)
At constant T & P the solution property m must lie between m1 and m 2 . If

the mole fraction weighting rule was obeyed, then equation (C.1) would hold
(the equation of a straight line). If the solution property m were plotted against
x1 , then equation (C.1) must plot as a straight line joining m1 with m 2 :
Constant T & P
X m1
m X
m
m2 X
0
0 x1 1.0
x1
Using this simple mole fraction weight approach, the solution property m lies
between the pure component values m1 & m 2 and approaches them in the
limit as x1 1 and x1 0 respectively.
However, the question is, what experimental evidence is there that the solution
property m varies linearly with composition x1 ?
The answer is, except for special cases (that will be discussed later), the
experimentally-determined variation in solution property is typically a non-
linear function of composition.

For a binary mixture 1 and 2 at constant T & P , a typical experimentally-

determined variation in solution property m versus composition x1 is shown
below by the blue curve:
Constant T & P
I1
X m1
m X
m I
2
m2 X
0
0 x1 1.0
x1
Clearly, the solution property m cannot be found using equation (C.1), as

shown again below:
m = m 2 + x1 (m1 m 2 )
However, if a tangent line is drawn to the curve (at the concentration of

interest), then the mixture property m could be found using the same rule, but
now the y-axes intercepts of the tangent line (that is I 1 & I 2 ) must be used in
place of m1 & m 2 and this leads to
m = I 2 + x1 (I 1 I 2 )
This in turn may be re-arranged into the same form (as the original mole
fraction weighting formula), to obtain the following expression for the mixture
property m as a function of concentration and the y-axes intercepts:
m = x1 I 1 + x 2 I 2 ..(C.2)
Clearly it is important to identify these y-axes intercepts I 1 & I 2 but first the
partial molar property mi must be defined and explored.

C-2 Definition of Partial Molar Property
Any solution property m may be thought of as being apportioned between

the various components in the solution. The rule for quantifying how a solution
property is apportioned to any component i is given by
(nm )
mi = ..(C.3)
ni T , P ,n j
Where,
mi is the partial molar property of species i in the mixture at T & P .
m is the overall solution property at the same T & P .
mi is the pure-component property of i at the same T & P .
In equation (C.3) the partial molar property mi of species i is defined as the

partial derivative of the total solution property M = nm with respect to the
mole number of species i :
T & P must be held constant.
Constant n j means that the mole number of all other species except
the i th must be held constant.
If any solution property can be apportioned into partial molar properties, then
it should be possible to recombine partial molar properties back into their
corresponding solution property, see representation below:
T &P
Solution
Property - m
Recombining Apportioning
Route Route
T &P
Partial Molar
Property - mi
The partial molar property mi is the property that component "i" seems to
have in the solution at T & P . The pure component property mi is the
property that the pure species actually has at the same T & P .

In the pure state like molecules are completely surrounded by like species
and, as a result, each molecule has an identical size, shape and charge
distribution the forces of attraction on a single molecule are the same in all
directions:
In mixtures like molecules are partially surrounded by unlike species.
The different species all have a different molecular size, shape and
charge distribution.
Thus, the partial molar property (the property it seems to have) is
shifted either above or below its corresponding pure component value
(the property it actually has in the pure state).
The direction in which the partial molar property is shifted (above or
below the pure component value) depends on whether like-unlike
species bond more or less tightly together than like-like species.
The extent to which partial molar properties differ from their pure
component values depends on the extent of these differences
(between like/unlike species) and the concentration of the mixture.
It follows that partial molar properties are not only a function of T & P ,
but also a function of composition.
Pure component properties, on the other hand, depend only on T & P .
It is clear that the partial molar property of any component must approach its
pure component value as the mixture approaches purity in that species.
Equation (C.3) can be used to define the partial molar property of component
1 m1 in a binary mixture as follows:
(nm )
m1 =
n1 T , P ,n2
In the same way we may define the partial molar property of component 2
m 2 in a binary mixture as follows:
(nm )
m2 =
n 2 T , P ,n1
Notice that for component 1 the partial derivative must be with respect to the
mole number of species 1 (while maintaining a constant T & P and a
constant mole number for species 2).
The same situation applies to m 2 in reverse. Notice that it is the mole number
and not the mole fraction that appears in these definitions it is important to
bear this in mind for what follows.

C-3 Partial Molar Properties from Correlations
The purpose of this section is to derive expressions for the partial molar
properties of each species (in a binary mixture) in terms of mole fractions.
As before, the partial molar property of component 1 in a binary mixture of

species 1 and 2 was defined as
(nm )
m1 =
n1 T , P ,n2
Expand out the partial derivative on the right
m n
m1 = n + m
n1 T , P ,n2 n1 T , P ,n2
For a binary mixture, the total moles of the mixture n is simply the moles of
component 1 n1 plus moles of component 2 n 2 . Thus, by material balance
n = n1 + n2
Since the number of moles of species 2 must be held constant, it follows that
n
=1
n1 T , P ,n2
Substituting this result into the expanded expression for m1 leads to
m
m1 = m + n
n1 T , P ,n2
Re-arrange the above to get
m m m
= 1
n1 T , P ,n2 n
However, the objective of this derivation is to express m1 as a function of mole

fractions and not mole number. A second partial derivative will be needed to
show how the mole number changes with mole fraction.

Start with the definition of the mole fraction of species 1
n1
x1 =
n
Now differentiate this with respect to the mole number of component 1 using
the product rule (a little easier than using the quotient rule)
x 1 n 1 n
1 = 12 = 1 1
n1 T , P ,n2 n n n n
n n
1 =
x1 T , P ,n2 x 2
Use the chain rule for differentiation but recognise that (for a binary), while the
mole number of species1 can vary (keeping the mole number of species 2
constant), the same is not true for mole fractions.
m n dm
1 =
n1 T , P ,n2 x1 T , P ,n2 dx1
The partial derivative notation is dropped because if T & P are always

constant, then the binary solution property m only depends on x1 . Substitute
the two partial derivatives into the chain rule to get
dm m1 m n
=
dx1 n x 2
Cancel and re-arrange the above to get
dm
m1 = m + x 2 .(C.4)
dx1
The partial molar property m2 is obtained in exactly the same way
dm
m2 = m x1 ...(C.5)
dx1
Given a correlation of m versus x1 , equations (C.4) and (C.5) may be used

find the partial molar properties as algebraic function of concentration.

C-3 Partial Molar Properties from Diagrams
Consider once again the relationship between any tangent to the solution
property curve at concentration x1 and the intercepts of this tangent line
I 1 & I 2 at the y-axes:
Constant T & P
I1
X m1
m X
m I
2
m2 X
0
0 x1 1.0
x1
From co-ordinate geometry the slope of this tangent line is given by
dm m I 2
=
dx1 x1 0
dm
I 2 = m x1 ......(C.6)
dx1
Again, using co-ordinate geometry, the slope of the tangent line at

concentration x1 is also given by
dm I 1 I 2
=
dx1 1.0 0
dm
I1 = I 2 +
dx1

Substitute equation (C.6) into above leads to
dm
I1 = m + x2 ...(C.7)
dx1
Compare equation (C.4) with equation (C.7) and equation (C.5) with equation
(C.6) and it is clear that the intercepts of any tangent line (on the y-axes) must
correspond to the partial molar properties as follows:
m1 = I 1
(C.8)
m2 = I 2
Thus, for a binary mixture (at constant T & P ) the following graphical
relationship exists between all the properties mentioned thus far:
Constant T & P
m1
X m1
m X
m m
2
m2 X
0
0 x1 1.0
x1
It follows that the intercepts at either end of any tangent line are the partial
molar properties ( m1 and m2 ) of each species
The partial molar properties ( m1 and m2 ) are the properties that the species
seems to have in the binary mixture. These partial molar properties differ from
the pure component values ( m1 and m 2 ) at the same T & P .
Given a plot of solution property m versus concentration x1 , the partial molar

properties of each species may be found by laying off a tangent to the solution
property curve (at the concentration of interest) and reading m1 and m2 from
intercepts at either end of the tangent line this applies to any property m .

Notice that the intercepts change as the concentration varies which means
that, even when T & P are held constant, the partial molar properties mi
change with concentration xi . The pure component properties remain fixed.
Consider limiting values of m1 and m2 as x1 1.0 :
Constant T & P
I1
I1
I1 m1 m1
X X m1
X
m2 m2 I2
X
I2
I2
m2 X
0
0 x1 x1 x1 1.0
As x1 1.0
The following holds true for all partial molar properties as the concentration
changes:
1. As: xi 1.0 mi mi (pure component value)(C.9)
2. As: xi 0 mi mi (infinite dilution value)..(C.10)
Hence, the partial molar properties always vary between the pure-component
value and the infinite dilution value:
mi mi as species i becomes completely surrounded by like
molecules, i.e. xi 1 .
mi mi as species i becomes completely surrounded by unlike

molecules, i.e. xi 0 .
The largest departure of the partial molar property m1 from its pure component
value m1 is its infinite dilution value m1 . This is because at this concentration
component 1 is completely surrounded by unlike species (component 2).

C-4 Recombination Rule
Now, if solution property m can be apportioned into its constituent partial

molar properties mi . The question is how can the constituent mi be
recombined so that the original solution property m is recovered?
(conceptually represented by the drawing below)
T &P
Solution
Property - m
Recombination Apportioning
Rule Rule
(nm )
? mi =
T &P ni T , P ,n j
Partial Molar
Property - mi
In fact the answer is already available. Return to equation (C.2) where it has
been shown that the solution property depends graphically on the y-axes
intercepts of the tangent line.
m = x1 I 1 + x 2 I 2
However, it has already been shown that these intercepts I 1 & I 2 are the
partial molar properties m1 & m 2 of components 1and 2 respectively. Thus,
for a binary mixture the recombination rule is simply
m = x1 m1 + x 2 m 2 (C.11)
In words, the solution property m is the mole fraction weighted partial molar
properties of the components making up the solution. However, there is no
reason why the mixture should be limited to a binary mixture.
Thus, for a multicomponent mixture where i identifies the species and there
are N such species in total ( i = 1, 2.....i......N ), the mixture property m is
given by the more general recombination rule
N
m = xi mi .(C.12)
i =1

C-5 Gibbs-Duhem Equation
Equation (C.12) is the recombination rule. It indicates how, if the apportioning

rule is used to find mi from the solution property m , then equation (2.12) must
be used to recover the original solution property m .
This procedure may be conceptually represented using the diagram below:
T &P
Solution
Property - m
Rule Rule
(nm )
m = xi mi mi =
i T &P ni T , P ,n j
Partial Molar
Property - mi
There is yet a third rule which is very useful when using partial molar
properties and it is known as the Gibbs-Duhem equation.
An important concept to grasp is that any solution property m can be

apportioned between the components in a solution:
And that these apportioned partial molar properties mi are known to
vary with composition, as shown by the intercepts at either end of the
tangent line.
Thus, the partial molar properties of the constituent species mi in the
mixture cannot vary independently of each other.
This dependence is shown by the intercepts of the tangent line as
one partial molar property decreases (say the right hand intercept) the
other partial molar property must increase (the left hand intercept).
To find the exact relationship, start with the fact that any property m may be
expressed in terms of T & P and the mole number of each species
n1 , n 2 ......ni as follows.
nm = f (T , P, n1 , n2 ......ni )

The total differential can be expressed in terms of all the partial changes in
nm caused by each variable considered separately
(nm ) (nm ) (nm )

d (nm ) = dP + dT + dni
j
But the last term on the right (within the summation) is defined as the partial
molar property of species i
m m
d (nm ) = n dP + n dT + mi dni
P T , x T P , x i
But,
ni = x i n
dni = xi dn + ndxi
Expand d (nm ) = ndm + mdn on the left and substitute the above expression
into the right to get
m m
ndm + mdn = n dP + n dT + mi ( xi dn + ndxi )
P T , x T P , x i
Bring all the terms to the left and group terms in n and dn together
m m
dm dP dT mi dxi n + m xi mi dn = 0
P T , x T P , x i i
For any system n and dn are independent and arbitrary, so that each of the
terms within the square brackets must individually be zero:
m m
dm dP dT mi dxi = 0 (C.13)
P T , x T P , x i
and
m x i mi = 0 >>>> m = x i mi
i i
This last expression is the recombination rule see equation (C.12). This is
just a more general way of deriving it.

Using the recombination rule, find differential changes in solution property m

as follows:
dm = xi dmi + mi dxi
i i
Substitute this into equation (C.13)) to get
m m
x dm + m dx P
i
i i
i
i dP dT mi dxi = 0
T P , x i
T ,x
Cancelling like terms leads to the general Gibbs-Duhem expression
m m
P dP + T dT xi dmi = 0 (C.14)
T ,x P, x i
However, if T & P are held constant, then equation (C.14) reduces to the
most widely used form of the Gibbs-Duhem equation
x dmi
i i = 0 ....(C.15)
It will be shown later just how useful this relationship can be. It is sufficient now
to understand that it restricts how partial molar properties can change relative
to each other.
A conceptual diagram of the main rules associated with partial molar

properties is shown below it includes the apportioning rule (the definition of a
partial molar property), the recombination rule and the Gibbs-Duhem equation:
T &P
Solution
Property - m
Rule Rule
m = xi mi (nm )
mi =
ni T , P , n j
i
T &P
Partial Molar
Property - mi
Gibbs-Duhem
Equation
x dm
i
i i =0
Notice that all these rules apply at a mixture at constant T & P .

C-6 Partial Molar Property Relations
Any total solution property may be expressed as a function of temperature,

pressure and composition. As an example, take the total solution property ng
ng = f (T , P, n1 , n2 ......ni )
In which n1 , n 2 ...ni is the number of moles of each species. The total

differential is given below, where the summation term includes all the partial
changes in ng caused by changes in the mole number of each component
(ng ) (ng ) (ng )

d (ng ) = dP + dT + dni
j
At constant T & P the first two terms drop out
(ng )
dGT , P = dni
i n i T , P , n
j
In exactly the same way d (H )T , P and d (S )T , P are given by
(nh )
dH T , P = dni
i n i T , P , n
j
and
(ns )
dS T , P = dni
i ni T , P , n
j
Now, take definition for the Gibbs free energy and differentiate it at constant
T & P as follows:
G = H TS
dGT , P = dH T , P TdS T , P
Substitute dGT , P , dH T , P and dS T , P into the above equation as follows:
(ng ) (nh ) (ns )

dni = dn T dni
i ni T , P ,n
j
i ni T , P , n
j
i ni T , P , n
j

According to equation (C.3) each term within the square brackets is a partial
molar property, given by g i hi and s i respectively. Substitute these partial
molar properties into the previous expression to get
g dn = h dn
i
i i
i
i i T s i dni
i
For each component individually
g i dni = hi dni Ts i dni
g i = hi Ts i ..(C.16)
Compare this to the pure component intensive (molar) property relation
g i = hi Ts i ..(C.17)
The on-to-one correspondence between equations (C.16) and (C.17) is

immediately apparent. In fact, all pure component property relations have their
equivalent partial molar property analogues.
The left hand column below is a selection of common pure component

property relations, the right hand column are the corresponding partial molar
property analogues:
Pure Component Property Partial Molar Property

Relation Relation
g i = hi Ts i g i = hi Ts i
hi = u i + Pvi hi = u i + Pvi
dg i = vi dP s i dT dg i = vi dP s i dT
d i = RTd ln f i d i = RTd ln fi
This is fortunate because partial molar property relations are, therefore, the
same as pure component property relations.

The circumflex above fi denotes that it is the fugacity of component i in the

mixture. One might well ask, why not use f i instead of fi ? The answer is
that the fugacity of a component in a mixture is not a partial molar property.
Thus, in general if a partial molar property is denoted by mi then it follows
fi mi
In actual fact the relation below is a partial molar property
( )
ln fi / xi = mi
But the fugacity of a component in a mixture fi often appears by itself and to

distinguish it from the pure component fugacity f i the circumflex is added.
Likewise the fugacity coefficient in a mixture is denoted by

and the
i
argument for doing this is similar to that above, namely

m
i i
In actual fact the relation below is a partial molar property
( )
ln
=m
i i
But the fugacity coefficient of a component in a mixture

i often appears by
itself and to distinguish it from the pure component fugacity i the circumflex
is added Thus, if the solution property is m ln , then its corresponding
partial molar property must be mi ln
i and
T &P
Solution
Property - m = ln
Rule Rule
ln = xi ln i (n ln )
mi = = ln i
ni T , P ,n j
i
T &P
Partial Molar
Property - mi
Gibbs-Duhem
Equation x d ln
i
i i =0


TOPIC 3A: VLE for Polar Liquid Mixtures 1
Topic 3A
VLE for Polar Liquid Mixtures
Contents
TOPIC 3A: VLE FOR POLAR LIQUID MIXTURES
3A.1 INTRODUCTION ............................................................................................................... 3
3A.2 MODELS OF IDEALITY AND STANDARD STATES .................................................................. 4
3.2.1 Lewis-Randall Model of Ideality ........................................................................... 4
3.2.2 Henrys Law Model of Ideality .............................................................................. 5
3A.3 REAL MIXTURES ............................................................................................................. 7
3.3.1 Excess Properties ............................................................................................... 7
3.3.2 Activity Coefficients ............................................................................................. 8
3A.4 THE GAMMA-PHI APPROACH TO VLE ............................................................................. 12
3.4.1 Rigorous Gamma-Phi Approach to VLE ............................................................ 13
3.4.2 Vapour Phase as an Ideal Solution.................................................................... 14
3.4.3 Vapour Phase as an Ideal Gas .......................................................................... 15
3.4.4 Both Phases as Ideal Solutions ......................................................................... 15
3.4.5 Raoults Law...................................................................................................... 16
3A.5 PHASE EQUILIBRIUM DIAGRAMS ..................................................................................... 17
3.5.1 P-x-y Diagrams.................................................................................................. 18
3.5.2 T-x-y Diagrams .................................................................................................. 19
3.5.3 Example 3A.1 .................................................................................................... 22
3.5.4 The x-y Diagram ................................................................................................ 24
3.5.5 Example 3A.2 .................................................................................................... 26
3A.6 MODELS OF SOLUTION BEHAVIOUR ................................................................................ 28
3.6.1 The One Parameter Model ................................................................................ 28
3.6.2 Margules Two Parameter Model ....................................................................... 31
3.6.3 Van Laars Two Parameter Model ..................................................................... 33
3.6.4 Wilsons Local Composition Model .................................................................... 35
3.6.5 Example 3A.3 .................................................................................................... 36
3A.7 TUTORIAL 3A ................................................................................................................ 37
3A.8 BIBLIOGRAPHY .............................................................................................................. 40
APPENDIX A.......................................................................................................................... 41
Real Mixtures ............................................................................................................. 41
A.1 Property Changes on Mixing ................................................................................ 42
A.2 Property Changes on Mixing for Ideal Solutions ................................................... 43
A.3 Activity of a Species in a Mixture.......................................................................... 45
A.4 Excess Properties ................................................................................................ 47
APPENDIX B.......................................................................................................................... 49
B.1 Positive & Negative Deviation from Ideality .......................................................... 49
B.2 Formation of Azeotropes ...................................................................................... 52
APPENDIX C ......................................................................................................................... 54

Gamma-Phi VLE Worksheet for Azeotrope ................................................................ 54
Learning objectives
After studying this VLE for Polar Liquids topic you should be able to:
Define standard states used for VLE purposes.
Define excess property and write down expressions ideal solution
excess properties.
Define the activity coefficient and demonstrate how it is used to correct
the fugacity of a component in a mixture between the ideal solution
state and the real fluid state.
Explain the gamma-phi approach to VLE (polar systems) and compare
and contrast with the phi-phi approach to VLE (non-polar systems).
Derive rigorous VLE relationship for gamma-phi approach to VLE and
successively simplify it to derive Modified Raoults Law and Raoults
Law.
Define K-values in relation to gamma-phi method.
Construct P-x-y, T-x-y and x-y diagrams.
Outline reasons for positive and negative deviation from ideality and
relate this behaviour to activity coefficients.
Draw P-x-y, T-x-y and x-y diagrams involving maximum and minimum
temperature azeotropes and relate this behaviour to activity
coefficients.
Define the relative volatility, calculate and average value over the
boiling point range and use this to plot an x-y diagram.
Use different models of solution behaviour to solve practical problems
involving polar systems. Differentiate between power-law models
Margule and Van Laar and Wilson local composition model.
Show how problems may be solved (involving azeotropes) using a
maths package.

3A.1 Introduction
In Topic 2 the following general criterion for VLE was derived
fiV = fi L (i = 1,2......N )
For non-polar (usually hydrocarbon) systems the phi-phi approach to VLE

leads to the fugacity coefficient of a component in the mixture
i being used
to model the non-idealities in both liquid and vapour phases.
The fugacity coefficient of a component in the mixture (which uses the

i
ideal gas as a reference state) was defined as follows:

= fi = fi
i
fi ig y i P
As can be seen, the fugacity coefficient corrects the fugacity of a species from
the ideal gas reference state to the real fluid state at the same temperature,
pressure and composition.
i is best found using an EOS, see example phase equilibrium calculation in

Appendix B of Topic 2. This approach works well for non-polar systems, such
hydrocarbon mixtures.
Less rigorously de Priester charts may be used or, if the system can be
assumed to be an ideal gas in equilibrium with an ideal solution, Raoults Law
can be used.
Cubic EOS cannot (as yet) adequately model complex polar liquid mixtures
due to different interactions between like/unlike species in the liquid phase.
Thus, for polar liquid mixtures an empirical approach, called the gamma-phi
approach to VLE has been developed.
The gamma-phi approach to VLE uses an empirically-based correction factor,

namely the activity coefficient i , to model non-idealities in the liquid phase
while retaining
to model non-idealities in the vapour phase.
i
Thus, the gamma-phi approach to VLE uses different correction factors for the
vapour and liquid phases: is predicted using an EOS; is determined
i i
experimentally and the data is regressed to an empirical correlation.

3A.2 Models of Ideality and Standard States
The ideal solution state requires a model of ideality. In Topic 2 the Lewis-
Randall Rule was developed to find fi id . However, the Lewis-Randall Rule
implicitly uses a standard state which cannot be applied in all cases.
3.2.1 Lewis-Randall Model of Ideality
Some simple expression is needed to link the fugacity of a component in an

ideal mixture fi id to the fugacity of pure i in its standard state f i 0 . The
simplest relationship for an ideal gas is fi ig = y i P (Daltons Law). Similarly for
an ideal solution we might write::
fi id = xi f i0 ..(3.1)
However, equation (3.1) is only valid when molecular interactions between two
components are weak. In the figure below the experimental behaviour of f1L
(liquid 1 in liquid 2) is described.
The fugacity of component 1 in the real solution f1L is shown by the solid
line. The fugacity of component 1 in an ideal solution f1id ( ) L
is shown by the
dotted line:
Constant T
With the Lewis-Randall

model of ideality f i 0 = f i
f10 = f1 f 1 is the actual fugacity

f1L of pure 1 as a liquid at
(kPa) f1L the same T & P as the
mixture.
( f )
1
id
L
The slope of the ideal

solution line f 1L is
0 affected by interactions
0 x1 1.0 with species 2.
1.0 x2 0

Using this standard state equation (3.1) becomes
fi id = xi f i ....(3.2)
Notice that the real solution behaves increasingly as an ideal solution as it

approaches purity in that species. That is fi L x i f i as x i 1 .
The reason for this is easy to understand. As purity is approached in any

component, then like species are increasingly surrounded by like species and
the species in question behaves as if it were in an ideal solution.
The is why the real solution curve (solid line) becomes a tangent to the ideal
solution line (dotted line) as x1 1 .
3.2.2 Henrys Law Model of Ideality
Sometimes, as shown below, the right hand side of the real solution curves
stops short of the y-axis intercept; e.g. when a gas dissolves in a liquid, pure
i is not a liquid at mixture T & P and f i does not exist:
Constant T
f10 = k1 The ideal solution line is

drawn as a tangent to
the real solution curve at
xi 0 .
f1L
(kPa)
( f )
1
id
L
The y-axis intercept is

now f i 0 = k i which is
the Henry Law constant.
f1L
The slope of the ideal
0 solution line is now k i .
0 x1 1.0
1.0 x2 0
Using this standard state equation (3.1) is equivalent to Henrys Law:
fi id = x i k i ....(3.3)

Notice that the real solution behaves increasingly as an ideal solution as the
solution approaches infinite dilution in that species. That is f1L x1 k1 as
x1 0 (this applies to any species that approaches infinite dilution).
Gases such as oxygen or nitrogen are supercritical components, which mean

that the prevailing system temperature (assuming room temperature) is well
above their critical temperature; thus O 2 and N 2 cannot exist as pure liquids
under these conditions.
However, oxygen and nitrogen can still dissolve in a liquid (say water) to form
a liquid solution. Henrys Law is a good model of ideality for these
components, because Henry Law constants are available in the literature.
It can be easily shown that when Henrys Law applies to the solute (the
supercritical species), then the Lewis-Randall rule applies to the solvent. As
the solvent approaches purity, the solute approaches purity infinite dilution.

3A.3 Real Mixtures
There are two ways of working out the properties of real mixtures:
1. By measuring the property change relative to the pure components
properties at the same T & P as the mixture.
2. By measuring the property change relative to the property of an ideal
solution at the same T , P & xi as the mixture.
The first case leads to so called property changes on mixing. A property

change on mixing m (this could be v , u , h , s or g ) is defined as
m = m xi mi
i
The property change on mixing is the (experimentally determined) departure of

the real solution property m relative the pure component properties mi at the
same T & P as the mixture see Appendix A.
The second case leads to a class of properties known as excess properties.

It is by studying these excess properties that the new correction factor for the
liquid phase fugacity can be found.
3.3.1 Excess Properties
An excess property m E (this could be v E , u E , h E , s E or g E ) is defined in a

similar way. However, it is the (experimentally determined) real solution
property m minus the corresponding ideal solution property m id at the same
T , P & xi . Thus, by definition
m E = m m id .....(3.4)
The partial molar analogue of this expression is given by
miE = mi miid ......(3.5)
Of particular interest are the expressions for the Gibbs free energy
g E = g g id
And,
g iE = g i g iid ..(3.6)

3.3.2 Activity Coefficients
The chemical potential of a species in a mixture i was defined in Topic 2 by

equation (2.25). However, it has been shown that i = g i and equation (2.25)
becomes
g i = RT ln fi + i
In which i is the species specific constant of integration that depends only on

temperature. Writing this for a component in an ideal solution at the same
T & P gives
g iid = RT ln fi id + i
Subtract the latter equation from the former and, after cancelling out the
constant of integration i , this becomes
fi
( )
g i g iid = RT ln
f id
i
In view of equation (3.6) the left side must be g iE which leads to
g iE fi
= ln ....(3.7)
RT f id
i
The ratio of the fugacity of a component in a real solution to the fugacity of a

component in an ideal solution at the same T , P & xi is known as the activity
coefficient and is denoted by i . Thus, by definition
fi
i = ..(3.8)
f id
i
or
fi
i = ....(3.9)
xi f i 0
For VLE the usual standard state is the pure component, as a liquid at the
same T & P as the liquid mixture.

In this case, f i 0 = f i and Lewis-Randall rule may be substituted to get
fi
i = (3.10)
xi f i
From equation (3.8) it is clear:

The activity coefficient i is a correction factor.
It corrects the fugacity of a species from the ideal solution state to the
real liquid state the same T , P & xi as the ideal mixture.
It uses the ideal solution as a reference state.
Activity coefficient i are used to find fi in polar liquid mixtures (many
accurate empirical correlations for i are available in the literature).
To find the fugacity of a component on the real solution curve (see below):
1. Find the corresponding point on the ideal solution straight line, at some
concentration, using the Lewis-Randall Rule fi id = x i f i ; see point 1 on
the diagram below.
2. The activity coefficient shifts point 1, on the ideal solution line, to point
2, on the real solution curve. This is in accordance with a re-arranged
version of equation (3.10), namely fi L = i x i f i .
Constant T
f1L = P1sat
If i > 1 then the real

f1L 2 solution curve will lie
above the ideal solution
f1L
( f )
1
id
L line (as shown).
(kPa) 1
If i < 1 , then the real
solution curve lies
below the ideal solution
line (not shown).
0
0 x1 1.0
1.0 x2 0
If i > 1 (as shown) the component will have a greater escaping tendency
from the real solution fi L = i x i f i than it would have from a corresponding
ideal solution fi id ( )
L
= x i f i at same T , P & xi .

If i < 1 the component will have a lesser escaping tendency from the real
solution fi L = i x i f i than it would have from a corresponding ideal solution
( f )
i
id
L
= xi f i , at same T , P & xi .
Note that as the binary mixture heads towards purity in any component (as
xi 1 ) then the solution moves towards an ideal solution and i 1 . The
solution curve becomes a tangent to the Lewis-Randall line.
Since most binary mixtures do not have i = 1 , the problem will need to be
solved by looking at g iE . Substitute equation (3.8) into (3.7) to get
gE
= ln g i ....(3.11)
RT
From equation (3.11) it is clear that ln i is a partial molar property of

g E / RT . Using the general notation, where m is any solution property and mi
is its corresponding partial molar property, then it is clear that
g E / RT m and ln i mi
Thus, the apportioning rule (applied to the activity coefficient) is given by
( )
ng E / RT
ln g i = (3.12)
ni T , P ,n j
And the recombination rule (applied to activity coefficients) is
g E / RT = xi ln g i (3.13)
i
Since ln i is a partial molar property it must also conform to the Gibbs-

Duhem equation.
x d ln
i
i i = 0 (3.14)
Activity coefficients are measured experimentally from the results of phase

equilibrium experiments. This data is converted into g E / RT , using equation
(3.13), and then this is regressed to fit some model of solution behaviour.

Knowing the fitted parameters of this g E / RT function, equation (3.12) may

then be used to calculate the activity coefficients at other conditions.
This g E / RT function is known as models of solution behaviour. Many such

models have been developed and much experimental data is available
(Gmehling, J, Onken, U, Arlt, W, Chemistry Data Series, 1977-1990).
The most popular models of solution behavior are listed below:

One Parameter model.
Margule model.
Van Laar model.
Wilson model.
NRTL model.
UNIQUAC model.
UNIFAC model.
Most of these (except UNIFAC) are empirically-based correlations. Different

models apply in different situations and the first four of these will be discussed
in Topic 3A. The NRTL model will be reviewed in Topic 3B.
UNIQUAC and UNIFAC are more complex models (not really suitable for
detailed hand calculations). However, simulation packages offer these as more
advanced options.

3A.4 The Gamma-Phi Approach to VLE
Any VLE problem can be solved using the general criterion for phase
equilibrium derived earlier:
fiV = fi L .......(i = 1,2,3.....N )
In Topic 2, the fugacity coefficient of a component in a vapour mixture (using

the ideal-gas as a reference state) was defined as
V
Vi = f i

fi ig
But, the fugacity of a species in an ideal gas mixture is given by fi ig = y i P ,

which leads to
fiV =
Vi y i P
The activity coefficient of a component in a liquid mixture (using the ideal

solution as a reference state) was also expressed earlier as
fi L fi L
i = =
fi id xi f i
L
Leading to
fi L = i x i f i L
Substitute these expressions for fi L and fiV into the general criterion for
phase equilibrium leads to the general gamma-phi approach to VLE:
V y P = x f L .......(i = 1,2,3.....N ) .....(3.15)

i i i i i
It is necessary to step back and recognise that equation (3.15) is completely

different from the phi-phi approach to VLE:
With the phi-phi approach to VLE, the is used to model the non-
i
idealities in both phases
With the gamma-phi approach, the Vi models non-idealities in the
vapour phase while the i is used to model liquid phase non-idealities.
With gamma-phi, different reference states are used for each phase.

3.4.1 Rigorous Gamma-Phi Approach to VLE
In Topic 1 an expression for the pure-component liquid fugacity f i was

developed
ln
fiL
=
(
v Lsat P P sat )
sat
f RT
(
v sat P P sat
f i L = f i sat exp L
)

RT
(
v sat P P sat
f i L = isat Pi sat exp L
)
..(3.16)
RT
Substitute equation (3.16) into equation (3.15) to get

sat
(
sat
)
Vi y i P = i xi isat Pi sat exp v L P P .......(i = 1,2,3.....N )
RT
..(3.17)
Equation (3.17) can be re-arranged in terms of K i -values to give
i Pi
sat sat
exp
(
v Lsat P Pi sat )
i
RT
Ki = ..(3.18)
i P
V
If the Poynting correction factor (see Topic 1) is set to unity and the phase
equilibrium relation, equation (3.17), reduces to
Vi y i P = i x i isat Pi sat .......(i = 1,2,3.....N ) .(3.19)

While the corresponding K i -value becomes
i isat Pi sat
Ki = ..(3.20)

Vi P

Both fugacity coefficients (

V and sat ) may be obtained using a single EOS
i i
(such as PR) but the activity coefficients i must be found from a suitable
empirical correlation.
Equations (3.19) and (3.20) constitute the rigorous gamma-phi approach to

VLE and are used to model polar liquid mixtures at low to moderate pressure.
By moderate pressure it is meant well removed from the critical point.
Of course it is rigorous only in the sense that the Poynting correction factor is
set to unity. In practice this term adds little to the overall accuracy of the result.
3.4.2 Vapour Phase as an Ideal Solution
If the vapour phase mixture is an ideal solution

V = V or
i i
V = f / P and,
i i
making use of equation (3.16), then equation (3.17) reduces to
f iV y i = i xi f i L .......(i = 1,2,3.....N )
Further, if f i L = f i sat (Poynting correction factor is unity), then this becomes
f iV y i = i x i f i sat .......(i = 1,2,3.....N ) .(3.21)
Equation (3.21) may again be re-arranged in terms of K i -values to give
i f i sat
Ki = ...(3.22)
f iV
In terms of fugacity coefficients equations (3.21 and (3.22) become
Vi y i P = i x i isat Pi sat .......(i = 1,2,3.....N ) .(3.23)
And,
i isat Pi sat
Ki = .(3.24)
Vi P
Of course these are the same as the previous case, except that
Vi = Vi .

3.4.3 Vapour Phase as an Ideal Gas
At low pressure
Vi 1 & isat 1 and equation (3.19) reduces to
y i P = i x i Pi sat .......(i = 1,2,3.....N ) .....(3.25)
While the corresponding K i -values are given by
i Pi sat
Ki = ...(3.26)
P
Equation (3.25) is called modified Raoults Law and equation (3.26) defines
corresponding modified Raoults Law K i -values.
The assumption here is that the pressure is low enough for the gas phase to
be an ideal gas. However, the liquid phase remains a non-ideal solution
(because most polar liquid mixtures are non-ideal).
3.4.4 Both Phases as Ideal Solutions
Since not only is the gas phase an ideal solution, but also the liquid phase is
an ideal solution, return to equation (3.21) and set all i = 1
f iV y i = x i f i sat .......(i = 1,2,3.....N ) ....(3.27)
In terms of K i - values
f i sat
Ki = ...(3.28)
f iV
Reverting back to fugacity coefficients equations (3.27) and (3.28) become
Vi y i P = x i isat Pi sat .......(i = 1,2,3.....N ) ...(3.29)
And,
isat Pi sat
Ki = .(3.30)
Vi P
Both of these fugacity coefficients can be obtained from a generalised fugacity

coefficient chart or an appropriate EOS.

3.4.5 Raoults Law
If the pressure is very low then the vapour phase is an ideal gas, then all

V 1 , all sat 1 . As always, if the pressure is low, then the Poynting
i i
correction factor will be unity.
In addition, for an ideal gas in equilibrium with a liquid phase that is an ideal
solution, all the activity coefficients i = 1 and modified Raoults Law reduces
to Raoults Law:
y i P = x i Pi sat .......(i = 1,2,3.....N ) ......(3.31)
With corresponding Raoults Law K i -values
Pi sat
Ki = .....(3.32)
P
However, Raoults Law often fails to accurately model polar systems (even
when the pressure is low). This is because the interactions between like/unlike
species in polar liquid mixtures can be very different.
Most polar liquids exhibit either positive or negative deviation from Raoults
Law and in some cases these deviations are strong enough to lead to the
formation of azeotropes see Appendix B (which is examinable).
In the case of positive deviation from ideality the lack of affinity between unlike
species may be large enough for the liquid phase to spit into two partially
miscible phases see Topic 3B for VLE, LLE and VLLLE.

3A.5 Phase Equilibrium Diagrams
For a binary system that obeys Raoults Law the phase equilibrium
relationships, for both species, are as follows:
y1 P = x1 P1sat y 2 P = x 2 P2sat ....(3.33)
The left hand side of each expression are the partial pressures of component
1 and 2 respectively. Adding both equations together gives the total
pressure P as follows:
P = x1 P1sat + x 2 P2sat
P = x1 P1sat + P2sat (1 x1 )
P = P2sat + x1 ( P1sat P2sat ) ..(3.34)
Thus, for ideal VLE systems the partial pressures of each species and the
total pressure must all plot as straight lines. This type of behaviour is shown
below:
Constant T
P1sat
P
(kPa) y1P = x1P1sa t
P2sat
y2 P = x2 P2sa t
0
0 x1 1.0
1.0 x2 0
According to the Phase Rule F = 2 + N for binary VLE = 2 and N = 2

which means that there are two degrees of freedom F = 2 . The Gibbs Phase
Rule variables are T , P, x1 & y1 (taking 1 to be the independent mole
fractions).
Thus, any two out of these four variables must be fixed in order to solve the
two independent Raoults Law expressions.

3.5.1 P-x-y Diagrams
By specifying a sequence of x1 values, while always keeping the temperature

T fixed, then a sequence of pressures P and corresponding compositions y1
may be found; see equation (3.34) and Raoults Law for species 1.
For Raoults Law a plot of P versus x1 produces a straight line, called the
bubble point line this joins P2sat with P1sat . A plot of P versus y1 produces a
curve, called the dew point curve which also joins P2sat with P1sat .
On a P x1 y1 diagram the bubble point curve ( P x1 curve) is above the

dew point curve ( P y1 curve). As before, both curves meet at the same y-
axis intercepts, corresponding to the vapour pressure of each pure species.
Constant T
Bubble Point
Curve sat For an ideal system,
X P1
P x1 the bubble point
curve will plot as a
P y1
straight line, as
P2
sat
X Dew Point given by equation
Curve
P (3.34)
Most systems are

non-ideal and will
deviate from this.
0 x1 y1 1.0
1.0 x2 y 2 0
If the system is non-ideal, then the P x1 line will either lie above or below the
straight line case: above is positive deviation from ideality; below is negative
deviation from ideality. The dew point line P y1 is always curved.
Positive deviation is associated with activity coefficients greater than one and
is caused by unlike species have less affinity (lower forces of attraction) than
like species. The opposite is the case for negative deviation from ideality.
Ideal solutions will have activity coefficients equal to one and like and unlike
molecules all have the same size and forces of attraction.
If the dew point and bubble point curves meet at some intermediate
composition, forming either a minimum or a maximum, then this is called and
azeotrope more details are available in Appendix B.

3.5.2 T-x-y Diagrams
In the case of a T x1 y1 diagram, it is necessary to fix the system pressure

P and then carry out either a sequence of bubble point (fixing x1 finding y1 ),
or a sequence of dew point (fixing y1 finding x1 ) calculations. It is also
possible to fix both T & P (see Example 3A.1).
A typical T x y diagram is shown below. Notice that the bubble point curve
T x1 is below the dew point curve T y1 and, for an ideal system (Raoults
Law) neither curve is straight:
Constant P Heat the mixture from

A at constant overall
composition and
T2sat Dew Point D constant pressure.
Line
C The mixture must

T x1 follow a vertical line
L V A BC D
M T y1
T
B
Point A corresponds
A Bubble Point T
1
sat to some initial
Line subcooled liquid state.
0
At this point the liquid
0 x1 y1 1.0
is below its BP.
1.0 x2 y 2 0
Point B corresponds to the bubble-point of the mixture: reading horizontally

(to left) gives the bubble point temperature T , while reading horizontally (to
right) and then vertically down gives the bubble point composition y1 .
Further heating leads to point C on the dew point curve. Reading horizontally
(to the left) gives the dew point temperature T , while reading horizontally (to
the left) and then vertically down gives the dew point composition x1 .
Further heating leads to point D , a superheated vapour state with the same
composition y1 . Notice that all states on A B C D have the same overall
composition, even when the system splits into a liquid phase, composition x1
and a vapour phase, composition y1 , between points B & C .
In order to flash a single phase system into the two-phase region, at a fixed
pressure P , the temperature must lie somewhere between the bubble point
and the dew point temperatures.

The mixing point M lies on a horizontal tie-line L V . It has the same overall
composition as points A, B, C , D , and lies between points B & C . Therefore,
the mixture will split (or flash) into L moles of liquid and V moles of vapour.
The concentration of the saturated liquid x1 is read vertically down from L .

The concentration of saturated vapour y1 is read vertically down from V .
To find the amounts of each phase present use the lever rule. The lever rule is
an expression of how the horizontal tie-line L V is bisected by the mixing
point M and is given by
LM V
= (3.35)
MV L
The ratio ( mol V / mol L ) is the same as the ratio ( LM / MV ), i.e. the length
of the tie-line either side of the mixing point. This may also be expressed as
the ratio ( LM / LV ) which leads to
LM V V
= = (3.36)
LV V + L F
In which F = V + L is the amount of both phases present (when is set F = 1

then this is vapour fraction). The T x1 y1 plot starts with T2sat on the left
and ends with T1sat on the right. The bubble point and dew point curves
connect to these two points.
Fix the pressure P and split the temperature range up into increments
between T1sat & T2sat ; calculate P1sat & P2sat at each chosen temperature
increment using the Antoine equation.
The total pressure P was initially fixed and is the sum of the partial pressures
and, for an ideal system, each is given by Raoults Law. Now solve for x1 as
follows:
P = x1 P1sat + x 2 P2sat
P = x1 P1sat + (1 x1 ) P2sat
x1 ( P1sat P2sat ) = P P2sat
P P2sat
x1 = ..(3.37)
P1sat P2sat

It only remains to find the vapour phase composition y1 at each temperature

increment from Raoults Law:
x1 P1sat
y1 =
P
Plot both x1 & y1 versus temperature T , join up all the points that lie on the
bubble point curve T x1 , join up all the points that lie on the dew point curve
T y1 and then connect the two curves at T2sat & T1sat .
Notice that the two degrees of freedom have been fixed by specifying P & T
( P is fixed and then T is incremented). Thus, it is possible to solve for
x1 & y1 directly without iteration.
However, it is important to make sure that, at the fixed pressure P , the

temperature T is incremented between the pure species BP, that is
T2sat & T1sat . This is the quickest way of plotting a T x y diagram.
The MVC has the lower BP at pressure P , while the LVC has the higher. In
the case of the T x y diagram shown previously, component 1 is the
MVC while 2 is LVC.
If this convention is adopted, then the T x y diagram will slope downwards

as the concentration of MVC increases; not all textbooks follow this
convention.

3.5.3 Example 3A.1
Use Raoult's Law and the vapour pressure given below to produce a
T x1 y1 diagram for the ethanol/water system at 2000 mmHg.
Vapour Pressure Data.
Temperature Ethanol Water

o
( C) Vapour Pressure (mmHg)
104.8 2000.00 899.70
110.0 2372.95 1074.41
115.0 2786.81 1267.84
120.0 3257.91 1488.98
125.0 3792.01 1740.72
129.6 4344.22 2000.00
Assume that Ethanol is component (1) and Water is component (2).


3.5.4 The x-y Diagram
A T x1 y1 diagram is actually a plot of different tie-lines versus

temperature. The right hand end of all tie lines join up to form the upper dew
point curve, the left hand end all join up to form the lower bubble point curve.
Constant P
1.0
On the diagonal
line:
y1
x1 = y1
x2 = y 2
0
0 1.0
x1
An x1 y1 diagram is simply a plot of liquid concentration x1 against vapour

concentration y1 . A quick way of sketching an x y diagram is as follows:
The K i - value for component 1 is given by
y1
K1 =
x1
While the K i -value for component 2 is given by
y2
K2 =
x2

Now for any system the relative volatility 12 of component 1 with respect to
component 2 is defined as
K1 y /x
12 = = 1 1 ..(3.38)
K 2 y 2 / x2
Re-arrange this to get
y1 x
= 12 1
y2 x2
For a binary mixture this is the same as
y1 x1
= 12
(1 y1 ) (1 x1 )
Further re-arrangement gives y1 as a function of x1 and 12 as follows:
12 x1
y1 = ..(3.39)
1 + ( 12 1)x1
A numerical value may be obtained for the relative volatility 12 (assuming an

ideal VLE system) using equation (3.31)
P1sat
12 = sat
P2
While the volatility of any species is a strong function of both temperature and
pressure, the relative volatility of an ideal VLE system is independent of
pressure and only a weak function of temperature.
In order to construct the x y diagram a single value is needed for 12 . It

is common practice to use an average value over the pure component
boiling point range.

3.5.5 Example 3A.2
Use Raoult's Law to produce an x1 y1 plot for the propane (1)/n-butane (2)
system at a pressure of 400 mmHg.
Vapour pressures can be evaluated using the Antoine equation
B
lnP sat = A
T +C
Where P sat = vapour pressure (mmHg)

T = temperature (K)
A B C
propane (1) 15.7260 1872.46 -25.16
n-butane (2) 15.6782 2154.90 -34.42


3A.6 Models of Solution Behaviour
When the vapour phase is an ideal gas and the liquid phase is non-ideal
solution then use Modified Raoults Law for VLE calculations, as follows:
y i P = i x i Pi sat .......(i = 1,2,3.....N )
Solving this equation requires knowledge of the activity coefficients (these

quantify the deviation from ideal solution behaviour). It has already been
shown that ln i mi is a partial molar property of g E / RT m :
gE
= ln g i
RT
This means that activity coefficients may be recovered from g E / RT using the
apportioning rule that applies to any partial molar property:
(
ng E / RT
ln g i =
)

ni T , P ,n j
The solution property g E / RT depends on T and liquid phase composition

xi only, that is g E / RT does not depend on pressure P (so long as the
pressure is not too high).
One approach is to express the functional dependency of g E / RT as a power

series; the one below is called a Redlich-Kister expansion.
gE
= B + C (x1 x 2 ) + D(x1 x 2 ) + ........
2
x1 x 2 RT
3.6.1 The One Parameter Model
The simplest power-series model that could be used is to drop all the higher
order terms, i.e. C = D = .... = 0 , as follows:
gE
= B ..(3.40)
x1 x 2 RT
In which B is a constant that is fitted to the experimental data.

Thus, if the correlating function g E / x1 x 2 RT were plotted as a function of

composition x1 , then it would plot as a horizontal straight line. Notice that as
purity is approached in either component x i 1 , then g E / RT 0 and
solution becomes ideal.
In order to apply the apportioning rule (and recover the activity coefficients),
then this expression must be re-arranged in terms of mole number
gE nn
= x1 x 2 B = 1 2 2 B
RT n
and
ng E n1 n 2
= B
RT n
Substitute the last expression into the apportioning rule leads to
n1 n 2

n
ln 1 = B
n1

T , P , n2
Upon differentiation this becomes
n nn n n
ln 1 = B 2 1 2 2 = B 2 1 1
n n n n
After differentiation move back into mole fractions
ln 1 = Bx 2 (1 x1 )
ln 1 = Bx 22 (3.41)
In exactly the same way
ln 2 = Bx12 .(3.42)
Notice that as purity is approached in either component x i 1 then ln i 0

and i 1 . Once again as the solution approaches purity, in whichever
species, then it starts to behave as an ideal solution.

On the other hand, as x1 0 then component 1 approaches infinite dilution

and ln 1 ln 1 . Thus, from equation (3.41) the infinite dilution activity
coefficient for species 1 is given by
ln 1 = B ..(3.43)
In exactly the same way, the infinite dilution activity coefficient for species 2
is found from equation (3.42) and is given by
ln 2 = B ..(3.44)
Thus, the infinite dilution activity coefficients (the intercepts of the activity
coefficient curves as x i 0 ) are given by equations (3.43) and (3.44) while
each activity coefficients change with x i in accordance with (3.41) and (3.42).
This model leads to symmetrical activity coefficient curves. In addition, setting

x1 = x 2 = 0.5 it is easy to show that ln 1 , ln 2 and g E / RT must all pass
through the same point corresponding to a y-axis value of 0.25 B :
Constant T
Few systems will

/ x1 x2 RT = B
E actually behave in
ln 1 = B X X ln 2 = B
this way.
This model has

ln 2
only one adjustable
ln 1 parameter B .
Dimensionless
Values all Curves The adjustable
E / RT parameter is called
a Binary Interaction
Parameter BIP.
0
0 1.0
x1
The activity coefficient curves are parabolic and, through equations (3.41) or
(3.42), a single VLE measurement is sufficient to find B .
Positive deviation from ideality produces g E / x1 x 2 RT > 0 , g E / RT > 0 ,

ln i > 0 and i > 1 . The reverse is the case for negative deviation from
ideality see Appendix B for further details.

3.6.2 Margules Two Parameter Model
The symmetric behaviour shown on the previous diagram is restrictive and it is

unlikely that both components will behave in the same way. Other, more
complicated models are needed for asymmetric activity coefficients.
If B & C are retained in the Redlich-Kister expansion (and all other higher
order terms are set to zero), then it becomes
gE
= B + C ( x1 x 2 )
x1 x 2 RT
For a binary this is the same as
gE
= B( x1 + x 2 ) + C ( x1 x 2 )
x1 x 2 RT
Gathering terms together terms in x1 & x 2 leads to
gE
= (B + C )x1 + (B C )x 2
x1 x 2 RT
Which is usually written as
gE
= A21 x1 + A12 x 2 ....(3.45)
x1 x 2 RT
Equation (3.45) is known as the Margule model of solution behaviour. Notice

that this expression has two adjustable BIPs, namely A12 & A21 .
As before, differentiate equation (3.45) and rearrange to find the Margule

activity coefficient relations first for component 1:
ln 1 = (2 A21 A12 )x 22 + 2( A12 A21 )x 23 (3.46)
And then for component 2:
ln 2 = (2 A12 A21 )x12 + 2( A21 A12 )x13 (3.47)

As before, equations (3.46) and (3.47) yield infinite dilution activity coefficients
ln i as follows:
ln 1 = A12 (3.48)
ln 2 = A21 (3.49)
Plotting everything on one diagram gives
Constant T
X ln 2 = A21

E / x1 x2 RT
ln 2
ln 1 = A12 X
E / RT
Dimensionless
Values all Curves
ln 1
0
0 1.0
x1
Notice that ln i curves join the correlating function g E / x1 x 2 RT at the infinite

dilution activity coefficients; thus, ln 1 = A12 and ln 2 = A21 .
Alternatively, if one set of ln 1 & ln 2 measurements have been made, then

equations (3.46) and (3.47) may be solved simultaneously:
(x 2 x1 ) 2 ln 2
A12 = 2
ln 1 + (3.50)
x 2 x1
(x1 x 2 ) 2 ln 1
A21 = 2
ln 2 + (3.51)
x 1 x2
Although BIPs may be found in this way for examination purposes, it is poor
industrial practice to use only a single VLE measurement. Better practice is to
regress the BIPs to a complete VLE data set (after it has been checked for
consistency).

3.6.3 Van Laars Two Parameter Model
The Van Laar model may also be used to develop the Van Laar activity
coefficient expressions ln i as follows:
2
A x
ln 1 = A12 1 + 12 1 .....(3.52)
x2
A21
2
A x
1 + 21 2
ln 2 = A21 ..(3.53)
A12 x1
In which A12 & A21

are the two adjustable Binary Interaction Parameters
(BIPs). These parameters are again best found by regressing experimental
binary VLE data to this model.
As before, the infinite dilution activity coefficients for Van Laar are given by
ln 1 = A12 (3.54)
(3.55)
ln 2 = A21
If one set of ln 1 & ln 2 measurements have been made, then equations

and A21
(3.52) and (3.53) could be solved simultaneously to find A12
2
x ln 2
A12 = ln 1 1 + 2 .....(3.56)
x1 ln 1
2
x ln 1
= ln 2 1 + 1
A21 .(3.57)
x 2 ln 2
However, using one set of experimental measurements is prone to error and

(as was the case for Margules model) it is recommended that a full data set
be regressed to the Van Laars model in order to find these BIPs.

Finding BIPs from Minimum Data
Although not recommended, the expressions below show how the BIPs might
be estimated using a single VLE experimental result. First consider modified
Raoults Law written for species 1
y1 P = 1 x1 P1sat
y1 P
1 = ..(3.58)
x1 P1sat
In order to solve equation (3.58) for 1 the T & P will need to be measured, in
addition both liquid and vapour samples must be taken and analysed for
x1 & y1 respectively.
Likewise the activity coefficient 2 is found using the same VLE data
y2 P
2 = ..(3.59)
x 2 P2sat
Or,
(1 y1 ) P
2 =
(1 x1 ) P2sat
If the system forms an azeotrope, then by measuring the T & P and the liquid
phase composition x1 (at the azeotrope), a pair of 1 & 2 values can be
found from the expressions below (at the azeotrope y1 = x1 and y 2 = x 2 ):
P
1 = ..(3.60)
P1sat
P
2 = .....(3.61)
P2sat
Once 1 & 2 are known at liquid composition x1 the BIPs for Margules
model may be found from equations (3.50) and (3.51).
Likewise the BIPs for the Van Laar model may be estimated from equations
(3.56) and (3.57).

3.6.4 Wilsons Local Composition Model
Margule and van Laar equations are simple power-series models that assume
random mixing; although they have been extended there is no theoretical
basis for extending them to multi-component systems at different temperature.
Local composition models take into account that the microscopic local
composition differs from the macroscopic composition. This non-random
mixing effect is caused by:
Differences in molecular size and shape.
Differences in inter-molecular forces of attraction.
A key advantage of all local composition models is that binary pair data at one
temperature may be extrapolated to multi-component systems at a different
temperature the simplest local composition model is Wilsons model:
gE
= x1 ln ( x1 + x 2 12 ) x 2 ln ( x 2 + x1 21 ) ...(3.62)
RT
The Wilson model activity coefficient expressions are then
12 21
ln 1 = ln ( x1 + x 2 12 ) + x 2 .(3.63)
1
x + x
2 12 x 2 + x
1 21
12 21
ln 2 = ln ( x 2 + x1 21 ) x1 .(3.64)
x1 + x 2 12 x 2 + x1 21
The infinite dilution activity coefficients for the Wilson model are given by
ln 1 = ln 12 + 1 21 (3.65)
ln 2 = ln 21 + 1 12 (3.66)
The two BIPs for the Wilson model 12 & 21 are obtained using non-linear
regression techniques; once obtained they are readily extrapolated to multi-
component systems at different temperatures.
Wilson equations are only suitable for VLE applications. If Liquid-Liquid

Equilibrium LLE is to be modelled, then use the three-parameter NRTL
model (this will be introduced in Topic 3B) or UNIQUAC/UNIFAC models.
A fully worked out solution using the gamma-phi approach to VLE and the
Wilson equation appears in Appendix C. This example features an azeotrope.

3.6.5 Example 3A.3
Using the van Laar method and the following vapour pressure data determine
the composition of the vapour stream in equilibrium with a liquid containing a
water mole fraction of 0.5.
Vapour Pressure
Temperature
o (mmHg)
C
Water (1) Ethanol (2)
78.15 330 751
78.4 334 760
80 355 805
90 526 1153
100 760 1648
Take it that an azeotrope exists at 78.13oC and that, for this T x y data
set, the total pressure remains constant at 760 mmHg.
Take it that, under these conditions, the system obeys Modified Raoults Law.

3A.7 Tutorial 3A
1. Use Raoults law to evaluate vapour-liquid equilibrium data for the

water-phenol system at 1 bar. Compare your calculated data with the
x y diagram provided (below) and comment on the assumptions
implied in the method of calculation.
B
log 10 ( Psat ) = A P sat (mmHg) T (C)
T +C
Component A B C
Water 8.10765 1750.226 235.000
Phenol 7.13301 1516.790 174.954
(Answer: see summary answers)
2. Benzene and toluene form an ideal liquid solution. Calculate the

compositions of the vapour and liquid in equilibrium at 760 mmHg total
pressure and at temperatures of 80oC, 92oC, 100oC and 110.4oC.
Vapour Pressure Data (mmHg)

Temperature (oC) Benzene Toluene
80.0 760 300
92.0 1078 432
100.0 1344 559
110.4 1748 760

3. Using the Van Laar equations, calculate the x y equilibrium

compositions at atmospheric pressure for a mixture of chloroform and
acetone. Chloroform and acetone form an azeotrope containing 0.666
mole fraction of chloroform at 64.5oC.
Vapour Pressure Data (mmHg)

Temperature (oC) Acetone Chloroform
45.0 510.5 439.0
50.0 612.6 526.0
55.0 --- 625.2
56.3 760.0 ---
60.0 860.6 739.2
60.9 --- 760.0
70.0 1190.0 1019.0
80.0 1611.0 1403.0
4. Experimental investigation shows that at 100.7oC and 760 mmHg, a

liquid of mole fractions 0.568 n-heptane and 0.432 toluene exists in
equilibrium with a vapour of mole fractions 0.637 n-heptane and 0.363
toluene.
Use Margules model to evaluate the composition of vapour in
equilibrium with a liquid containing mole fractions of 0.185 n-heptane
and 0.815 toluene.
Vapour pressure data at 100.7oC:
n-heptane 780 mmHg
toluene 432 mmHg
(Answer: y 1 = 0.4347 and y 2 = 0.5653)
5. Di-chloroethane and methanol form an azeotrope at 760 mmHg and

41.9oC containing a di-chloroethane mole fraction of 0.769. A
laboratory experiment carried out at 760 mmHg gave the following
equilibrium point:
x DCE = 0.294 and y DCE = 0.653
(Cont..)

Use the azeotrope data and the Van Laar equations to predict the
composition of the vapour in equilibrium with the liquid above (0.294
mole fraction). Compare the predicted and experimental data and
comment on their accuracy.
B
log 10 ( Psat ) = A
T +C
In which P sat (mmHg) and T (C)

Component A B C
Di-chloroethane 6.96513 1141.984 231.93
Methanol 7.87863 1473.110 230.00
(Answer: y 1 = 0.711, Comment.)
6. Using a laboratory still operating at 760 mmHg, experimental vapour-

liquid equilibrium data for the propanol-water system has been found
for two different liquid compositions as shown below:
x (Propanol) y (Propanol) Temperature (oC)

0.36 0.42 87.7
0.52 0.44 87.9
Using the Van Laar equations evaluate the activity coefficients of both
components and hence the vapour composition in equilibrium with a
liquid with a propanol mole fraction of 0.60 at 760 mmHg.
B
log 10 ( Psat ) = A
T +C
In which P sat (mmHg) and T (C).
Component A B C
Water (1) 7.96681 1668.21 228.0
Propanol (2) 7.99733 1569.70 209.5
(Answer: 1 = 2.001 , 2 = 1.148 , y 1 = 0.518 and y 2 = 0.482)

3A.8 Bibliography




ed., 1998.


Appendix A
Real Mixtures

A.1 Property Changes on Mixing
There are two ways of calculating the properties of real mixtures:

1. By measuring the property change relative to the pure components
properties at the same T & P as the mixture.
2. By measuring the property change relative to the property of an ideal
solution at the same T , P & xi as the mixture.
The first case leads to the so-called property change on mixing, the second
case leads to the so-called excess properties.
Both sets of properties have been studied experimentally and extensive

reference works have been compiled for many different mixtures. These data
sets have been regressed to fit a variety of correlations.
Any property change on mixing is defined as
m = m xi mi .(A.1)
i
In which m is the real solution property at T , P & xi and mi are the pure
component properties of each species (outside the mixture), but at the same
T & P as the mixture.
The property change on mixing m is the property change observed

(experimentally) when 1 mol of the mixture is formed at constant T & P from
the pure components that make-up the mixture.
However, in Topic 2 it was shown that the real solution property m is the mole
fraction weighted partial molar properties mi of the species making up the
mixture:
m = x i mi
i
Substitute this into equation (A.1) to get
m = xi m xi mi
i i
m = xi (mi mi ) .(A.2)
i

Equation (A.2) is identical to the recombination rule for partial molar

properties, in which m is any property (in this case a property difference)
___ ___
and mi is its corresponding partial molar property ( mi = mi mi )
___
m = xi mi .(A.3)
i
Given that m is any property change on mixing, it follows that
v = v xi vi v = xi (vi vi )
i i
u = u xi u i u = x i (u i u i )
i i
h = h xi hi (
h = x i hi hi )
i i
s = s xi s i s = xi (s i s i )
i i
g = g xi g i g = xi ( g i g i )
i i
The equations on the left enable the various property changes on mixing to be
experimentally measured. The equations on the right may be used in the
section below.
A.2 Property Changes on Mixing for Ideal Solutions
An ideal solution is one where there are no volume change or heat effects
during the mixing process. It follows that the partial molar volume, internal
energy and enthalpy must be equal to their respective pure component values:
viid = vi (A.4)a
u iid = u i (A.4)b

and,
hi id = hi .(A.4)c
The two properties for which this is not true are entropy and Gibbs free energy.
Standard texts (Smith, Van Ness, Abbott, 2005) show that these expressions
are given by
s iid = s i R ln xi ..(A.4)d
g iid = g i + RT ln xi .(A.4)e
Substitute each partial molar property expression miid (listed above) into its
corresponding m expression (listed on right side of table) to get the following
ideal solution property changes on mixing m id :
v id = 0 (A.5)a
u id = 0 (A.5)b
h id = 0 (A.5)c
Again, the two ideal solution properties that behave differently are the entropy
and the Gibbs free energy as given below:
s id = R x i ln xi ........(A.5)d
i
g id = RT xi ln xi ..(A.5)e
i
An ideal solution is one where there is no volume change on mixing effects

(when pure species are mixed together); there are also no enthalpy and
internal energy change on mixing effects (heat effects on mixing).
Mixing, however, is an irreversible process and equation (A.5d) shows that in

accordance with the second law s id > 0 (each term within the summation is
negative).
It is also to be expected that the Gibbs free energy should decrease as the
pure components are irreversibly mixed together; equation (A.5e) confirms that
for an ideal solution g id < 0 .

A.3 Activity of a Species in a Mixture
The definition of the fugacity of a component in a solution was shown to be
g i = RT ln fi + i
In which i is the species specific constant of integration that depends only on

temperature. Whereas the same expression for a pure species in some
standard state is given by
g i0 = RT ln f i 0 + i
Subtract the latter equation from the former and, after cancelling out the
constant of integration i , this becomes
fi
( )
g i g i0 = RT ln (A.6)
f i0
The activity a i is defined as the ratio of the fugacity of species i in the mixture
fi at T , P & xi to the standard state fugacity f i 0 of pure i . Thus, a i is
defined as follows:
fi
a i = ..(A.7)
f i0
And, equation (A.6) becomes
( )
g i g i0 = RT ln a i .(A.8)
Equation (A.7) is a general definition for the activity of a species. However, for
VLE the standard state is chosen to be the pure species at the same T & P
as the mixture. Thus, for VLE equation (A.7) becomes
fi
a i = ..(A.9)
fi
It must be emphasised that equation (A.9) is a special case of equation (A.7).

This is because other standard states are used (as will be seen later when
studying chemical equilibrium).

It has already been shown that the fugacity of component i in an ideal solution
fi id is given by the Lewis-Randall rule; this is one model of ideality.
fi id = xi f i
Substitute the Lewis-Randall rule into equation (A.9) to get the activity of a
species in an ideal solution a iid as follows:
xi f i
a iid =
fi
a iid = xi ..(A.10)
However, if a different standard state is chosen so that f i 0 f i the more

general relationship for a iid is obtained:
xi f i 0
a =
id
i ...(A.11)
fi
Equation (A.10) leads to the suggestion that the activity is an effective

concentration but it is better to think of activity in terms of a ratio of two
fugacities; see defining expression for a i - namely equation (A.7).
Fugacity is often thought as the escaping tendency of a species from a liquid.

In this sense the activity of a mixture is the escaping tendency of the
component from the mixture, relative to its escaping tendency from the pure
liquid in some standard state.
The two (escaping tendencies) are different because in a mixture any single
molecule is partially surrounded by like and unlike components. Thus, its
escaping tendency is different (from the pure species) because it is packed
differently and experiences different forces of attraction in the mixture.

A.4 Excess Properties
A property change on mixing was defined by equation (A.1) as follows:
m = m xi mi
i
The property change on mixing is the (experimentally determined) departure of

the real solution property at T , P & x i relative the pure component properties
at the same T & P .
An excess property m E is defined in a similar way. However, it is the

(experimentally determined) real solution property m minus the corresponding
ideal solution property m id at the same T , P & x i . Thus, by definition
m E = m m id .....(A.12)
The recombination rule also applies to ideal solutions as follows:
m id = xi miid
i
Substitute equations (A.4a) through (A.4e) into the equation above to get the
various ideal solution properties listed below:
v id = xi vi ...(A.13)a
i
u id = xi u i ..(A.13)b
i
h id = xi hi ...(A.13)c
i
Thus, the previous three ideal solution properties are just the mole fraction
weighted pure component values. However, entropy and Gibbs free energy
behave differently (due irreversibilities associated with the mixing process):
s id = xi s i R xi ln xi ..(A.13)d
i i
g id = xi g i + RT xi ln xi ..(A.13)e

Substitute each of these into the general definition of an excess property,

given by equation (A.12), leads to the following excess properties:
v E = v xi vi .....(A.14)a
i
u E = u xi u i ....(A.14)b
i
h E = h xi hi .(A.14)c
i
s E = s xi s i + R x i ln xi .(A.14)d
i i
g E = g xi g i RT xi ln xi ...(A.14)e
The above equations may be used to find experimental values for m E . Notice
that in every case (on the right of each expression) there is a common group
of terms given by m i
x i mi .
Comparing this common group to equation (A.1), it is clear that this

combination of terms is the property change on mixing m . Thus, as
expected, there is a relationship between m and m E as follows:
v E = v .....(A.15)a
u E = u .....(A.15)b
h E = h ..(A.15)c
In the first three cases m E = m , but the next two are different:
s E = s + R xi ln xi ..(A.15)d
i
g E = g RT x i ln xi .....(A.15)e
Using these relations experimental values for m (mostly available in older

literature) may be converted into m E values. The benefit of using excess
properties is that for an ideal solution they are all are zero.

Appendix B
This appendix is examinable.
B.1 Positive & Negative Deviation from Ideality
If the system obeys modified Raoults Law (an ideal gas in equilibrium with a
non-ideal liquid solution), then the partial pressures are given by modified
Raoults Law and may be added together in the usual way:
P = 1 x1 P1sat + 2 x 2 P2sat (B.1)
The P x1 curve is no longer straight; this is because activity coefficients are

non-linear functions of composition. Whether the P x1 curve lies above or
below the ideal (straight line) case depends on whether i > 1 or i < 1 .
It must be emphasised that Raoults law (on its own) cannot be used to model
either positive or negative deviation from ideality; activity coefficients are
needed.
When the activity coefficients i > 1 the solution exhibits positive deviation
from Raoults Law; the partial pressures of each species exceed the Raoults
Law case which means the P x1 curve must lie above the straight line case:
Constant T
If i > 1 , then the
Bubble Point
Curve sat Py i = i x i Pi sat
and this
X P1
P x1 must now be larger
P y1 than Py i = x i Pi sat .
P2sat X Dew Point
P
Curve i > 1 means that the
escaping tendency
from a non-ideal
solution exceeds that
0 from an ideal solution.
0 x1 y1 1.0
1.0 x2 y 2 0
If i > 1 (escaping tendency is larger than ideal case), then unlike molecules
attract less strongly (have weaker interactions) than like molecules.

On the other hand, when the activity coefficients i < 1 the solution exhibits
negative deviation from Raoults Law. This leads to the P x1 y1 diagram
having a completely different shape.
In this case, the partial pressures of each species are below the Raoults Law
partial pressures and the P x1 curve must lie below the straight line case:
Constant T
If i < 1 , then the

P y1 X P1
sat
Py i = i x i Pi sat and this
Bubble Point
Curve must now be smaller
P x1 Dew Point
Curve
than Py i = x i Pi sat .
P2sat X
P If i < 1 the escaping
tendency from a non-
ideal solution is less
0 that from an ideal
0 x1 y1 1.0
solution.
1.0 x2 y 2 0
If i < 1 (escaping tendency is smaller), then unlike molecules attract more

strongly (have stronger interactions) than like molecules.
Consider a T x y diagram. In this case, the bubble point line T x1 is

below the dew point line T y1 :
Constant P Now heat the mixture

at constant overall
composition and
T2sat Dew Point D pressure.
Line
C The mixture must

T x1 follow a vertical line
L V A BC D
M T y1
T
B
Point A corresponds
A Bubble Point T
1
sat to some initial
Line subcooled liquid state.
0
At this point the liquid
0 x1 y1 1.0 is below its BP.
1.0 x2 y 2 0

Point B corresponds to the bubble-point of the mixture: reading horizontally

(to left) gives the bubble point temperature T ; reading horizontally (to right)
and then vertically down gives the bubble point composition y1 .
Further heating leads to point C on the dew point curve. Reading horizontally
(to the left) gives the dew point temperature T ; reading horizontally (to the left)
and then vertically down gives the dew point composition x1 .
Further heating leads to point D , a superheated vapour state with the same
composition y1 as the following states: initial subcooled liquid x1 at A ; the
saturated liquid x1 at B ; and the saturated vapour y1 at C .
To flash any two-phase mixture (at this fixed pressure P and the same overall
composition) the temperature must be set somewhere between the (lower)
bubble point temperature and the (higher) dew point temperature.
Notice that the mixture point M (same overall composition as A, B, C , D ) lies

between these two temperatures; it will separate (or flash) into L moles of
liquid and V moles of vapour, joined by horizontal tie-line L M V .
The concentration of the saturated liquid (that separate out from the mixture) is
read vertically down from L . The concentration of saturated vapour (that
separates from the mixture) is read vertically down from V .
To find the amounts of each phase present use the lever rule. The lever rule
depends on how exactly the horizontal tie-line L M V is bisected by the
mixing point M . The lever rule is given by
LM V
= (B.2)
MV L
The ratio ( mol V / mol L ) is the same as the ratio ( LM / MV ), i.e. the length
of the tie-line either side of the mixing point. This may also be expressed as
the ratio ( LM / LV ) which leads to
LM V V
= = (B.3)
LV V + L F
In which F is total of both phases present. If F = 1 , then V / F represents the

vapour fraction.

Flashing is usually thought of as sudden drop of pressure which releases a

two-phase mixture (from the initial subcooled liquid) into a heated flash drum;
heating (in turn) then fixes the constant temperature T of the flash.
However, the opposite may also occur: a superheated vapour may be

compressed and then split a two-phase mixture within a cooled drum; cooling
now fixes the constant temperature T of the flash.
Referring back to the previous diagram and the location of mixing point M up
and down the vertical line BC :
As M approaches the bubble point B ; LM 0 and V / F 0 , at
which point L / F 1 . Thus, if the vapour fraction is zero, then the
liquid fraction must be one (remember mass balance F = V + L ).
As M approaches the dew point C ; LM LV and, V / F 1 at
which point L / F 0 . As the vapour fraction approaches one, then
the liquid fraction must approach zero.
B.2 Formation of Azeotropes
It has already been said that negative deviation from ideality reflects a lower
releasing tendency (greater forces of attraction) of a species from a mixture
than from an ideal solution.
If the boiling points of the pure species are not too far apart, then this can lead
to the mixture (at some composition) having a higher boiling point than either
of the pure components.
At the azeotrope, the boiling point of the mixture will be greater than the boiling
points of the two pure species and will, therefore, go through a maximum on a
T x y diagram as shown below:
Constant P
Dew Point
NOTE:
T y1 Line
A maximum temperature
T x1 azeotrope on a T x y
Bubble Point diagram.
T2sat Line
T
..Will plot as a minimum
pressure azeotrope on a
0
T1sat P x y diagram.
0 x1 y1 1.0
1.0 x2 y 2 0

It has also been said that positive deviation from ideality reflects a greater
releasing tendency (lower forces of attraction) of a species from a mixture than
from an ideal solution.
If the boiling points of the pure species are not too far apart, then this can lead
to the mixture (at some composition) having a lower boiling point than either of
the pure components.
At the azeotrope, the boiling point of the mixture will be less than the boiling
points of the two pure species and will, therefore, go through a minimum on a
T x y as shown below:
Constant P
NOTE:
T2sat
T y1 A minimum temperature
azeotrope on a T x y
Dew Point
diagram..
Curve
T T x1
T1sat ....Will plot as a maximum
Bubble Point pressure azeotrope on a
Curve P x y diagram.
0
0 x1 y1 1.0
1.0 x2 y 2 0
From the above diagram, clearly both species have a K value equal to one.
Thus the composition of species 1 is the same in both phases, likewise for
species 2 ( y1 = x1 and y 2 = x 2 ). It follows that liquid and vapour
compositions must be the same.
An azeotrope is known as a constant boiling mixture; both the vapour and the
liquid phases have identical compositions. Such a mixture cannot be
separated by distillation because its relative volatility 12 = K 1 / K 2 = 1 .
Remember, distillation is a process that separates species due to differences

in volatility. If the relative volatility is one, then there is no difference in volatility
between the components and distillation alone cannot separate them.
From this discussion it is clear that the formation of azeotropes depends both
on activity coefficients 1 & 2 and the boiling points T1sat & T2sat . For this
reason it is important to develop suitable correlations for i such as the
Margule, Van Laar and Wilson expressions.

Appendix C
Gamma-Phi VLE Worksheet for Azeotrope
Modified Raoults Law
This file is rather long to be included in these lecture notes but may be
obtained from the course lecturer:
gamma-phi Worked
Example.pdf

TOPIC 3B: VLE, LLE and Miscibility 1
Topic 3B
Vapour-Liquid Equilibrium, Liquid-

Liquid Equilibrium and Miscibility
Contents
TOPIC 3B: VAPOUR-LIQUID EQUILIBRIUM, LIQUID-LIQUID EQUILIBRIUM
AND MISCIBILITY
3B.1 INTRODUCTION ............................................................................................................... 3
3B.2 PURE SPECIES ............................................................................................................... 4
3.2.1 P-T Diagrams ...................................................................................................... 4
3.2.2 P-v Diagrams ...................................................................................................... 6
3B.3 BINARY SYSTEMS AND MISCIBLE VLE .............................................................................. 7
3.3.1 P-T Diagrams ...................................................................................................... 8
3.3.2 P-x-y Diagrams.................................................................................................. 13
3.3.3 T-x-y Diagrams .................................................................................................. 14
3.3.4 x-y Diagrams ..................................................................................................... 15
3.3.5 Azeotropes ........................................................................................................ 15
3B.4 PARTIALLY MISCIBLE SYSTEMS ...................................................................................... 20
3.4.1 Liquid-Liquid Solubility Diagrams ....................................................................... 20
3.4.2 VLLE and P-x-y Diagram ................................................................................... 22
3.4.3 VLLE and T-x-y Diagram ................................................................................... 24
3.4.4 VLLE at Higher Temperature ............................................................................. 26
3.4.5 Quantitative Treatment of VLLE ........................................................................ 28
3.4.6 Example 3B.1 .................................................................................................... 34
3B.5 IMMISCIBLE SYSTEMS .................................................................................................... 36
3.5.1 Quantitative Treatment ...................................................................................... 37
3.5.2 Example 3B.2 .................................................................................................... 39
3B.6 TUTORIAL 3B ................................................................................................................ 40
3B.7 BIBLIOGRAPHY .............................................................................................................. 41
APPENDIX A.......................................................................................................................... 42
Rigorous Approach to LLE ......................................................................................... 42
A.1 Liquid-Liquid Equilibrium ...................................................................................... 43
APPENDIX B.......................................................................................................................... 46
Heating or Cooling a VLLE System. ........................................................................... 46
B.1 Isobaric Cooling from A ..................................................................................... 47
B.2 Isobaric Cooling from B ..................................................................................... 48
B.3 Isobaric Cooling from C ..................................................................................... 49
B.4 Isobaric Vaporisation from X .............................................................................. 50

Learning objectives
After studying this VLE, LLE and Miscibility topic you should be able to:
Differentiate phase diagrams for pure species from phase diagrams for
mixtures.
Sketch P-T-x-y diagram and show how P-x-y and T-x-y diagrams are
related to the three dimensional P-T-x-y surface.
Sketch diagrams, including x-y diagrams, where one or both of the
species are supercritical.
Sketch P-T diagrams for mixtures and explain retrograde condensation
and the position of the critical point in relation to the cricondenbar and
the cricondentherm.
Explain how changing the mixture composition affects the shape of the
P-T diagram and the importance of these diagrams to Petroleum
Engineers.
Identify which systems can form azeotropes, the location of the
azeotrope and then demonstrate an ability to sketch azeotropic
behaviour on P-x-y, T-x-y and x-y diagrams.

3B.1 Introduction
The following types of system will be studied in this topic: fully miscible
systems; partially miscible systems; completely immiscible systems. The
objective is to review, qualitatively and quantitatively, both VLE and LLE.
VLE was studied quantitatively, for fully miscible systems, in Topic 3A. This
was the so-called gamma-phi approach to VLE. The assumptions and the
complexity of the system depended on the pressure and the chemical make-
up of the system.
In real situations, large deviations from ideal behaviour occur in both gas and
liquid phases this is because of forces that exist between molecules:
If forces between like and unlike molecules are the same, then the
solution behaves as an ideal solution.
If these forces are not the same, then deviations from ideality will
occur.
The magnitude of the molecular forces depend on both the nature of
the molecules themselves and the distances between them.
In the liquid phase, the molecules are close together and these forces have a
large effect. Thus, the molecules that make up a liquid phase mixture must be
very similar in shape and size for them to behave as an ideal solution.
Most liquid mixtures have molecules that are sufficiently different to cause
deviations from ideal solution behaviour. Aqueous solutions are nearly always
non-ideal.
In gaseous mixtures, force fields are less strong and ideal solution behaviour
is a good approximation at low pressures. At higher pressures, the molecules
are closer together and difference in molecular shape and size exert an
influence on intermolecular forces.
The situation may be summarised as follows:

1. At low pressures, deviations from ideal solution behaviour will occur
primarily in the liquid phase. The gas phase will approach an ideal
solution and may even behave as an ideal gas.
2. At high pressures, both liquid and gas phases may deviate from ideal
solution behaviour. The extent of the deviation will depend on the
dissimilarity of the components.
3. The most difficult problems are dissimilar species at high pressures,
where there are large deviations from ideality in both phases.
Partially miscible and immiscible systems will be discussed later in this

topic, but first it is necessary to complete a qualitative review of fully miscible
systems (this was started for polar systems in Topic 3A).

3B.2 Pure Species
For any pure-component, pressure-volume-temperature data plots as a three

dimensional P-v-T surface, see below:
C
Liquid Gas
V-L
Psat
Vapour
Three dimensional diagrams are difficult to work with and it is normal to project
the above diagram onto a plane, as in P-T or P-v diagrams shown in topic 2.
Of course these were for pure component systems.
3.2.1 P-T Diagrams
A P-T diagram is a result of projecting the P-v - T surface on to the

pressure-temperature plane, for a pure species this is given by:
Supercritical C is the critical point.

P A region
C
Liquid TP C is the
region vaporisation or boiling
Solid
region Gas point curve.
region
B
TP TP A is the fusion
Vapour or freezing point curve.
region
T TP B is the
sublimation curve.
The areas on this P-T diagram represent single phase regions, while the lines
represent two phases in equilibrium. The Triple Point, denoted by TP , is the
single point on the diagram where three phases are in equilibrium.

Line TP C is actually the experimental vapour pressure P sat curve; values

of P sat at any temperature T (along this curve) may be estimated using the
Antoine vapour pressure correlation as discussed in Topic 1.
Line TP C represents the only possible combinations of T & P where

liquid and vapour may co-exist in equilibrium; any point on this curve must
satisfy f V = f L . Thus, there is one degree of freedom.
Line TP A represents the only combinations of T & P where liquid and solid
may co-exist in equilibrium; any point on this curve must satisfy f L = f S .
Again there is one degree of freedom.
Line TP B represents the only combinations of T & P where vapour and

solid may co-exist in equilibrium; any point on this curve must satisfy
f V = f S . Once again there is one degree of freedom.
At the triple point there are three phases in equilibrium, so that f V = f L = f S .

At the triple point there are zero degrees of freedom. Only one fixed T & P
will simultaneously satisfy the two independent phase constraints.
Consider a process starting at point 1 in the solid region and undergoing a

constant pressure heating process. As heat is added the solid temperature
rises until point 2, the melting point, is reached:
Supercritical
P A region
C
Liquid
region
Solid
region Gas
1 2 3 4
region
TP
B
Vapour
region
Further addition of heat causes the solid to melt; melting, like vaporisation and
sublimation, takes place at constant temperature and pressure.
Eventually all the material is liquid. The amount of heat needed to melt all the
solid is called the latent of fusion or the latent heat of melting. These phase
transitions were mentioned in earlier courses (Process Industries).

From this point onwards, as more heat is added, the temperature of the liquid
rises from 2 to 3. Liquid between these two points is termed subcooled or
unsaturated liquid.
Point 3 corresponds to the boiling point of the liquid and this is the only
temperature, at this pressure, where vapour and liquid can co-exist in
equilibrium.
Boiling continues at constant temperature and pressure until vaporisation is

complete. The heat needed to completely vaporise all the liquid is called the
latent heat of vaporisation. From 3 to 4 the vapour becomes superheated.
3.2.2 P-v Diagrams
If the three dimensional P-v - T surface is projected on to the pressure-

volume plane, then the result is a P-v diagram:
P
TC Supercritical Curve A C is saturated
Fluid liquid curve.
PC C
Gas Curve B C is saturated

vapour curve.
Liquid
Vapour
Point C is critical point. The
highest T & P at which
PTP A vapour and liquid can co-
B
exist in equilibrium
The saturated liquid curve A C and the saturated vapour curve B C start
at the triple point TP and meet at the critical point C . At the critical point both
liquid and vapour phase properties become identical.
To the right of the saturated vapour curve (but at temperatures below the fluid
critical temperature TC ) is the superheated vapour region. To the left of the
saturated liquid curve is the subcooled liquid region.
Above the critical temperature TC , but below the critical pressure PC , the fluid
is called a gas because it cannot be liquefied by the action of pressure alone.
Above the critical temperature TC and above the critical pressure PC , the fluid
is called a supercritical. At these high pressure states, gaseous properties
grade into liquid properties without any phase splitting.

3B.3 Binary Systems and Miscible VLE
For a two-component system another variable must be added to the phase

diagram. Thus, the diagram becomes a three dimensional P T x y
diagram. The diagram has thee axes pressure temperature and composition.
Often component 1 is the MVC, so that component 1 will have the higher
vapour pressure P sat , as is shown below:
The upper surface

represents all saturated
liquid states (bubble
point states).
P Above this are all the

C2 subcooled liquid states.
C1
The lower surface
represents all saturated
sat P1sat
T vapour states (dew
P 2 point states).
0
x1 y1 Below this are all the
1 superheated vapour
states.
The region of space enclosed by these two surfaces is the two-phase region.
The right hand vertical plane is the vapour pressure curve for pure 1; notice
that it terminate at the critical point for component 1, denoted by C1 .
Likewise the left hand vertical plane is the vapour pressure curve for pure 2;
it terminates at the critical point for component 2, denoted by C 2 .
Three dimensional diagrams are not convenient. Thus, it is common practice

to slice through this 3-D surface using planes where one of the variable is
held constant, as follows:
A P x y diagram is produced by a vertical slice at constant
temperature
A T x y diagram is produced by a horizontal slice at constant
pressure.
A P T diagram is produced by a vertical slice at constant
composition.

3.3.1 P-T Diagrams
A vertical slice at constant composition produces the pressure-composition

diagram shown below (the pure component vapour pressure curves are added
for convenience):
CP
Curve X C1 is the vapour
P
C1 CM pressure curve for pure
species 1 (MVC).
Liquid CT
Curve S C 2 is the vapour
pressure curve for pure
C2 species 2 (LVC).
X W
V Envelope W-CP-CM-CT-V
S Vapour corresponds to the mixture
phase envelope; all points
T have the same composition.
A pure component critical point (in this case either C1 or C 2 ) is the highest
temperature and pressure at which a pure substance can co-exist as vapour
and liquid in equilibrium.
However, this is not the case for a mixture: the highest pressure is CP which
is called the cricondenbar. On the other hand, the highest temperature is CT
which is called the cricondentherm. Now mixture VLE is represented by an
envelope and not a curve.
The upper and lower parts of this phase envelope W-CP-CM-CT-V represent
all the saturated liquid (bubble point) and saturated vapour (dew point)
states respectively note that these all have the same composition.
The two saturation curves join at the mixture critical point CM. At this point,
saturated vapour and and saturated liquid properties become identical:
Above and to the left of the envelope is the liquid region. Below and to the right
is the vapour region. Inside the envelope is the two-phase region.
Quality lines are lines of constant liquid fraction notice that they are drawn
inside the envelope and all converge at the mixture critical point CM.

As seen before in Topic 1, for a pure component phase changes occur at a

constant T & P . However, this is not the case for a mixture (in this case a
binary mixture) where the initial and a final boiling points are different.
Consider the constant composion system below. It is maintained at a constant

pressure and gradually heated from a subcooled liquid through to a
superheated vapour (as indicated by crosses):
The initial boiling point is the

CP bubble point temperature
CM Tbp .
Liquid
CT
Bubble Point
The final boiling point is the
Curve dew point temperature Tdp .
P
Vapour Unlike pure components,
mixtures boil over a definite
Dew Point temperature range, not at a
Curve
fixed temperature.
Tbp T
Tdp
Consider a subcooled liquid undegoing an isothermal decrease of pressure

starting starting from point A:
The decrease in pressure

moves the sytem to point
P CP B on the bubble point line.
A CM
Liquid This is also called the
B CT saturated liquid line and the
Bubble Point mixture is 100% liquid.
Curve
Vaporisation begins at a
Dew Point point fractionally below B.
D
Curve
Vapour
Hence, vaporisation or
E boiling of the mixture starts
at B.
T
If the pressure is decreased further (following vertical line B-D), then more
and more liquid vaporises until the system arrives at point D which is on the
dew point curve.

Fractionally above point D the last drop of liquid will have vaporised, so that
point D itself is 0% liquid hence, vaporisation or boiling stops at D.
As the pressure is decreased further the system moves along path D-E. At
point E, the system is a superheated vapour indeed all states below D are
superheated vapour states (point D itself is a saturated vapour state).
The reverse process starts at E and the system moves up from E-D with
increasing pressure. When the system pressure rises to D it is on the dew
point curve corresponding to 0% liquid.
As the system moves fractionally away from D the vapour starts to condense
and dew will form on any exposed surface (thus dew point) hence, this is
the initial condensation point.
As the pressure increases from D-B, there is a steady increase in the fraction
of liquid until, at the bubble point B where there is 100% liquid. The bubble
point B is where condensation stops.
As the pressure is increased further the system moves along path B-A. At
point A, the system is again a subcooled liquid indeed all states above B
are subcooled liquid states (point B itself is a saturated liquid state).
Thus, the bubble point is the initial boiling point (when pressure is decreasing)
or the final condensation point (when pressure is increasing).
Likewise, the dew point is either the final boiling point (when pressure is
decreasing) or the initial condensation point (when pressure is increasing).
Retrograde Condensation
It has been shown what happens as a subcooled liquid undergoes an

isothermal decrease in pressure starting from point A along path A-E:
Vaporisation starts
A at B on the
CM
B bubble point curve.
Vaporisation stops
at D on the dew
point curve.
The liquid fraction

changes smoothly
from 100% to 0%
D Vapour Gas from B to D.
E

Note the subcooled liquid A is to the left of the mixture critical point CM.
On the other hand, starting with a fluid at point 1, which is a little to the right
of the mixture critical point CM, following a similar vertical path (now 1-2-3),
something completely different happens:
Although point 1 is still above the phase envelope, it does not represent a
subcooled liquid since it lies above the dew point curve. It is really a
supercritical fluid.
The temperature must still be less than the cricondentherm (as shown),
otherwise the vertical line will not intersect the two-phase envelope, but it is
now greater than the mixture critical temperature CM (if the line is to the left
of CM, then the previous scenario will occur).
The main point to notice is that as the pressure is reduced the superheated
vapour, being to the right of the mixture critical point, will now intersect the
two-phase envelope at the dew point curve.
Once fractionally below the dew point curve condensation will start. It is
unusual for a fluid to condense when its pressure is reduced, but this must
happen over a limited temperature range (see shaded region)
Further pressure reduction will result in substantial quantities of liquid being

formed up to some maximum. In this case, as the pressure drops, the liquid
fraction increases from 0% (at the dew point curve) to a maximum of about 5%
(see point 2 on the diagram).
Futher pressure reduction below 2 causes this condensed liquid to

re-evaporate until the mixture again intersects the lower part of the dew point
curve. At the dew point there is again 0% liquid (or 100% vapour).
Further expansion below the dew point pressure leads to point 3 which again
produces a superheated vapour state.
As a two phase mixture expands within the two-phase envelope, it usually

vaporises. This phenonena, of the mixture condensing with decreasing
pressure, is the opposite of what one would expect and leads to the
expression Retrograde Condensation.
Petroleum Engineers are interested in high pressure situations (near the

critical point) and phenomena such as retrograde condensation is important to
field development plans; thus, the importance of P T diagrams to them.
Chemical Engineers tend to operate at processes at lower pressure and are

more interested in P x y , T x y and x y diagrams; these are
mainly used in the design and operation of distillation columns and flash
drums.

The P T diagram below represents a binary ethane/n-heptane system at a

variety of compositions (the pure ethane and n-heptane vapour pressure
curves are also shown):
Shown are three separate

phase equilibrium
envelopes for different
mixtures.
The locus of the mixture

critical points is shown by
the outer curve.
The mixture critical points

are each a tangent to this
outer curve.
Notice that the mixture

critical properties need not
lie between the pure
component values.
The envelope A 1 B 1 has mixture critical point C 1 and is 90.22% (w/w)

ethane. The envelope A 2 B 2 has a mixture critical point C 2 and is 50.25%
(w/w) ethane. The envelope A 3 B 3 has mixture critical point C 3 and is 9.78%
(w/w) ethane.
The curves labelled A are bubble point curves, while all the curves labelled
B are dew point curves.
Also, the relationship between the mixture critical point, the cricondenbar and
the cricondentherm (for each envelope) changes as the composition of the
envelope changes.
Wherever the bubble point curve of one envelope intersects with the dew point
curve of another envelope, then the two phases are in VLE at the same
T & P ; this situation corresponds to two phases in equilibrium at either end of
a tie line on a T x y or a P x y diagram.

3.3.2 P-x-y Diagrams
Slicing vertically through the 3-D surface, at constant temperature, produces

the well-known P x y diagram. There are actually three cases depending
on the selected temperature:
Constant T With a P x y diagram:

Case 3
The bubble point curve
is above and the dew
Case 2 point curve is below.
P2sat
The liquid region is
sat
Bubble Point X P1
Curve
P P x1 above the two phase L-
P y1 Case 1
V region.
P2sat X Dew Point
Curve
0 The vapour region is
0 1.0 below the two phase
x1 & y1 L-V region.
Case 1 is where the temperature is below the critical temperature of both pure
species. Case 2 is where the temperature is above the critical temperature of
component 1, but below that of 2. Case 3 is where the temperature
exceeds the critical temperatures of both pure species.
Take a subcooled liquid A, composition x1 and pressure P , and reduce its

pressure while keeping T constant. The process will follow a vertical path
A-B-E-F with no change in the overall composition, as shown:
Constant T
When the pressure

P A P x1 P1sat drops to Pbp , the
B mixture reaches the
Pbp D
Pdp C bubble point line at B.
E
P y1
P
F At B the liquid just
starts to vaporise.
0 The saturated liquid still

x1
0 1.0 has composition
x1 & y1 composition x1 .
However, the composition of the first bubble of vapour y1 is given by the other
end of the horizontal tie line at D (read y1 vertically below D).

Notice that this vapour at D is richer in component 1 (the MVC) than the
original saturated liquid at B.
Further reduction of pressure from Pbp Pdp follows vertical path B-E.
However, the liquid composition will now follows the bubble point curve from
B-C, while the vapour composition follows the dew point curve from D-E.
When the pressure drops to Pdp , the mixture reaches the dew point curve at
E where the saturated vapour is in equilibrium with the last drop of liquid.
Any further pressure reduction below Pdp , say to point F, will result in the
vapour becoming superheated. While individual phase compositions may vary,
the overall composition of the mixture always remains the same.
3.3.3 T-x-y Diagrams
Slicing horizontally through the 3-D surface, at constant pressure, produces

the T x y diagram. Tie-lines are again horizontal and link together phases
in equilibrium. There are three cases depending on the pressure:
Constant P
Case 1: the pressure is
below the critical pressure
of both pure species.
Case 3
Case 2: the pressure is
above the critical pressure
T1sat of component 2, but
Dew Point Case 2
T2satX Curve below the critical pressure
T
T y1 of component 1.
T x1
Bubble Point X T1sat Case 3: the pressure is
Curve Case 1
0 above the critical pressure
0 1.0 of both pure species.
x1 & y1
In a T x y diagram, the dew point curve is above and the bubble point
curve is below. The vapour region is now above the two phase L-V region
while the liquid region is below the two-phase L-V region.
If a subcooled liquid on a T x y (composition x1 ) is gradually heated while

keeping the pressure P constant. The process will follow a vertical path
upward with no change in overall composition. The process is analogous to
that already discussed in the previous P x y diagram.

3.3.4 x-y Diagrams
Using the data obtained from a T x y diagram, at constant pressure, or a

P x y diagram, at constant temperature, a plot of liquid phase mole
fraction x1 versus vapour phase mole fraction y1 may be prepared.
For the previous T x y diagram, an x y diagram may be prepared

simply by plotting the compositions from either end of the tie-lines against
each other. The diagonal x = y line is added for convenience:
Constant P
Plotting a succession of
1.0 points, from a sequence
of tie-lines, produces
the entire case 1
x y diagram.
Case 1
This can be repeated
y1 y1 for case2.
Case 3
And again for Case 3.
Case 2 The result is an x y
diagram at constant
0 pressure.
0 x1 1.0
x1
These x y diagrams correspond to three different pressures. The same

exercise could be repeated (starting with a P x y diagram) to get an x y
diagram at constant temperature.
It is not possible to construct T x y , P x y or x y diagrams for

multicomponent mixtures (more than two components). However, it is still
possible to plot a P T diagram for such a system.
3.3.5 Azeotropes
Azeotropes are formed when there is a strong departure (both positive and
negative) from ideal solution behaviour. Thus, i will either be i > 1 (positive
deviation) or they will be i < 1 (negative deviation), this was discussed in
Topic 3A and Appendix B.

If the pressure is low enough for the vapour phase to be an ideal gas then all
V = 1 and, for the liquid phase, f = f sat = P sat . Thus, the appropriate VLE
i i i i
relationship is Modified Raoults Law, as developed earlier in topic 3A:
y i P = i x i Pi sat .......(i = 1,2,3.....N )
The species K i value was previously defined as
yi
Ki =
xi
Thus, modified Raoults Law K i values are given by
i Pi sat
Ki = ..(3.1)
P
The relative volatility of component i with respect to component j , denoted

ij , is simply the volatility of i divided by the volatility of j , previously defined
in Topic 2 as
Ki
ij =
Kj
Modified Raoults Law relative volatility ij may be found by substituting

equation (3.1) into the previous expression
i Pi sat
ij = .(3.2)
j Pjsat
For a binary mixture that forms an azeotrope, the volatilities of each species
are the same K 1 = K 2 = 1.0 (literally the mixture boils with a constant
composition y1 = x1 and y 2 = x 2 ).
This means that the relative volatility at the azeotropic composition is one
( ij ) AZ = 1 and distillation alone can no longer separate the mixture; some
other technique, usually in combination with distillation, must be used.
At the azeotrope the T x y diagram will either exhibit a maximum

(negative deviation from ideality) or a minimum (positive deviation from
ideality); vice versa for P x y diagrams.

The diagrams below are taken from Appendix B of Topic 3A where there is a
full discussion of both positive and negative deviation from ideality as well as
the formation of azeotropes please read this appendix carefully:
Constant P Constant P
Dew Point
T y1 Line T2sat
T y1
T x1
Bubble Point Dew Point
T2sat Line Curve
T T T x1
T1sat
Bubble Point
T1sat Curve
0 0
0 1.0 0 1.0
x1 & y1 x1 & y1
Thus, for a binary mixture of 1 and 2 equation (3.2) becomes
1 P1sat
12 = ..(3.3)
2 P2sat
An azeotrope will form when the liquid and vapour have the same composition.
Thus, the relative volatility must be equal to one 12 = 1 , or from (3.3)
1 P1sat
=1
2
sat
AZ P2
P sat
1 = 2sat .(3.4)
2 AZ P1
In other words azeotropes form in a binary mixture at composition x AZ only

when equation (3.4) is satisfied. Thus, azeotropic behaviour depends not only
on the activity coefficients but also on the vapour pressure of the pure species.
If the boiling points of the species are not too far apart, and the deviation from
ideal solution behaviour is large enough, and if equation (3.4) is satisfied, then
an azeotrope will form.
There are two limiting values of 12 - the first is ( 12 ) x1 =0
1 P1sat
( 12 )x =0 = ...(3.5)
1
P2sat

The second limiting value is ( 12 ) x1 =1.0
P1sat
(12 )x =1.0 = .....(3.6)
1
2 P2sat
Thus, if one limiting value exceeds one while the other limiting value is less
than one, then an azeotrope must exist somewhere in the composition range
0 x1 1.0 .
If this test is satisfied, then an azeotrope must exist at some composition. To

find the exact azeotropic composition simply search for a value for x1 such
that equation (3.4) is satisfied:
1 P sat
= 2sat
2 AZ P1
In Appendix C of Topic 3A a 2-propanol (1)/water(2) system at 80oC was

shown to have an azeotrope at x1 = 0.7174. The P x y diagram shows a
maximum pressure azeotrope (minimum temperature azeotrope):
At the maximum pressure azeotrope the tie line has zero length and the
mixture boils with a constant composition. At this point x1 = y1 and x 2 = y 2 .

From this diagram the maximum pressure azeotrope occurs at 99.82 kPa, but
this valve can be found with more accuracy by setting (in modified Raoults
Law) either y1 = x1 or y 2 = x 2 thus for species 1:
P AZ = 1AZ P1sat .(3.7)
And for species 2:
P AZ = 2AZ P2sat .(3.8)
Of course, exactly the same pressure P AZ must be obtained from either

equation since, at the azeotrope, equation (3.4) must be satisfied.
If an azeotrope occurs, then the result is a constant boiling mixture where

x1 = y1 and x 2 = y 2 . For an x y diagram the curve must, therefore, cross
the diagonal line. The three x y diagrams below correspond to three cases
listed below:
1. The mixture is an ideal solution.
2. The mixture has positive deviation from ideality leading to a maximum
pressure azeotrope.
3. The mixture has negative deviation from ideality leading to a minimum
pressure azeotrope.
The x y diagram for these three cases is shown below:
Constant P
1.0
Curve 1 is the ideal
solution case (no
1 azeotope can occur).
Curve 2 is positive
2
deviation from ideality
case (maximum pressure
3
y1 azeotrope).
Curve 3 is negative
deviation from ideality
case (minimum pressure
0 azeotrope).
0 1.0
x1
Notice that in case 2 before the azeotrope species 1 is MVC, while after the
azeotrope species 1 is LVC; at the azeotrope the volatilities are the same.

3B.4 Partially Miscible Systems
The liquid phase will only split into two immiscible or partially miscible phases
if the system has strong positive deviation from ideality (where unlike species
have little affinity for each other), see Appendix B of Topic 3A.
It is interesting to consider the phase behaviour of some alcohol /water

systems as the non-polar end of the alcohol gets longer, i.e. methanol/water,
ethanol/water, propanol/water and finally butanol/water systems:
All show positive deviation from ideality with increasing activity
coefficients as the non-polar end of the alcohol gets longer.
The methanol/water system does not form an azeotrope, while
ethanol/water and propanol/water do form maximum pressure
azeotropes. However, all of these mixtures are miscible.
As the non-polar end of the alcohol gets longer, the affinity between the
different alcohol/water systems gets less and less.
With the butanol/water system the affinity between the two species is
so low that, depending on the composition, the liquid phase may split
into two partially miscible liquid phases; one butanol-rich, the other
water-rich.
Depending on the temperature of the mixture, butanol/water mixture can also

form a so called vapour-liquid-liquid-equilibrium, or VLLE, system.
3.4.1 Liquid-Liquid Solubility Diagrams
Below is a system where the miscibility between the two liquid phases
(denoted L 1 and L 2 phases) increases with temperature (phenol/water is an
example). Point U is called the Upper Critical Solution Temperature UCST:
Constant P
Above UCST the two
components are completely
U miscible.
TU
At temperature T < TU , if

Two Phase overall composition x1 < x1
L2 Region L1
T only the L2 - phase exists.
T
L 2 L1
At high overall compositions
REGION
x1 > x only the L1 - phase
exists.
0
x1a
x1
x1 At temperature T two liquid
1.0
1.0 x2
phases L2 L1 exist between
0
these two compositions.

Occasionally the system exhibits a Lower Critical Solution Temperature

LCST here miscibility decreases as the temperature rises (a typical system
is trimethylamine/water):
Constant P
Below the LCST the two
components are completely
miscible.
This is explained by the
L 2 L1 two components forming
REGION complexes below LCST.
L2 L1
T
T These complexes help
maintain miscibility below
LCST.
TL L
Above LCST complexes

break down (increased
x1a x1 thermal motion) and
0 x1 1.0
1.0
miscibility decreases.
x2 0
Occasionally an island diagram, shown below, represents how the solubility

of species 1 in each phase ( x1 and x1 ) varies with temperature T :
Constant P
The left side of this binodal
curve (U-L2-L) is the
TU U solubility of species 1 in
the L2 - phase.
L 2 L1
REGION The right side of the
T
L2

M
L1
binodal curve (U-L1-L) is
T the solubility of species 2
in the L1 - phase.
TL
L
LLE compositions x1 and
x1 are found either side of
x1a x1
0 x1 1.0 the horizontal tie-line.
1.0 x2 0
The L 2 - phase is rich in component 2, while the L 1 - phase is rich in

component 1. The relative amounts of each phase, found using the lever
rule, depends on how the mixing point M bisects the horizontal tie line.

Temperature TU is the UCST (also called the upper consulate temperature).

Temperature TL is the LCST (also called the lower consulate temperature).
For any temperature between TL and TU , the mixture will split into two
partially miscible liquid phases. At points U and L the L 2 - phase and the L 1
- phase will have identical properties.
Above U, and below L, the system is completely miscible (typical system is

nicotine/water). Island diagram behaviour is rare, usually either an UCST or a
LCST is observed, the former being the more common.
In addition (to whichever region is missing), if the remaining region intersects

an upper or lower phase boundary (boiling or freezing), then the remaining
UCST or the LCST nodes may also be absent from the phase diagram.
Of particular interest is where the LLE region exhibits increasing miscibility

with temperature, but the UCST is missing because, at lower pressure, the
system boils at temperatures before it reaches an UCST.
3.4.2 VLLE and P-x-y Diagram
On a P x y diagram the shape of the miscibility curves are reversed. The

diagram below represents a system where the USCT is missing; this is
because the system boils before reaching this point:
Constant T
At constant temperature T :
The lines are nearly vertical here because Three-phase pressure
liquid phase properties are only weak P0 is above the pure
functions of pressure.
component vapour
L 2 L1 pressures.
L2 L1
P0
D E F
HE and GE are the L2
V L2
P V L1
and L1 phase dew point
curves respectively.
H
V HD and GF are the L2
G and L1 phase bubble
point curves respectively.
0 x1 y1 1.0
1.0 x2 y 2 0
Above the three-phase pressure P0 the system has only two liquid phases.

Below the three-phase pressure P0 the system may either be: single phase,
namely L 2 - liquid, L 1 - liquid or vapour V ; there are also two two-phase
regions, namely V L 2 and V L 1 .
Consider that the system is at the three-phase pressure P0 , but that there is
only pure 2 present (i.e. pure L 2 liquid). Now add small quantities of liquid 1
gradually to this mixture:
1. Referring to the previous diagram, it is clear that species 1 dissolves

in the L 2 - liquid; up to point D. At this stage, there is only a single L 2
rich liquid phase present and the system is miscible.
2. At point D the mixture becomes a saturated solution of 1 in 2, i.e.

no more will dissolve. At this point, addition of a little more 1 will
cause a second liquid phase to appear (the L 1 -liquid phase). In addition
a small amount of vapour will form.
3. Applying the phase rule to the line segment DEF. There are three
phases present (L 2 - liquid phase, L 1 - liquid phase and the vapour
phase V ). According to the Phase Rule the system has one degree of
freedom F = 2 + N = 1 (it is univariant).
4. With three phases present, and with the system temperature T fixed,
this means that the state of the system is fixed. Thus, P = P0 must be
fixed, as must the compositions of all three phases; x1 , x1 and y1* .
5. Here y1* is the vapour phase composition at point E. The two liquid
phases will both be saturated solutions with compositions x1 (vertically
below point D) and x1 (vertically below point F).
6. So long as three phases are present, the compositions of all three

phases and the pressure P = P0 must be fixed. This applies anywhere
along line segment DEF.
7. Further addition of 1, will not change the compositions of the two

liquids or the vapour phase. The overall composition will change and
thus the quantities of the two liquid phases must change: the L 2 - liquid
phase will decrease, while the L 1 - liquid phase will increase.
8. When the overall composition reaches point F the L 1 - liquid phase

becomes saturated and L 2 - liquid phase completely dissolves into the
L 1 - liquid phase.
9. Further addition of species 1, beyond point F, leads to the solution
becoming unsaturated. The overall composition approaches pure 1.

Only the L 1 - liquid phase is present, both L 1 - liquid and vapour

phases are absent. The system is now miscible.
10. In the single phase regions there are three degrees of freedom
F = 2 + N = 3 ; with the temperature fixed, both pressure and
composition can be varied independently.
11. If the process is reversed and we start with 2 gradually being added
to pure 1, then the process will be repeated, but in reverse.
12. A similar situation will occur at a different fixed temperature, but then
the three-phase pressure P = P0 and the compositions of the two
saturated liquid phases (and the vapour phase) will all be different.
Above the three-phase pressure P > P0 is the liquid phase region, no vapour
is present and only LLE is evident. When the pressure is less than the three-
phase pressure P < P0 , then only VLE phase behaviour is evident. VLLE is
observed only when P = P0 .
3.4.3 VLLE and T-x-y Diagram
The T x y diagram below represents the same system as before. Again,

the UCST is missing because the LLE region intersects the VLE region before
the UCST is reached (an example would be aniline/water system):
Constant P
Below the three-phase
G temperature T0 the system
V is entirely liquid:
H
Within the L2 L1 region
there are two degrees of
V L2 V L1 freedom and only two
T0 L D E F liquid phases are present.
T 2 L1
L 2 L1 These split into an L2 -
liquid phase and an L1 -
liquid phase.
x1a y1* x1
No vapour is present.
0 x1 y1 1.0
1.0 x2 y 2 0

Above the three-phase temperature T > T0 is the VLE region. There are three
single-phase regions (two liquid regions and a vapour region). In addition,
there are two two-phase regions; the - V L 2 region and the V L 1 region.
Whenever T > T0 , and there is only a single phase present,

F = 2 + N = 3 , all intensive properties (except the fixed pressure) may be
changed. Thus, temperature and composition can all be varied independently.
In addition to these single phase regions, whenever T > T0 , there are two
vapour-liquid regions. In these two-phase regions F = 2 + N = 2 and there
are two degrees of freedom.
Given that the pressure is fixed, this means that the temperature may
be changed, in which case the composition is fixed, or vice versa;
graphically this situation is represented by horizontal tie-lines.
Here the gamma-phi approach may be used to solve bubble point or
dew point problems. VLE will result in either an L 2 - liquid phase or an
L 1 - liquid phase in equilibrium with the vapour phase (which liquid
phase is present depends on the overall composition).
At T = T0 and for overall compositions that fall within the line segment DEF,
the mixture will split into two liquid and one vapour phase in equilibrium:
The L 2 - liquid phase has a composition x1 , corresponding to a
composition on the x-axis vertically below point D.
The L 1 - liquid phase has a composition x1 , corresponding to a
composition on the x-axis vertically below point F.
The vapour phase has a composition y1* , corresponding to a
composition on the x-axis vertically below point E.
As before, within line segment DEF the number of degree of freedom

F = 2 + N = 3 = 1 which means the system is univariant. Given that the
pressure is already fixed, this means the temperature and the phase
compositions cannot change.
The removal of heat at constant pressure will cause the system temperature to
drop. How exactly the system will behave depends on the starting point and
the overall composition see Appendix B (this appendix is examinable):
See Appendix B.1: if cooling starts in superheated vapour region with
overall composition to the right of F.
See Appendix B.2: if cooling starts in superheated vapour region with
overall composition that passes exactly through E.
See Appendix B.3: if cooling starts in the superheated vapour region
with overall composition anywhere between E and F (the same
applies to cooling from a superheated vapour state with overall
composition anywhere between D and E).

3.4.4 VLLE at Higher Temperature
T x y diagrams are drawn at constant pressure, but if the pressure is

increased, then both pure component boiling points will rise.
On a T x y diagram this means that the intercepts on the temperature

axis, saturation temperatures T1sat & T2sat , will both rise. Effectively the entire
VLE region will move upwards.
The same effect will occur with a VLLE system. The VLE region will slide up
the LLE region. Thus, the three-phase temperature T0 will also rise.
Due to the higher temperature, the LLE region miscibility will increase (for
systems of this type) and line segment DEF will become shorter see below:
Constant P2 Apart from the higher

G temperatures and miscibility
V in the liquid region:
H
The three-phase VLLE
behaviour remains the
V L2 V L1 L1
same.
T0 D E F
L2 An azeotrope occurs
L 2 L1 along DEF since the
L 2 L1 mixture boils with
T constant composition.
The azeotrope is called

heterogeneous since
there are two liquid
x1a x1E x1 phases present.
0 x1 y1 1.0
1.0 x2 y 2 0
Heterogeneous azeotropes often occur when distilling systems like

butanol/water and ethyl acetate/water. Separating these components requires
specialised techniques and equipment.
The overhead product from one column is the heterogeneous azeotrope.

When this product is totally condensed, it will cool isobarically with to a
temperature below its three-phase temperature T0 .
Thus, a subcooled liquid is formed which, inside a decanter, splits into two
liquid phases. One liquid phase is returned as reflux to the first column while
the other liquid phase is separated in a second column.

An x y diagram for the VLLE system just discussed is shown below. Notice
the heterogeneous azeotrope and mixture boiling with constant composition:
On the previous T x y
Constant P1
1.0 diagram, HDE is the

V L 2 region. Notice that
1 is MVC
On the previous T x y
D E F
y1 *
y
1
diagram, GEF is the
V L1 region. Notice that
here 1 is LVC
0 These facts are also

0 a
x x E
x 1.0 evident on x y diagram
1 1 1
x1 to the left.
This switch in volatility is again evidence of an azeotrope. However, because

the azeotrope is heterogeneous the two liquid phases, in equilibrium with the
vapour phase, have different compositions compared to the vapour phase.
At still higher pressure P3 > P2 the boiling points of the pure species will rise
and the VLE and the LLE regions can separate from each other:
Constant P3
The LLE occurs at lower
Dew Point G
temperatures.
V Curve
H The VLE occurs at higher

Bubble Point temperatures above
Curve L UCST.
Miscible Region
UCST Now there is a minimum
temperature azeotrope
T L2 L 2 L1 L1 and a fully miscible
region between the LLE
and VLE regions.
0 x1 y1 1.0
1.0 x2 y 2 0
At this higher pressure there is no VLLE effect and clearly a heterogeneous

azeotrope cannot form. Any distillation column will produce the minimum
temperature homogeneous azeotrope overhead.

For VLLE systems, it is important to review the qualitative condensation and

vaporisation effects that are detailed in Appendix B (please note that this entire
appendix is examinable). As seen, these effects can have important
consequences in terms of the design of process plant.
More complex types of phase behaviour can occur and, at a fixed pressure,
this depends not only on the relationship between T0 , T1sat & T2sat but also on
whether or not homogeneous azeotropes can form.
These more complicated systems are sometimes categorised as system 2,

system 3 and system 4 type behaviour. Only system 1 type phase
behaviour is discussed here (and in Appendix B).
3.4.5 Quantitative Treatment of VLLE
It has been said that in the VLE region bubble point and dew point calculations
can be handled using the gamma-phi approach. At low pressure all V =1
i
and all isat = 1 , the rigorous approach reduces to modified Raoults Law:
y i P = i x i Pi sat (1.2.3.......N ) ..(3.9)
For VLLE systems, the activity coefficients are best found using the Non-
Random Two Liquid NRTL model. This model, unlike the Wilson model, has
three adjustable BIPs and is recommended for both VLE and LLE.
As stated, in the VLE region use the gamma-phi approach, while in the LLE
region use the approach outlined in Appendix A. This appendix also reviews all
the key NRTL expressions.
In the three-phase region the phase behaviour may be modelled rigorously

using a combination of these methods; again, NRTL is the preferred model of
solution behaviour the Wilson model cannot be used.
In rare circumstances, when the miscibility of each species in the other is very
limited, and when Henry Law constants are available, then simplifying
assumptions may be made. This is what follows in the section below.
If the pressure is very low (ideal gas) and if the liquid phase is also an ideal
solution, then all i = 1 and Raoults Law is obtained which, for species 2 in
a binary mixture may be written
y 2 P = x 2 P2sat ......(3.10)

A VLE system described by equation (3.9) will only behave in accordance with
equation (3.10) as species 2 approaches purity, that is x 2 1 . At the same
time as species 2 approaches purity then, in a binary mixture, species 1
must approach infinite dilution, that is x1 0 .
Thus, for a binary mixture as species 2 is increasingly modelled by Raoults

Law, species 1 must increasingly be modelled by Henrys Law
y1 P = h12 x1 .(3.11)
In which h12 is the Henry Law constant of species 1 (the low concentration
species often called the solute) in species 2 (the high concentration
species often called the solvent).
1. VLE Region of VLLE System
Thus, in the V region, species 2 partial pressures can be found from

Raoults Law while, at the same time, species 1 partial pressures should be
found from Henrys Law:
y 2 P = x 2 P2sat
y1 P = h12 x1
Adding these two equations together leads to
P = x 2 P2sat + h12 x1
P = x 2 P2sat + h12 (1 x 2 )
Re-arranging this in terms of x 2 leads to
P h12
x2 = (LINE HD)..(3.12)
P2sat h12
The vapour phase composition y 2 in equilibrium with this liquid composition is

given by Raoults Law
x 2 P2sat
y2 = (LINE HE)....(3.13)
P
Equations (3.12) and (3.13) represent equilibrium concentrations located at

either-end of a horizontal tie line within the V L 2 region.

In the V L 1 region, if the two components have very little solubility, then
species 1 partial pressures may be found using Raoults Law, while species
2 partial pressures are then found using Henrys Law as follows:
y1 P = x1 P1sat
y 2 P = h21 x 2
In which h21 is the Henry Law constant of species 2 (now the low
concentration species) in species 1 (now the high concentration species).
Adding these two equations together leads to
P = x1 P1sat + h21 x 2
P = P1sat (1 x 2 ) + h21 x 2
P P1sat
x2 = (LINE GF)..(3.14)
h21 P1sat
The vapour phase composition y 2 in equilibrium with this liquid phase in

equilibrium is now given by Henrys Law
h21 x 2
y2 = (LINE GE).......(3.15)
P
Now, equation (3.14) and (3.15) represent equilibrium concentrations either-

end of a horizontal tie line within the V L 1 region.
Only where the system behaves ideally (in the limit of very low and high
concentrations) will these expressions apply. It must be noted that not many
systems in practice will behave in this way.
2. LLE Region of VLLE System
Given some fixed pressure, if the temperature is below the three-phase

temperature, then only liquid will be present. In order to solve such LLE
problems rigorously please refer to appendix B.
The more difficult case is to solve the VLLE case where the mixture is at the
three-phase temperature and the overall composition falls somewhere along
the line segment DEF.

In this scenario there are three phases in equilibrium:

1. The saturated L 2 - liquid phase; composition x1 corresponding to point
D on the T x y diagram.
2. The saturated L 1 - liquid phase; composition x1 corresponding to point
F on the T x y diagram.
3. The saturated vapour phase V ; composition y1* corresponding to point
E on the T x y diagram.
Dew point lines HE and GE meet at point E. Where these curves meet
defines the vapour phase composition y 2* . Thus equating (3.13) with (3.15)
leads to
x 2a P2sat h21 x 2
y =
*
2 =
P P
In which x 2 and x 2 are the liquid phase compositions of species 2 in the L 2

- liquid and L 1 liquid phases respectively. Cancelling and re-arranging gives
x 2a h

= 21 (3.16)
x2 P2sat
In a similar way two expressions for y1* may be equated to get
y =
* (
h12 1 x 2a
=
)
P1sat 1 x 2 ( )
1
P P

(1 x ) = h

2 12
(1 x ) P
a sat
...(3.17)
2 1
Equations (3.16) and (3.17) may be solved by trial-and-error for x 2 and x 2 ,

or they be combined to solve for x 2 and x 2 explicitly. Substitute equation
(3.16) into equation (3.17), so as to eliminate x 2 , as follows:
(1 x )
2
=
h12
h P1sat
1 21 x 2
sat
P2

Or,

(
1 x 2 = ) h12

h
1 21 x 2

P1sat P 2
sat

h21 h12 h
sat sat
x 2 x 2 = 12sat 1
P2 P1 P1
h h12 h
x 2 21sat sat
1 = 12sat 1
P2 P1 P1
h12
1
P1sat
x 2 = ..(3.18)
h21 h12
sat sat 1
P P
2 1
Thus, once a value is available for x 2 from equation (3.18), this may be
substituted into equation (3.16) to find x 2 without any trial-and-error
calculations. A re-arranged version of equation (3.16) is given again below:
h21
x 2 = x 2 ..(3.19)
P2st
The last step is to find the relationship for y 2* . First write down equation (3.12),
representing dew point line HD for the V L 2 region and then re-arrange it
in favour of the total pressure P to get
P h12
x2 =
P2sat h12
P h12
x2 =
P2sat h12
P = x 2 P2sat x 2 h12 + h12
Now do the same for equation (3.13) representing bubble point curve HE
x 2 P2sat
y2 =
P

Or,
x 2 P2sat
P =
y2
Now eliminate the total pressure P to find the desired relationship between
x & y anywhere in the V L 2 region. These are corresponding points at
either end of any tie-line in the V L 2 region:
x 2 P2sat
= x 2 P2sat x 2 h12 + h12
y2
x 2 P2sat
y2 =
x 2 P2sat x 2 h12 + h12
x2
y2 = ..(3.20)
h
x 2 + 12sat (1 x 2 )
P2
In exactly the same way, starting with equations (3.14) and (3.15) the
relationship between x & y anywhere in the V L 1 region is given by
x
y2 = sat
2
..(3.21)
P
x2 + (1 x 2 )
1
h21
Now substitute x 2 from equation (3.19) into equation (3.20) to find

corresponding y 2* (at point E) in equilibrium with this L 2 - liquid.
Alternatively, substitute x 2 from equation (3.18) into equation (3.21) to find

corresponding y 2* (at point E) in equilibrium with the L 1 - liquid.
If the value of y 2* occurs at a very high concentration, then the latter approach
is correct. If the value of y 2* occurs at very a low concentration, then the
former approach is correct.
With little loss in accuracy, the ratios h12 / P2sat and P1sat / h21 may be treated as
constants.

3.4.6 Example 3B.1
Construct a T-x-y diagram for the partially miscible system ether/water system
at 1 atm total pressure using the following data for saturated (i.e. three phase)
vapour-liquid equilibrium at various pressures.
Saturated Saturated
Total
Temperature Water Phase Ether Phase
Pressure
(oC) Composition Composition
(atm)
Mole fraction ether
34.15 1.005 0.0122 0.9451
40.00 1.250 0.0116 0.9416
50.00 1.744 0.0103 0.9348
60.00 2.381 0.0093 0.9271
70.00 3.195 0.0075 0.9212
80.00 4.229 0.0069 0.9158
90.00 5.514 0.0058 0.9107
100.00 7.040 --- ---
Vapour Vapour
Temperature Pressure Pressure
(oC) Water Ether
(atm) (atm)
34.15 0.053 0.983
40.00 0.073 1.212
50.00 0.121 1.679
60.00 0.196 2.271
70.00 0.306 3.018
80.00 0.467 3.935
90.00 0.691 5.040
100.00 1.000 6.390
Assume that Raoults Law is valid for ether in the ether-rich phase and for
water in the water-rich phase.
Note that an azeotrope forms at 34oC between water and ether.


3B.5 Immiscible Systems
There are probably no completely immiscible systems, but some systems

have such positive deviation from ideality, and so little affinity to each other,
that they may (with little loss in accuracy) be considered completely
immiscible:
Constant P
X Y G
H
V
V L2 V L1
T0 F
T D E
L 2 L1
0 x1 y1 1.0
1.0 x2 y 2 0
A system with limited miscibility may be changed into a completely immiscible

system by converting HD and GF into vertical lines HD and GF , these lines
also extend (along the two y-axes) vertically into the LLE region.
There is still a VLE region above and a LLE region below the three-phase line
segment DEF. However, the L 2 - liquid is now just pure species 2, while the
L 1 - liquid is now just pure species 1.
In the V L2 region, horizontal tie-lines connect saturated vapour

compositions (curve HE) to pure liquid 2 while in the V L 1 region,
horizontal tie-lines connect saturated vapour compositions (curve GE) to pure
liquid 1.
At the three-phase temperature T0 , the pure L 2 - liquid at D is in equilibrium

with pure L 1 - liquid at F and both of these are in equilibrium with the vapour
phase (composition E).
Isobaric cooling of a superheated vapour mixture at Y eventually intersects

the dew point curve GE at a point vertically below Y. At this point pure L 1 -
liquid (pure species 1) condenses from the vapour mixture.

Further cooling and the vapour follows the dew point curve GE while pure L 1
liquid continues to condense out in increasing quantities. This continues until
the vapour composition reaches composition E.
From this point onwards both L 2 - liquid (pure species 2) and L 1 - liquid (pure
species 1) condense out together, until the last bubble of vapour has
condensed. Further cooling leads to both liquids becoming subcooled.
Isobaric cooling of a superheated vapour mixture from X also follows the

vertical path and behaves in much the same way. Only now L 2 - liquid
condenses out first until the three-phase temperature T0 .
Once cooled to T0 , as before, both liquid phases condense together with all
intensive properties remaining constant until the last bubble of vapour
condenses. Further cooling subcools both liquid phases.
If the vertical line on the previous diagram were to pass through E, then the
same would apply, only now condensation of pure L 2 - liquid and pure L 1 -
liquid will occur simultaneously.
3.5.1 Quantitative Treatment
For the completely immiscible system in the V L 2 region, the L 2 - liquid is

pure 2 and the total pressure above the liquid is simply its vapour pressure
P2sat , this is also equal to its partial pressure Py 2 which leads to
P2sat = Py 2
Thus, the vapour composition, corresponding to curve HE is given by
P2sat
y2 = ...(3.22)
P
Likewise for a completely immiscible system in the V L 1 region, the L 1 -

liquid is pure 1, so that the same results are obtained in terms of species 1:
P1sat = Py1
Thus, the vapour composition, corresponding to curve GE, is given by
P1sat
y1 = .(3.23)
P

The pressure is fixed; hence, the three-phase equilibrium pressure P0 at

temperature T0 is just the sum of the individual partial pressures at T0
P0 = Py1 + Py 2
For an immiscible system at temperature T0 this is equivalent to
P0 = P1sat + P2sat ..(3.24)
The value of P0 is known because the pressure is fixed. If a T P sat data set
is available, then T0 may be estimated by interpolation; varying T0 and hence
Pi sat until equation (3.24) is satisfied (within some tolerance).
The final step is to substitute equation (3.24) into equation (3.23) to get an
expression for y * which is the dew point vapour composition at point E
consistent with T = T0
P1sat (T0 )
y1* = ....(3.25)
P0
This is the vapour composition corresponding to the point where curve HE

meets curve GE. In addition P1sat (T0 ) is the vapour pressure of species 1 at
the interpolated three-phase temperature T0 .

3.5.2 Example 3B.2
Using the vapour pressure data given below, construct a T x y diagram

for the toluene/water system at 760 mm Hg pressure.
Vapour Pressure Data:
B
ln P sat = A
T +C
Where,
P sat is the vapour pressure (mm Hg)
T is the temperature (K)
The Antoine constants for each component are listed below:
A B C
Water: 18.3036 3,816.44 -46.13
Toluene: 16.0137 3,096.52 -53.67

3B.6 Tutorial 3B
1. Construct a T x y diagram for benzene-water mixtures at a total

pressure of 760 mmHg using the following vapour pressure data and
assuming complete immiscibility.
Vapour Pressure Data

Temperature (oC)
Benzene (mmHg) Water (mmHg)
50 269 93
60 389 149
70 547 233
75 640 285
80 754 355
90 1016 526
100 1344 760
110 1748 1075
2. Toluene and water are totally immiscible. Use the vapour pressure data
supplied to construct a P x y diagram for the system at 80oC.
Vapour pressure data can be evaluated using:
ln (Psat ) = A
B
T +C
Where P sat is the vapour pressure in (mmHg) and T is the

temperature in (K)
Component A B C
Water 18.3036 3816.44 -46.13
Toluene 16.0137 3096.52 -53.67

3B.7 Bibliography




ed., 1998.


Appendix A
Rigorous Approach to LLE
The material in this appendix could prove useful for Design Projects and
Research Projects. It deals with non-ideal liquid systems where a combination
of Raoults Law and Henrys Law is invalid.

A.1 Liquid-Liquid Equilibrium
The criterion for liquid-liquid-equilibrium LLE is the same as for VLE
fi = fi .( i = 1,2,3.....N ).(A.1)
In which fi is the fugacity of component i in the L 2 - phase and fi is the

fugacity of component i in the L 1 - phase (both phases are now liquid
phases). The activity coefficient for a species in the L 2 - phase is defined as
fi
i =
x i f i
Re-arranging this expression leads to
fi = i xi f i .(A.2)
Likewise for the L 1 - phase
fi
i =
xi f i
fi = i xi f i .(A.3)
Substitute equations (A.2) and (A.3) into (A.1) and, if both species exist as
liquids, then at the same T & P the pure component fugacity must be the
same f i = f i . Thus, for LLE
xi i = xi i .( i = 1,2,3.....N )(A.4)
The liquid phase activity coefficients are functions of temperature, pressure

and liquid phase composition:
i = i (xi , T , P )..(A.5)
i = i (xi , T , P ) .(A.6)
For an N component system, there are 2 N intensive variables ( T & P and

N 1 independent mole fractions in each phase). Given that there are N
equilibrium relations, it follows that there must be N degrees of freedom.

This is confirmed by the phase rule F = 2 + N which again leads to

F = N . Thus, N intensive variables must be fixed in order to solve the
equilibrium relationships.
For the special case of a binary LLE system there are two degrees of freedom
F = 2 and equation (A.4) may be written for each species as follows:
x1 1 = x1 1 ...(A.7)
(1 x )

1

2 ( )
= 1 x1 2 ...(A.8)
Often the pressure is fixed or low enough that the activity coefficients are
regarded as being functions of liquid composition and temperature only.
1 = 1 (x1 , T )....(A.9)
2 = 2 (x2 , T ) .(A.10)
If empirical data is available for the activity coefficients, and the temperature T
is fixed, then equations (A.7) through (A.10) may be solved for the liquid phase
compositions ( x1 and x1 ).
In order to avoid tedious iteration schemes, equations (A.7) through (A.10) are
best solved using a mathematics package. To solve equations (A.9) and
(A.10) some model of solution behaviour is needed.
Neither the Wilson equations nor any of the power law models are suitable for
LLE. However, NRTL, UNIQUAC and UNIFAC expressions do model systems
with limited miscibility.
The Non-random-Two-Liquid NRTL model of solution behaviour uses three

adjustable BIPs and the excess Gibbs function is given by
gE G 21 21 G12 12
= + ...(A.11)
x1 x 2 RT x1 + x 2 G 21 x 2 + x1G12
The NRTL model is widely regarded as one of the most accurate models of
solution behaviour. The adjustable BIPs are found by regressing experimental
data to equation (A.11).

Partially differentiating the above expression leads to the following expressions

for ln 1 & ln 2 :
G 21
2
G12 12
ln 1 = x 21
2
+ (A.12)
2
x1 + x 2 G 21 (x 2 + x1G12 )2
G12
2
G 21 21
ln 2 = x 12
2
+ ...(A.13)
1
x 2 + x1G12 (x1 + x 2 G21 )2
Where,
G12 = exp( 12 ) ...(A.14)
G21 = exp( 21 ) .(A.15)
b12
12 = (A.16)
RT
b21
21 = ...(A.17)
RT
The infinite dilution activity coefficients are readily obtained, from equations
(3.12) and (3.13) by setting xi = 0
ln 1 = 21 + 12 exp( 12 ) ..(A.18)
ln 2 = 12 + 21 exp( 21 ) ..(A.19)
The 3 BIPs for the NRTL model are b12 , b21 and . These BIPs are
independent of temperature and composition.
The species-specific and the temperature dependencies are expressed

through equations (A.14) through (A.17) while the composition dependence is
expressed through (A.12) and (A.13).
Activity coefficients are empirically-based correction factors. Thus, in order to

use this approach to LLE, experimental values for the BIPs are needed. Many
values are listed in the literature (Chemical Data Series, Dechema).

Appendix B
Heating or Cooling a VLLE System.
This material is examinable, so please read through carefully.

B.1 Isobaric Cooling from A
When the overall composition is A (to the right of point F) the system starts
as a single phase superheated vapour. If this vapour is then cooled (at
constant pressure) it will follow a vertical line indicated below:
Constant P
Further cooling and the
A G system will eventually
V intersect the dew point
H curve (line GE), leading to
the first drop of L1 - liquid.
V L1
V L2
Further cooling and the
T0 L D E F mixture will enter the two-
T 2 L1
phase V L 1 region.
L 2 L1
The compositions of
saturated liquid and vapour
x1a y1* x1
are then located either end
0 x1 y1 1.0 of horizontal tie-lines.
1.0 x2 y 2 0
When the mixture is cooled further, the system will eventually intersect the
bubble point curve (line GF) and the last bubble of vapour will condense.
Further cooling leads to a single phase subcooled L 1 - liquid. This behaviour is

identical to a miscible VLE system and calculations may be tackled in the
usual way using the gamma-phi approach to VLE. Use either the rigorous or
Modified Raoults Law method depending on pressure.
When the system is either in the superheated vapour region or the subcooled
liquid region it will have three degrees of freedom. Thus, temperature and
composition can be varied independently of each other (assuming a fixed
pressure.
The system will have two degrees of freedom in the V L 1 two-phase region.
Therefore, assuming a fixed pressure, the temperature and composition
cannot be varied independently of each other; they are linked by tie-lines.
In summary, cooling from A involves no VLLE effects. At no time are there

ever three phases in equilibrium. The same effect would occur when passing
through the V L 2 region, so long as the starting point was a superheated
vapour state to the left of point D.

B.2 Isobaric Cooling from B
A different behaviour is observed if a superheated vapour mixture is cooled at

constant pressure starting from B, which has the same overall composition
as point E:
Constant P
The superheated vapour

G
will cool until point E.
V
H
B At the three-phase
temperature T0 the vapour
V L2 V L1
will start to condense into
T0 L D E F an L2 - liquid at D and a
T 2 L1
L1 - liquid at F.
L 2 L1
Liquids D and F and
vapour E are all in
x1a y1* x1 equilibrium at constant
temperature T0 .
0 x1 y1 1.0
1.0 x2 y 2 0
Now there are three phases in equilibrium; thus, only one degree of freedom.
Since the pressure is already fixed, this means that neither the temperature
nor any of the three phase compositions may change.
The vapour mixture will start to condense (composition corresponding to E)

into two liquid phases with compositions corresponding to D & F. The
condensation will continue right up to the last bubble of vapour without any
change in any of the intensive variables.
Further cooling will eventually condense the last bubble of vapour and the
system moves into the subcooled region where there are now only two liquid
phases present.
In the L 2 L 1 region, the system will have two degrees of freedom. Therefore,
assuming a fixed pressure, the temperature and composition cannot be varied
independently of each other; horizontal tie-lines link these variables together.
In summary, cooling at E leads to both liquid phases condensing together out

of the vapour phase. This continues from the first drop of liquid right up to the
final bubble of vapour.

B.3 Isobaric Cooling from C
If a superheated vapour mixture is cooled at constant pressure starting from

C, then the behaviour is different again:
Constant P
G First the system cools to

the dew point line GE
C
H where L1 - liquid only starts
V to condense.
V L2 V L1
Further cooling leads to a
T0 L D E F vapour following the dew
T 2 L1
point curve GE and L1 -
L 2 L1 liquid following the bubble
point curve GF.
x1a y1* x1 Only L1 - liquid condenses

out until the vapour
0 x1 y1 1.0 composition reaches E
1.0 x2 y 2 0
Once the horizontal tie line drops to EF, then the first L 2 - liquid starts to
condense together with the L 1 - liquid. There are now three phases in
equilibrium (two liquid and one vapour) and one degree of freedom.
The vapour mixture now condenses, at constant temperature T0 , into two

liquid phases; one is the L 2 - liquid with composition corresponding to D; the
second liquid is a L 1 - liquid with composition corresponding to F.
The condensation will continue right up to the last bubble of vapour without
any change in any of the intensive variables.
Further cooling will eventually condense the last bubble of vapour and the
system moves into the subcooled region where there are now only two liquid
phases present (and two degrees of freedom).
This same situation prevails if the starting point is superheated vapour with
overall composition between points D and E. Now the L 2 - liquid condenses
first until the tie-line drops to DE and then the two liquid phases condense
together until the last bubble of vapour.

B.4 Isobaric Vaporisation from X
Consider isobaric heating of subcooled liquid starting from X, well below the
three-phase temperature T0 . Two liquid phases are present at X and initially
there is no vapour phase:
Constant P
G As the system is heated

the tie line tracks up within
H V the LLE region.
V L2 V L1
L2 - liquid is in equilibrium
T0 L with the L1 - liquid.
D E F
T 2
X
L1
Y
When T = T0 there is liquid
L 2 L1 at D in equilibrium with
liquid at F and both are in
x1a y1* x1 equilibrium with first bubble
of vapour at E
0 x1 y1 1.0
1.0 x2 y 2 0
At this point vaporisation starts. Both the L 2 - liquid and the L 1 - liquid vaporise
together but there is more L 1 - liquid than L 2 - liquid, so that inevitably the L 2 -
liquid is depleted first.
Once the final drop of L 2 - liquid has vaporised, the tie-line is at EF and there
are now two degrees of freedom. As heating continues the temperature is now
free to increase.
Vaporisation continues with saturated vapour composition tracking up EG

and the saturated liquid composition tracking up FG. Once the saturated
vapour composition is the same as the overall composition, then the last drop
of L 1 - liquid will vaporise (this is the dew point).
The mixture then becomes a superheated vapour with the same overall
composition as the original two-phase subcooled liquid at X.
In summary, initially both phases vaporise (with all intensive properties

constant) then, once the L 2 - liquid phase has vaporised, vaporisation of L 1 -
liquid continues (with rising temperature) up to the dew point curve.

The same applies if the overall composition lies between point D and E.
Initially both phases vaporise, but now L 1 - liquid is depleted first leaving the L 2
- liquid to vaporise on its own until the mixture reaches the dew point curve
HE.
A special case is heating a mixture with overall composition corresponding to

point E. In this case both phases vaporise (with constant intensive
properties) until the last drop of liquid then the vapour becomes superheated.


TOPIC 4: Chemical Equilibrium 1
Topic 4
Chemical Equilibrium
Contents
TOPIC 4: CHEMICAL EQUILIBRIUM
4.1 INTRODUCTION .................................................................................................................. 3
4.2 REACTION STOICHIOMETRY ............................................................................................... 5
4.2.1 Extent of Reaction ............................................................................................... 5
4.2.2 Example 4.1 ........................................................................................................ 7
4.3 EQUILIBRIUM CRITERION .................................................................................................... 9
4.3.1 Equilibrium Constant ......................................................................................... 11
4.3.2 Approximate Temperature Dependence of K ..................................................... 14
4.3.4 Accurate Temperature Dependence of K........................................................... 17
4.3.5 Example 4.2 ...................................................................................................... 20
4.4 GAS PHASE CHEMICAL EQUILIBRIUM ................................................................................ 22
4.4.3 Ideal Gas Mixture .............................................................................................. 25
4.4.4 Effect of Pressure on Chemical Equilibrium ....................................................... 25
4.4.6 Calculation of Equilibrium Conversion ............................................................... 26
4.4.7 Example 4.3 ...................................................................................................... 29
4.5 LIQUID PHASE CHEMICAL EQUILIBRIUM ............................................................................. 31
4.6 HETEROGENEOUS REACTIONS ......................................................................................... 36
4.6.1 Heterogeneous Reactions with Solids ............................................................... 37
4.7 TUTORIAL 4..................................................................................................................... 38
4.8 BIBLIOGRAPHY ................................................................................................................ 44
Introduction to Kinetics and Thermodynamics (Chemistry)

Learning objectives
After studying this Chemical Equilibrium topic you should be able to:
Differentiate between chemical conversion and rate of reaction.
Define the extent of reaction and show how it can be used to carry out
a material balance within a chemical reactor.
Review and derive the criterion for chemical equilibrium.
Define activity and equilibrium constant.
Write down expressions for enthalpy change and Gibbs free energy
change of reaction.
Outline and derive expressions for the temperature dependence of the
equilibrium constant.
Explain and write down expressions that indicate how the equilibrium
state is affected by pressure and composition.
Solve gas phase chemical equilibrium problems involving ideal and
non-ideal systems then calculate equilibrium composition and the
conversion of limiting reactant.
Solve liquid phase chemical equilibrium problems involving ideal and
non-ideal systems then calculate equilibrium composition and the
conversion of limiting reactant.
Explain how pressure and temperature affect chemical equilibria and
write down the equations that support these arguments.
For heterogeneous reactions write down solution strategies and related
equilibrium relationships.
Explain how solids are handled in a heterogeneous reaction system.

4.1 Introduction
In a chemical reaction the rate of reaction and extent of conversion are both
important. These two variables do not necessarily vary with reaction conditions
(such as temperature and pressure) in the same way.
For instance consider the oxidation of sulphur dioxide to sulphur trioxide as

shown below:
2SO 2 + O 2 2SO 3 H 298

o
= 98,890 J/mol
The reaction is exothermic and an increase temperature will tend to shift the
reaction to the left, which leads to a decrease in the equilibrium conversion of
reactants to products.
However, increasing temperature will increase the rate of reaction. Thus, when
the conversion is low and the rate is high, a smaller reactor is possible but with
more downstream purification.
On the other hand, if the reaction is endothermic a decreasing temperature

would decrease the rate of reaction while, at the same time, increase the
equilibrium conversion. Thus, a larger reactor will be needed, but with less
downstream purification.
Hence rate of reaction and extent of conversion are critical parameters for
designing chemical reactors
Conversion is determined by chemical reaction equilibrium (the subject of this

topic). This has been discussed in various other courses and an overview is as
follows:
Find a numerical value for the equilibrium constant K (usually at
298.15 K).
Knowing the standard enthalpy change of the reaction H o find a
value for the equilibrium constant K at temperature T (the actual
reaction temperature).
Correct this value for the effect of reaction pressure P .
Express chemical equilibrium in terms of the extent of reaction .
Solve for the extent of reaction and find the equilibrium composition.
Calculate the equilibrium conversion.
The basic procedure has already been outlined (for ideal gases) in Process
Engineering B. This material will be reviewed and extended to cover non-ideal
gaseous systems by relating the reaction stoichiometry to the equilibrium
constant K .

Thermodynamics has nothing to say about the rate of reaction.

Thermodynamics is only concerned about the final equilibrium state, no matter
how fast or slow it takes to reach this state.
It should be noted that the rate of a chemical reaction is determined by

chemical kinetics and that these rates are empirically determined; see Y2
course, Introduction to Kinetics and Thermodynamics.

4.2 Reaction Stoichiometry
Consider any general reversible reaction
1 A1 + 2 A2 3 A3 + 4 A4
In chemistry lectures this is usually written as follows
aA + bB cC + dD
Thus,
i are the stoichiometric coefficients of reactant/product species.
Ai are the various chemical species taking part in the reaction.
The sign convention for i has been discussed in Process Engineering B:

i is negative for reactant species.
i is positive for product species.
i is zero for inert species.
During the course of a reaction, the sign of the stoichiometric coefficient

allows: reactants to be consumed (negative i ); products to be produced
(positive i ); and inerts to pass through the reactor unchanged (zero i ).
4.2.1 Extent of Reaction
In Process Engineering B it was shown that, for any species i in the reaction
mixture, the change in the number of moles of i (that is dni ) depends on its
stoichiometric coefficient i and the extent of reaction :
dni = i d ...... (i = 1,2,3.....N ) ..(4.1a)
The extent of reaction (mol) is also called the reaction co-ordinate and has
the same numerical value for all species taking part in the reaction.
Equation (4.1a) may also represent a flowing system. In this case, everything
stays the same except for dots above the following terms:
dn i = i d ...... (i = 1,2,3.....N ) ..(4.1b)
In this case both the extent of reaction and n will have units (mol/s).

Integrate (4.1a) from the unreacted state, where = 0 , to any later state
where the extent of reaction has a value of (mol)
ni
dni = i d ...... (i = 1,2,3.....N )

nii , 0 0
The initial amount of species i present is n = ni , 0 (mol) at any later time this
has changed to ni (mol). Thus, for any species i in the reaction mixture:
ni = ni , 0 + i ...... (i = 1,2,3.....N ) ...(4.2)
Summing over all the species leads to
n = n0 + ...... (i = 1,2,3.....N ) ..(4.3)
Where,
n = ni n 0 = ni , 0 = i ..(4.4)
i i i
The overall stoichiometric coefficient is . Dividing equation (4.2) by (4.3)

leads to the composition of each species y i as a function of :
ni ni , 0 + i
yi = = .....(4.5)
n n0 +
Taking N 2 to be species (1), H 2 to be species (2) and NH 3 to be species

(3), the following table may be prepared for the ammonia reaction:
N2 + 3H 2 2NH 3
Initially n1, 0 n 2, 0 n 3, 0 n0
Reacted 1 2 3
Remaining n1 = n1, 0 + 1 n 2 = n 2, 0 + 2 n 3 = n 3, 0 + 3 n0 +
n1, 0 + 1 n 2, 0 + 2 n 3, 0 + 3
y1 = y2 = y3 =
yi n0 + n0 + n0 + y i =1
Knowing the extent of reaction (and moles of each species initially), the
remaining number of moles and the composition are easily found.

4.2.2 Example 4.1
Assume 10 moles of H 2 reacting with 5 moles of N 2 to produce NH 3 . Find the

molar fraction of each component of the reaction exiting the system in terms of
extent of reaction.


4.3 Equilibrium Criterion
The equilibrium criterion is the same as before:
Process Entropy Change Gibbs Energy Change

Irreversible dS TOTAL > 0 dGT , P < 0
Reversible dS TOTAL = 0 dGT , P = 0
At Equilibrium dS TOTAL = 0 dGT , P = 0
For a reaction system, where T & P are fixed and uniform throughout the
system, the Gibbs free energy provides the criterion for chemical equilibrium
dGT , P = 0 ....(4.6)
Clearly the point of chemical equilibrium occurs at minimum Gibbs free energy
as shown in the diagram below:
Constant T & P
Equilibrium may be
approached from either
side.
In either case total Gibbs

free energy must
G decrease.
G The point of equilibrium
= 0
T ,P occurs when total Gibbs
EQUIL
0 free energy is a
minimum.
For a mixture it has been shown (Process Engineering B)
G = G (T , P, n1 , n 2 ,......ni )
G G G
dG = dP + dT + dni
P T , n T P , n i n i T , P , n
j
dG = VdP SdT + g i dni

i

At constant T & P this becomes
dGT , P = g i dni
i
Substitute equation (4.1a) into the above expression to get
dGT , P = g i i d
i
G
= g i i
T , P i
The derivative is clearly the gradient of the curve (on the previous diagram)
and the point of equilibrium coincides with this curve passing through a
minimum. Thus, at chemical equilibrium this gradient must be zero
g i i = 0
i
However, the chemical potential i (J/mol) of any species i in the mixture

has been shown to be the partial molar Gibbs free energy of the species in the
mixture i = g i .
Therefore, the general criterion for chemical equilibrium becomes

i
i i = 0 .......(4.7)
For equation (4.7) to be useful, the chemical potential of a species in a mixture

(for non-ideal systems) must be expressed in terms of the fugacity. As
introduced in Topic 1, equation (1.18) this is given by
g i = RT ln f i +
In which is the constant of integration that depends only on temperature T .

The Gibbs free energy of pure i in some standard state g i0 is therefore also
given by the same expression
g i0 = RT ln f i 0 +

It is helpful to define a standard state property of a species in a reacting

mixture as the pure species property at the same temperature as the mixture
T but at a fixed pressure P 0 (now 1 bar but formerly 1 atm), the reason for
this choice will become apparent later.
For each relation between pure component properties there exists a

corresponding relation between the same partial molar properties of a
component in a mixture as given by
g i = RT ln fi +
Subtract the previous equation from the last one leads to
fi
g i g i0 = RT ln
f i0
The fugacity ratio fi / f i 0 is defined as the activity of species i in the solution

and is given the symbol a i
fi
a i = ...(4.8)
f i0
Substituting this back into the previous expression leads to
g i = g i0 + RT ln a i
This shows the partial molar Gibbs free energy of a species in a mixture is a
function of its Gibbs free energy in a standard state (its pure form), plus an
adjustment based on its interaction with the other species in the mixture, as
represented by the species activity a i .
4.3.1 Equilibrium Constant
However, the chemical potential i of any species i in the mixture has been
shown to be identical to the partial molar Gibbs free energy of the species in
the mixture i = g i . Thus, the previous expression becomes
i = g i0 + RT ln a i .....(4.9)

Substituting equation (4.9) back into the general criterion for chemical
equilibrium (equation 4.7) gives
(g
i
i
0
i + RT ln a i = 0 )
Then, after some re-arrangement it can be separated into its pure-form
contribution and its interaction contribution:
i g i0 + RT i ln a i = 0
i i
g i
0
i
ln (a i ) =
i i
i RT
Any sum of the logs may be represented as follows:
ln x i = ln x1 + ln x 2 + .... + ln x n = ln ( x1 x 2 .....x n ) = ln ( xi )
i
i
Applying the same reasoning to left side of the previous equation gives
g i
0
i
ln (a i ) =
i i
i RT
or
i g i0
(a i ) = exp i .(4.10)
i
i RT

The thermodynamic equilibrium constant K is now defined as
i g i0
K = exp i .(4.11)
RT

Upon re-arrangement this leads to
g
i
i
0
i = RT ln K

And this may be written
G 0 = RT ln K ...(4.12)
In which G 0 is the standard Gibbs free energy change of reaction.
From equation (4.11), the definition of the equilibrium constant, it is clear that
the equilibrium constant K depends on g i0 (the standard state Gibbs free
energy of each of the pure species), which means that:
K is independent of composition, since g i0 are pure component
properties.
K is independent of pressure, since g i0 are at a fixed pressure of 1
bar (formerly 1 atm).
K is dependent only on temperature, since g i0 depend on the reaction
temperature only.
Notice that defining standard states (in the manner discussed earlier) means
that the equilibrium constant K is dependent on temperature only.
Substituting equation (4.11) into (4.10) gives
K = (a i ) i ........(4.13)

i
For the general reversible reaction aA + bB cC + dD equation (4.13) can

be written as follows
K = (a i )
i
i
K = (a C ) (a D ) (a A ) (a B )
c d a b
(a C )c (a D )d
K = .....(4.14)
(a A )a (a B )b
Equation (4.14) is the key to solving chemical equilibrium problems, so long as
a numerical value for K is available. If G 0 is known, then equation (4.12)
provides a value for K . However, G 0 is initially only known at 298.15 K.
Thus, using equation (4.12) usually only provides a numerical value for K at
298.15 K. The next step is to investigate the temperature dependence of the
equilibrium constant K .

4.3.2 Approximate Temperature Dependence of K
The fundamental property relation for dg of a pure species was developed in

Process Engineering B as follows:
dg = vdP sdT ...(4.15)
Now consider differential changes in g / RT
g 1 g
d = dg dT
RT RT RT 2
By substituting equation (4.15) into this expression it is clear that
g v s g
d = dP dT dT
RT RT RT RT 2
g
d =
v
dP
1
(g + Ts )dT
RT RT RT 2
Recalling that g = h Ts gives
g v h
d = dP dT
RT RT RT 2
The above expression is an exact differential which means that at constant

pressure
( g / RT ) h
T = RT 2
P
For species i in its standard state this becomes (drop partial derivative
notation because mi0 depend only on temperature T ):
h = RT
0 2 (
d g i0 / RT )
i
dT

Multiply both sides by i and then sum over all species in the mixture

d i g i / RT
0
i i hi0 = RT 2 i dT

Where,
h
i
i i
0
= H 0 (standard enthalpy change of reaction)
g
i
i
0
i = G 0 (standard Gibbs energy change of reaction)
Making these substitutions gives
H = RT
0 2 (
d G 0 / RT )

dT
But, in view of equation (4.12), it is possible to substitute G 0 / RT = ln K

into the previous equation.
d (ln K ) H 0
= .(4.16)
dT RT 2
This is the famous Vant Hoff equation which shows how the equilibrium
constant K depends on the reaction temperature.
From equation (4.16) it follow that:

If H 0 is negative the reaction is exothermic. In this event, as T
increases K decreases and the reaction shifts to the LEFT and the
conversion decreases.
If H 0 is positive the reaction is endothermic. In this event, as T
increases K also increases and the reaction shifts to the RIGHT and
the conversion increases.
If it is possible to assume that the standard enthalpy change of reaction H 0

is independent of temperature, then equation (4.16) may be integrated.

Separating the variables and then integrating leads to
H
K2 0 T2
d (ln K ) = R T 2
dT
K1 T1
K2 H 0 1 1
ln = ...(4.17)
K1 R T2 T1
K 1 is needed at T1 (this is usually standard conditions of 298.15 K) then K 2

may be found at any other temperature T2 . The equilibrium constant K 1 is
best found using equation (4.12), since G 0 is known at 298.15 K.
Alternatively (but less accurately) charts like the one below are available:
Locate 1 / T on the x-
axis (as the temperature
+VE Exothermic increases move to the
Reaction left).
Identify correct reaction

ln K 0 Endothermic
Reaction from the family of curves
(only two are shown).
-VE
Obtain ln K from y-axis
and solve for K .
4 8 12 20
1/ T
If the assumption that H 0 is independent of temperature were true then this

graph of ln K versus 1 / T would plot as a perfect straight line.
Experimental data (like the chart above) indicate that the points are nearly
linear over short temperature intervals, but tend to be non-linear over much
longer temperature ranges.
Equation (4.17) requires values for both H 0 , the standard enthalpy of

reaction and G 0 , the standard Gibbs free energy of reaction.

In Process Engineering B these were shown to be
H 298
0
= i h f i (4.18)
i
G 298
0
= i g f i ...(4.19)
i
In which h f i are the enthalpies of formation or each reacting species and

g f i are the Gibbs free energy of formation of each reacting species. All
of these are only listed at 298.15 K.
It has been said that equation (4.17) is only accurate over short temperature
ranges or when interpolating between two closely spaced values, it follows that
some more accurate expression is needed.
4.3.4 Accurate Temperature Dependence of K
The reason why straight lines are not obtained on the previous graph is
because H 0 is actually a function of temperature T . Integrating the Vant
Hoff equation leads to
H 0
ln K = dT + J .....(4.20)
RT 2
In which J is the constant of integration. If species i is in its standard

state, then pressure is fixed (at 1 bar) which means that the pure species
enthalpy is a function of temperature only and given by
dhi0 = C P0i dT ....(4.21)
Multiplying through by the species stoichiometric coefficient and then summing

over all reacting species gives

d i hi0 = i C P0i dT
i i
This may be written
dH 0 = C P0 dT ..(4.22)

In which C P0 is the standard heat capacity change of reaction. Pure species

heat capacities of constant pressure C Pi are often correlated (Smith, Van
Ness, Abbott, 2005) by polynomials such as:
(
C P0i = R Ai + Bi T + C i T 2 + Di T 2 )
With this type of correlation either the C i or Di coefficient is zero. Multiply
each species heat capacity by its stoichiometric coefficient and then sum over
all species to find C P0 as follows:
( )
C P0 = R A + BT + CT 2 + T 2 .....(4.23)
Where,
A = i Ai ...(4.24a)
i
B = i Bi ...(4.24b)
i
C = i C i ......(4.24c)
i
DD = i Di ......(4.24d)
i
Now substitute equation (4.23) into equation (4.22) and integrate

DH 0 = R (DA)T +
(DB ) T 2 + (DC ) T 3 (DD ) + I ..(4.25)
2 3 T
This expression is very important since it allows the enthalpy change of

reaction H 0 to be found at any temperature T .
The procedure for solving equation (4.25) is as follows:

Calculate H 298 0
from equation (4.18), knowing the enthalpy of
formation of each compound at 298.15 K.
Then solve equation (4.25) for the constant of integration I at a
temperature of 298.15 K.
Knowing the constant of integration I solve equation (4.25) for H 0
at any other temperature T .

Substitute (4.25) into equation (4.20) and integrate

R (DA)T +
(DB ) T 2 + (DC ) T 3 (DD ) + I
T
ln K =
2 3
dT + J
RT 2
ln K =
I
+ (DA) ln T +
(DB ) T + (DC ) T 2 + (DD ) + J ..(4.26)
RT 2 6 2T 2
In view of equation (4.12) a similar expression to (4.26) can readily be derived

for G 0 at any temperature simply multiply through by RT to get

G 0 = I RT (A) ln T +
(B ) T + (C ) T 2 + ( ) + J ..(4.27)

2 6 2T 2
There are at least three ways to find the constants I & J that appear in
equation (4.26):
1. If values of K are known at two different temperatures, then two
versions of equation (4.26) may be solved simultaneously for I & J .
2. If one value of K is known at any T and if a value of H 0 is known at
standard state, then solve equation (4.25) for I , followed by equation
(4.26) for J .
3. If H 0 is known at one T and G 0 is known at another (or the same)
temperature T , then use equation (4.25) to find I followed by
equation (4.27) to find J .
Once the equilibrium constant is known (at the reaction temperature T ) the
chemical equilibrium relationship can be solved to find the equilibrium extent
of reaction E .
At this point the equilibrium concentrations and/or the equilibrium conversion

may be found; this is what follows in the next section.

4.3.5 Example 4.2
Develop expressions for the equilibrium constant and the standard free
energy change as a function of temperature for the reaction:
CO 2 (g) + H 2 (g) CO(g) + H 2 O(g)
Dimensionless CO H2O CO 2 H2
a 3.376 3.470 5.457 3.249
b 10 3 0.557 1.450 1.045 0.422
c 10 6 --- --- --- ---
d 10 5 -0.031 0.121 -1.157 0.083
using:
(a) The experimentally determined values of the equilibrium constant.
T (oC) K
686 0.534
1309 2.930
(b) The first of the above values and the standard enthalpy change of
reaction.
H 0 = +9838 cal/mol = +41190 kJ/kmol at 298 K
(c) The standard enthalpy change of reaction and the standard free
energies of formation.
g 0f at 298K
Species
(cal/mol)
CO 2 -94,260
CO -32,808
H2O -54,635


4.4 Gas Phase Chemical Equilibrium
Take the standard reversible reaction aA + bB cC + dD as the basis for

discussion and assume that all species are gaseous, then from equation (4.8)
the activity of a species in a mixture a i was defined as
fi
a i =
f i0
Substituting equation (4.8) into equation (4.14) gives
c d
fC fD

f0 f0
K=
C D
...(4.28)
a b
fA fB

f0 f0
A B
The fugacity of a component in a mixture fi was shown in Topic 3A to be
fi = i y i f i
Substitute this last expression into the previous one
c d
C yC f C D yD f D

0 f D0
K=
f C
a b
A yA f A B yB f B

f0 f B0
A
For a gas phase reaction, the standard state is the pure species as an ideal-
gas at the reaction temperature T and P 0 = 1 bar. If the gas is ideal, then
= 1 and it follows that f i 0 = P 0 = 1 (bar).
Making this substitution into the previous expression gives
( C y C f C )c ( D y D f D )d
K=
( A y A f A )a ( B y B f B )b

Grouping like terms together leads to
Cc Dd f Cc f Dd y Cc y Dd
K = a b a b a b
f f y y

A B A B A B
The definition of a pure species fugacity coefficient was presented in Topic 1
fi
i = fi = i P
P
Substituting f i = i P into the last expression for the equilibrium constant K

gives
c d Cc dD y Cc y Dd P c P d
K = Ca Db a b

a b
y y
a b
P P (4.29)
A B A B A B
c d Cc dD y Cc y Dd c + d a b
K = Ca Db a b a b P
y y
A B A B A B
Noting that the overall stoichiometric coefficient is i

i = c + d a b , then
this may be written
c d Cc dD y Cc y Dd i
K = Ca Db a b a b P
y y
A B A B A B
A more compact way of writing this is given by
K = K K K y P i .(4.30)

This is the most compact way of representing non-ideal chemical equilibrium.

However, it must be understood that the pressure P is dimensionless
(Process Engineering B). That is the pressure is actually given by
P P P(bar)
0
0
fi P 1 (bar)

4.4.1 Rigorous Non-Ideal Mixture
Equation (4.30) is the rigorous approach to non-ideal chemical equilibrium:
K = K K K y P
i
The term K takes into account the different interactions between like and
unlike species and only applies to the vapour phase when the pressure is high
enough to bring the molecules close together.
At lower pressure K = 1 , the molecules are further apart, so that any

differences in interaction between like and unlike species become negligible.
This situation actually applies in the majority of cases.
The pure species fugacity coefficients i are best found using a cubic EOS,
such as Redlich-Kwong-Soave (RKS) or Peng-Robinson (PR).
Solving equation (4.30) is not easy as it looks. This is because K is a

complex function to temperature, pressure and composition and K is a
function of temperature and pressure. All of this means that lengthy iteration
procedures will be needed to obtain an answer by hand.
Although equation (4.30) is tedious (and time consuming) to solve by hand, it

is easy to solve using a mathematics package, such as Excel.
4.4.2 Ideal Solution Mixture
If the pressure is higher that atmospheric, then it would be inaccurate to

assume that the mixture is an ideal-gas. However, the molecules may still be
far enough apart for it to behave as an ideal mixture. For this assumption to be
valid the pressure must not be so high that the mixture is near its critical point.
As stated already, at lower pressure the molecules are further apart and any
differences in interaction between like and unlike species become negligible.
In this event K = 1 and equation (4.30) becomes
K ( i =1) = K K y P i (4.31)

Remember, this equation must be dimensionally consistent. This arises

because the pressure is the ratio of two pressures: the numerator is the
system pressure P (bar); the denominator is the standard state pressure of 1
(bar); thus the pressure, as it appears in (4.30) and (4.31) is dimensionless.

4.4.3 Ideal Gas Mixture
If the pressure is very low (below atmospheric pressure), then the mixture will
behave as an ideal gas and both K = 1 and K = 1 . Thus, equation (4.30)
assumes its simplest form:
K ( i =1, i =1) = K y P i ..(4.32a)

This is often referred to as K p
K p = K y P i ..(4.32b)

It is clear from equation (4.32b) that K P is numerically equal to K but,

whereas K is dimensionless, K P has units. This was fully explained in
Process Engineering B:
Remember the pressure that appears in (4.32a) and (4.32b) is actually
a dimensionless pressure ratio P / P 0 . The numerator goes with K y ,
as in K y P i , to give K P with units.

The denominator of this pressure ratio P / P 0 , as in ( P 0 ) i is equal

to unity. However, its units cancel with K P to keep K dimensionless.
Equation (4.32a) is very easy to solve because the equilibrium constant K

depends only on temperature, ratio of mole fractions K y depends only on
composition and the final term on the right depends only on pressure.
4.4.4 Effect of Pressure on Chemical Equilibrium
The effect of pressure acts through the final term in equation (4.32a) or
equation (4.32b). There are three cases that can arise:
1. Positive Overall Stoichiometric Coefficient

i
i +VE
At equilibrium, equation (4.32a) shows, that as the pressure increases, K y

must decrease to keep K constant. Thus, reaction shifts to the left and
conversion decreases.

2. Negative Overall Stoichiometric Coefficient i

i VE
At equilibrium equation (4.32a) shows that as the pressure increases, the

pressure term must decrease and K y must increase in order to keep K
constant. Thus, reaction shifts to the right and conversion increases.
3. Zero Overall Stoichiometric Coefficient

i
i =0
At equilibrium equation (4.32a) shows that pressure term must then be equal
to one (its exponent is zero). Thus, K y must be constant and, therefore,
pressure can have no effect on the equilibrium composition.
All of this is in accord with Le Chateliers Principle. If a change in reaction

conditions occur (temperature, pressure or composition), then the equilibrium
shifts to counteract the imposed change to the system.
Although the discussion has focussed on an ideal system, the same general
trends hold for non-ideal systems, except that now K or K may also
change K y and affect the equilibrium.
However for non-ideal systems, if the overall stoichiometric coefficient is zero,

then pressure will have a small effect on equilibrium composition through
these non-ideal adjustment parameters K or K .
4.4.6 Calculation of Equilibrium Conversion
The first step in finding the equilibrium conversion (and/or equilibrium

concentration) is to find K y from either one of the three expressions below:
1. Rigorous Non-Ideal System:
K = K K K y P
i
2. Ideal Solution System:
K ( i =1) = K K y P
i

3. Ideal Gas System:
K ( i =1, i =1) = K P = K y P
i
At this point a numerical value for K y should be available 9using any of the
above expressions). For a gas phase reaction aA + bB cC + dD the ratio of
mole fractions is given by the expression below:
( y C )c ( y D ) d
Ky =
( y A )a ( y B )b
At first glance, the previous equation appears to have too many unknowns.
However, each species mole fraction is actually linked to the extent of reaction
as given by equation 4.5:
ni , 0 + i
yi =
n0 +
Substitute this last expression into the previous one leads to
nC , 0 + C n D , 0 + D
c d

n + n +
= K y .(4.33)
0 0
n A, 0 + A n B , 0 + B
a b

n0 + n0 +
Solve equation (4.33) for the equilibrium extent of reaction E . A trial-and-

error approach is often needed depending on the stoichiometric coefficients,
alternatively a mathematics package can be used.
Once E is available, equation (4.5) may be solved for each component to find
the equilibrium concentration (these mole fractions must all sum to one,
irrespective of the value of ).
The equilibrium conversion, with respect to reactant species i , is the number

of moles of species i reacted divided by the total number of moles of
species i initially present:
Thus, by definition the equilibrium conversion is given by
i e
Conversion i = ..(4.34)
ni , 0

Where,
i is stoichiometric coefficient of species i expressed as a modulus.
Usually (but not always) the fractional conversion is based on the limiting
reactant that is initially present.
Of course, the equilibrium conversion may be expressed either as a fraction or

as a percentage.

4.4.7 Example 4.3
For the following reaction:
C 2 H 4 + 2 H 2 2CH 4
The reaction temperature and pressure are 1,500K and 10 bar respectively. In
the feed stream to the reactor there are 3 moles of ethene to 1 mole of
hydrogen. From the plot of ln K vs. 1 / T for the following two formation
reactions we find:
C + 2 H 2 CH 4 ln K = -6.0
2C + 2 H 2 C 2 H 4 ln K = -12.5
Calculate the extent of reaction and the percentage equilibrium conversions of

both C 2 H 4 to CH 4 and H 2 to CH 4 . Use the K P approach.


4.5 Liquid Phase Chemical Equilibrium
Liquid phase chemical equilibrium is also given by equation (4.13)
K = (a i ) i

i
Less generally for liquid phase reaction aA + bB cC + dD this is given by
(a C )c (a D )d
K=
(a A )a (a B )b
The activity a i is defined in the same way as before
fi
a i =
fi0
For liquid species, the standard state is the pure liquid at the system
temperature T and a pressure of P = 1 bar. In order to find fi start with the
definition of an activity coefficient
fi
i =
f id
i
Then substitute in the Lewis-Randall Rule to get the required expression for fi
fi = i xi f i
Thus, the activity a i for a non-ideal liquid species becomes
fi
a i = i xi ....(4.35)
fi0
Notice that the standard state fugacity f i 0 is at reaction T and a P = 1 bar,

while f i is at reaction T & P (both are pure component fugacities).
The ratio f i / f i 0 may be evaluated but, except for high pressure, pure
component liquid phase fugacities are not strongly pressure-dependent.

Thus, equation (4.35) becomes (with little loss in accuracy)
a i = i xi ...(4.36)
Substitute the above expression into equation (4.13) to get
K = ( i xi ) i ......(4.37)

i
4.5.1 Rigorous Non-Ideal Mixture
As before (for gas phase reactions) the right hand side of equation (4.37) may
be split up as follows:
K = ( x i ) i ( i )
i
i i
i
K = K x K .(4.38)
In which K x is the ratio of liquid phase mole fractions and K is the ratio of
activity coefficients, defined (respectively) as follows:
K x = ( xi ) i ....(4.39)

i
K = ( i ) i ....(4.40)

i
Like gas phase reactions (knowing the equilibrium constant K and the ratio of
activity coefficients K ) the idea is to solve equation (4.38) for K x and then
find the equilibrium composition or equilibrium conversion as before.
Activity coefficients (and the ratio of activity coefficients) can be estimated

using some model of solution behaviour; preferably a local composition model
such as Wilson, NRTL, UNIQUAC or UNIFAC.
However, hand calculations are highly iterative since activity coefficients are
themselves complex functions of temperature and composition and the
pressure if it is very high.
A better approach is to use a mathematics package which can solve all the
equations simultaneously even when some of the variables have complex
dependencies. Often the time spent writing the worksheet is less than the time
spent trying to execute a highly iterative hand-calculation.

4.5.2 Ideal Solution Mixture
If the liquid phase mixtures can be considered an ideal solution, then all the
activity coefficients are individually equal to one, and the ratio of activity
coefficients is also equal to one, so that equation (4.38) reduces to
K = K x = ( xi ) i ..(4.41)

i
Less generally for liquid phase reaction aA + bB cC + dD , this may be

written as
( x C )c ( x D ) d
K= .(4.42)
(x A )a (x B )b
Equation (4.42) is known as The Law of Mass Action. Although widely used in
standard textbooks, it rarely models liquid phase reaction equilibrium
accurately; particularly polar liquid mixtures.
Even when the mixture is non-ideal, there are two extreme cases where
ideality may be assumed:
1. When a species is at a high concentration.
2. When a species is at a low concentration.
High Concentration
At high concentration (of a species) the species approaches ideality, the real
solution curve (of that species) becomes a tangent to the Lewis-Randall Rule
and the activity coefficient (of that species) approaches one:
As: xi 1 i 1 .....(4.43)
Equation (4.36) shows that, under these conditions, the activity approaches
the composition
As: xi 1 a i xi .....(4.44)
In many textbooks the activity is understood as an effective composition; that

is as the solution becomes more non-ideal, the activity departs more and more
from the true composition.

Low Concentration
At low concentrations the Lewis-Randall Rule is not even approximately valid;

at low concentrations Henrys Law is valid:
Constant T
The Henry Law model
of ideality (dotted line) is
a tangent to the real
solution curve (solid
line) in the limit of very
low concentration
fi fi id = ki mi
(kPa) As: x i 0 the real
fi solution fugacity tends
to the ideal solution
fi 0 = ki
fugacity or fi fi id .
B
0
0 mi0 = 1
molality mi
Notice that the x-axis is the molality of the component in the mixture (not the
mole fraction). The molality is the concentration of the species expressed in
(mol/kg); that is moles of solute per kilogram of solvent.
The law is actually only valid up to about point B, but the procedure is to
assume that this law (dotted line) is valid up to a molality of 1 and then to
choose this y-axis intercept as the standard state fi 0 ; call this intercept k i .
It is important to understand that the standard state is not the pure species but
is taken to be an ideal solution at a composition of 1 molal. Therefore, activity
must now be modified to take into account this new standard state, as follows:
fi
a i =
f 0
i
Clearly from the diagram fi 0 = k i . The slope of the dotted line is then simply
k i so that its equation must be
fi id = k i mi

As the species approaches infinite dilution and, bearing in mind that the
standard state for the species in the mixture was defined as fi 0 = k i , we get
As: xi 0 fi fi id
As: xi 0 fi k i mi
k i mi
As: xi 0 a i
ki
As: xi 0 a i mi (4.45)
This produces a simple result, but the procedure is only useful when G
0
data is available in terms of this fictitious standard state; the state is fictitious
because the real mixture is not ideal at a one molal concentration.
This approach is only accurate for species at very low concentrations when
Henrys Law is valid. Often one component will exist at higher concentration at
which point the accuracy may decline significantly.
If Henrys Law is invalid (high concentration) and an accurate answer is

required, then there is no option but to use the rigorous approach outlined
earlier (see section 4.5.1). Equation (4.38) is shown again below:
K = K x K
Less generally for liquid phase reaction aA + bB cC + dD , this may be

written as
(xC )c (x D )d ( C )c ( D )d
K=
(x A )a (x B )b ( A )a ( B )b
If K is known, then the previous expression is used to find the equilibrium

conversion. The added complication is that all the activity coefficients are
needed and the local composition models all have a complex functional
dependency.

4.6 Heterogeneous Reactions
There are a number of ways of tackling heterogeneous chemical equilibrium

problems. For instance, consider a gas B reacting with a liquid A to form a
dilute aqueous solution C:
aA(l) + bB(g) cC(aq)
How this is handled depends on what information is available. If the standard

Gibbs free energy change of reaction G 0 is based on the following mixed
standard states:
B as a pure ideal gas at reaction temperature T and standard state
pressure of f B0 = P 0 = 1 bar.
A as a pure liquid at reaction temperature T and standard state
pressure of P 0 = 1 bar.
C as solute in a hypothetical ideal 1-molal aqueous solution.
(a C )c
K=
(a A )a (a B )b
Where,
a C = mi from equation (4.45)
a A = A x A from equation (4.36)
fB
a B = ..from equation (4.8)
f B0
a B = B B y B P
In view of what has already been said about the choice of standards states this
becomes
(mC )c
K= ....(4.46)
( A x A )a ( B B y B P )b
In order to solve equation (4.46), in conjunction with mole fractions relations, a

value for G 0 (and thus K ) will be needed based on these standard states.

4.6.1 Heterogeneous Reactions with Solids
The reaction equilibrium constant is always given by
K = (a i ) i

i
Less generally for a heterogeneous reaction aA + bB cC + dD this

becomes
(a C )c (a D )d
K=
(a A )a (a B )b
The activity a i is always defined in the same way
fi
a i =
fi0
The standard state for a solid component is the pure solid at reaction
temperature but a pressure of 1 bar. The fugacity of the solid in the mixture is
at the reaction T & P and is equal to fsolid = f solid . Thus, the activity becomes
f solid
a solid = 0
f solid
There is a difference in pressure between these two fugacities, but the

properties of solids are very weak functions of pressure so that
a S = 1 .(4.47)

4.7 Tutorial 4
1. A gas mixture reacts according to the reaction:
1
SO2 ( g ) + O2 ( g ) SO3 ( g )
2
The equilibrium conversion of dioxide to trioxide at 1atm and 500C is
93%. Calculate the reaction equilibrium constant K under these
conditions assuming the gas to be ideal and given an initial
composition as follows:
Component Mole %
Sulphur dioxide 10
Oxygen 11
Nitrogen 79
Sulphur trioxide 0
(Solution: 51.8)
2. Use the thermodynamic data given below to estimate the chemical

equilibrium constant K for the following gas phase reaction occurring
at 1500 K:
2 H 2 ( g ) + 2 NO ( g ) N 2 ( g ) + 2 H 2 0( g )
Dimensionless H2O N2 H2 NO
a 3.470 3.280 3.249 3.387
b 10 3 1.450 0.593 0.422 0.629
c 10 6 --- --- --- ---
d 10 5 0.121 0.040 0.083 0.014

The standard enthalpies and Gibbs free energy of formation of the

compounds taking part in this reaction are listed below:
h 0f g 0f
Component
J/mol
NO 90,374 86,688
H2O -241,827 -228,593
(Solution: 2.69 x 1016)
3. The water-gas shift reaction is set out below:
CO ( g ) + H 2 O( g ) CO2 ( g ) + H 2 ( g )
Calculate the fraction of the steam decomposed to hydrogen for each

of the different sets of conditions described below. Assume the mixture
to be an ideal gas.
4088
K can be calculated from the equation: ln K = 3.59
T
Reactant Mixture Temp Press

Case
H2O CO N2 CO 2 K bar
a) 1 1 0 0 700 1
b) 1 1 0 0 700 10
c) 1 1 1 0 700 1
d) 1 2 0 0 700 1
e) 2 1 0 2 700 1
f) 1 1 1 0 1366 1
(Solution: 0.755, 0.755, 0.755, 0.918, 0.401, 0.426)

4. In the Deacon process for the manufacture of chlorine, hydrochloric

acid is oxidised with 95% pure oxygen according to the reaction:
4 HCl ( g ) + O2 ( g ) 2 H 2 O( g ) + 2Cl 2 ( g )
The remaining 5% of the oxygen stream is nitrogen. The reaction

occurs at 1atm and at temperatures between 450 and 650 C. The
other variable involved is the ratio of the reactants. In view of the
above, determine the following:
(a) An equation relating ln K to the absolute temperature of reaction.

(b) The composition of the equilibrium mixture at both 450oC and
650C for feed ratios of both 4 and 12 moles HCl per mole of the
total oxygen stream (including the nitrogen).
(c) The purity of the chlorine produced on an HCl-free and H 2 O-free
basis for each condition in part (b).
(d) The conversion of HCl for each condition in part (b). Use the
thermodynamic data provided below:
H2O Cl 2 HCl O2
a 3.470 4.442 3.156 3.639
b 10 3 1.450 0.089 0.623 0.506
c 10 6 --- --- --- ---
d 10 5 0.121 -0.344 0.151 -0.227
h 0f g 0f
Component
J/mol
H2O -241,827 -228,593
HCl -92,307 -95,299
13905 41150
a) ln K = 0.439 Ln(T ) 0.040 10 3 T 13.01
T T2
(For part b), c), d)see summary answers)

5. One method for the manufacture of synthesis gas (CO/H 2 mixture)

depends on the vapour-phase catalytic reaction of methane with steam
according to the reaction:
CH 4 ( g ) + H 2 O( g ) CO ( g ) + 3H 2 ( g ) (1)
The only other reaction which ordinarily occurs to any extent is the
water-gas shift reaction:
CO ( g ) + H 2 O( g ) CO2 ( g ) + H 2 ( g ) .(2)
Chemical equilibrium may be assumed for both reactions. The following

enthalpies and free energies of formation are available:
T = 600K T = 1300K
Component h 0f g 0f h 0f g 0f
J/mol
CH 4 -83,212 -22,970 -91,713 52,300
H2O -244,722 -214,012 -249,450 -175,812
CO -110,165 -164,682 -113,847 -226,940

CO 2 -393,798 -395,137 -393,798 -396,141
(a) Would it be better to carry out the manufacture of synthesis gas at

1 atm or 100 atm?
(b) Would it be better to carry out the process so that the maximum
temperature reached in the reactor is 600 or 1300 K?
(c) Using the preferred conditions from parts (a) and (b) estimate the
molar ratio of H 2 to CO in the synthesis gas produced with a feed
ratio of 1 mole steam to 1 mole methane. (NOTE: this is most easily
solved in excel)
(d) Repeat (c) with a feed ratio of 2 moles steam to 1 mole methane.
(NOTE: this is most easily solved in excel)

(e) How would you alter the feed ratio to obtain a lower ratio of H 2 to
CO than is obtained in part (c) without altering the temperature or
pressure.
(Solution
a) 1 atm b) 1300 K c) 3.007 d) 3.600 e) Use a steam to methane
ratio of less than 1).
6. The gas phase bromination of methane to methyl bromide proceeds

according to the reaction:
CH 4 ( g ) + Br2 ( g ) CH 3 Br ( g ) + HBr ( g )
This reaction is to be carried out at 1 atm and 800 K with an initial ratio
of methane to bromine of 1:1. In view of the above, calculate the
following:
(a) The equilibrium composition of the product gas stream assuming
that no other reactions occur.
g 0f at 800K
Component
J/mol
CH 4 -2,230
CH 3 Br 27,916
HBr -59,367
In reality, when methane is brominated, four possible reactions occur:
CH 4 ( g ) + Br2 ( g ) CH 3 Br ( g ) + HBr ( g ) ..(1)
CH 3 Br ( g ) + Br2 ( g ) CH 2 Br2 ( g ) + HBr ( g ) .(2)
CH 2 Br2 ( g ) + Br2 ( g ) CHBr3 ( g ) + HBr ( g ) ..(3)
CHBr3 ( g ) + Br2 ( g ) CBr4 ( g ) + HBr ( g ) .(4)

(b) Construct a mass balance for this system assuming an initial

methane to bromine ratio of 1:1 and develop expressions for the
equilibrium constant for each reaction.
(Solution
a) CH 4 = 0.05, Br 2 = 0.05, CH 3 Br = 0.45, HBr = 0.45).
( 1 2 )( 1 + 2 + 3 + 4 )
b) K 1 =
(1 1 )(1 1 2 3 4 )
( 2 3 )( 1 + 2 + 3 + 4 )
K2 =
( 1 2 )(1 1 2 3 4 )
( 3 4 )( 1 + 2 + 3 + 4 )
K3 =
( 2 3 )(1 1 2 3 4 )
( 4 )( 1 + 2 + 3 + 4 )
K4 =
( 3 4 )(1 1 2 3 4 )

4.8 Bibliography



ed., 1998.


TOPIC 5 Volumetric Behaviour of Real Fluids & Solids 1
Topic 5
Volumetric Behaviour of Real

Fluids and Solids
Contents
TOPIC 5: VOLUMETRIC BEHAVIOUR OF REAL FLUIDS AND SOLIDS
5.1 INTRODUCTION .................................................................................................................. 3
5.1.1 P-T Diagram ........................................................................................................ 3
5.1.2 P-v Diagrams ...................................................................................................... 5
5.1.3 Relation between P-T & P-v Diagrams ................................................................ 7
5.2 VOLUME EXPANSIVITY AND ISOTHERMAL COMPRESSIBILITY ................................................. 8
5.2.1 Liquids and Solids ............................................................................................... 9
5.2.2 Gases................................................................................................................ 10
5.2.3 Worked Example 5.1 ......................................................................................... 11
5.2.4 Worked Example 5.2 ......................................................................................... 12
5.2.5 Reversible Volume Work ................................................................................... 13
5.2.6 Worked Example 5.3 ......................................................................................... 14
5.3 COMPRESSIBILITY FACTOR .............................................................................................. 15
5.3.1 Two-Parameter Generalised Correlations.......................................................... 16
5.3.2 Three-Parameter Generalised Correlations ....................................................... 17
5.4 VIRIAL EOS .................................................................................................................... 20
5.4.1 Truncated Virial EOS ......................................................................................... 22
5.4.2 Worked Example 5.4 ......................................................................................... 23
5.4.3 Worked Example 5.5 ......................................................................................... 24
5.4.4 Worked Example 5.6 ......................................................................................... 25
5.4.5 Worked Example 5.7 ......................................................................................... 26
5.5 SECOND VIRIAL COEFFICIENT B ..................................................................................... 27
5.5.1 Experimental Determination .............................................................................. 27
5.5.2 Generalised Correlations ................................................................................... 29
5.5.3 Worked Example 5.8 ......................................................................................... 30
5.5.4 Worked Example 5.9 ......................................................................................... 31
5.5.5 Worked Example 5.10 ....................................................................................... 32
5.5.6 Worked Example 5.11 ....................................................................................... 33
5.6 CUBIC EOS .................................................................................................................... 34
5.6.1 Manual Iteration Schemes ................................................................................. 34
5.6.2 Generalised Correlations for Cubic Parameters................................................. 36
5.6.3 Worked Example 5.12 ....................................................................................... 37
5.6.4 RKS EOS in Cubic Form ................................................................................... 39
5.6.5 Worked Example 5.13 ....................................................................................... 40
5.7 BIBLIOGRAPHY ................................................................................................................ 42

Learning objectives
After studying this Volumetric Behaviour of Real Fluids and Solids topic you
should be able to:
Relate the state of a fluid on a P-T and P-V diagram; identify critical
point and the shape of various isotherms in relation to the phase
boundaries (two-phase envelope).
Define expressions for both the coefficient of volume expansivity and
the coefficient of isothermal compressibility and relate these to the
volume change of a species as a function of temperature and pressure.
Devise strategies for solving problems involving the volume change of
pure solids, liquids and gases.
Define reversible work to compress or expand a fluid and devise
strategies for problem solving.
Derive both forms of the Virial EOS truncated to two terms. Derive both
forms of the Virial EOS truncated to three terms.
Differentiate between experimental determination of the second Virial
coefficient and generalised correlation for the second Virial coefficient.
Write down generalised correlation developed by Pitzer for the second
Virial coefficient in terms of the acentric factor.
Develop strategies and solve problems using the Virial EOS.
List advantages and disadvantages of both Virial EOS and cubic EOS.
Outline solution strategies and solve problems using both original RK
and RKS EOS.

5.1 Introduction
Both P T and P v diagrams have been discussed before; however, a

quick review of some major points of interest are listed below.
5.1.1 P-T Diagram
Point C is the critical point, a unique point corresponding to the highest

temperature TC and the highest pressure PC at which liquid and vapour
phases can co-exist together in phase equilibrium.
Point TP is the Triple Point, a unique point on the diagram where three
phases (solid, liquid and vapour) co-exist together in phase equilibrium.
TP C is the
Supercritical vaporisation or boiling
P
A 1 region
C point curve.
PC
Liquid
Solid
region TP A is the fusion or
region freezing point curve.
Gas
region
TP B is the
TP sublimation curve.
B 2
Vapour
region
TC T
The Gibbs phase rule is given by the well-known expression
F = 2 + N
In which F is the number of degrees of freedom, is the number of

phases and N is the number of chemical species present. For a pure species
(as above) there is only one component present and N = 1 .
Areas on this P-T diagram represent single phase regions, which may either
be solid, liquid, or vapour. Anywhere in these regions = 1 and F = 2 . For a
pure species the Gibbs phase rule variables are temperature and pressure.

Thus, areas on the diagram (everywhere except on the lines) correspond to

states that have two degrees of freedom. This means that both temperature
and pressure may be varied independently of each other:
This is because there are no phase equilibrium relations for a single
phase system.
Two variables (temperature and pressure) and no phase equilibrium
relation leads to two degree of freedom.
Lines on the diagram are where two phases co-exist in equilibrium = 2 .

From the phase rule, any point on these lines (except the triple point TP ) have
one degree of freedom F = 1 . Thus, either temperature or pressure may be
varied independently, but not both:
This is because there is now a single phase equilibrium relation linking
the two phases together; there is one phase constraint.
Two variables (temperature and pressure), one phase equilibrium
relation f = f leads directly to one degree of freedom.
The triple point is where three phases all co-exist in equilibrium = 3 . At this
point there are zero degrees of freedom (neither temperature nor pressure
may be varied). The system is said to be invariant:
This is because there are now two phase equilibrium relations linking
these three phases together; there are two phase constraints.
Two variables (temperature and pressure), two phase equilibrium
relations, f V = f L and f L = f S , leads to zero degrees of freedom.
Notice that there are three phase equilibrium relations in total, but only
two (any two) are independent.
Following the straight line between 1 and 2 (the dotted line), the process
line must intersect or cross the vaporisation curve. At this point (where they
intersect) there are two phases at equilibrium and the liquid phase changes
completely into a vapour phase as the pressure declines.
Following the curved path between 1 and 2 (the solid line), the process line
never intersects the vaporisation curve. There is always only one phase
present. The properties gradually change from liquid to vapour without the
system ever passing through a phase transition.
Supercritical carbon dioxide CO2 is an inert, nonflammable, and inexpensive

solvent used in the production of active pharmaceutical ingredients and in the
extraction of delicate essences from plant material.
Using supercritical CO2 means that no other hydrocarbon-based solvents

need be used, the CO2 may then be removed without any further processing
steps; it is simply a matter of changing the pressure and temperature profiles
to carry out the extraction and purification of complex organic molecules.

5.1.2 P-v Diagrams
A P-v diagram also differentiates single phase areas from two phase areas. It
also serves to delineate phase boundaries. The critical isotherm TC has a
point of inflection at P = PC (the critical point C ).
To the right of the saturated vapour curve (but at temperatures below the fluid
critical temperature TC ) is the superheated vapour region. To the left of the
saturated liquid curve is the subcooled liquid (compressed liquid) region.
P
TC Supercritical Curve A C is saturated
Fluid liquid curve.
PC C
Gas Curve B C is saturated

Liquid vapour curve.
Vapour Point C is critical point

where P = PC , T = TC
PTP A B
and v = vC .
vC v
To the right of the saturated vapour curve (but above the critical temperature
TC and below the critical pressure PC ) is the gaseous region, so-called
because a gas cannot be liquefied by the action of pressure alone.
Above the critical temperature TC and above the critical pressure PC , the fluid
is called a supercritical fluid.
TC
P Supercritical Curve 1-2-3-4 is called a
Fluid subcritical isotherm.
PC 4 C
Gas The isotherm is called

Liquid subcritical because the
Vapour temperature T < TC .
Psat 3 2
T < TC
If curve 3-4 were a
A B 1 perfect vertical line then
the liquid would be
vC v incompressible.

At these high pressure states, gaseous properties gradually change into liquid
properties without any phase transition. Notice the discontinuities at the phase
boundary, i.e. points 2 and 3 along the low temperature isotherm 1-2-3-4:
The fact that curve 3-4 is steep (but not perfectly vertical) means that the liquid
actually has small but nearly constant compressibility.
Different isotherms may be plotted on the P-v diagram as follows:
PC C
Q
Gas T4 > TC
N
Liquid
T3 = TC
Vapour
Psat
J K T2 < TC
A Liquid/Vapour B
T1 < TC
vC v
Note the four isotherms: T4 > TC > T2 > T1 . Any isotherm above T3 (say T4 ) is
known as a supercritical isotherm.
Vertical line Q-J represents a constant volume expansion process; notice that
as the pressure drops the vapour fraction increases.
Vertical line N-K represents another constant volume expansion process; as

the pressure drops the liquid fraction increases.
Within the two-phase region liquid and vapour co-exist in equilibrium. Any
mixture point along a horizontal tie-line will split into a saturated vapour phase
(less dense/upper phase) and a saturated liquid phase (more dense/lower
phase).
The properties of each of these co-existing phases (either end of the tie-line)
do not change as the mixing point changes, all that changes is the amount of
each phase that is present (as given by the lever rule).

5.1.3 Relation between P-T & P-v Diagrams
P T and P v diagrams may be linked to each other. Notice how the

saturated vapour pressure curve on right hand P T diagram, corresponds to
two-phase envelope on the left hand, P v diagram.
The saturated vapour pressure curve terminates at the upper end at the critical
point C , while at the lower end it terminates at the Triple Point TP .
P CONSTANT T P
C C
4 4
3 2 LIQUID
P sat 2&3
1 1
TP VAPOUR
A B
T
v
Consider that the fluid pressure is increased isothermally from 1-2-3-4:

P < P sat : fluid is a single-phase vapour (see state 1)
P > P sat : fluid is a single-phase liquid (see state 4).
P = P sat : fluid state is anywhere within the two-phase envelope, i.e.
anywhere along the tie-line between state 2 and state 3.
State 2 corresponds to a saturated vapour state; dryness fraction one, no

liquid is present. State 3 corresponds to a saturated liquid state; dryness
fraction zero, no vapour is present.
Between state 2 and state 3 the dryness fraction varies continuously from
one to zero; this corresponds to complete condensation of vapour.
Within two phase envelope, say from 2 3 , the state is continually

changing but both T & P are constant. Thus, any volume changes along a tie
line cannot be expressed as a function of T & P because they do not change.

5.2 Volume Expansivity and Isothermal Compressibility
If F is any property and x & y are two other properties then, because of the
two-property rule, F may be expressed as a function of x & y
F = f ( x, y )
For any single phase substance (not within the two phase region), the total
derivative dF is given as the sum of two partial changes: one with respect to
x ; the other with respect to y
F F
dF = dx + dy
x y y x
If v is molar volume (cm3/mol) of the substance then, in a similar way, dv may

be expressed as a function of T & P using the total derivative
v v
dv = dT + dP ..(5.1)
T P P T
Multiplying through by total number of mole n of the substance yields the

same equation expressed in terms of V , the total volume (cm3) of the
substance (notice V = nv)
V V
dV = dT + dP ...(5.2)
T P P T
The coefficient of volume expansivity (K-1) is defined as
1 V
= .........(5.3)
V T P
The coefficient of isothermal compressibility (atm-1) is defined as
1 V
= ......(5.4)
V P T
Substitute these last two expressions into equation (5.2) to get
dV
= dT dP ...(5.5)
V

5.2.1 Liquids and Solids
If state 1 corresponds to T1 & P1 and if state 2 corresponds to T2 & P2 , then

equation (5.5) may be integrated choosing some combination of constant
temperature/constant pressure paths.
The exact path chosen between the same two ends states is immaterial since
all paths must yield the same change in volume given by
V2 T
2 P 2
dV
V V T= dT P dP
1 1 1
2T 2P
V2
ln = dT dP (5.6)
V1 T1 P1
For most liquids and solids (not under extreme conditions) & are small
and nearly constant and equation (5.6) is readily integrated to get
2 T 2P
V2
ln = dT dP
V1 T1 P1
V2
ln = (T2 T1 ) (P2 P1 ) ....(5.7)
V1
A) Constant Temperature Process
For liquids and solids undergoing an isothermal process, volume changes are
caused by changes in pressure only and the first term drops out
V2
ln = (P2 P1 ) ....(5.8)
V1
B) Constant Pressure Process
For liquids and solids undergoing an isobaric process, volume changes are
caused by changes in temperature only and the second term drops out
V2
ln = (T2 T1 ) ......(5.9)
V1
& values are experimentally determined and available in the literature.

These values are often needed for Design Projects.

5.2.2 Gases
Equation (5.6) may be applied to gases; however, & can no longer be

regarded as small and constant.
2 T 2 P
V2
V1 T1
ln = dT dP
P1
However, the above expression may still be used if the coefficient of volume
expansivity is a known function of temperature = (T ) and/or if the
coefficient of isothermal compressibility is a known function of pressure
= (P)
A) Constant Temperature Process
For gases undergoing an isothermal process, volume changes are caused by

changes in pressure only and the first term drops out
2 P
V
ln 2 = dP ..(5.10)
V1 P1
This may be integrated, but only if the coefficient of isothermal compressibility

is a known function of pressure = (P ) .
B) Constant Pressure Process
For gases undergoing an isobaric process, volume changes are caused by

changes in temperature only and the second term drops out
T
2
V
ln 2 = dT ..(5.11)
V1 T1
Again, this may be integrated, but only if the coefficient of expansivity is a

known function of temperature = (T ) .
In summary, all these expressions (in sections 5.2.1 and 5.2.2) are only useful
in the following circumstances:
1. When constant (or average) values for & are known for solids or
liquids.
2. For gases when & are known functions of temperature and
pressure respectively.

5.2.3 Worked Example 5.1
The isothermal compressibility of lead is 2.21 10 6 atm-1. Calculate the

pressure that must be applied to increase its density by 0.08%.

Calculate the volume expansivity of gas X at a constant Pressure ( P ) of

10 atm. The initial volume is 1 m3. Experimental measurements are:
Temperature T (K) Volume V (m3)

100 1.0
250 2.2
390 3.5
460 4.2
510 4.8

5.2.5 Reversible Volume Work
If a fluid is compressed within a piston/cylinder device, then work must be

supplied to the fluid (positive work); the reverse is true if the fluid is expanded.
This type of work is called displacement work or volume work.
Work done on the system by the surroundings is positive and work done by
the system on the surroundings is negative.
For reversible compression or expansion of a fluid the volume work is given

by
dw = Pdv (5.12)
In which w is the work done on or by the fluid (kJ/mol), P is the absolute

pressure (kPa) and v is the molar volume (m3/mol) of the fluid. If a fluid is
compressed dv will be negative and then the negative sign is sign is needed
to keep the work term positive (work transferred to the system.
Alternatively, multiplying through by n (mol) converts to total quantities as

follows:
dW = PdV (5.13)
In which W is the total work done on or by the fluid (kJ), again P is the
absolute pressure (kPa) and V is now the total volume (m3) of the fluid.
In order to calculate the reversible work done on or by the fluid equations

(5.12) or (5.13) will need to be integrated. There are two strategies depending
on how the problem is framed:
1. Eliminate dV and then integrate to find W . This involves using either
equation (5.3) or (5.4). To use this approach for gases or must
be known functions of either temperature or pressure respectively.
2. Eliminate P and then integrate to find W . This approach involves
using an Equation of State (EOS). This will be discussed in the next
section (immediately following this worked example).
In the section that follows the next worked example, the Virial EOS will be
examined. Although the Virial EOS is used here to investigate the volumetric
behaviour of gases v = v(T , P ) , it could equally be used to calculate the
fugacity coefficient behaviour of gases = (T , P ) .

Calculate the reversible work to compress 5 m3 of gas A at constant

temperature (298 K) from 1 atm to 200 atm. The isothermal compressibility of
the gas is given as the following linear function of pressure:
= a bP
= 4 10 6 (1.5 10 10 P)
In which P and are quoted in atm and atm-1 respectively.

5.3 Compressibility Factor
The simplest EOS, shown below, only applies to ideal gases:
Pv ig = RT
In the limit of very low pressure (as pressure approaches zero) real gases
approach ideal gas behaviour. A simple way of finding molar volume v for a
real gas is to use a correction factor Z , as follows:
Pv = ZRT ..(5.14)
Z is known as the compressibility factor and after re-arranging equation

(5.14) it is clear that Z = Pv / RT . If v is measured in units (cm3/mol), P in
(bar), T in (K) and if R = 83.1447 (cm3 bar/mol K), then Z is dimensionless.
Ideal gases are assumed to consist of point particles occupying zero volume
and with zero forces of attraction. However, Van der Waal reasoned (correctly)
that the departure from ideal gas behaviour was caused by two factors:
1. Gases particles (molecules/atoms) do have a definite volume which
exerts a repulsive force at short-range distances as particles approach
each other.
2. There are longer-range forces of attraction between gas particles which
drop off as particle separation is increased.
However, for real gases, as P 0 the separation between gas particles

increases and these very short-range repulsive forces become insignificant
(the particles effectively act as if they were point particles with zero volume).
The force of attraction for real gases follows an inverse square law. Thus, for
real gases as P 0 the separation between gas particles increases and
these longer-range forces of attraction decline in strength (the particles
effectively act as if they had zero force of attraction).
For real gases (at any temperature) as P 0 , then Z 1 and equation

(5.14) reduces to the ideal gas law. It is clear that if a gas is behaving as a
perfect ideal gas, then the compressibility factor Z = 1 .
On the other hand, the larger the departure of a real gas from ideal gas
behaviour the more the compressibility factor will deviate from unity. In these
cases, the compressibility factor may be greater or less than unity.

The definition of Z may be found by re-arranging equation (5.14) as follows:
Pv
Z= .(5.15)
RT
If no experimental data is available, then the compressibility factor may be

obtained from some sort of Z -factor correlation. Two types of Z -factor
correlation have been developed:
1. Two-Parameter correlations.
2. Three-Parameter correlations.
5.3.1 Two-Parameter Generalised Correlations
A two-parameter Z -factor correlation takes the general form
Z = Z (Tr , Pr )
Reduced temperature Tr is defined as follows:
T
Tr = ..................................................................(5.16)
TC
Reduced pressure Pr are defined as follows:
P
Pr = ..(5.17)
PC
Where,
T is absolute temperature of the system (K).
TC is absolute critical temperature of fluid (K).
P is absolute pressure of the system (bar).
PC is absolute critical pressure of the fluid (bar).
Two-parameter correlations are based on the two-parameter Law of

Corresponding States (also known as the Corresponding States Principle
CSP):
All fluids at same Tr and Pr deviate from ideality to the same extent
and, therefore, have the same compressibility factor Z .

The above statement holds for simple gases with symmetrical charge
distribution (such as noble gases Ar, Kr and Xe), but other non-simple gases
depart from the two-parameter Law of Corresponding States.
5.3.2 Three-Parameter Generalised Correlations
For other non-simple gases, a three-parameter correlation is needed. A

three-parameter generalised correlation relates some property to reduced
temperature Tr , reduced pressure Pr and a third parameter called the
acentric factor .
Thus, a three-parameter generalised correction for the compressibility factor

takes the general form
Z = Z (Tr , Pr , )
Three-parameter correlations are based on the three-parameter Law of

Corresponding States:
All fluids with the same acentric factor and at the same Tr and Pr
deviate from ideality to the same extent and, therefore, have the same
compressibility factor Z .
Experimental results indicate that the data for each fluid falls nearly on a
straight line see below:
1 / Tr
Tr = 0.7
- 1.0 Simple Fluids Ar

Kr, Xe
log Prsat
Non-Simple
Fluids

sat
In order to understand the acentric factor plot log( Pr ) versus 1 / Tr for a
number of different fluids; the reduced vapour pressure Prsat is just the vapour
pressure P sat (or the saturation pressure) divided by the critical pressure PC
P sat
Prsat = .....(5.18)
PC
It is also found that Ar, Kr and Xe all lie on a common line (the red line) and
that, for all these noble gases, log Prsat = 1 at a Tr = 0.7 (these gases are
called simple fluids because they have a uniform charge distribution).
Data for other non-simple fluids fall on other straight lines (the dotted blue
lines); the more non-uniform the charge distribution the more the line departs
from the simple fluid case.
Thus, an experimental value of log Prsat at a reduced temperature of Tr = 0.7

may be used to characterise the departure of any non-simple gas from simple
gas behaviour see below:
1 / Tr
Tr = 0.7
- 1.0 Simple Fluids Ar

Kr, Xe
log Prsat
Non-Simple
Fluids
Thus, the acentric factor, defined below is a measure of how asymmetric the
charge distribution is for any gas:
= log10 ( Pr sat ) T =0.7 1 .(5.19)

t

The acentric factor is easily determined by a single vapour pressure

measurement P sat at a specified reduced temperature of Tr = 0.7 .
For simple gases (Ar, Kr, Xe) the acentric factor is zero. The three-parameter
correlations are more accurate than two-parameter correlations and have a
good accuracy over a full range of states (particularly when using tables).
Three-parameter graphical correlations use two charts: the Z 0 component

chart (sometimes written as Z ( 0 ) ) and the Z 1 component chart (sometimes
written as Z (1) ). These two component parts of Z are combined in a simple
way:
Z = Z 0 + Z 1 .(5.20)
Z 0 is the major component of Z due to simple gas behaviour, while Z 1 is the

minor adjustment to Z due to non-simple behaviour. Charts for the Z 0 and
Z 1 have been discussed in earlier topics.

5.4 Virial EOS
Charts and tables are useful ways of finding the compressibility factor.
However, algebraic relationships are often more convenient.
For real gases it is possible to model the shape of isotherms (well to the right
of the critical point) using a power series:
Pv = a + bP + cP 2 + .......
Let b = aB , c = aC , etc., then substitute these into previous equation to get
Pv = a (1 + B P + C P 2 + .......) ...(5.21)
In which a, B , C are constants that depend only on temperature and the

nature of the chemical species. The Virial coefficient reflects the chemical
character of species and is a function of molecular interactions: the second
Virial coefficient depends only on two-molecule interactions; the third on three-
molecule interactions. As such they provide a correction for non-interactive
ideal gases.
In the limit as P 0 it follows that
Pv = Pv * a = RT
Notice it is only possible to set a = RT when the Kelvin scale of temperature

is used and this incidentally is defined as follows:
T
=
( )
Pv *
TTP ( )
*
PvTP
By definition the triple point of water is chosen to have a precise numerical

value TTP 273.16 K (this value is chosen to align a degree on the centigrade
scale with that on the Kelvin scale). Thus, to find T on the Kelvin scale of
temperature
T = 273.16
(Pv ) ..(5.22)
*
(Pv )
*
TP
An ideal gas thermometer may be immersed in a triple point mixture of pure

*
water to find PvTP . If an ideal gas thermometer is then place in boiling water
(at 101.325 kPa) and Pv * is measured, then equation (5.22) will provide the
boiling point of water on the Kelvin scale T = 373.15 K.

Bearing this in mind a = RT (using the Kelvin scale) and the value of the
universal gas constant may be taken as R = 8.31447 (J/mol K). In this course
it is more convenient to use R = 83.1447 (cm3 bar/mol K).
If a = RT is substituted into equation (5.21), then the following power series is

obtained:
Pv = RT (1 + B P + C P 2 + .......) ...(5.23)
In order to express equation (5.23) as a function of the dimensionless

compressibility factor Z , simply divide through by RT to obtain one form of
the extended Virial EOS
Z = (1 + B P + C P 2 + .......) .(5.24)
An alternate form of the Virial EOS is given by
B C
Z = 1 + + 2 + ........... ......(5.25)
v v
Where,
B
B ....(5.26)
RT
C B2
C ..(5.27)
R 2T 2
Notice that equations (5.26) and (5.27) are only approximately true. The
equality only hold in the limit of an infinite number of terms.
Either form of the Virial EOS, equation (5.24) or equation (5.25), may be used;
however, the higher the system pressure the more terms that are needed.
Earlier it was shown that Z is dimensionless and this means that each term
on the right of (5.24) and (5.25) must be dimensionless. In which case:
B , equation (5.24) must have the same units as pressure P which
would be (bar-1).
B , equation (5.25) must have the same units as molar volume v which
would be (cm3/mol).
Thus, the units of the second Virial coefficient depend on whether the
expansion is with respect to P or v .
The Virial EOS is always truncated. Often the truncation is to the first two
terms, but sometimes it is truncated to the first three terms.

5.4.1 Truncated Virial EOS
The Virial EOS is usually truncated to two terms and as such applies only to
gases or vapours at low to moderate pressure; notice that the molar volume of
liquids cannot be found from these expressions.
Equation (5.24) may be truncated to two terms to get one form of the truncated
Virial EOS
Z = 1 + B P
Equation (5.28) is usually expressed in terms of B . Substitute equation (5.26)

into the previous expression to get
BP
Z = 1+ ...(5.28)
RT
Equation (5.25) may be truncated to two terms to get a second form of the
truncated Virial EOS
B
Z = 1+ ..(5.29)
v
In this course problems are frequently solved by re-arranging equation (5.29)

as follows
Pv B
= 1+
RT v
BRT
Pv = RT +
v
Pv 2 RTv BRT = 0 ..(5.30)
This leads to a quadratic in molar volume v (cm3/mol), but care must be taken
to use the correct units: if universal gas constant is taken to be R = 83.1447
(cm3 bar/mol K); then P must be expressed in (bar); and T must be in (K).
The second Virial coefficient B will have the same units as molar volume, see
equation (5.29) and, therefore, will have units (cm3/mol).

At still higher pressure (to get a more accurate answer) equation (5.25) may
be truncated to the first three terms as follows:
B C
Z = 1+ + ....(5.31)
v v2
It is clear from equation (5.23) that in the limit as P 0 all higher order terms
drop out and, thus, all forms of the Virial EOS reduce to the ideal gas law:
Pv = RT Z =1
As a rule of thumb the first two terms are reasonably accurate (depending on
temperature) up to about 15 bar, while the first three terms will provide
reasonable accuracy to about 50 bar.
Calculate Z and v for chemical species X at a temperature of T = 90 o C

and a pressure of P = 10bar .
Use the Virial EOS truncated to the first two terms and the second Virial
coefficient B is listed in the experimental literature as
B = 194 cm 3 mol 1

Calculate Z and v for SF 6 at 75oC and 15 bar using the Virial EOS truncated
to two terms. The second Virial coefficient B is listed in the experimental
literature as
B = 194 cm 3 mol 1
Then calculate v at the same conditions using the Virial EOS truncated to
three terms. The third Virial coefficient C is listed in the experimental literature
as
C = 15300 cm 6 mol 2

Given that the second Virial coefficient B for N 2 is 21.7 cm 3 mol 1 , at a

temperature of 273 K and a pressure of 100 bar, calculate molar volume v
using:
(i) Ideal gas EOS

(ii) Virial EOS truncated to two terms

0.01 kg of steam at 260oC expands reversibly and isothermally from an initial

volume of 2 dm3 to a final volume of 20 dm3. Given the following Virial
coefficients:
B = 142.2 10 6 m 3 mol 1
C = 7140 10 12 m 6 mol 2
Using the Virial EOS truncated to three terms, calculate how much work is
done by the steam during the expansion

5.5 Second Virial Coefficient B
There are two approaches to finding the second Virial coefficient B :

1. The scientific approach is to find B experimentally or by referencing
peer-reviewed publications and/or data handbooks.
2. The engineering approach is to estimate B using some sort of
generalised correlation.
The scientific approach is more accurate and rigorous. On the other hand, the
engineering approach may be less accurate but is more convenient. Engineers
can allow for inaccuracy by including factors of safety into their designs or
estimates.
5.5.1 Experimental Determination
For a particular chemical species consider the Virial EOS as given by equation
(5.24), as shown again below:
Z = (1 + B P + C P 2 + .......)
Differentiate Z with respect to P at constant T to get
dZ
= B + 2C P + .......
dP
This gives the slope at any point along a Z vs P isotherm. Now find the slope
of any isotherm at zero pressure, i.e. as P 0
dZ
= B
dP P =0
Thus, the second Virial coefficient B is the slope of an isotherm at zero

pressure and this provides a method of obtaining B experimentally:
Choose a temperature T which corresponds to one particular
isotherm.
Measure the dZ / dP at pressure P .
Repeat the procedure at successively lower pressures.
Plot dZ / dP on the y-axis against pressure P on the x-axis.
Extrapolate the trend to zero pressure to obtain the y-axis intercept.
This intercept is (dZ / dP ) P =0 which is equal to B .

At zero pressure all gases behave as idea gases and the compressibility factor
is Z = 1 . Clearly all isotherms must meet at the point Z = 1 and P = 0 :
Red curve represents

Z any isotherm.
Z=1
1.0 The blue line is
tangent to the
Slope = B
Isotherm isotherm at Z = 1 and
P = 0.
The slope of the

tangent line is B .
0 P
From co-ordinate geometry, the equation of the tangent line is clearly
Z = 1 + B P
This result is identical to the Virial EOS truncated to two terms as given by
equation (5.28) below:
BP
Z = 1+
RT
Thus, the graphical interpretation of the two-term Virial EOS, equation (5.28),
is that it represents the equation of a straight line which is a tangent to the
actual isotherm at Z = 1 and P = 0 .
From this diagram it is clear that the Virial EOS truncated to two terms is a
better fit to the isotherm, at finite pressure, than the ideal gas law ( Z = 1 ).
However, the accuracy of the Virial EOS drops off at higher pressure.
The straight line fits the shape of an isotherm reasonably well at low pressure;
up to about 15 bar.
To improve the accuracy the Virial EOS may be truncated to three terms.
However, more data is available for the second Virial coefficient B than is
available for the third Virial coefficient C .
The accuracy of the two-term Virial EOS (either version of it) also depends on
the gas temperature; the lower the temperature, at any pressure, the lower the
accuracy.

5.5.2 Generalised Correlations
Equation (5.28) can be re-arranged and put into generalised form
BP
Z = 1+
RT
BP P / PC
Z = 1 + C
RTC T / TC
BP Pr
Z = 1 + C .....(5.32)
RTC Tr
Pitzer and co-workers then proposed that BPC / RTC consists of two parts:
1. The first component is labelled B 0 and is the simple fluid component
based on Ar, Kr and Xe.
2. The second component is labelled B 1 and is the non-simple fluid
component based on the departure from simple fluid behaviour.
The two parts are combined as follows
BPC
= B 0 + B 1 ...(5.33)
RTC
In which is the acentric factor defined previously by equation (5.19):
= 1.0 log( Pr sat ) T =0.7

t
Substitute equation (5.33) into (5.32) to get
(
Z = 1 + B 0 + B 1 ) TP r
Pr P
Z = 1+ B0 + B 1 r ...(5.34)
Tr Tr
The B 0 and B 1 components have been correlated from a database of

experimental measurements (by Abbott and Van Ness) as follows:
0.422
B 0 = 0.083 1.6
..(5.35)
Tr

And,
0.172
B 1 = 0.139 4.2
.(5.36)
Tr
The Virial EOS may be summarised as follows:

The various forms of the Virial EOS may only be used for gases; they
cannot be used to model liquid phase properties.
The truncated two-term Virial EOS produces reasonable results,
depending on temperature, for gases at pressures up to about 15 bar.
The truncated three-term Virial EOS is reasonably accurate, depending
on temperature, for gases up to about 50 bar.
Since there is much less data available for the third Virial coefficient C
than for B , it is preferable to use a cubic EOS at high pressure.
Cubic EOS have the added advantage that they can model liquid
properties as well as vapour properties.
While the various forms of the Virial EOS are important and must be
remembered, it is not expected that candidates have to remember
correlations, such as equations (5.35) and (5.36).
Starting with the general volume change expression below:
dV dV
dV = dT + dP
dT P dP T
Calculate the % volume change when 10 m3 of fluid A is compressed from 1

atm to 500 atm at constant temperature (300K). The coefficient of isothermal
compressibility is known to be
= 4 10 4 atm 1

One mole of a gas undergoes an isothermal compression at 390 K in a piston/

cylinder assembly where the molar volume of the gas is given by
RT
v= +b
P
Where,
b = 42.6 cm 3 /mol
If the starting volume is 10 cm 3 /mol and the final volume is 2 cm 3 /mol ,

calculate the reversible work input per mole (J/mol) needed to compression
the gas.

Calculate Z and v for gas X at 70oC and 6 bar by applying the Pitzer
correlation to calculate B where:
TC = 330 K
PC = 42 bar
VC = 212 cm 3 /mol
= 0.273

Calculate Z and v for ethane at 50oC and 15 bar using the following
methods:
(a) using the truncated Virial EOS
B = 156.7 cm 3 /mol
(b) using the Pitzer correlation
TC = 305 K
PC = 48.7 bar
= 0.10

5.6 Cubic EOS
Cubic equations of state have been discussed in other courses as well as in

Topic 1 of this course. A brief summary is supplied below.
The Van der Waal (VdW) EOS is the root cubic from which all modern
versions are derived. The pressure-explicit form of the VdW EOS is given
below:
RT a
P= 2 ...(5.37)
(v b ) v
In which a & b are species specific parameters. The VdW EOS is now of
historical interest only. A more modern version of VdW is the Redlich-Kwong
(RK) EOS. The pressure-explicit form of the RK EOS is given by
RT a
P= 0.5 ...(5.38)
(v b ) T v(v + b)
In which a & b are again species specific parameters. Notice that VdW a & b
parameters differ from the RK a & b parameters. Notice that when a & b are
zero both of the above expressions revert to the ideal gas law.
5.6.1 Manual Iteration Schemes
Often pressure and temperature are fixed and the objective is to find the molar
volume. Of course all cubic EOS will yield three volume roots and various
manual iteration schemes have been proposed to find these roots.
For instance, multiply (5.38) by (v b ) / P and then re-arrange to get the

following expression which efficiently converges to the vapour root (the largest
root):
RT a(v b )
vi +1 = + b 0.5 i .(5.39)
P T Pvi (vi + b)
In which vi +1 is the next molar volume in the iteration cycle and vi is the
current value. A reasonable starting value v 0 is the ideal-gas molar volume
RT
v0 = ..(5.40)
P

Equation (5.38) may be expanded out as follows (notice that the cubic form of
the RK EOS is now evident):
RT 2 2 bRT a ab
v3 v b + v = 0 .(5.39)
P P PT 0.5 PT 0.5
An iteration scheme designed to converge to the liquid root (the smallest root)
has been developed from equation (5.39)
1 RT 2 ab
vi +1 = vi3 vi ...(5.40)
c P PT 0.5
Where,
bRT a
c = b2 + .(5.41)
P PT 0.5
In which vi +1 and vi are again the molar volumes at successive stages of the
iteration cycle. A good starting value v 0 is b which represents the volume of
the molecules themselves
v0 = b ....(5.42)
The RK EOS yields three roots, select the correct root as follows:
P < P sat : fluid must be a single-phase vapour, choose largest root.
P > P sat : fluid must be a single-phase liquid, choose smallest root.
P = P sat : fluid state may be anywhere along a tie-line, choose both the
largest and the smallest roots, since both saturated liquid and vapour
can be present.
The a & b parameters may be obtained experimentally by fitting the RK EOS

to volumetric P v T data; this would be the most accurate approach.

5.6.2 Generalised Correlations for Cubic Parameters
A more convenient approach for engineers would be to use generalised

correlations for the parameters a & b and, for the VdW EOS, these are given
by the expressions below:
2
27 R 2TC
a= ..(5.43)
64 PC
RTC
b= ..(5.44)
8 PC
While the generalised a & b parameters for the RK EOS are as follows:
2.5
R 2TC
a = 0.42748 .(5.45)
PC
RTC
b = 0.08664 ..(5.46)
PC
Equation (5.38) may be multiplied through by v / RT to get the compressibility

factor Z which can then be re-arranged to get
Pv v ab
=
RT (v b ) bRT (v + b)
3/ 2
1 a h
Z =
1 h bRT 1.5 1 + h
Where,
b bP
h= =
v ZRT
Substituting (5.45) and (5.46) into the last two equations then gives
1 4.934 h
Z = 3/ 2 ...(5.47)
1 h Tr 1 + h

And,
0.08664 Pr
h= ..(5.48)
ZTr
Equations (5.47) and (5.48) are solved together to find the vapour root. The
following procedure is used:
Set Z = 1 and then find then find h from equation (5.48).
Substitute this value of h into equation (5.47) and find a new value for
Z.
Repeat this procedure until Z converges within some tolerance.
The worked example below will not appear as an examination question since it
just involves substituting supplied data into supplied equations. However,
these equations are very useful (in Design projects) for finding properties of
real fluids.
Therefore, it is important to understand how cubic EOS work, when they are
appropriate and how they may be solved.
Calculate Z and v for gas X at 90oC and 10 bar. Use the Redlich-Kwong
EOS together with the following data:
TC = 319 K
PC = 38 bar
vC = 198 cm 3 mol 1


5.6.4 RKS EOS in Cubic Form
The original RK EOS was updated by Soave in 1972 and is now much more
accurate. In textbooks this is either called the Soave-Redlich-Kwong (SRK)
EOS or the Redlich-Kwong-Soave (RKS) EOS.
The two cubic EOS used most frequently by chemical engineers are the Peng-
Robinson (PR) EOS and the RKS EOS. The Peng-Robinson EOS was
discussed in detail towards the end of Topic 1.
The best approach for solving the RKS EOS is to re-arrange it into cubic form
(in terms of the compressibility factor Z ) and then find all three roots either
using an advanced hand calculator, an on-line calculator, or using any
mathematics package.
Find first step is to find the pure component aT & b parameter from the three
equations given below:
RTC
b = 0.08664 .(5.49)
PC
2
R 2TC
aT = 0.42748 a (T ) .....(5.50)
PC
[ (
(T ) = 1 + (0.48 + 1.57 0.17 2 ) 1 Tr 0.5 )]
2
....(5.51)
Notice that an alpha function has been incorporated into the aT parameter
making it a function of temperature and acentric factor (thus the subscript).
These refinements provide much better vapour pressure prediction.
The dimensional aT & b parameters are then converted into dimensionless

A & B parameters as follows:
aT P
A= .................(5.52)
R 2T 2
bP
B= ...(5.53)
RT

The cubic form of the RKS EOS is then expressed in terms of these
dimensionless A & B parameters:
Z 3 + (1) Z 2 + ( A B B 2 ) Z + ( AB) = 0 (5.54)
The final steps are to find all three roots, select the correct root and find the
molar volume from the definition of Z - see below:
Pv
Z=
RT
Repeat the previous problem using the approach outlined above. To make this
consistent with the original RK EOS, simply use the following expression for
the alpha-function.
1
(T ) = 0.5
Tr
If the same critical properties are used then the same answer should be
obtained. If the RKS alpha-function is used, then a more accurate result
should be obtained.
Assume the acentric factor to be 0.210.


5.7 Bibliography



ed., 1998.

TOPIC 1: Glossary 1
Topic 7
Glossary
Azeotrope
If a binary VLE system is in phase equilibrium, then an azeotrope will be

present at a particular composition only if the volatility of each species is the
same, which is the same as saying that the K-value of each species is unity, or
that the relative volatility of one species with respect to the other is unity.
Since the volatility of each species is the same, then the vapour and liquid
phases must both have the same composition. Distillation separates
component by exploiting difference in volatility and, therefore, distillation alone
cannot separate an azeotropic mixture.
Azeotrope form when the liquid phase deviates from ideality either positively or
negatively. Azeotropes form when the species have close boiling points (close
vapour pressures at prevailing temperature) and when there are sufficient non-
ideality for the relative volatility to be unit.
A maximum pressure azeotrope is caused by strong positive deviation from

ideality and corresponds to a minimum temperature azeotrope. On the other
hand, a minimum pressure azeotrope is caused by strong negative deviation
from ideality and this corresponds to a maximum temperature azeotrope.
Chemical Equilibrium
Chemical equilibrium is achieved when there is no tendency for any change to

occur within a reaction system. This means that the temperature, pressure and
composition are all absolutely uniform throughout the system.
Such a system will experience no heat transfer, no work transfer and no net
change in composition. Of particular interest in a chemical reaction, where the
temperature and pressure are constant, is that the composition is then fixed
and cannot change.
Chemical equilibrium for a reaction system does not imply that conversion
between reactants and products has stopped; it implies a dynamic equilibrium
where the rate of inter-conversion of species is the same in either direction.
That is the forward and the reverse reactions are operating at the same rate
and, therefore, the composition cannot change.
HERIOT-W ATT UNIVERSITYB49CEAugust 2015 v1

GLOSSARY 2
Chemical equilibrium occurs at a point where the total Gibbs free energy of the
reaction mixture is a minimum. This in turn leads to the species activity, raised
to the power of its stoichiometric coefficient, summed over all species taking
part in the reaction, is equal to the equilibrium constant.
Chemical Potential
The chemical potential of a species in a mixture is the partial molar Gibbs free
energy that that component has in a mixture (this is different from the molar
Gibbs free energy that a component has as a pure species). The chemical
potential, like all potentials must be an intensive property. Thus, unlike the
Gibbs free energy, it does not have an extensive form and this is why,
although equal to partial molar Gibbs free energy, the chemical potential has a
different symbol.
Closed system
A closed system in one where the same quantity of matter remains within the
boundary for all times that is matter cannot cross the boundary. In this case,
only energy in the form of heat or work can cross the boundary.
Processes occurring within closed systems are called non-flow processes. In

the US a closed system is often referred to as a Control Mass.
Compressed Liquid
A compressed liquid is when, at a fixed temperature, the species pressure is

larger than the vapour temperature. Thinking about it the other way around, it
also means that, at a fixed pressure, the temperature is less than the species
boiling point. Thus, a compressed liquid is the same as a subcooled liquid.
Compressibility Factor
The compressibility factor is the ratio of the molar volume of the real fluid to
the molar volume of an ideal gas at the same temperature and pressure. The
compressibility factor corrects the molar volume between the ideal gas state
and the real fluid state. Both liquids and gases may have compressibility
factors and the ideal gas is the reference state for both. As pressure tends to
zero, the compressibility factors tend to unity.
HERIOT-W ATT UNIVERSITY..June 2014 v1

GLOSSARY 3
Energy Balance
An energy balance can be applied to any item of process equipment. The

general principle is that energy is conserved and the total energy entering a
system must be equal to the total energy leaving a system (whatever the
system). As well as energy associated with inlet and exit streams, energy can
also enter or leave the system as either heat or work. The energy content of
streams is expressed in terms of enthalpy.
Batch systems are handled in much the same way as continuous systems, so
long as the energy content of the charge material and discharge material is
considered in place of streams. The energy content charge and discharge
material is expressed in terms of internal energy.
Enthalpy
The enthalpy represents the energy content of a system and as such is the
energy that is stored in the system. The enthalpy is the sum of the internal
energy plus the combination Pv . Both terms have units (J/kg), in the case or
specific enthalpy or (J/mol) in the case of molar enthalpy.
Enthalpy is often associated with steady-flow processes, but also occurs when
the heat transferred to or from a non-flow constant pressure process.
Enthalpy balance
For the special case of a steady-flow process, without any work or heat and
with negligible kinetic and potential energy effects, the steady-flow energy
equation reduces to an enthalpy balance.
An enthalpy balance simple states that the total rate at which enthalpy enters
the device, summed over all inlet streams, is equal to the total rate at which
enthalpy leaves the device, summed over all exit streams.
The rate at which enthalpy enters or leaves with a stream (kW) is the product
of the stream specific enthalpy (kJ/kg) times the mass flow (kg/s) of the
stream. Alternatively it can also be the product of the stream molar enthalpy
(kJ/kmol) times the mole flow (kmol/s).
Enthalpy balances can be applied to mixing and splitting points, flow through
expansion valves or any item of equipment where no heat or work is either
supplied or removed.

GLOSSARY 4
Enthalpy of formation
The enthalpy of formation h of is the enthalpy change associated with forming

one mole of a compound from its constituent elements. The elements should
be as they commonly occur in nature, e.g. Cl 2 rather than Cl.
Pure elements are defined as having an enthalpy of formation of zero. The

standard enthalpies of formation are usually listed for pure species at 298.15 K
and at the reference pressure.
Enthalpy of reaction
The enthalpy of reaction is the pure component enthalpy of each species, in its
standard state, times its stoichiometric coefficient and then summed over all
reacting species.
The standard state of the pure species corresponds to a particular state of

aggregation of the pure species at the reaction temperature and a reference
pressure of 1 bar. Thus, the standard enthalpy of reaction depends on
temperature only (and is independent of pressure).
The standard enthalpy of reaction may be calculated in terms of the enthalpies

of formation of individual species and these are usually listed in tables at
298.15 K.
Exothermic reactions have a negative enthalpy of reaction. Endothermic

reactions will have a positive enthalpy of reaction. Whether heat is actually
given out or taken in also depends on the design of the reactor. The enthalpy
of reaction varies depending on the temperature and the state of aggregation
of both reactants and products. Although over short temperature ranges it may
be considered constant.
It also depends on how the reaction as written. Thus, if the stoichiometric

coefficients are doubled or halved then, through Hesss Law, the same must
apply to the standard enthalpy of reaction.
The enthalpy of reaction has units of (J/mol) and while this looks like a molar
property it is not one in the usual sense. The enthalpy of reaction is the heat
effect pre mole of reaction, that is to get the heat released, or absorbed (J), the
enthalpy of reaction must be multiplied by the extent of reaction, not the total
number of moles present or flowing through the system.
It is not good practice to refer to the enthalpy of reaction as the heat of

reaction. This is because heat does not reside within a system, it is an energy
exchange term.

GLOSSARY 5
Entropy
Entropy is commonly understood as a measure of disorder within a system:

the more disorder the larger the system entropy; the less disorder the lower
the system entropy.
The second law of thermodynamics states that the entropy change of an

isolated system at best (reversible process) stays constant or increases
(irreversible process). As a system spontaneously moves towards its
equilibrium state, the entropy change is always positive and when it reaches
equilibrium the entropy of the system will be a maximum. Therefore, at
equilibrium the entropy change will be zero.
The same law applies to systems that are not isolated, but now both the
entropy change of the system and the surroundings must be considered
together. For a reversible process the total entropy change (both system and
surroundings) is zero. However, for irreversible processes the total entropy
change (system and surroundings) is greater than zero. An increase in entropy
during any process is associated with energy being degraded and a loss in
capacity to do work.
Once this capacity to do work has been lost it may never be recovered and
this is the meaning of the expression that the process is irreversible, literally-
speaking something irrevocable has occurred that can never be undone. Once
work has been lost as random thermal energy it may never be completely
recovered.
In classical thermodynamics the entropy change of the system is found by

integrating the incremental heat added (positive entropy change) or removed
(negative entropy change) from the system divided by the absolute
temperature of the system at the instant that the heat is added or removed.
The requirement is that this integration must be carried out from the initial state
to some final state using an imaginary reversible process that connects the
two states.
Entropy was first identified when it was discovered that this integral summed to
zero over a cycle. Entropy is, therefore, a thermodynamic property (all state
functions, thermodynamic properties, must sum to zero over a cycle) and its
existence is a consequence of the second law of thermodynamics.
The entropy change of the surroundings is found in the same way. However,
for an infinite heat reservoir the surroundings temperature may be taken as
fixed. Thus the entropy change of the surroundings is just the heat added
(positive entropy change) or removed (negative entropy change) divided by
the absolute fixed temperature of the surroundings. The second law then
becomes the entropy change of the system plus surroundings must be greater
(irreversible) or equal to (reversible) zero. It is impossible to have any process
that results in a total entropy change of less than zero.

GLOSSARY 6
Equation of State
An Equation of State EOS is simply a relationship between pressure, total

volume, temperature and the number of moles. The simplest EOS is the ideal
gas law. A better Equation of State, although now out of date, is the Van der
Waal EOS. Many accurate EOS have been developed over the years.
Chemical Engineers often use Redlich-Kwong-Soave (RKS) EOS or the Peng-
Robinson (PR) EOS.
These are popular because they can model both liquid and vapour properties,
although vapour properties tend to be more accurately modelled than liquid
properties. Polar liquids are especially difficult to model and empirical
correlations are used to find the properties of these mixtures.
Equilibrium Conversion
Equilibrium Conversion is defined as moles of reactant species in the feed

that has reacted, once the system is at chemical equilibrium, divided by
moles of the same component in feed. It may be expressed as a percentage or
as a fraction.
The equilibrium conversion can be calculated for any reactant species. However,
it is usually the conversion of the limiting reactant that is of interest (see also
limiting reactant and excess reactant in this glossary). The moles of a
reactant which has reacted is equal to the extent of reaction times the species
stoichiometric coefficient.
In the case of industrial reactors, particularly steady-flow reactors, there may

be insufficient residence time for the reaction system to reach equilibrium. In
this event the fraction conversion, defined the same way, will be less than the
equilibrium fractional conversion.
Equilibrium Extent of reaction
The equilibrium extent of reaction is the amount of a species that has reacted,
when the reaction system is at chemical equilibrium, divided by its
stoichiometric coefficient. The equilibrium extent of reaction is always positive
if the reaction proceeds in the forward direction. It either has units of (mol), for
non-flow processes, or (mol/s), for steady-flow processes. It is the same for all
reacting species in the mixture.
Naturally, the equilibrium extent of reaction must lie somewhere between zero
and some maximum value and this in turn is moles of limiting reactant initially
present divided by its stoichiometric coefficient.

GLOSSARY 7
The equilibrium extent of reaction depends on the equilibrium constant. If the

equilibrium constant is very small (several orders of magnitude less than one)
then the equilibrium extent of reaction will be close to zero. If the equilibrium
constant is very large (several orders of magnitude greater than one), then the
equilibrium extent of reaction will be close to its maximum value (mentioned
above).
If the equilibrium constant is around one, then the equilibrium extent of

reaction will have a value somewhere near the middle of its possible range.
This information is very useful when a starting value is needed for the
equilibrium extent of reaction (often the case when a trial-and-error approach
is adopted).
Excess reactant
Excess means that one reactant is supplied in excess of the theoretical

stoichiometric proportion. In practice one reactant is always supplied in
excess to force the reaction to the right and to ensure that as much as
possible of the limiting reactant actually reacts at equilibrium.
Extent of reaction
The extent of reaction is the amount of a species that has reacted divided by
its stoichiometric coefficient. The extent of reaction is always positive (so long
as the reaction proceeds in the forward direction) and either has units of (mol),
for non-flow processes, or (mol/s), for steady-flow processes.
The extent of reaction is the same for all species and is the key variable
needed when balancing all material flows entering and leaving a chemical
reactor. It is also a vital component when balancing all the energy flows
entering and leaving a reactor (see also equilibrium extent of reaction in this
glossary).
Fractional Conversion
Fractional conversion is defined by the following fraction: moles of reactant

species that has reacted divided by moles of the same component in feed; it
may also be expressed as percentage conversion. The conversion of any
reactant may be found, but it is usually the conversion of the limiting reactant
that is of interest.

GLOSSARY 8
Fugacity
The fugacity is an auxiliary property of the chemical potential. The chemical

potential tends to minus infinity as pressure tends to zero. Changes in fugacity
are directly linked to changes in chemical potential but, because of the way it
is defined, the fugacity tends to the system pressure as pressure tends to
zero, because of this the fugacity is often called an effective pressure. The
fugacity of an ideal gas is equal to the ideal gas pressure.
Fugacity Coefficient
The fugacity coefficient is defined as the fugacity of the real fluid divided by its
pressure. But due to what has been discussed under fugacity (this glossary)
the fugacity coefficient is also the fugacity of the real fluid divided by the
fugacity of an ideal gas.
Thus, the fugacity coefficient is a correction factor that corrects the fugacity
from the ideal gas state to the real fluid state at the same temperature and
pressure. As the fluid pressure approaches zero the fugacity coefficient
approaches unity. For an ideal gas the fugacity coefficient is unity. Fugacity
coefficients may be applied to both vapour and liquid phases.
The fugacity coefficient does exactly the same job for the fugacity as the
compressibility factor does for the molar volume.
Gas
A gas cannot be liquefied by the action of pressure alone, whereas a vapour

may be liquefied by the action of pressure. If a gas is at a temperature above
its critical temperature it may not be liquefied by the action of pressure and is
called supercritical.
Gibbs Free Energy
The Gibbs free energy (now recommended by IUPAC to be called Gibbs

energy) is the maximum amount of non-expansion work that may be
recovered from a closed system, undergoing a reversible process at constant
temperature and pressure. The energy is free only in the sense that it is the
energy left over or available (after accounting for expansion/compressive
work).
When a system reaches equilibrium, at constant temperature and pressure,

then the total Gibbs free energy is a minimum. Thus, changes in the total
Gibbs free energy must be equal to zero at equilibrium.

GLOSSARY 9
Gibbs free energy is a chemical potential that, at constant temperature and

pressure, must be minimised at equilibrium. The direction for spontaneous
change, as the system heads towards equilibrium, is always in the direction of
decreasing Gibbs free energy. These facts apply to both phase and chemical
equilibrium.
The derivative of the total Gibbs free energy with respect to the extent of
reaction (at constant temperature and pressure) is zero and this condition
defines the equilibrium extent of reaction.
Gamma-Phi Approach to VLE
The gamma-phi approach to VLE is used for polar mixtures which exhibit
strong differences in forces between like/unlike species in the liquid phase.
With this approach to VLE the ideal gas state is used as the reference state for
the real vapour phase and the ideal solution state is used as the reference
state for the real liquid phase. The fugacity coefficient corrects the fugacity in
the vapour phase, while the activity coefficient corrects the fugacity in the
liquid phase.
The fugacity coefficients are usually found using an EOS (often a cubic EOS),
while the activity coefficients are obtained from an empirical correlation (called
a model of solution behavior).
Gibbs free energy change of reaction
The Gibbs free energy change of reaction is the pure component Gibbs free
energy of a species, in its standard state, times its stoichiometric coefficient
and then summed over all reacting species.
The standard state of the pure species corresponds to a particular state of

aggregation of the pure species at the reaction temperature and a reference
pressure of 1 bar. Thus, the standard Gibbs free energy of reaction depends
on temperature only (and is independent of pressure).
The equilibrium constant is defined directly in terms of the Gibbs free energy
change of reaction and thus also depends on temperature only.
The standard Gibbs free energy of reaction can be calculated in terms of

Gibbs free energy of formation of individual species and these are usually
listed in tables at 298.15 K. Thus, it is always possible to find the equilibrium
constant at a temperature of 298.15 K.

GLOSSARY 10
Gibbs free energy of formation
The standard Gibbs free energy of formation g of is the Gibbs free energy
change associated with forming one mole of a compound from its constituent
elements. The elements should be as they commonly occur in nature, e.g. Cl 2
rather than Cl.
Pure elements are defined as having a Gibbs free energy of formation of zero.
The standard Gibbs free energy of formation is usually listed for pure species
at 298.15 K and at the reference pressure.
Heat
Heat is energy transferred between the system and surroundings. It is an

energy exchange term, not an energy storage term. Heat is energy crossing
the boundary only while the system is actually changing state.
In this sense heat is energy in transit and is only observed when there is a
temperature difference across the boundary. Since heat does not reside in the
system, it is not good practice to use the term heat content. What is stored in
the system is energy, either as internal energy or enthalpy.
Heat is positive when transferred to and negative when transferred from the
system.
Heterogeneous Azeotrope
When a VLE system has large positive deviation from ideality like and unlike
species will have little affinity with each other. If this lack of affinity is large
enough, the liquid phase can split into two co-existing liquid phases.
When, at a fixed pressure, the system temperature is at the three-phase

temperature then a heterogeneous azeotrope can form (depending on the
overall composition). Vaporisation or condensation, under these conditions,
will occur with no change in liquid or vapour composition, and at constant
temperature, until one or other of the phases disappear.
The system is azeotropic because the compositions of vapour and liquid are
fixed and heterogeneous because the liquid splits into two phases.

GLOSSARY 11
Ideal VLE System
An ideal VLE system is one that obeys Raoults Law. The vapour phase
mixture is an ideal gas while the liquid phase mixture, in equilibrium with this
vapour, is an ideal solution. The fugacity coefficients of all species in the
vapour phase are unity, the activity coefficients of all species in the liquid
phase are unity and the fugacity of the pure liquid is equal to its pressure (the
vapour pressure of the species).
Immiscible Systems
When two species have large positive deviation from ideality they may have so
little affinity for each other that they can split into two liquid phases. Depending
on this lack of affinity the two liquid phases may be partially miscible or
immiscible: partial miscibility is where unlike components may dissolve a little
in each phase (up to the point that the phase becomes saturated in the unlike
species); immiscibility is where unlike species have so little affinity that they
cannot dissolve in each phase; each phase is a pure species.
It is probably the case that no system is truly immiscible, but the miscibility
may be so low that (from an engineering point of view) the two liquid phases
may be considered immiscible.
Internal energy
Internal energy is a consequence of the first law it arises because some

quantity sums to zero across a cycle and, any quantity that behaves in this
way, must be a thermodynamic property.
Internal energy is the energy stored in a system by virtue of the kinetic,

potential and vibrational energies of all the particles making up the system.
For a monoatomic ideal gas internal energy is the average kinetic energy of all
the gas particles, since a monoatomic gas can only store energy in the form of
translational kinetic energy.
For any species internal energy depends on the temperature and the pressure
of a pure species. For an ideal gas internal energy is a function of temperature
only this is because only temperature affects the translational kinetic energy
of ideal gas particles.

GLOSSARY 12
Limiting reactant
Limiting means that one reactant is supplied in quantities less than the
theoretical stoichiometric proportion this is often the more valuable reactant.
Le Chateliers Principle
Le Chateliers Principle states that for a chemical reaction at equilibrium,

whenever the temperature, pressure or concentration of the reacting mixture is
changed, the equilibrium will adjust to oppose the change until a new
equilibrium state is eventually established.
Molar
A mole is the amount of any substance that contains the same number of
entities (ions, electrons, atoms or molecules) as 12 g of 12 C. This is the same
as Avogadro's number (6.02210 23 entities).
A molar property is an intensive property which is independent of the amount of

the substance. This is because a molar property is based on 1 mole of a
substance (containing Avogadros number of particles), see also molar mass
(this glossary).
Molar Mass
The molar mass is the mass of a substance divided by the amount of

that substance. In SI units it is expressed as (g/mol). T hus, the molar mass,
in SI units, is the mass in (g) of 1 mole of the substance.
The molar mass of a species is found by adding the atomic masses of all the
atoms in a molecule. The molar mass of a species is often referred to be engineers
(incorrectly) as the molecular weight.
Number of Degrees of Freedom
A degree of freedom analysis is where a set of equations that describe a

system is examined. The number of equations less the number of variables (in
these equations) is the number of degrees of freedom of the system. It is the
number of variables that must be specified in order for the number of unknown
variables to equal the number of equations available.

GLOSSARY 13
When used in the context of the Gibbs Phase Rule the general meaning in the
above paragraph is the same, but now the number of degrees of freedom
applies to phase equilibrium problems (bubble point and dew point problems).
Now the number of equations is the number of independent phase equilibrium
relations, while the number of variable is the number of intensive variables of
the system (temperature, pressure, composition).
If these are counted correctly (not all variables and equations are
independent), then the famous Gibbs Phase Rule formula is obtained.
For flash calculations the number of equations is larger (mass balances are
included) and the number of variables is larger (the amount of each phase is
needed). Thus, if these are counted properly, the number of degrees of
freedom is two. Notice this does not come from the phase rule.
Open system
In an open system matter can cross the boundary. Therefore, both matter and
energy, in the form of heat or work, may cross the boundary.
Processes occurring within an open system are called flow processes; either
steady-flow or unsteady flow. Steady-flow processes arise when the mass
flow entering the system is equal to the mass flow leaving the system. In the
US an open system can be called a Control Volume.
Partially Miscible Systems
When two species have large positive deviation from ideality they may have so
little affinity for each other that they can split into two liquid phases. Depending
on this lack of affinity the two liquid phases may be partially miscible or
immiscible. Partial miscibility is where unlike components may dissolve a little
in each phase (up to the point that the phase becomes saturated in the unlike
species), see also immiscible systems (this glossary).
Phi-Phi Approach to VLE
The phi-phi approach to VLE is used for non-polar mixtures (often hydrocarbon
mixtures). With this approach to VLE the ideal gas state is used as the
reference state for both the real vapour phase and the real liquid phase. The
fugacity coefficient corrects the fugacity in both vapour and liquid phases.
This approach works well for hydrocarbons because like/unlike species are
more similar in the liquid phase that fugacity coefficients (usually found from
an EOS) adequately model liquid phase properties.

GLOSSARY 14
Poynting Correction Factor
The Poynting correction factor is the name often given to the exponential term
that correct the pure species liquid fugacity between vapour pressure of the
pure species and the system temperature. In many calculations this correction
factor has a negligible on the overall accuracy of the result. If this is the case
then it may be ignored.
Process
A system undergoes a process when its state is altered in some way. The
state may be altered in some arbitrary way or it may be altered in some well-
defined way, i.e. isothermal, isobaric, isochoric isenthalpic or adiabatic
processes.
The process may involve a definite change in state (two different end states)
or may trace out a cycle where, after passing through a sequence of states,
the initial and final states are the same.
Quasi-static Process
Work is done by moving an external force through some distance under the
action of an unbalanced force. A quasi-static process is when the system
exchanges work with the surroundings but the unbalanced force becomes
infinitesimally small (which is not to say that it is zero).
Under these circumstances internal and external forces acting over the
(slowly) moving boundary can be equated. Thus, the work done by moving the
external force through a distance becomes equal to a combination of internal
system properties.
Such a quasi-static process is a reversible processes. This just means that if

the system changes state and is then returned to its original state, there will be
no evidence (in either the system or surroundings) that the process ever
occurred. In essence no work done on or by the system will be lost as heat at
any time during a quasi-static (reversible) process.

GLOSSARY 15
Reversible Process
A reversible process is one with only infinitesimal driving forces. This means
for work to be transferred reversibly the unbalanced force, causing the
boundary to move (the external force) through a distance, must be
infinitesimal. For heat transfer to be reversible, the temperature difference
across the boundary must be infinitesimal.
For a reversible process all intermediate states are equilibrium states and in
most textbooks such processes are represented by a solid line. Taking a solid
line to be an infinite series of dots, then any single dot represents one of these
intermediate equilibrium states.
If a reversible process is stopped and returned to its original state, then there
will be no evidence, either in the system or surroundings that the process had
ever occurred. All the heat and work exchanged in one direction will be exactly
equal to the heat and work exchanged when the system is returned to its
original state in the other direction.
Thus, in a reversible process there is no loss in capacity to do work. Nothing

irrevocable has been done during a reversible process, see quasi-static
processes (this glossary).
Saturated liquid
A saturated liquid, in the sense of VLE, is a liquid mixture at its boiling point.
The mixture is still completely liquid but has a tendency to generate vapour.
When moved incrementally (in the correct direction) the liquid phase will
generate this first bubble of vapour. This is why the saturated liquid curve is
also known as the bubble point curve. In this sense the liquid is saturated
because, at the prevailing temperature and pressure, it can no longer dissolve
or contain the vapour phase, thus, two different phases may co-exist at this
point.
At any prevailing pressure, the liquid mixture becomes saturated at its bubble
point temperature (this depends not only on the pressure, but also on the
overall composition). Below the bubble point temperature the liquid is
subcooled or unsaturated.
This sense of a saturated liquid also applies to LLE. The system is miscible
until the liquid is saturated. At this point the single liquid phase can no longer
dissolve the unlike species and (upon further addition of the unlike species)
the single liquid splits into two liquid phases, each of which is saturated in the
unlike species.

GLOSSARY 16
Saturated Vapour
A saturated vapour, in the sense of VLE, is a vapour mixture at the point of

condensation. The mixture is still completely vapour but has a tendency to
generate liquid. When moved incrementally (in the correct direction) the
vapour phase will generate this first drop of liquid. This is why the saturated
vapour curve is also known as the dew point curve. In this sense the vapour is
saturated because, at the prevailing temperature and pressure, it can no
longer contain the vapour phase, thus, two different phases may co-exist at
this point.
At any prevailing pressure, the liquid mixture becomes saturated at its dew
point temperature (this depends not only on the pressure, but also on the
overall composition). Above the dew point temperature the vapour is
superheated (thus also unsaturated).
Spontaneous process
Any process that is spontaneous, or occurs spontaneously, is a process that

will occur naturally by itself without any outside assistance. Thus heat flows
from a high to a low temperature occurs spontaneously, while a wheel that is
sent spinning will spontaneously slow down and eventually stop.
Spontaneous processes are by their nature irreversible processes, and with

any real process, there must be an increase in entropy (both system and
surrounds considered together)
Chemical reactions proceed spontaneously in the direction of decreasing

Gibbs free energy and ultimate reach equilibrium when this thermodynamic
potential is minimised.
Standard conditions
Standard conditions, as applied to a chemical reaction, are when the reactants

and products are in specified states of aggregation at the reaction
temperature, but a reference pressure of 1 bar. Thus, standard state
properties only depend on temperature.

GLOSSARY 17
Steady-flow
A steady-flow process is associated with open systems where mass can cross
the boundary. If an open system changes its state, then it properties must
change, such a process is called a flow process.
A flow process is called a steady-flow process if the flow is steady and this
means that at any point within the system the thermodynamic properties do
not change with time although, in all likelihood, they will change with position.
Steady-flow conditions prevail when the total mass flow entering the device is
equal to the total mass flow leaving the device. Obviously, when starting up or
shutting down an item of equipment, the operation will be transient and, during
such time, flow will be unsteady.
Stoichiometric Coefficients
Stoichiometric coefficients are the numbers that appear beside each reactant
or product in a properly balanced chemical equation. Stoichiometric
coefficients have no units and simply relate the number of reactant molecules
that are consumed to the number of product molecules that are produced.
Stoichiometric coefficients are positive for products, negative for reactants and
zero for inert species. This allows, using algebraic equations, products to be
produced, reactants to be consumed and inerts to pass through unchanged.
In a balanced chemical equation the stoichiometric coefficients appear as

absolute quantities (without any sign). This is because it is known that
reactants are on the LHS, while products are on RHS. Thus, in chemical
reactions signs are omitted.
The overall stoichiometric coefficient is the stoichiometric coefficients of each

species summed over the reaction mixture. This is an important quantity since
it determines whether the number of moles in the reaction mixture will
increase, decrease or stay the same as the reaction proceeds.
In general, as a reaction a progress, while mass is always conserved, moles

are generally not conserved; moles will only be conserved if the overall
stoichiometric coefficient is zero. This means that the number of moles of
reactants being consumed exactly equals the number of moles of products
being produced.

GLOSSARY 18
Surrounding
The surrounding is the immediate environment around the system which is

affected by changes occurring within the system.
Thermodynamic system
A thermodynamic system, or just the system, is a quantity of matter which

has been separated from its environment and its behaviour is being
investigated.
By investigated it is meant that changes in the thermodynamic properties are

being linked to the quantities heat and work.
The boundary envelops only the system itself and should exclude any
mechanical components that support or contain the system. Thus, for most
chemical engineering applications, the system is restricted to the fluid itself
(and not the equipment it is passing through).
Thermodynamics properties
If a system is repeatedly returned to the same state, by any route, on attaining

this state all the thermodynamic properties must return to their original values.
Therefore, thermodynamic properties do not depend on the path taken to

reach a particular end state. It is said that the properties are not path
functions. Since thermodynamic properties depend only on the current state
they are called state functions or more commonly state variables.
Three-Phase Pressure
When, at a fixed temperature, the system pressure is at the three-phase

pressure then a heterogeneous azeotrope can form (depending on the overall
composition). Vaporisation or condensation, under these conditions, will occur
with no change in liquid or vapour composition, and at constant pressure, until
one or other of the phases disappear.

GLOSSARY 19
Above the three-phase pressure is the LLE region, below the three-phase
pressure is the VLE region. VLLE can only occur at the three-phase pressure
and when the overall system composition lies within a band corresponding to
the saturated liquid compositions of each phase.
Three-Phase Temperature
When, at a fixed pressure, the system temperature is at the three-phase

temperature then a heterogeneous azeotrope can form (depending on the
overall composition). Vaporisation or condensation, under these conditions,
will occur with no change in liquid or vapour composition, and at constant
temperature, until one or other of the phases disappear.
Above the three-phase temperature is the VLE region, below the three-phase
temperature is the LLE region. VLLE can only occur at the three-phase
temperature and when the overall system composition lies within a band
corresponding to the saturated liquid compositions of each phase.
Work
Work is force times distance moved. Hence, work can only be transferred
when the system boundary is actually moving and, thus, while the system is
changing state. Whenever the boundary stops moving then no work can be
done
Therefore, work is energy in transit and is only observed at a moving

boundary. Work cannot be exchanged if the boundary is fixed or rigid.
Movement of the boundary only occurs when there is an unbalanced force.
Thus, work is only done by moving the boundary against an external force
while under the action of an unbalanced force. Work is an energy exchange
term. Work does not reside within the system and is only manifest at a moving
boundary as the system is changing state. When the state is fixed the system
is in equilibrium and work transfer stops.
Work is positive when transferred into the system and negative when
transferred out of the system.

GLOSSARY 20

B49CE Student Guide v1 PDF

Uploaded by

Copyright:

Available Formats

B49CE Student Guide v1 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

B49CE Student Guide v1 PDF

Uploaded by

Copyright:

Available Formats

CONTENTS 1

Edinburgh EH14 4AS, United Kingdom

Produced by Heriot-Watt University, 2015

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

Copyright 2015 Heriot-Watt University

All rights reserved. No part of this publication may be reproduced, stored in

This material is prepared to support the degree programmes in Chemical

Distributed by Heriot-Watt University

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

TOPIC 1: PURE SPECIES PHASE EQUILIBRIUM

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

TOPIC 2: VLE FOR NON-POLAR MIXTURES

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

TOPIC 3A: VLE FOR POLAR LIQUID MIXTURES

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

TOPIC 3B: VAPOUR-LIQUID EQUILIBRIUM, LIQUID-LIQUID EQUILIBRIUM

TOPIC 4: CHEMICAL EQUILIBRIUM

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

4.4.1 Rigorous Non-Ideal Mixture ............................................................................... 24

TOPIC 5: VOLUMETRIC BEHAVIOUR OF REAL FLUIDS AND SOLIDS

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

5.6.4 RKS EOS in Cubic Form ................................................................................... 40

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

Pure Species Phase Equilibrium

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

In previous courses (Process Engineering B) a general chemical reaction was

At chemical equilibrium the concentration of each species [ A], [B ], [C ], [D ]

In science equilibrium means that there are no changes. However, in

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

1.2 Criterion for Equilibrium

For mechanical equilibrium there can be no pressure gradient (this is the

The potential that drives interphase mass transfer is not immediately

1.2.1 Pure Species Phase Equilibrium

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

Which may be written as follows:

Therefore, the total entropy change of the surrounding dS SURR becomes

The second law is given by (Process Engineering B)

If T & P remain constant this may be written

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

The Gibb's free energy (Process Engineering B) is defined as

Substitute this last expression into equation (1.1) leads to

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

After expanding out the differential this becomes

From a previous course (Process Engineering B) the fundamental property

dg = vdP sdT ..(1.5)

For a pure species which initially is not in equilibrium but is distributing

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

1.2.2 Pure Species Chemical Potential

In a previous course (Process Engineering B) it was shown that, for a pure

HERIOT-W ATT UNIVERSITY B49CE August 2015 v1

If the pressure is specified, then this condition is satisfied at some fixed

Comparing Gibbs free energy and entropy:

Situation Gibbs Free Energy Entropy

1.2.3 Pure Species Gibbs Energy Relation

The total enthalpy is defined as

Substitute dH into the previous equation gives