B49CE

Multiphase
Thermodynamics
Topics in VLE, LLE and Miscibility

Topic 2

VLE for Non-Polar Mixtures

© Heriot-Watt University, B49CE,

December 2017 v3 1
Multiphase Thermodynamics
 Learning Objectives
After studying this “VLE for Non-Polar Mixtures” topic you
should be able to:
• Write down the 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 Dalton’s Law.

© Heriot-Watt University, B49CE, Multiphase

December 2017 v3 2
Thermodynamics
• 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.

© Heriot-Watt University, B49CE,

December 2017 v3 3
Multiphase Thermodynamics
• Derive expressions for Raoult’s Law and Raoult’s Law K-
values.
• Define and explain the Antoine vapour pressure
correlation.
• Explain and execute bubble/dew point and flash
calculations.
• Outline the rigorous phi-phi approach to VLE and how
problems may be solved using an EOS and a maths
package.

© Heriot-Watt University, B49CE, Multiphase

December 2017 v3 4
Thermodynamics
 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 extend phase equilibrium
expressions from pure species to mixtures.
• The property that a component seems to have in a mixture
is called a partial molar property.
• The partial molar property is different from the pure
component property because in a mixture any one
molecule is surrounded by both like and unlike species.
• With pure components any one molecule is always
surrounded by like species.
>>> A detailed review of partial molar properties is given in
Appendix C……..only basic relations appear here.
© Heriot-Watt University, B49CE,
December 2017 v3 5
Multiphase Thermodynamics
 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“ in the mixture the apportioning rule,
which is the definition of a partial molar property, is given by

 nm 
mi   
 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.
© Heriot-Watt University, B49CE,
December 2017 v3 6
Multiphase Thermodynamics
 The partial molar property mi of a species in a mixture is
defined as the partial derivative of the total solution property
M  nm with respect to the mole number of that species:
• T & P must be held constant.
• Constant n j means that the mole number of all other
species except the one that is being varied must be held
constant.
• The partial molar property mi is generic, it could be
vi ui hi si g i .
• Other properties such as chemical potential, fugacity and
fugacity coefficient may also be apportioned between the
components in the solution.
>>> If M  nm can be “apportioned” to find mi for each
species, then it should be possible to mathematically
“recombine” all mi to recover the overall property M  nm .
© Heriot-Watt University, B49CE,
December 2017 v3 7
Multiphase Thermodynamics
 The recombination rule for partial molar properties is given by
N
m   xi mi
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 the
expression above, 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 for derivation) it is given by

 m   m 
 P  dP   T  dT   xi dmi  0
 T ,x   P, x i

© Heriot-Watt University, B49CE,

December 2017 v3 8
Multiphase Thermodynamics
 However, if T & P are held constant, then this expression
reduces to the most widely used form of the Gibbs-Duhem
equation

 x dm
i
i i 0

• 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.

© Heriot-Watt University, B49CE,

December 2017 v3 9
Multiphase Thermodynamics
 The three important rules that apply to partial molar
properties may be summarised as follows:
T &P
Solution
Property - m
2 Recombination 1 Apportioning
Rule Rule

m   xi mi    nm  
mi   
  ni  T , P ,n j
i
T &P
Partial Molar
Property - mi

3 Gibbs-Duhem
 x dm i i 0
Equation i

• These rules apply to ANY partial molar property mi .

• For instance, the partial molar property in question could
be any one of those listed below:
vi , ui , hi , si gi
© Heriot-Watt University, B49CE,
December 2017 v3 10
Multiphase Thermodynamics
 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  Tsi & hi  ui  Pvi
• There exits an identical relation between their
corresponding partial molar properties. Thus, it is possible
to immediately write
g i  hi  Tsi & hi  ui  Pvi

Pure Component Property Partial Molar Property

Relation Relation

dg i  vi dP  si dT dg i  vi dP  si dT >>> Other important

relations that may be
d i  RTd ln f i dˆ i  RTd ln fˆi found by inspection.

© Heriot-Watt University, B49CE,

December 2017 v3 11
Multiphase Thermodynamics
 2.2 Vapour-Liquid Equilibrium
The Gibbs free energy is of fundamental importance in both
phase and chemical equilibrium and it needs to be examined
more closely.
• Any extensive property M may be expressed in terms of
T & P and the mole number of each species n1 , n2 ......ni
as follows:
M  f (T , P, n1 , n2 ......ni )

2.2.1 Gibbs Free Energy of a Mixture

• This applies especially to G where these are its “canonical”
variables; the special variables on which G fundamentally
depends.
G  f (T , P, n1 , n2 ......ni )
© Heriot-Watt University, B49CE,
December 2017 v3 12
Multiphase Thermodynamics
 Therefore, 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 the total Gibbs free energy is given by G  ng

 ng   ng   ng 

d ng     dP    dT     dni
 P  T ,n  T  P ,n i  ni  T , P , n
j

• The first two partial derivatives have been shown (Process

Engineering B) to be nv and ns .
 ng 
d ng   nv dP  ns dT     dni
i  ni  T , P , n
j

 nm  Compare with definition   ng  

mi    of a partial molar property
   gi
 n i  T , P ,n j  ni  T , P ,n j
© Heriot-Watt University, B49CE,
December 2017 v3 13
Multiphase Thermodynamics
 Making this substitution leads to

d ng   nv dP  ns dT   g i dni

i
• Divide through by n , the total number of moles of all
species in the mixture, gives
dg  vdP  sdT   g i dxi
i
• Thus, the molar Gibbs free energy is fundamentally a
function of T , P & xi ; these are the intensive variables of
the system.
• Hence, Chemical Engineers can alter phase equilibrium by
changing any combination of these properties.
• In the previous topic it was shown that for a pure species
g i  i >> Thus for a component gi  ̂i
In a mixture this becomes >>

© Heriot-Watt University, B49CE,

December 2017 v3 14
Multiphase Thermodynamics
 Thus, the chemical potential of a component in a mixture ̂i
must also be defined as follows:
  ng  
ˆ i   
  ni  T , P ,n j

• The total change in molar Gibbs free energy may also be

written in the following way:
dg  vdP  sdT   ̂ i dx
i
• The meaning of the circumflex over the chemical potential
will become apparent later.
• At constant T & P it will be shown that this expression leads
to a general criterion for Vapour Liquid Equilibrium (VLE).
>>>> However, it is better to review some fugacity and fugacity
coefficient relationships first.
© Heriot-Watt University, B49CE,
December 2017 v3 15
Multiphase Thermodynamics
2.2.2 Dalton’s Law
From earlier modules (Process Industries) the partial pressure
of a species in an ideal gas mixture pi 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

© Heriot-Watt University, B49CE,

December 2017 v3 16
Multiphase Thermodynamics
 Dividing the former expression by the latter leads to Dalton’s
Law
p i ni
  yi
P n
 pi  Py i
• This is known as Dalton’s Law. In many texts the partial
pressure is denoted by pi and the total pressure by P .
• However, the difference between uppercase P and
lowercase p can lead to confusion.
>>> Therefore, in this text the partial pressure will be referred
to as yi P and uppercase P will be reserved to denote the
system (or total) pressure.

© Heriot-Watt University, B49CE,

December 2017 v3 17
Multiphase Thermodynamics
2.2.3 Fugacity Coefficient of a Component in a
Mixture
In Topic 1 the pure component fugacity coefficient  i was
defined as follows:
fi
i 
P
• As P  0 , then  i  1 and f i  P. Thus, for an ideal gas it
is clear

 ig
i 1 &  f i ig  P
• Substitute this into the definition of the fugacity coefficient

fi The fugacity coefficient is a correction

 i  ig factor that corrects the fugacity between
fi the ideal gas and the real fluid states.
© Heriot-Watt University, B49CE,
December 2017 v3 18
Multiphase Thermodynamics
 The fugacity coefficient of a component in a mixture ̂ i is the
partial molar analogue of  i and is defined as follows:
ˆ
f
ˆ  i
 i
yi P
• In a completely analogous way as P  0 , then  ˆ  1 and
i
fˆi  yi P . Thus, for an ideal gas it is clear

ˆ
i  1
ig
& ˆ
 f i  yi P
ig

• Substitute this into the definition of the fugacity coefficient

of a component in a mixture leads to the analogous result
ˆ
f This shows that ̂ i is a correction factor that corrects
ˆ  i
 the fugacity of a component in a mixture from the
i
ˆf ig ideal gas state to the real fluid state at the same T , P & yi
i

>>> Notice the ideal gas state is being used as a reference state.
© Heriot-Watt University, B49CE,
December 2017 v3 19
Multiphase Thermodynamics
 Notice that as P  0 :

f i  P ….......(Pure component)
fˆi  yi P….......(component in a mixture)
• As P  0 the fugacity of a pure species approaches the
system pressure, while the fugacity of a component in a
mixture approaches its partial pressure
2.2.4 Ideal Solution 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.

© Heriot-Watt University, B49CE,

December 2017 v3 20
Multiphase Thermodynamics
 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

v i
id
 vi  vi
• The superscript “id” is used to differentiate an ideal
solution from an ideal gas (superscript “ig”). 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 same size and shape.
2. Like and unlike molecules have same forces of attraction.

© Heriot-Watt University, B49CE,

December 2017 v3 21
Multiphase Thermodynamics
 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.

© Heriot-Watt University, B49CE,

December 2017 v3 22
Multiphase Thermodynamics
 The question is: for an ideal solution, how does the fugacity of
a component in a mixture fˆi id 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
ˆ 1
ln  i  i  viig dP
RT 0

>>> This applies to a REAL solution where differences

between like and unlike species are significant.

© Heriot-Watt University, B49CE,

December 2017 v3 23
Multiphase Thermodynamics
 But for an IDEAL solution viid  vi . Making this substitution
gives the fugacity coefficient of an ideal solution ̂ idi :

 v 
P
ˆ 1
ln  id
i  i  viig dP
RT 0

• An ideal gas is more restrictive than an ideal solution.

Hence, an ideal gas mixture is automatically an ideal
solution and viig  viig

 v  v dP
P

 v  v dP
P
ˆ id  1 1
 ln  i i
ig
i Compare ln  i  i
ig
i
RT 0 RT 0

• Thus for an ideal solution

̂ idi   i
>>> The fugacity coefficient of a component in an ideal
solution is equal to its pure component value.
© Heriot-Watt University, B49CE,
December 2017 v3 24
Multiphase Thermodynamics
 But, the fugacity coefficients are defined as follows:

ˆ id
f
ˆ  i
 id
i
yi P Substitute both….
fi
i 
P
In here……
̂   i
id
i

The LEWIS-RANDALL rule applies to real

To fˆi id fi solutions but only as solution approaches
get
 purity
yi P P
• After cancelling and re-arrangement this leads to:

fˆi id  yi f i
The LEWIS-RANDALL rule which gives the
fugacity of a species in an ideal solution.
© Heriot-Watt University, B49CE,
December 2017 v3 25
Multiphase Thermodynamics
 2.2.5 Criterion for Multicomponent Phase
Equilibrium
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.
• 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.
• 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 for each
component in either phase.
>>> The derivation for multicomponent phase equilibrium
follows the same steps up to the point where it was shown
that, for phase equilibrium, dG  0 .
© Heriot-Watt University, B49CE,
December 2017 v3 26
Multiphase Thermodynamics
 The multicomponent system under consideration for mixtures
consists of  phases and N components (these are 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 the change in total
Gibb’s energy G within any phase, denoted as the  -phase,
is given by the expression below:

d ng   nv  dP  ns  dT   ̂i dni

   At constant T & P
i

     
d ng  nv dP  ns dT   ̂i dni
   Likewise for any other
phase, denoted the 
i
© Heriot-Watt University, B49CE,
Phase.
December 2017 v3 27
Multiphase Thermodynamics
 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 Gibb’s 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   ˆ i dni   ˆ i dni  
i i

>>> However, for the moment there are only two phases
present within this closed system.
© Heriot-Watt University, B49CE,
December 2017 v3 28
Multiphase Thermodynamics
 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
• And if both phases are in equilibrium it is known that dG  0 .
Thus, for equilibrium between these two phases

 ˆi

i 
 ˆ i dni  0

>>> However, dni is arbitrary. It follows that the only way this
summation can always be zero is when the chemical potential
(of each species) in both phases is the same at equilibrium.

© Heriot-Watt University, B49CE,

December 2017 v3 29
Multiphase Thermodynamics
 Thus, for phase equilibrium between any pair of phases at
constant T & P, the general criterion for equilibrium may be
written

ˆ i  ˆ i ..........(i  1,2......N )

• This reasoning holds for any other pair of phases (taking

care to keep one phase of each pair the same).
• This leads to the following general criterion for phase
equilibrium between all  phases (this criterion must hold
for each of the N components in the mixture):
ˆ i  ˆ i  ˆ i  ........ˆ i ..........(i  1,2......N )

>>> When this condition is satisfied net mass transfer of all

species, between all  phases, stops and the phases are all in
equilibrium.
© Heriot-Watt University, B49CE,
December 2017 v3 30
Multiphase Thermodynamics
 2.2.6 Fugacity as a Criterion for 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
ˆ i  RT ln fˆi  
• In which  is the temperature dependent constant of
integration.
After cancelling this……
Substitute
ˆ i  ˆ i  ˆ i  ........ˆ i

fˆi  fˆi   ............  fˆi ..........(i  1,2......N ) Leads to………

>>> This is an equally general criteria for phase equilibrium,

but now in terms of fˆi which is better behaved than ̂i .
© Heriot-Watt University, B49CE,
December 2017 v3 31
Multiphase Thermodynamics
 2.2.7 Finding the Fugacity of a component in a
Mixture
In Topic 1 it was shown that the pure component fugacity
coefficient could be found from

 v  v dP
P P
1 1
ln    ln  
R ig
v dP <<->>
RT 0 RT 0

• Which could also be expressed (often more conveniently) in

terms of the dimensionless Z-factor
P
ln    Z  1
dP
0
P
• The partial molar analogues of these two expressions are

 v dP
P P
ˆ 1 ˆ  Z  1 dP
ln  i  v ln  0 i P
ig
i i & i
RT 0

© Heriot-Watt University, B49CE,

December 2017 v3 32
Multiphase Thermodynamics
 Solving multicomponent phase equilibrium problems is in
principle (although not in practice) quite straightforward:
1. Solve equation below (or similar) to find the fugacity
coefficient of a component in the mixture ̂ i .
P
ln  i   Z i  1
ˆ dP
0
P
2. Use the definition of ̂ i to find the fugacity of all the
components in the mixture fˆi - see below:
ˆ
f
ˆ  i
 i
yi P
3. Adjust the intensive properties T , P, xi or yi (depending on
the problem) to satisfy the phase equilibrium condition:

fˆi  fˆi   ............  fˆi ..........(i  1,2......N )

© Heriot-Watt University, B49CE,
December 2017 v3 33
Multiphase Thermodynamics
 In principle this is easy. However, ̂ i are complex functions of
T , P, xi or yi and this complicates the problem. There are two
general approaches:
1. Make simplifying assumptions that lead either to simpler
algebraic equations or to some generalised chart.
2. Retain the rigorous approach and use a Mathematics
package to help solve all the equations simultaneously.
• The use of the circumflex above the fugacity coefficient, as
in ̂ i , is now apparent. The partial molar analogue is

 v 
P
ˆi  1
ln  i  viig dP
RT 0

• In other words the partial molar property includes the

logarithm. Thus, m  ln  and mi  ln  ˆ .
i

>>> The circumflex denotes that ̂ i is a property of a

component in a mixture, but not a partial molar property.
© Heriot-Watt University, B49CE,
December 2017 v3 34
Multiphase Thermodynamics
 The same argument applies to fˆi . Again, this is a property of
a component in a mixture, but not (by itself) a partial molar
property.
• In fact the partial molar property in this case is ln( fˆi / 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 in either intensive or
extensive form, the same is not true of ̂i which, being
a potential, may only be intensive.
• Remember that the other potentials, temperature T
and pressure P , are intensive.

>>> In section 2.3 (following next) we see how to solve these

phase equilibrium relationships for non-polar systems.

© Heriot-Watt University, B49CE,

December 2017 v3 35
Multiphase Thermodynamics
 2.3 VLE for Non-Polar Mixtures
VLE problems in the hydrocarbon industry have typically been
solved using so called “K  value” charts.
• The original charts have largely been superseded by EOS
and computer simulation. However, K  value charts are
still helpful in terms of visualising problems.
• The K  value of any species in the mixture is defined as
follows:
yi
Ki 
xi
2.3.1 Rigorous Phi-Phi Approach to VLE
The most useful criterion for VLE is in terms of fugacity……

fˆiV  fˆi L ..........(i  1,2......N )

© Heriot-Watt University, B49CE,
December 2017 v3 36
Multiphase Thermodynamics
 Where the vapour and liquid phases are indicated by
superscripts applied to each fugacity. The vapour phase
fugacity coefficient ̂Vi is defined as
ˆV
f
ˆV  i
 >>> fˆiV  
ˆVyP
fˆ ig
i i i
i

• Likewise ̂ iL for the liquid phase is Note the difference in the

mole fraction notation.
ˆL
f
ˆ  i
 L
>>> fˆi L  
ˆ Lx P
fˆ ig
i i i
i

• When two phases are in Vapour-Liquid Equilibrium fˆiV  fˆi L

ˆVy 
 ˆ Lx ..........(i  1,2......N )
i i i i

>>> Notice that the ideal-gas state is being used as reference

state for BOTH vapour & liquid phases.
© Heriot-Watt University, B49CE,
December 2017 v3 37
Multiphase Thermodynamics
 Or, in terms of K  value notation:
ˆL Substitute
K i  Vi
ˆ
 both in here
i

• VLE problems for non-polar liquids are usually solved using

one or other of the last two equations. However, the
previous expression may be recast as follows:

 v 
P
ˆ 1
ln  i 
L
i
L
 viig dP
RT 0

 v 
P
ˆ 1
ln  i 
V
i
V
 viig dP
RT 0

ˆL
 v 

P

 v 
P
To 1 1
ln Vi  L
 viV dP >> ln K i  L
 viV dP
get ˆi
 RT 0
i
RT 0
i

© Heriot-Watt University, B49CE,

December 2017 v3 38
Multiphase Thermodynamics
 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).
• The most direct approach to solving VLE problems is to use
dimensionless fugacity coefficients. In the example above
these are found, for both phases, using the PR EOS.
• The “phi-phi” approach to VLE is used for non-polar
systems, such as hydrocarbons, where liquid phase non-
idealities can be modelled by an EOS.
• The phi-phi approach to VLE is not really suitable for polar
liquid systems – particularly aqueous systems.
>>> Due to the complex functional dependency of both
ˆ V   (T , P, y ) and 
 ˆ
i   (T , P, xi )
i i
L
several special cases will
be examined.
© Heriot-Watt University, B49CE,
December 2017 v3 39
Multiphase Thermodynamics
 2.3.2 Both Liquid & Vapour are Ideal Solutions
If both the liquid phase and the vapour phase are ideal
solutions, then ̂ idi   i and the rigorous VLE relations may be
written as follows:
Vi yi   iL xi ..........i  1,2,3.....N 
These are preferred
• Or, in terms of K  values since cubic EOS are
set up to solve for
 iL i
Ki  V
i
• Notice that the  i values do not depend on composition.
Alternatively they may be written in terms of fugacity:
f iV yi  f i L xi ..........i  1,2,3.....N 

fiL
Ki  V
fi
© Heriot-Watt University, B49CE,
December 2017 v3 40
Multiphase Thermodynamics
 2.3.3 Raoult’s Law: Liquid is an Ideal Solution,
Vapour is an Ideal Gas
In general whenever two phases are in VLE fˆiV  fˆi L :
ˆVy 
 ˆ L x ..........(i  1,2......N ) Substitute
i i i i
both in here
• If vapour is an ideal gas:
ˆ V 1
 i

• If liquid is an ideal solution:

L
f
̂ iL   iL  i
P
To yi P  xi f i L ..........(i  1,2......N )
get
• Or, in terms of K  values
fi
Ki 
P
© Heriot-Watt University, B49CE,
December 2017 v3 41
Multiphase Thermodynamics
 At very low pressure (ideal gas) the fugacity tends to the
system pressure. In this case the pressure exerted by a liquid is
its vapour pressure f i L  Pi sat .
 yi P  xi Pi sat ..........(i  1,2......N )
• Or, in terms of K  values
Pi sat
Ki 
P
• This is Raoult’s Law and it applies to the case of an ideal gas
in equilibrium with an ideal liquid solution.
• Raoult’s Law should be used with caution, since many
systems depart from Raoult’s Law and it cannot model
azeotropic behaviour.
>> However, as shown later in this topic, Raoult’s Law is very
efficient for VLE involving non-polar mixtures.

© Heriot-Watt University, B49CE,

December 2017 v3 42
Multiphase Thermodynamics
 2.3.4 VLE Summary
1. Completely rigorous VLE:
ˆVy 
 ˆ Lx ..........(i  1,2......N )
i i i i

And,
ˆL
K i  Vi
ˆ
 i

2. Both liquid and vapour phases as ideal solutions:

Vi yi   iL xi ..........(i  1,2......N )

And,
 iL
Ki  V
i
© Heriot-Watt University, B49CE,
December 2017 v3 43
Multiphase Thermodynamics
 3. Vapour phase, as an ideal gas, in equilibrium with a liquid
phase, as an ideal solution (Raoult’s Law):

yi P  xi Pi sat ..........(i  1,2......N )

And,
Pi sat
Ki 
P
• There is also a 4th case. This is the where the vapour phase
is an ideal gas, but the liquid phase behaves as a non-ideal
solution.
• As explained before, this is often due to differences in
strong short-range attractive forces between like and unlike
species in the liquid phase.
>>> Polar liquid mixtures rarely behave as ideal solutions. The
next topic deals with such systems.
© Heriot-Watt University, B49CE,
December 2017 v3 44
Multiphase Thermodynamics
 2.3.5 K-Value Charts
The fugacity coefficients needed to solve non-polar VLE
problems are usually obtained from an EOS.
• In the past, the Benedict-Webb-Rubin “BWR” EOS proved
very accurate for hydrocarbon mixtures – the equation is
given below:
RT  C0  1   1  a 1  / v2   
P   B0 RT  A0  2  2   bRT  a  3   6  c 3 2 exp   2 
v  T  v  v  v vT  v 

• Using this EOS K  values were calculated for 12 light

hydrocarbons at pressures up to 250 bar.
• The results were presented as 324 charts known as “Kellogg
charts” with K  values plotted a function of T , P and molar
average boiling point.
>>> However, the charts are laborious to use and simpler
methods have been developed.
© Heriot-Watt University, B49CE,
December 2017 v3 45
Multiphase Thermodynamics
 De Priester prepared a set of nomographs and these are often
used in place of the more complicated Kellogg charts.
• In the de Priester nomographs, the composition
dependence of the K  values has been completely
removed using an averaging process.
• Copies of these charts are given in Appendix A.
• There are two charts – see next two slides:

© Heriot-Watt University, B49CE,

December 2017 v3 46
Multiphase Thermodynamics
 1. High Temperature Chart (10oC to 200oC):
To use the chart:
• Specify temperature.
• Specify pressure.
 • Draw a straight line
between them.
• Read of the K 
values for each
species.


© Heriot-Watt University, B49CE,

December 2017 v3 47
Multiphase Thermodynamics
 2. Low Temperature Chart (-70oC to 20oC):
• Notice that both
these K  value
charts are functions
of T & P only.
• The original
Kellogg's charts are
more accurate, but
more laborious.
• Cubic EOS are
better than charts,
but often too
difficult to solve by
hand.

© Heriot-Watt University, B49CE,

December 2017 v3 48
Multiphase Thermodynamics
 2.3.6 Example 2.1
Determine the equilibrium constant 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:

Tr  1.0 Pr  0.35 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.
• Hence, the liquid phase fugacity coefficient  iL must be
found by extrapolation from the real liquid to this fictitious
liquid state.
© Heriot-Watt University, B49CE,
December 2017 v3 49
Multiphase Thermodynamics
 © Heriot-Watt University, B49CE,
December 2017 v3 50
Multiphase Thermodynamics
 Answer…….

© Heriot-Watt University, B49CE,

December 2017 v3 51
Multiphase Thermodynamics
 2.3.7 Vapour Pressure Correlation
If a VLE system is “ideal”, then it will conform to Raoult’s Law.
A P  x  y diagram, for such a binary system, is shown below:
Constant T
• The system pressure P is
Subcooled X P1
sat plotted on the y- axis.
Liquid P  x1
• Species “1” is the MVC, while “2”
P  y1
P2sat X
L-V is the LVC.
P
Superheated
• The P  x1 line (bubble point
Vapour curve) is the saturated liquid line.
0
0 1.0
• The P  y1 line (dew point curve)
x1 & y1 is the saturated vapour line.

• If the temperature is below the critical temperature of both

components, then the system pressure will vary from P1sat
to P2sat.
>>> In addition, for an ideal system, the P  x1 line is straight.
© Heriot-Watt University, B49CE,
December 2017 v3 52
Multiphase Thermodynamics
 To understand why an ideal VLE system behaves in this way
write down Raoult’s Law twice; first for species “1” and then
for species “2”:
y1 P  x1 P1sat Given y1  y2  1 and x1  x2  1 ,
add the two partial pressures
y 2 P  x2 P2sat together………………..

• The terms on the left are the partial pressure of component

“1” and component “2” respectively. Adding the two
equations gives the total pressure.
To get the total pressure……
P  y1 P  y 2 P  x1 P1sat  x2 P2sat

 P  x1 P1sat  1 x1 P2sat

This is the equation for the P  x1
P  P 2
sat

 x1 P 1
sat
P 2
sat
 line (saturated liquid line).
>>> It clearly has the form y  mx  c . Thus, for an ideal
system, the P  x1 line MUST be straight.
© Heriot-Watt University, B49CE,
December 2017 v3 53
Multiphase Thermodynamics
 For non-ideal systems the P  x1 line is not straight, but the
system pressure P still depends on both P1sat & P2sat (assuming
there are no supercritical components present).
• Thus, some sort of correlation is needed to predict the
vapour pressure Pi sat. One such correlation is called the
Antoine equation.

B >> Sometimes the B

ln P sat
 A log P sat  A 
T C correlation is given by T C

• In which A, B & C are species specific constants – these are

available in standard reference books (see Appendix A.3).
>> Clearly the numerical value of the constants A, B & C
depend on which correlation is used as well as the units used.
>>> Notice that Pi sat depends ONLY on temperature, which
must lie between the triple point and critical point.
© Heriot-Watt University, B49CE,
December 2017 v3 54
Multiphase Thermodynamics
 2.3.8 Temperature-Composition Diagram
A P  x  y diagram was shown previously, in this case the
temperature was held constant.
• A T  x  y diagram, for a similar ideal system, is shown
below (in this case the pressure is fixed):
Constant P

T2sat Dew Point

Superheated
Vapour Region
• Again “1” is MVC and “2” is LVC.
Curve

• The T  x1 curve is saturated

T  x1 T  y1 liquid curve (bubble point curve).
L-V Two-phase
T Region
• The T  y1 curve is saturated
Subcooled
Liquid Region Bubble Point T1sat
vapour curve (dew point curve).
Curve
0
0 1.0
x1 & y1

>>> Notice that, even though system is ideal, neither curve is a

straight line. This is because the vapour pressure of each
component vary with temperature.
© Heriot-Watt University, B49CE,
December 2017 v3 55
Multiphase Thermodynamics
 The saturated liquid curve, and the subcooled liquid region,
now lie below the two-phase region.
• The saturated vapour curve, and the superheated vapour
region, now lie 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 is used in reverse.
• Set the vapour pressure (in the Antoine equation) equal to
the fixed pressure, i.e. set Pi sat  P , then solve the Antoine
equation for temperature – this is now Ti sat (the BP).
>>> Distillation columns operate at a more or less constant
pressure. Thus, for a binary system, this type of process is best
represented by a T  x  y diagram.

© Heriot-Watt University, B49CE,

December 2017 v3 56
Multiphase Thermodynamics
 2.3.9 Significance of K-Values
Raoult’s Law for an N component VLE system is given by
yi P  xi Pi sat..........(i  1,2......N )
• The Raoult’s Law K  values are then
Pi sat
Ki 
P
• There are three factors that increase the K  value:
1. Pressure: as pressure decreases more of the
component is released from the liquid phase and K  .
2. Temperature: as temperature increases more of the
component is driven from the liquid phase and K .
3. MW: For similar compounds, as MW of species decreases
it will have higher Pi sat at the same T ; thus, more of the
component will be driven from the liquid phase at the
same T & P and K  .
© Heriot-Watt University, B49CE,
December 2017 v3 57
Multiphase Thermodynamics
This is better to visualise all this on a de Priester chart:
Notice K  values increase as:
1. Pressure decreases.
2. Temperature increases.
3. At the same T & P as the
MW of the species
decreases.
The K  value is the volatility
of a component in a mixture.
• The more volatile the more
the species accumulates in
the vapour phase and K  .
• Distillation works by
separating components of
different volatility.
© Heriot-Watt University, B49CE,
December 2017 v3 58
Multiphase Thermodynamics
 2.4 Phase Equilibrium Calculations
For bubble point or dew point calculations usually T or P will
be fixed and usually xi or yi will be fixed; the other unknowns
may then be found from the phase equilibrium relationships.
• Thus, there are usually four possible combinations
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 then to find T , followed by the composition
of the first “bubble” of vapour yi in equilibrium with this
liquid. Thus, there are 2 equations & 2 unknowns.
© Heriot-Watt University, B49CE,
December 2017 v3 59
Multiphase Thermodynamics
 However, the calculations are iterative and may be laid out in
the same way irrespective of whether Raoult’s Law or a de
Priester chart is used to find the K - values.
1. Fix P & xi .
2. Set each Pi sat  P and use the Antoine equation in reverse
to calculate Ti sat (see below). These are the boiling points
of each component at fixed pressure P:
Bi
Ti sat
  Ci
Ai  ln Pi sat

3. An initial trial temperature T( 0 ) may be estimated using

the expression below:

T( 0)   xi Ti sat
i

© Heriot-Watt University, B49CE,

December 2017 v3 60
Multiphase Thermodynamics
 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 estimated temperature T .
• Now read off all K-values from the chart.
• Alternatively if Raoult’s Law is used find all the K-values,
using the expression below:
Pi sat
Ki  ..........(i  1,2......N )
P
• In order to find these K-values, Pi sat will be needed
for each component (use the Antoine equation at
temperature T( 0 ) ).
5. Calculate all yi using phase equilibrium relationship below.
This holds whatever approach is used to find K -values:
y i  K i xi
© Heriot-Watt University, B49CE,
December 2017 v3 61
Multiphase Thermodynamics
 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

• Check is y
i
i  1  0.05 ?

7. In all likelihood this inequality will not be satisfied.

• Return to step 3 and adjust the guessed temperature
to new value T(1) . It will be fairly obvious whether to
increase or decrease the temperature.
• If the inequality sums to less than one, then K-values
need to be increased.
• This means a higher volatility is required; at a fixed
pressure, this means a higher temperature is needed.
© Heriot-Watt University, B49CE,
December 2017 v3 62
Multiphase Thermodynamics
 8. With this new temperature T(1) repeat steps (4) through (6)
and keep repeating the procedure until the inequality is
satisfied.
• At the end of the process the unknown temperature T
and vapour phase composition yi (bubble point
composition) will have been found.
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 Raoult’s Law (instead of charts) more efficient
two-stage iteration schemes have been developed.
>>> Ultimately the best approach is to obtain the K-values
using the rigorous approach and solve the bubble point
calculation using a maths package (instead of a calculator).
© Heriot-Watt University, B49CE,
December 2017 v3 63
Multiphase Thermodynamics
 Appendix B shows how to execute a bubble point calculation
using the rigorous phi-phi approach. In this case the
temperature is fixed and the pressure is unknown.
• The fugacity coefficients and thus the K-values are found
using the PR EOS. Lengthy iteration schemes are avoided by
using a “Solve Block” (methane is supercritical).
• Appendix B may prove useful when it comes to
Design/Research Projects.
2.4.2 Dew Point Calculations
If the pressure P and the vapour phase compositions yi are
fixed, then this is known as a dew point calculation.
• The objective is then to find T , followed by the composition
of the first drop of liquid xi in equilibrium with this vapour.
Thus, there are again 2 equations & 2 unknowns.

© Heriot-Watt University, B49CE,

December 2017 v3 64
Multiphase Thermodynamics
 Apart from the fact that P and yi are now fixed, step 1, step
2, step 3 and step 4 are all the same.
• In step 5 the equation is reversed since yi is now fixed and
the purpose of a dew point calculation is to find xi as
follows:
yi
xi 
Ki
• The equations in step 6 are also reversed (what is needed is
 xi ). Thus, these two equations become:
 yi 
i xi  i  K   1.0
 i
• The inequality check now becomes  x  1  0.05
i
i ?
>>> Repeat steps 4 through 6 until the liquid phase mole
fractions all sum to one within some tolerance.
© Heriot-Watt University, B49CE,
December 2017 v3 65
Multiphase Thermodynamics
 A dew point calculation may also be carried out by fixing the
temperature T and finding unknown pressure P . This type
sat
P
of calculation is easier since all i are now fixed.
Summary:
1. Bubble Point Calculation:
Fix xi Find yi Either T or P may be unknown
2. Dew Point Calculation:
Fix yi Find xi Either T or P may be unknown
For bubble point and dew point calculations, the number of
variables to be fixed is given by the phase rule F  2    N.
For a binary system F  2 . Thus, it is also possible to
3. Fix T Fix P Find BOTH xi and yi .
>>> There are five usual combinations and this last one is the
easiest way of plotting a T  x  y diagram.
© Heriot-Watt University, B49CE,
December 2017 v3 66
Multiphase Thermodynamics
 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.

© Heriot-Watt University, B49CE,

December 2017 v3 67
Multiphase Thermodynamics
 Answer……

© Heriot-Watt University, B49CE,

December 2017 v3 68
Multiphase Thermodynamics
 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.

© Heriot-Watt University, B49CE,

December 2017 v3 69
Multiphase Thermodynamics
 Answer……

© Heriot-Watt University, B49CE,

December 2017 v3 70
Multiphase Thermodynamics
 2.4.5 Flash Vaporisation & Condensation
Consider a subcooled liquid mixture with an overall composition
zi and then consider that this mixture is gradually expanded at
constant temperature and constant overall composition.
• From 1 - 2 pressure decreases, but fluid remains entirely
liquid.
• At some pressure (point 2), the
Constant T liquid becomes “saturated” at
z1 its bubble point:
1 
sat
P  x1 X P1
• Notice that the saturated liquid
2 
P P  y1 composition is still the same as
P2sat X subcooled liquid, i.e. x1  z1 .
P
• Read off to left to find P then
right and down (as shown) to
0
find y1 .
0 x1 y1 1.0
x1 & y1

© Heriot-Watt University, B49CE,

December 2017 v3 71
Multiphase Thermodynamics
 Notice that T & xi are fixed, while P & yi are read from the
diagram. This is identical to the previous bubble point
calculation, only now the problem is solved graphically.
• At point 2 there is actually no vapour present, however,
moving fractionally away from 2 produces the first bubble
of vapour with composition yi .
Constant T • A dew point problem works in
z1
exactly the opposite sense.
1  • A superheated vapour is
sat
P  x1 X P1

P 2 
P  y1
compressed isothermally until
P2sat X
it intersects the saturated
P vapour curve.
• Now T & yi are fixed while
P & xi are read from the
0
0 x1 y1 1.0 diagram.
x1 & y1

© Heriot-Watt University, B49CE,

December 2017 v3 72
Multiphase Thermodynamics
 On the saturated vapour curve itself there is actually no liquid
present, but move fractionally away from this point and the
first drop of liquid, composition xi , will be formed.
Constant T
• With “flash calculations” the
PF 1 P  x1 X P1sat temperature and/or pressure
PBP is altered.
L P  y1
P   2  V • In such a way that a two-phase
P2sat X PDP mixture can be separated.
P
• The mixture is separated inside
the drum with V mol/s of
vapour leaving the top.
0
0 x1 z1 y1 1.0
• And L mol/s of liquid leaving
x1 & y1 the bottom.
>>> With flash calculations the unknowns are the vapour
flowrate and composition V & yi and the liquid flowrate and
composition L & xi .
© Heriot-Watt University, B49CE,
December 2017 v3 73
Multiphase Thermodynamics
 As well as more unknowns, flash calculations require more
equations; in addition to phase equilibrium relations, mass
balances are needed.
• For bubble/dew point
Vapour
calculations, the degrees
V, yi
of freedom is found from
the Gibbs phase rule.
T, P
Feed • The Gibbs phase rule
Steam
F, zi, T, CANNOT be used for flash
PF
calculations.
If F  1, then • This is because the
Fz i  ni , i.e. T
variables are no longer
moles “i ” Condensate
purely intensive.
entering in the
Liquid • The AMOUNT of each
feed.
phase is needed.
L, xi

>>> It is easy to show that a flash calculation requires two

variables to be fixed; in this case they will be taken to be T & P .
© Heriot-Watt University, B49CE,
December 2017 v3 74
Multiphase Thermodynamics
 Two important types of flash are “isothermal” and “adiabatic”
flash.
Vapour
• With an isothermal flash
both T & P are fixed.
V, yi
• The vapour offtake valve
T, P controls the pressure
Feed
within the drum.
Steam
F, zi, T, • Here the steam valve
PF
controls the temperature
within the drum.
T
Condensate • While heating is shown, this
type of flash may also
Liquid
involve cooling.
L, xi

>>> An “adiabatic” flash involves no heat transfer. Thus, the

flash depends entirely on the enthalpy of feed stream and the
flash pressure.
© Heriot-Watt University, B49CE,
December 2017 v3 75
Multiphase Thermodynamics
 For an isothermal P  T flash, the pressure P must lie
between the bubble-point pressure Pbp and the dew-point
pressure Pdp . • A subcooled liquid feed
Constant T stream, state 1, enters with
temperature and pressure
PF 1 P  x1 X P1
sat T & PF .
PBP
• The overall composition is
L P  y1
P   2  V z1 , or n1 moles of species
P2sat X PDP “1” given that F  1 .
P • Pressure drops from PF to
P while the temperature
is kept constant at T .
0 • Liquid flashes to V kmol/h
x1 z1 y1 1.0
0
of vapour and L kmol/h
x1 & y1
of liquid.
>>> The amounts of each phase is given by the lever rule. The
compositions of each phase are different despite that fact that
the overall composition is fixed at z1 (or n1 moles of “1”)
© Heriot-Watt University, B49CE,
December 2017 v3 76
Multiphase Thermodynamics
 A bubble point calculation may thought of as a special case of
a more general flash calculation:
yi Vapour

V=0 V, yi • A bubble point calculation

is where the liquid exits
T, P the drum with the same
Feed
Steam
flow and composition as
F, zi, T,
PF
the feed entering.
• The drum is full of liquid
T and no vapour leaves.
Condensate • The first bubble of vapour
that liquid tends to generate
xi = zi Liquid
will have composition yi .
L=F L, xi

>>> If subcooled liquid leaves the drum, flash vaporisation of

the liquid into a two-phase mixture cannot occur and there is
no tendency to generate any vapour (liquid is unsaturated).
© Heriot-Watt University, B49CE,
December 2017 v3 77
Multiphase Thermodynamics
 A dew point calculation may also be thought of as a special
cases of a more general flash calculation:
yi= zi Vapour

V=F V, yi • A dew point calculation is

where the vapour exits the
T, P drum with the same flow
Feed
Steam
and composition as the
F, zi, T,
PF
feed entering.
• The drum is full of vapour
T and no liquid leaves.
Condensate • The first drop of liquid that
vapour tends to generate will
xi Liquid
have composition xi .
L=0 L, xi

>>> If superheated vapour leaves the drum, flash condensation

of the vapour into a two-phase mixture cannot occur and there
is no tendency to generate any liquid (vapour is unsaturated).
© Heriot-Watt University, B49CE,
December 2017 v3 78
Multiphase Thermodynamics
 Flash Calculations
It has been shown how bubble point and dew point
calculations can be carried out graphically (for two
components) or algebraically (for any number).
• Flash calculations may also be carried out graphically, as
just shown, or may be completed algebraically. Now we
need mass balances as well as equilibrium relations.
• An overall material balance with the feed flow initialised to
one ( F  1 kmol/h) is given by
1  V  L …….(using “dot” notation)
• If there are N components present, then there are N  1
independent component mass balances
>>>
zi  Vyi  L xi ni  Vyi  L xi
Fzi  n1
© Heriot-Watt University, B49CE,
December 2017 v3 79
Multiphase Thermodynamics
 There are N independent equilibrium relations given by
y i  K i xi
• The K -values may be from de Priester charts or from
Raoult’s Law (as below):
sat
If the rigorous approach is used,
Pi then there will be a more complex
Ki  functional dependency here.
P
• Take the general component material balance ni  Vyi  L xi
and substitute in the overall material balance V  1  L and
then the equilibrium relation yi  K i xi , to get
ni  1  L K i xi  L xi These may be solved by hand for de
Priester and Raoult’s Law. The
ni
 xi  rigorous approach for K i is best
1  L K i  L solved using a maths package.

© Heriot-Watt University, B49CE,

December 2017 v3 80
Multiphase Thermodynamics
 A P  T flash calculation entails the following trial-and-error
approach:
1. Fix F , n , T , P and initialise F  1 kmol/h.
1 F

2. Ensure flash pressure P lies between bubble point

pressure Pbp and the dew point pressure Pdp at prevailing
temperature.
3. Knowing T & P find all K i from a de Priester chart, or
calculate from Raoult’s Law.
4. Guess a starting value for L .
• A value of 0.5 is sensible.
• If the pressure is much nearer the bubble point, then a
value around 0.5 - 1.0 is better.
• If the pressure is much nearer the dew point, then a
value around 0 – 0.5 is better.
© Heriot-Watt University, B49CE,
December 2017 v3 81
Multiphase Thermodynamics
 5. Compute all xi from the expression below:
ni
xi 

1  L K  L  i

6. Calculate  xi and check if x i  1.0  0.05.

i

7. If  xi  1.0  0.05 , then go back to step 4, adjust L and

continue iterating until the tolerance condition is satisfied.
8. Once tolerance condition is satisfied the calculation has
ended and the final values of L & xi are correct to the
required precision.
• Calculate the vapour rate from the overall material
balance:
1  V  L

© Heriot-Watt University, B49CE,

December 2017 v3 82
Multiphase Thermodynamics
 Calculate all the vapour phase compositions yi from each of
the phase equilibrium relationships:
y i  K i xi
• For de Priester charts and Raoult’s Law the K i -values do
not change throughout the procedure temperature and
pressure are fixed.
• Finally scale-up all the flowrates based on the actual feed
flowrate.
• If needed the heat balance may be solved for Q (kW). All
the enthalpies are fixed if T & P and the stream
compositions are known. For “adiabatic” flash the left
hand side of the heat balance is
Q  VhV  L hL  FhF set to zero and pressure is fixed.
>>> An adiabatic flash is more complicated because now the
temperature “floats” and all relations must be solved
simultaneously (a maths package is advisable).
© Heriot-Watt University, B49CE,
December 2017 v3 83
Multiphase Thermodynamics
 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
CH4 0.05

C2H8 0.10

C3H8 0.30

i-C4H10 0.55

Calculate the fractional flowrates and compositions of both

the exit liquid and vapour streams leaving the drum using the
following conversions:
C   F  32 
o o 5 1 psi  6894.8 Pa  6.9 kPa
9
© Heriot-Watt University, B49CE,
December 2017 v3 84
Multiphase Thermodynamics
 Answer……

© Heriot-Watt University, B49CE,

December 2017 v3 85
Multiphase Thermodynamics