Thermodynamics 2

Generalized Phase Equilibria Models, Colombia, Summer 2000 -
Author: Dr. Maria Barrufet - Summer, 2000 Page 1/77

Instructional Objectives
Understand the Principle of Corresponding States.
Calculate the compressibility factor using different correlations and models.
Understand phase equilibrium.
Determine the number of variables required to define a system in equilibrium (Phase
Rule).
Evaluate energy relationships using the First and Second Law of thermodynamics.
Evaluate dew and bubble points given pressure or temperature as independent
variables.
Generalized Phase Equilibria Models

The Principle of Corresponding States. Correlations and Models.
Extension of Corresponding States to Mixtures.
Phase equilibrium. Phase rule.
Thermodynamic Properties of Homogeneous and Heterogeneous Systems.
Phase Equilibrium: Vapor-Liquid-Equilibrium (VLE), Liquid-Liquid Equilibrium (LLE),
Solid-Liquid-Equilibrium (SLE).
Phase Equilibrium Models: Single Components. Reduced Equations of State (EOS).
Multicomponents. Mixing Rules.
Types of VLE Computations: Dew Point and Bubble Point Calculations. Multiphase
Flash.
Low Pressure Phase Equilibria Computations (Surface Separators).
Ideal Systems.
K-value correlations. Empirical methods to determine equilibrium ratios (K-values).

Suggested reading: EL, WM, MAB


Evaluate flash separation processes.

The Principle of Corresponding States
The compressibility factor, or Z factor, of all pure species (C
1
, C
2
, N
2
, CO
2
etc) can be
read from charts which are presented as a function of reduced properties T
r
and P
r
.

Any correlation, or model, which expresses the Z factor as function of T
r
and P
r
is said
to be generalized. Modern equation of state (EOS) can be put into this form, thus
providing a generalized correlation for the compressibility factor.

One needs only the critical temperature and the critical pressure of the fluid. This is the
basis for the two-parameter theorem of corresponding states.

All fluids when compared at the same reduced temperature and
reduced pressure, have approximately the same compressibility
factor, and all deviate from ideal gas behavior to about the same
degree

The Principle of Corresponding states (POC) originated with single component fluids.
We, engineers, stretched it to multicomponent systems.

Generalized Corresponding States
The principle of corresponding states says that all material properties when expressed
in terms of reduced parameters such as:


Reduced Temperature:
( )
c r
T T T /
Reduced Pressure:
( )
c r
P / P P ,
Reduced Molar Volume :
( )
c r
V
~
/ V
~
V
~

obey a similar or corresponding behavior.

Note that below T
c
and P
c
the reduced properties are lower than one and at these
conditions, for a single component fluid, there are two coexisting phases. The Z-factor
charts DO NOT provide information for saturation properties.

Nomenclature:
V
~
molar volume [=] cm
3
/gmol, ft
3
/lbmol (an intensive property.)
The Van der Waals (VdW) and the Redlich-Kwong equations of state are two-parameter
corresponding state equations, these two parameters are the critical temperature and
the critical pressure of the component in question. The critical volume is determined
once these two parameters ) P , T (
c c
are fixed.

Further improvement, in terms of describing the fluid behavior for a broad spectrum of
pressures and temperatures is achieved by adding a third parameter. These models are
called three-parameter corresponding state equations. This third parameter is called the
acentric factor and was introduced by Pitzer and coworkers. It takes into account the
non-spherical nature of molecules. The Peng Robinson and the Soave Redlich Kwong
equations of state (EOS) are examples of three parameter corresponding states
models.


The acentric factor is tabulated and is defined as
( )
7 0
1
. T
sat
r
r
P log

(1)

For ideal gases such as He, Ar, and Kr, the acentric factor is zero. For methane, which
is a nearly spherical molecule the acentric factor is nearly zero (0.0104).
1/T
r
log(P
r
Sat
)
-1
1.43 1.0
Slope = -2.3 ( Ar, Kr, Xe)

Figure 1 - Acentric factor definition.

Compressibility Factor Charts
Following the POC only one compressibility factor chart can be used to determine
volumetric properties of any pure fluid using its reduced properties. The shape of this
chart is in general.



T
r
P
r
Z
1

Figure 2 - Compressibility factor as a function of reduced properties..

Corresponding States Correlations & Models
Equations of State for Gases: Virial Equations
The objective is then to find a model (models) to predict the Z factor. The ideal gas
behavior is described from the ideal gas Equation of State (EOS) with a compressibility
factor of 1.

1
RT
V
~
P
(2)

For real gases the analogous expression is



Z
RT
V
~
P
(3)

where Z is the compressibility factor. The compressibility factor can also be defined as
the ratio of the real molar volume over the ideal molar volume of a substance
measured at the same pressure and temperature.

Z
Z
V
~
P
RT
RT
V
~
P
id

,
_
,
_
1
(4)

Now the deviation of the compressibility from ideal behavior (Z = 1) can be expressed in
terms of an infinite series (in practice only two or three terms are used). Two equations
were devised for this purpose
The pressure virial equation is a polynomial expansion in pressure

(1) Pressure Virial Equation
... P ' D P ' C P ' B
RT
V
~
P
+ + + +
3 2
1 (5)
with B, C, D called pressure virial coefficients. These are determined from
experimental data and are temperature dependent. This equation is used for moderate
pressures (P < 15 bar at subcritical temperatures). Only two pressure-virial coefficients
are enough.
For higher pressures we would require more terms in the series, but these would be
difficult to determine experimentally, thus other models are used instead.



(2) Density Virial Equation
...
V
~
D
V
~
C
V
~
B
RT
V
~
P
+ + + +
3 2
1 (6)

This equation is used for higher pressures (P between 15 and 50 bar, with three
density-virial coefficients being enough for this pressure range).
The coefficients from the two expansions (pressure and density virials) are related.
The higher the pressure the higher the deviation from ideality and the more terms are
required (in either) expansion to describe the compressibility factor.
These virial coefficients have been obtained experimentally for various substances and
are a function of temperature. Additionally, several correlations exist for them. (see EL).

Cubic Equations of State
Cubic polynomials in molar density (or molar volume) are the most popular
EOS for many industrial processes (petroleum production, reservoir compositional
simulation, refining, distillation, separation processes, petroleum recovery, etc.) These
equations describe reasonably well the pressure-volume-temperature (PVT) behavior of
fluids in both the gas and the liquid region. They can also be used for mixtures,
provided certain mixing rules are applied (these will be seen later).
The most well-known and older EOS is the Van der Waals equation, which is:

2
V
~
a
b V
~
RT
P
(7)



The parameter a corrects for attraction forces between molecules, while the parameter b
corrects for repulsion forces by taking into account the volume of the molecules. Recall
that the ideal gas assumption was that gas molecules had zero volume.
Probable the most widely used EOS in the gas and petroleum industry is the Peng-
Robinson EOS

V
~
) b V
~
( ) b V
~
( V
~
a
b V
~
RT
P
+ +
(8)

The two parameters (a and b) in these EOS can be determined from least squares
regression (fitting) of data at a broad range of Pressure and Temperature for the
substance in question. However most of the time, that is not affordable.
Critical properties, however, are known for a variety of substances and these have been
used to determine the constants a and b using theoretical constraints.
The critical point observed in a ) V
~
P ( diagram or a (PT) diagram exhibits a maximum in
pressure and an inflection point on the critical isotherm (Tc). Figure 3 shows these
conditions for a pure substance in a ) V
~
P ( diagram.



0
1
20
0
30
0
40
0
50
0
60
0
70
0
1
2
1
4
P
r
e
s
s
u
r
e
Molar Volume
T
c
T
2
T
1
P
1
v
L
2 - Phases
CP
V
L
V
0
1
20
0
30
0
40
0
50
0
60
0
70
0
1
2
1
4
P
r
e
s
s
u
r
e
Molar Volume
T
c
T
2
T
2
T
1
T
1
P
1
v
L
2 - Phases
CP
V
L
V

Figure 3 - Pressure-volume behavior indicating isotherms for a pure component system.

These two conditions are expressed mathematically in two equations that are used to
solve the constants a and b in terms of the critical properties.
For example lets use the VdW EOS.

2
V
~
a
b V
~
RT
P
(9)



The first and second derivatives of pressure with respect to volume at constant
temperature are:

3 2
2
V
~
a
) b V
~
(
RT
V
~
P
T
+

,
_
(10)

4 3 2
2
6 2
V
~
a
) b V
~
(
RT
V
~
P
T
,
_
(11)

These two derivatives must vanish at the critical point

3 2
2
0
c c
c
T
T
V
~
a
) b V
~
(
RT
V
~
P
c
+
,
_
(12)

4 3 2
2
6 2
0
c c
c
T
V
~
a
) b V
~
(
RT
V
~
P
,
_
(13)

Now we have two equations and two unknowns.

3 3
2
2
3
2
2
3 2
3 3
3 2
c
c
c
c
c
c
c
c
c
c
c c
c
V
~
b V
~
) b V
~
(
RT
) b V
~
(
RT

V
~
a
V
~
) b V
~
(
RT
V
~
a

) b V
~
(
RT

(14)

Once b has been found, a can be obtained from either Eq. (12) or (13).



( )
8
9
8
9
3 2
2
2
3
2
3
3
2
c c
c
c c
c c
c c
c
c
c
V
~
RT
V
~
V
~
RT
/ V
~
V
~
V
~
RT
V
~
) b V
~
(
RT
a
(15)

The critical compressibility factor is

c
c
c
Z
RT
V
~
P
(16)

Replacing the values found for a and b and using the EOS at the critical point we obtain

c
c
c
c
c
c
c
c c
c c
c
c
V
~
RT
V
~
RT
V
~
RT
V
~
V
~
RT
V
~
V
~
RT
P
8
3
8
9
2
3
8
9
3
2

(17)

And the critical compressibility calculated from VdW EOS is

375 0
8
3
. Z
RT
V
~
P
c
c
c
(18)

Using this we can express the a and b constants as

( ) ( )
c
c
c
c
c
c
c c c c
P
RT
P
RT
Z
P
RT RT V
~
RT
a
64
27
8
3
8
9
8
9
8
9
2 2

,
_
(19)

and



c
c
c
c
c
c
c c
P
RT
P
RT
Z
P
RT V
~
b
8 8
3
3 3 3

,
_
(20)

The Z factor is then evaluated as

RT V
~
a
b V
~
V
~
RT
V
~
P
Z
(21)

Note that this EOS is a cubic polynomial in volume, therefore three possible real roots
could be obtained from the equation. The root selection will be discussed in future
lectures. To solve cubic equations there are analytical techniques. The following web
site will provide you the computer codes to solve the roots of polynomials up to quintic
degree.
http://www.uni-koeln.de/math-nat-fak/phchem/deiters/quartic/quartic.html

Since web sites change addresses quite frequently I recommend to copy and test the
course codes as soon as possible.



Correlations for the Compressibility Factor
The simplest correlation for the compressibility factor is expressed in terms of the
second virial coefficient.

r
r
c
c
T
P
RT
BP
RT
BP
Z

,
_
+ + 1 1 (22)

And the term (BP
c
/RT
c
) is determined using Pitzers Correlation as follows,

1 0
B B
RT
BP
c
c
+
,
_
(23)

with

6 . 1
0
422 . 0
083 . 0
r
T
B (24)

2 . 4
1
0172
139 . 0
r
T
B (25)

Therefore the compressibility factor is expressed as:

r
r
r
r
T
P
B
T
P
B Z
1 0
1 + + (26)



The following example illustrates the use of the compressibility factor (Z) in a design
problem

ExampleProblem ExampleProblem ExampleProblem ExampleProblem
Mr. Jones wants to usesome30 liter cans to ship ethanefrom CollegeStation to Conroe. He
would liketo fill each of thesecylinders with 10 kg of ethane, but hedoes not know the
pressureat which heneeds to fill thesetanks or if thewalls of thetanks will beableto
withstand that kind of pressure. Theshipping should bedoneat an averagetemperatureof 25
o
C.

Usethreedifferent methods:
(a) (a) (a) (a) Ideal gas EOS
(b) (b) (b) (b) Z factor compressibility correlations given in class
(c) (c) (c) (c) Z factor charts using chart given in class notes (you will do this one)
(d) (d) (d) (d) Z factor from a cubic EOS (you will do this)
(e) (e) (e) (e) Z factor using properties evaluated from NIST websiteseen in Module1 (you will do
this)
TheCritical properties and acentric factor for ethaneare:
Mw = 30 g / mol
Tc = 305.5 K
Pc = 48.8 bar
= 0.098

(a) Ideal Gas EOS (a) Ideal Gas EOS (a) Ideal Gas EOS (a) Ideal Gas EOS


( )
( )
) K . (
mol K
bar cm .
g/mol
g ,
Mw
mRT
nRT PV 25 15 273
14 83
30
000 10
3
+
,
_

bar .
cm
bar cm
,
.
MwV
mRT
P 424 275
000 30
10 26273 8
3
3 6
,
_

(about 4,000 psia)

(b) Z (b) Z (b) Z (b) Z - -- -factor Correlations factor Correlations factor Correlations factor Correlations
To usePitzer correlation wemust calculatethereduced temperaturewhich is

9755 . 0
5 . 305
15 . 273 25
,
_
+

c
r
T
T
T

Thecorrelation is

r
r
c
c
T
P
RT
BP
Z

,
_
+ 1 (A)

wealso know that

mRT
VMw P P
nRT
PV
Z
c r

r
r
P
P
Z 177181 . 0
) 25 15 . 273 )( 14 . 83 )( 000 , 10 (
) 30 )( 000 , 30 )( 8 . 48 (
+
(B)



with

1 0
B B
RT
BP
c
c
+
,
_

Theacentric factor for ethanecan beobtained from thetables provided with properties for pure
components.

( )
3561 . 0
9755 . 0
422 . 0
083 . 0
422 . 0
083 . 0
6 . 1 6 . 1
0

r
T
B
( )
0519 . 0
9755 . 0
0172
139 . 0
0172
139 . 0
2 . 4 2 . 4
1

r
T
B
( ) 3612 . 0 0519 . 0 098 . 0 3561 . 0
1 0
+ +
,
_
B B
RT
BP
c
c

and
,
_
,
_
+
9755 . 0
3612 . 0 1 1
r
r
r
c
c
P
T
P
RT
BP
Z

Combining the2 equations A and B, wehave
r
P Z 37027 . 0 1

and
r
P Z 177181 . 0


Equating thesetwo equations (A) & (B) wesolvefor thereduced pressure, and then for the
pressureP.

bar . . . P P P
. P
c r
r
12 89 8 48 8263 1
8263 1

Thereis a substantial differencefrom theideal gas model!

(c) (c) (c) (c), (d) (d) (d) (d) and (f) (f) (f) (f) arepart of your homework assignment # 2 homework assignment # 2 homework assignment # 2 homework assignment # 2.

Extension of Corresponding States to Mixtures
We, engineers, would love to stretch the corresponding states principle to mixtures and
we do.
Z factor charts (all built from EOS) are also used for multicomponent systems in this
case the coordinates used are pseudo-reduced properties. You can use the same
charts for a pure component.
For mixtures the same type of charts apply but using pseudoreduced properties which
are defined similarly as the ratio of pressure (or temperature) with pseudoreduced
critical pressure" (or temperature). These pseudocritical properties are an average of
the critical properties of the components in the mixture. Charts for mixtures can also be
used for single component fluids.
A typical chart using an EOS is shown in Figure 4.



Figure 4- Compressibility factor Z as a function or pseudoreduced pressure.

The same models are used to determine the gas compressibility factor for mixtures. The
extension is through some mixing rules

The accuracy will depend largely from model used and information input to the model.



Pseudocritical Properties of Natural Gases
Pseudoreduced Pressure
pc
pr
P
P
P (27)
Pseudoreduced Temperature
pc
pr
T
T
T (28)

If only the specific gravity of the gases is known then charts are available to estimate
these pseudocritical properties (undergraduate material, review McCain).
Naturally the degree of accuracy is reduced substantially. We well see methods when
compositional information is available, in this case:

( )
ci i
N
i
pc
P y P
c
1
(29)

( )
ci i
N
i
pc
T y T
c
1
(30)

Once Z is evaluated you can find the gas density as

( )
3
/ ft lbm
V
M
g
(31)
g
nMw M

(32)



(Mw
g
is the average molecular weight of the gas evaluated as)

( )
Nc
i
i i g
Mw y Mw
1

(33)

and the total volume (extensive property is)

P
ZnRT
V

(34)

Therefore,

ZRT
PMw
V
M
g
g

(35)

So far we were just determining properties either for a gas or a highly compressed
fluid (liquid like density) in the SINGLE PHASE REGION.

Notice that Z- factor charts DO NOT HELP AT ALL IN DETERMINING
PROPERTIES OF GAS AND LIQUID COEXISTING PHASES

Figure 5 shows a chart of the compressibility factor for low reduced pressures.



Figure 5- Compressibility factor chart for low reduced pressures.

There is an undefined region that corresponds to the two-phase region.



Phase Equilibrium and the Phase Rule
Equilibrium indicates static conditions, the absence of change. In thermodynamics is
taken no mean not only the absence of change, but the absence of any tendency to
change. Therefore a system that is in equilibrium is one in which under such conditions
that there is no tendency for a change to state to occur.
Tendencies toward a change are caused by a driving force of any kind, the absence of
such a tendency indicates also the absence of any driving force, or that all forces are in
exact balance.
Typical driving forces include mechanical forces such as pressure on a piston tend to
cause energy transfer as work; temperature differences tend to cause the flow of heat;
chemical potentials tend to cause mass transfer from one phase to another or cause
substances to react chemically.
In reservoir engineering applications we assume that reservoir fluids are at equilibrium,
we do not say how long the equilibrium will last. Therefore as a reservoir block
changes pressure due to production (injection) we assume that equilibrium is reached
instantly. Fluid properties in reservoir cells are evaluated using a sequence of
connected equilibrium stages.

The Phase Rule
As mentioned earlier, the state of a system is determined when all intensive properties
are defined. The intensive properties are related, for example for a single component in
the single phase region providing P and T is enough to define the state of the system,
since the molar volume and the compressibility can be evaluated as a function of these
two variables.


For multicomponent systems, we need to find out what is the minimum number of
properties (variables) is required to define the state of the system. The phase rule
provides the answer. Lets begin by describing different cases.

Single component Single Phase
To define the state of the system need to provide two coordinates. The number of
independent variables is two, and these are usually pressure and temperature. The
molar volume is another variable, but that is not independent, because a relationship
exists between P, T, and V.
We have two degrees of freedom. Once a pressure and a temperature are selected, the
state of our single component system is defined (i.e. all intensive properties).

Single component Two Phases
Assume the system exhibits VLE (vapor-liquid-equilibrium). Here we just need to define
either the saturation pressure or the saturation temperature. Only one variable is
needed to specify the state of the system. We have one degree of freedom. These two
are related through the vapor pressure equation.

Single component Three Phases
Here we have a unique point in space named the triple point. The system is an
invariant. We cannot specify any variable. We have zero degrees of freedom.

Two components Single Phase
To define the state of the system need to provide three coordinates usually (P ,T & z
1
).
We have three degrees of freedom. The state of our binary system is defined (i.e. all
intensive properties).



Two components Two Phase
Assume the system exhibits VLE (vapor-liquid-equilibrium). Here we just need to define
two variables to define the state of the system The choices could be (P,y
1
) or (P,x
1
),or,
(T,x
1
), or (T,y
1
). Note that the overall composition is not phase rule variables when more
than a phase are present. We have three degrees of freedom. Following this reasoning,
Table 1 shows the degrees of freedom, or number of independent variables required to
define the system, of different non-reacting components and number of phases

Number of components Number of phases Degrees of Freedom
1 1 2
1 2 1
1 3 0
2 1 3
2 2 2
2 3 1

N
c
N
p
(N
c
-N
p
)+2
Table 1 - Generalization of the phase rule for N
c
non reacting components.

Thus for non-reacting systems

2 +
p c
N N F (36)

This is the so called phase rule presented by an American mathematical physicist, J.
Willard Gibbs (1839-1903).


The number of independent variables that must be arbitrarily fixed to establish the
intensive state of any system is called the degrees of freedom F of the system. Another
way to derive this equation (phase rule) in a more general way is:

F = # of variables # of Independent equations relating these variables

These independent equations are: an EOS (equation of state), material constraints
(sum of mole fractions = 1), chemical reactions.
Note that the phase rule does not tell which variables to chose, it only says how many
and these must be independent. The choice of independent variables depends upon the
type of model available and the simplicity, or complexity of the calculations.

Thermodynamic Properties of Homogeneous and
Heterogeneous Systems
The objective of this section is to present the most widely used thermodynamic
properties of homogeneous and heterogeneous systems. These properties are
functions of primary variables such as pressure, temperature, molar volume, and
compositions (the last for multicomponent systems). These properties are based upon
the first and the second law of thermodynamics and are used to evaluate energy
requirements for a variety of processes and to derive models to evaluate phase
equilibrium.

First Law and Fundamental Thermodynamic Relationships
Closed Systems
The system does not exchange matter with the surroundings, but it can exchange
energy.


The first law is a generalization of the conservation of energy and can be defined by the
following equation

dW dQ dU
t
(37)

where dU
t
is the change in internal energy as a result of dQ, heat absorbed (or
released) by the system, and dW, work done (or provided) by the system on the
surroundings. By convention work done on the system is negative, and heat released by
the system is positive. Figure 6 indicates this convention, a gas contained in a vessel
with a movable piston. Compressing the gas (-) will cause the system to increase its
temperature and heat will be released by the system (+). The opposite process involving
gas expansion has the opposite signs for work and heat.
- dW
+dQ
Compression
+dW
- dQ
Expansion
- dW
+dQ
Compression
+dW
- dQ
Expansion



Figure 6 - Compression and expansion work in a gas container indicating the
convention used for heat and work.

For a reversible process, dQ = TdS
t
thus,

dW TdS dU
t t
(38)

If the work of expansion or compression is the only kind of work allowed then:

t
PdV dW (39)

Replacing Eq. (39) into Eq. (38)

t t t
PdV TdS dU (40)

Thus

( )
t t t t
V S U U , (41)

Equation (40) applies to any process in a closed PVT system that results in a
differential change from one equilibrium state to another.

Equation (40) could also be written as:



( ) ( ) ( ) nV Pd nS Td nU d (42)

again this equation applies to ANY CLOSED single or multicomponent system.
Equations (40) or (42) are exact differentials therefore one can identify,
( )
( )
T
nS
nU
n nV
,
_
,
and
( )
( )
P
nV
nU
n nS

,
_
,
(43)

The primary thermodynamic properties are internal energy, volume and entropy (U
t
, V
t
,
S
t
, respectively). For convenience a set alternate thermodynamic properties are
defined. These are the Entalphy, the Gibbs's energy and the Helmholtz energy (H
t
, G
t
,
A
t
, respectively). The enthalpy is used for flow processes while the Gibbss energy is
used for equilibrium computations. The Helmholtz energy does not have a lot of use in
Petroleum engineering type calculations.

By definition:
M
t
, = nM with M = U, H, A, G, S (A also known as F in European notation). The
relationship among these properties is:

t t t
PV U H + (44)

t t t t t t
TS U TS PV H F (45)

and

t t t
TS H G (46)


The same relationship holds for the intensive properties (M = M
t
/n)
Expressions similar to Eq. (40) can be derived for equations (44) to (46).

( ) P S H H dP V TdS dH
t t t t t t
, i.e. + (47)

( )
t t t t t t
V T F F PdV dT S dF , i.e. (48)

( (( ( ) )) ) T P G G dP V dT S dG
t t t t t
, i.e. + ++ + (49)

The (U
t
,H
t
,F
t
,G
t
,S
t
) are STATE properties which means independent of path.

Open Systems
For an open system, the basic thermodynamic functions U
t
, H
t
, F
t
, and G
t
in addition to
the two independent variables outlined above, will also depend on the concentration of
each of the components.
The number of moles of each specie may change due to:
Chemical reaction within system
Interchange of matter with surroundings
Interchange and chemical reaction.

In this course we will not consider chemical reactions. However; the treatment for these
is similar.



The functional form of U
t
, H
t
, F
t
, and G
t
for open systems are,

( )
c
N t t t t
n n n V S U U ..., , , ,
2 1
(50)
( )
c
N t t t
n n n P S H H ..., , , ,
2 1
(51)
( )
c
N t t t
n n n V T F F ..., , , ,
2 1
(52)
( )
c
N t t
n n n P T G G ..., , , ,
2 1
(53)

The differential form of the above equations are,

i
i n V S
N
i i
t
t t t
dn
n
U
PdV TdS dU
j t t
c
,
_
+
, ,
1
(54)
i
i n P S
N
i i
t
t t t
dn
n
H
dP V TdS dH
j t
c
,
_
+ +
, ,
1
(55)
i
i n V T
N
i i
t
t t t
dn
n
F
PdV dT S dF
j t
c
,
_
+
, ,
1
(56)
i
i n P T
N
i i
t
t t t
dn
n
G
dP V dT S dG
j
c
,
_
+ +
, ,
1
(57)

let the chemical potential of component " i " ,

be defined by

i n P T
i
t
i n V T
i
t
i n P S
i
t
i n V S
i
t
i
j j t j t j t t
n
G
n
F
n
H
n
U

,
_
,
_
,
_
,
_
, , , , , , , ,
(58)



The chemical potential was introduced first by Gibbs.
From all the alternative expressions in Eq. (58) for the chemical potential the last is the
most useful for phase equilibrium computations. The reason is that the independent
variables (P, and T) are readily measured.

Second Law and the Equilibrium Criteria
One form to state the second law is that for an isolated system, all real processes occur
with a zero or positive entropy change. Figure 7. shows the entropy evolution with time
and Eq, (59) puts the above statement in mathematical form.

0
t
dS (59)

In Eq. (59) dS
t
is positive for an irreversible process and zero for a reversible process.



Equilibrium
dS = 0
Time
E
n
t
r
o
p
y
,

S

Figure 7 - Entropy versus time for any physical process.

Figure 7 shows the trend to equilibrium. It then follows that for an isolated system to be
at equilibrium, the entropy must have reached the maximum value. Therefore, at
equilibrium,

0
t
dS (60)

subject to the constraints

0
t
dU (61)


0
t
dV (62)
c j
,...N , i dn 2 1 0 (63)

Note that if the system is composed of several phases (and it is a
heterogeneous system) S
t
,U
t
,V
t
, and n
i
in Eqs. (54) to (57) are the
summation over the values in all parts or phases.

The criteria of equilibrium of a system can also be stated in terms of U
t
, H
t
, F
t
, and G
t
as
follows
The internal energy, U
t
, must be a minimum at constant S
t
, V
t
, and n
i
.
The enthalpy, H
t
, must be a minimum at constant S
t
, P, and n
i
.
The Helmholtz free energy, F
t
, must be a minimum at constant T, V
t
, and n
i
.
The Gibbs free energy, G
t
, must be a minimum at constant T, P, and n
i
.

Therefore the equilibrium problem is evaluated by minimizing either one of these
thermodynamic functions. We require a thermodynamic model to evaluate these
functions and EQUATIONS OF STATE are these models.
The choice of the function will depend upon the selection for dependent/independent
variables. The most popular one is the Gibbss energy because of its natural dependent
variables. Its the easier from a computational view point.

Chemical and Phase Equilibria Criteria for an Open
System Using Intensive Properties
Consider a closed PVT system consisting of two phases in equilibrium. Each phase
may be considered a single-phase OPEN system.


Vapor
P
v
T
v
n
i
v
Liquid
P
l
T
l
n
i
l
Vapor
P
v
T
v
n
i
v
Liquid
P
l
T
l
n
i
l

Figure 8 - Two phases in equilibrium.

The differential equation for the internal energy applied to each phase is,

( ) ( ) ( )
v
i
N
i
v
i
v v v
dn nV Pd nS Td nU d
c
+
1
(64)
( ) ( ) ( )
l
i
N
i
l
i
l l l
dn nV Pd nS Td nU d
c
+
1
(65)

And the total energy is the sum of Eqs. (25) and (26).

( ) ( ) ( )
l
i
N
i
l
i
v
i
N
i
v
i
dn dn nV Pd nS Td nU d
c c

+ +
1 1

(66)



Comparing Eq. (67) with Eq. (42) for a closed system we must have that at equilibrium;

0
1 1
+

l
i
N
i
l
i
v
i
N
i
v
i
dn dn
c c
(67)

but from mass conservation;

l
i
v
i
dn dn
(68)

Thus

0
1 1

v
i
N
i
l
i
v
i
N
i
v
i
dn dn
c c
(69)

replacing Eq. (58) into Eq. (59),

( ) 0
1

v
i
l
i
N
i
v
i
dn
c
(70)

For more than two phases in equilibrium, successive applications of Eqs. (64) and (65)
lead to.

c i i i i
N i ,... 2 , 1 ...

(71)



Where the phases could be two liquids, solid-liquid, vapor-solid-liquid etc.
Additional primary thermodynamic functions are the mole fractions defined as,

and
1 1
l
l
i
N
i
l
i
l
i
i v
v
i
N
i
v
i
v
i
i
n
n
n
n
x
n
n
n
n
y
c c

(72)

conventionally "vapor" and "liquid" molar compositions of component "i" are denoted as
"x
i
" and "y
i
" respectively.
Now rewrite Eq. (65) using (72). (We could have chosen a gas phase as well).

( ) ( ) ( )
l
i
N
i
l
i
l l l
n dx nV Pd nS Td nU d
c
+
1
(73)

by using the chain rule for differentiation, expanding and collecting terms, we obtain:

0
1 1

,
_
+ +
,
_
+

l
i
N
i
l
i
l l l l
i
N
i
l
i
l l l
dn x PV TS U n dx PdV TdS dU
c c
(74)

Since n
l

and dn
l
are arbitrary, the terms inside the brackets must both be zero, and this
provides the following identities in terms of intensive properties instead of total
properties.

i
N
i
l
i
l l l
dx PdV TdS dU
c
+
1
(75)



and

i
N
i
l
i
l l l
x PV TS U
c
+
1
(76)

Similar expressions are obtained with the gas phase.

Recall that for a constant composition fluid

( )
t t t t t t t
V S U U PdV TdS dU , i.e. (77)

( ) P S H H dP V TdS dH
t t t t t t
, i.e. + (78)
( )
t t t t t t
V T F F PdV dT S dF , i.e. (79)
( ) T P G G dP V dT S dG
t t t t t
, i.e. (80)

The (U
t
,H
t
,F
t
,G
t
,S
t
) are STATE properties which means independent of path. These
sets of equations are exact differential expressions.

For example lets take the last set

and
t
T
t
t
P
t
V
P
G
S
T
G
,
_
,
_
(81)

From math relations we can see that cross derivatives are the same regardless of the
order of differentiation.


1 2
2
2 1
2
x x
y
x x
y

(82)

therefore

T P
G
P T
G
t t

2 2
(83)

That means

P
t
T
t
T
V
P
S
,
_
,
_
(84)

These identities are very handy and are used frequently in deriving suitable expressions
for a variety of processes (heat transfer, evaluation of energy requirements in flow
processes, phase equilibria, etc).
These identities are called the The Maxwell Equations.

Some Mathematical Relations of Thermodynamics
The next part of this handout will give you a flavor of some of the mathematical
manipulations that will be used in many of the mathematical derivations that will follow
in the course.
Appendix A from Van Ness and Abbot (Classical Thermodynamics of Non Electrolyte
Solutions, Mc Graw Hill is also a helper for some of these math rules


In the derivation of a thermodynamic identity, you start with one or more equations (i.e.,
a fundamental eq. or definition) and apply the following mathematical relations:

Maxwell Equations.
P
t
T
t
T
V
P
S
,
_
,
_
(85)
t
V
T
t
t
T
P
V
S
,
_
,
_
(86)
t t
V
t
S
t
S
P
V
T
,
_
,
_
(87)
and
P
t
t
S
S
V
P
T
t
,
_
,
_
(88)

The Maxwell identities can be written also for constant composition fluids, in this case
we replace M
t
by M, with M any of the functions in Eq.(85) to (88).

Minus One Rule
If P (T, V) then:
(P/T)
v
(T/!V)
p
(V/P)
T
= - 1
If S(T, V) then
(S/T)
v
(T/V)
s
(V/S)
T
= -1
or
(C
v
T) (T/V)
s
= - (S/V)
T
= - (P/T)
v



Straight Chain Rule
(P/T)
v
(T/S)
v
= (P/S)
v

(the partial differentials may be cancelled but always the same property must be held
constant)

Change of Variable Rule
if P (U, V)
dP = (P/U)
v
dU + (P/V)
U
dV
or
(P/V)
T
= (P/U)
V
(U/V)
T
+ (P/V)
U

You can use this rule to switch from (P/V)
U
to (P/V)
T
.

Inversion Rules
First derivatives can be inverted:
[1/(P/T)
V
] = (T/P)
V

Second derivatives cannot be directly inverted:
[1/(
2
P/T
2
)
V
] ! (
2
T/P
2
)
V

rather
(
2
P/T
2
)
V
(T/P)
2
V
= - (
2
T/P
2
)
V

Integration by Parts.
( ) [ ]

2
1
2
1
2
1
VdP PV PdV


Measurable Thermodynamics Quantities & Their
Definitions

1. C
i
(Q/T)
i
Isobaric Heat Capacity

C
p
= (Q/T)
p
= T (S/T)
p
= (H/T)
p
Isochoric Heat Capacity

C
v
= (Q/T)
v
= T(S/T)
v
= (U/T)
v
Saturated Heat Capacity

C
= (Q/T)
= T(S/T)
= (dH/dT)
- V
(dP
/dT)

2. Compressibilities
Isobaric Compressibility
V
-1
(V/T)
p
= -
-1
(
-
/T)
p
Isothermal Compressibility
K
T
V
-1
(V/P)
T
=
-1
(/P)
T
Isentropic Compressibility
K
s
V
-1
(V/P)
s
=
-1
(/P)
s
= (U
2
s
)
-1
, where U
s
= sonic velocity



3. J oule Thomson and Isothermal Throtling coefficients
(H/T)
p
(T/P)
H
(P/H)
T
= - 1,
where
C
p
(H/T)
p
,

JT
(T/P)
H
the Joule Thomson coefficient

T
(H/P)
T
, the isothermal throttling coefficient.

4. More Relations
C
p
- C
v
= TV
2
/K
T
= TV
2
v
/K
T
,

V
(P/T)
V
, the isochoric slope.
(C
p
/P)
T
= - T(
2
V/T
2
)
p

and (C
v
/V) T
2
v
.K
s
- K
T
= - VT
2
/C
p
.
For a perfect gas (P.G)
U
2
s
= RT
o
,
o
= (C
p
o
/C
v
o
)


Phase Equilibria Models
There are many models to describe fluid phase equilibria these could be classified
according to the type of fluids (hydrocarbons, alcohols, electrolytes, water and other
non-hydrocarbon species), and to the pressure and temperature ranges of interest. In
this session we will describe models that are used in petroleum engineering
applications. These models are for low-pressure ranges, such as those of separator and
surface conditions and models for high pressures which apply to the reservoir. The type
of reservoir fluid, whether a black oil or a volatile oil, also determines the type of model
that can be used.
We will start with the simpler models first, the ones for lower pressures.

Residual Properties
To derive the phase equilibria models we define the residual properties for
mathematical convenience as the difference between the actual (real) property minus
the same property, evaluated at the same pressure, temperature, and composition, but
evaluated using the ideal gas equation. That is

M
R
= M-M
ig
M=U, H, G, S, F (F is A in American Notation)
M: Real Property @ (T, P) of the system
M
R
: Residual Property
M
ig
: Property @ (T, P) of the system evaluated as if the fluid were an ideal gas

Note: there is no T
R
or P
R



Recall for a constant composition closed system
SdT VdP dG (89)
dT S dP V dG
ig ig ig
(90)
dT S dP V dG
R R R
(91)

Note that the properties used in these equations are intensive properties, that is the
volume is the molar volume, G and S are expressed in BTU/lb-mol and BTU/lb-mol-R,
respectively, (or in cal/g-mol, cal/g-mol K in the SI system of units).

At constant temperature,

dP V dG
R R
(92)

we divide by RT

P R G R R R
dP
RT
V
RT
dG
dP
RT
V

RT
dG
R
0 0
(93)

From previous lectures we had:

1 ,
RT
PV
z
RT
PV
ig
(94)

Thus,



P
z
RT
V
R
) 1 (

(95)

and,

( )
P R
P
dP
z-
RT
G
0
1 (96)

Phase Equilibrium of a Single Component
We will start deriving the general expressions for phase equilibrium of a pure
component.

Recall

l v
l l l
v v v
G G
dT dP-S V dG
dT dP-S V dG

(97)

We also know that,

vl vl
l l l
v v v
S -T H
-TS H - G
-TS H G

0

(98)

At saturation, the pressure and temperature of the liquid and gas phases are the same


l
dG dT S dP V dT S dP V dG
l l v v v

(99)

Thus,

vl
vl
l v
l v
V T
H
V V
S S
dT
dP
(100)

This is nothing else but the Clapeyron Equation seen without derivation previously.

Reduced Equations of State and Maxwell Equal Area Rule
If states 1 and 5 in Figure 9 are in equilibrium, then

5 1
G G (101)



-
10
0
0
10
0
20
0
30
0
40
0
50
0
60
0
70
0
2 4 6 8 1
0
1
2
1
4
A1
A2
P
r
e
s
s
u
r
e
Molar Volume
T
c
T
2
T
1
P
1
v
L
2 - Phases
CP
V
L
V
1
2
3
4
7
6
5
0 >
,
_
T
V
~
P
-
10
0
0
10
0
20
0
30
0
40
0
50
0
60
0
70
0
2 4 6 8 1
0
1
2
1
4
A1
A2
P
r
e
s
s
u
r
e
-
10
0
0
10
0
20
0
30
0
40
0
50
0
60
0
70
0
2 4 6 8 1
0
1
2
1
4
A1
A2
P
r
e
s
s
u
r
e
Molar Volume
T
c
T
2
T
1
P
1
v
L
2 - Phases
CP
V
L
V
1
2
3
4
7
6
5
0 >
,
_
T
V
~
P

Figure 9 - Predicted isotherms from a cubic EOS.

Or

5
1
0 dG (102)

For constant temperature,

0 ) ( PdV PV d VdP dG (103)



5
1
5
1
1 1 5 5
0 dV P V P V P dG (104)

At equilibrium P
1
=P
5
=P
5
1
1 5
0 ) ( dV P V V P (105)

By inspection,

) 1 7 6 5 3 1 ( ) (
1 5

Area V V P (106)

And also,

) 1 7 6 5 4 3 2 1 (
5
1

Area PdV (107)

) 3 5 4 3 ( ) 1 2 3 1 ( ) 1 7 6 5 3 1 (
) 1 7 6 5 4 3 2 1 (
+

Area Area Area
Area
(108)

Combining Equations (106), (107) and (108), we get

) 3 5 4 3 ( ) 1 2 3 1 ( Area Area

(109)



The liquid and vapor saturated volumes
l
V V V
1 7
and
g
V V V
6 5
are at a specified
isotherm and at p
.

VLE in Dimensionless or Reduced Form

Write the EOS in dimensionless form using T
r
=T/T
c
, P
r
=P/P
c
, V
r
=V/V
c
, and the values for
a and b found from the critical constraints

0 , 0
2
2
,
_
,
_
c T c T
V
P
V
P
(110)

For Van der Waals EOS

2
V
a
b V
RT
P
(111)

with

c c
V RT a
8
9
(112)
c
c c
P
RT V
b
8 3
(113)

c
c c
c
RT
V P
z
8
3
(114)


c
c
c
P
RT
V
8
3
(115)
2 2
8
9
3
r c
c c
c
c r
c r
c r
V V
V RT
V
V V
T RT
P P
(116)

After rearranging and simplifying

2
3
1 3
8
r r
r
r
V V
T
P
(117)

Equation (117) applies to the gas and liquid. For the liquid, we will use a reduced
volume V
rl
=V
l
/V
c
, and for the gas V
rg
=V
g
/V
c
.

Apply equal area rule

0 dG (118)

At T constant

0 ) ( dV P PV d VdP dG (119)

or

( )

rg
rl
V
V
r r rl rg r
dV P V V P (120)


Replace Equation (117) in the integrand of Equation (120) and integrate

( )
1
1
]
1
+
1
]
1
rl rg rl
rg
r
rl rg r
V V V
V
T
V V P
3 3
1 3
1 3
ln
3
8
(121)

Since P is constant, 0 dP

Thus,
0 dP V

0
r

,
_

rg
rl r
rg
rl
V
V
r
T
r
r
V
V
r r
dV
V
P
V dP V
1
]
1
+
rg
rl
V
V
r
r r
r
dV
V ) - V (
Tr -
V
6
1 3
24

3 2
(122)

From integral tables,

1
]
1
+
+ +
+
bx a
a
bx a
b bx a
xdx
) ln(
1
) (
2
2
(123)

etc.,

0
1 1
4
9
1 3
1
1 3
1
1 3
1 3
ln
,
_
,
_
,
_
,
_
rg rl r rl rg rg
rg
V V T V V V
V
(124)


So, we have three equations to work with,

EOS (117)
Maxwell Equal Area (121) ! unknowns P
r
, V
rl
, V
rg
.
and
0 dP V (124)

In density form
,
_

r
r
V
1

( )( )
0
3 3
1
8
3
3
'
1
1
]
1
rl rg r
rl rg r
T
p (125)

combining
0 dP V and Maxwell equal area

( )( )
0 ) (
3 3
8
+

rg rl
rl rg
r
T
(126)

EOS for liquid and gas.

And Eq. (124) in density form

0
3 3
) (
4
9
3
3
ln
,
_
,
_
+ +
'
,
_

rl
rl
rg
rg
rg rl
r rl
rg
rl
rl
T
(127)


Change of Variables
rl rg
s +
rg rl
w
2
w s
rl
+

2
w s
rg

Write Equations (125), (126) and (127) in terms of s and w.

Advantage 1: Good starting point! At the critical point s = 2 and w = 0. VLE calculation
starts from the critical point and the algorithm steps down in temperature.
Tr =1 ! solution - s = 2 and w = 0
0.99
0.98
0.97
.
0.4
Keep lowering the T
r
until it gets close to the triple point, which in reduced coordinates is
pretty close to 0.4. For each T
r
, you will obtain s and w, which will give provide the
densities for the gas and the liquid.

Advantage 2: Reduced form makes it universal. VLE calculated only once.



Disadvantage: Algebraic manipulations.

Assignment Assignment Assignment Assignment
Read paper Generalized Saturation Properties of PureFluids via Cubic Equations of State
by Barrufet & Eubank Chemical Engineering Education Summer 1989. You can access
this paper through thelink in your Calendar of Events in theWebCT site.
Onederivation will beassigned following this paper plus thesenotes.

Systems of Variable Composition - Mixtures
We will start with the simplified assumption that the gas phase behaves as an Ideal Gas
and the liquid phase exhibits Ideal Solution Behavior.

We have seen in previous lectures that the equilibrium criteria between 2 phases and
was,

c i i
N i
T T
P P
,.... 2 , 1 ,

(128)

where and could be: vapor, liquid 1, liquid 2, solid 1, solid 2, etc.

We will mainly deal with vapor and liquid equilibria: VLE. Thus, at constant T and P,



c
c
n
i
l
i
l
i
l
n
i
v
i
v
i
v
dn nG d
dn nG d
1
1
) (
) (
(129)

The simplest model is to assume that the gas phase behaves as an ideal gas
(IG), and that the liquid phase behaves as an ideal solution (IS). The assumptions imply
that
IG: molecular interactions are zero, molecules have no volume.
IS: forces of attraction/repulsion between molecules are the same regardless of
molecular species. Volumes are additive (Amagats Law).

A A B B A B

Figure 10 - Forces between molecular species.

AB BB AA
F F F

Ideal Gas Mixture



P pk
T1 T1
n1,n2, nk
,nk

Figure 11 - Comparison of the pressure in a constant volume vessel between a mixture
of gases and a pure component.

The pressure in a vessel containing an ideal gas mixture (n) or a single gas component
(n
k
) is
t
k
k
t
V
RT n
P
V
nRT
P
(130)

k
k k
y
n
n
P
P
(131)

P
k
is the partial pressure of component k, and by definition, for ideal behavior

c
N
i
k
P P
1
(132)

Same vessel
volume


Equation (132) can be generalized to any thermodynamic property for an ideal gas
mixture.

c
n
k
k
ig
k k
ig
P T M n P T nM
1
) , ( ) , ( (133)
or,

A total thermodynamic property (nU, nG, nS, nH, nF) of an ideal gas
mixture is the of the total properties of the individual species each
evaluated at the T of the mixture and at its own partial pressure.

To derive the equilibrium relations we are interested in using equation (133) for H
ig
and
S
ig
because,

ig ig ig
TS H G (134)

For an ideal gas, the enthalpy is independent of pressure, thus,

) , ( ) , (
k
ig
k
ig
k
P T H P T H (135)

For the entropy, we must express ) , ( P T S S
ig ig

dT
T
S
dP
p
S
dS
P
T

,
_
,
_
(136)



Recall Maxwell Rules

dT
T
c
dP
T
V
dS
p
P
+
,
_
(137)

For ideal gas,

dT
T
c
dP
P
R
dS
ig
p ig
k
+ (138)

at constant temperature,

P
P
P
P
ig
k
ig
k
k k
P d R dS
P Rd dP
P
R
dS
ln
ln
(139)

k
k
k
ig
k
ig
k
y R
P
P
R P T S P T S ln ln ) , ( ) , (
,
_
(140)

We also know from equation (133) applied to entropy,

c
N
k
k
ig
k k
ig
P T S n P T nS
1
) , ( ) , ( (141)
or


c
N
k
k
ig
k k
ig
P T S y P T nS
1
) , ( ) , ( (142)

Substitute ) , (
k
ig
k
P T S from equation (140) into equation (142) after over the
components,

c c
N
k
k k
N
k
k
ig
k k
ig
y y R P T S y P T S
1 1
ln ) , ( ) , ( (143)

which means that the entropy change of mixing the ideal gases is not zero, and it is
greater than zero,

>
c c
N
k k
k
N
k
ig
k k
ig
y
y R S y S
1 1
0
1
ln (144)

Now, we can build the expression for the Gibbs energy using equations (135) and (144).

+
c c c
N
k
k k
N
k
ig
k k
N
k
ig
k k
ig
y y RT P T S y T P T H y P T G
1 1 1
ln ) , ( ) , ( ) , ( (145)

Remember that expression (145) is for an ideal gas.

Now, recall the expression for the chemical potential,



i j
n P T
i
i
n
nG
,
_

, ,
(146)

Using equation (145) expressed in terms of n (y
k
=n
k
/n),

+
c c
N
k
k k
N
k
ig
k
k ig
n
n
n
n
RT G
n
n
P T G
1 1
ln ) , ( (147)

1
]
1
+

c c c
N
i
n
i
k k k
N
i
ig
k k
ig
n n n n RT G n P T nG
1 1 1
ln ) ( ln ) , ( (148)

Recall,
c
N
i
i
n n
1
(149)
k j
n
n
k j
n
n
n
n
j
k
j
k
k

, 1
, 0
1
(150)

Therefore,

1
]
1
+ +
,
_
n
n
n
n
n
n RT G
n
nG
k
i
i
i
ig
i
n p T
i
ig
ig
i
i j
ln ln
, ,
(151)



i
ig
i
ig
i
y RT G ln + (152)

Keep in mind equation(152) since it will be used later in the development of Raoults
Law for phase equilibria of IG+IS.

So far, we have worked with an ideal mixture of gases. Now, we will work with an ideal
solution of liquids.

Ideal Solution
We have seen that Amagats Law is followed.

i i
id
V x V @ same T and P of mixture (153)

Following the same reasoning as for gases, we have that,

i i i
id
x R V x S (154)
i i i i
id
x x RT G x G ln

+ (155)
i i
id
i
RTx G + (156)

Here, S
i
and G
i
are the properties of the pure species in the liquid state at the T and P of
the mixture.



Raoults Law
It is a combination of IG + IS models. If we consider VLE for a mixture made up of N
c

components

c
id l
i
ig v
i
l
i
v
i
N i ,..., 1 ; ) ( ) ( (157)

Thus, at T and P,

i
l
i i
ig
i
x RT G y RT G ln ln + + (158)

) , ( ) , ( ln P T G P T G
x
y
RT
ig
i
l
i
i
i
(159)
The right hand side of Eq. (159) indicates pure species properties evaluated at the
equilibrium T and P of the mixture

More simplifications
Assume negligible effect of pressure on G
i
l
, (nearly incompressible fluid far away from
the critical point).

) , ( ) , (

i
l
i
l
i
P T G P T G (160)

where P
i
is the pure species saturation pressure at T.

For an ideal gas, we have,


i i i
P
P
P
P
ig
i
P
P
ig
i
dP
P
RT
dP V dG @ T constant (161)

P
P
RT P T G P T G
i
ig
i i
ig
i
ln ) , ( ) , ( (162)

Now, combine Eqs. (159) to (162),

P
P
RT P T G P T G
x
y
RT
i
i
ig
i i
l
i
i
i

+ ln ) , ( ) , ( ln (163)

As we seen before for a pure component,

0 ) , ( ) , (

i
ig
i i
l
i
P T G P T G (164)

So, Eq. (163) leads to Eq (165), which is known as Raoults Law.

i i i
P x P y (165)

This equation will be used (as a warm-up) to evaluate all the phase equilibrium
computations (dew point, bubble point, and flashes).

Equation (165)must hold for all species if equilibrium exists. Its validity is for low
pressures (P < 100 psia) and high temperatures (T > 70F).


Types of Calculations
Given Variables
(independent)
Unknown Variables
(dependent)
Problem Type
Example
Application
P, z
i
= x
i
T, y
i
Bubble Point
T, z
i
= x
i
P,y
i
Bubble Point
Gas injection,
production
P, z
i
= y
i
T,x
i
Dew Point
T, z
i
= y
i
P,y
i
Dew Point
Gas
Condensates,
Production
P, T, z
i
x
i
, y
i
, f
v
Flash
Production
Separation

Bubble Point Evaluation given T, z
i
and
i
P
We need a model for each P
i
as a function of temperature.

i i
x z (166)

Find BP pressure and equilibrium gas compositions

i i i
i i i
P z P y
P x P y
or (167)


The bubble point pressure at a given T is

i i
bp
i
P z P y (168)

i i bp
P z P (169)

Under Raoults law, the bubble point has a linear dependence with the vapor pressures
of the pure components.
Once the bubble point pressure is found, the equilibrium vapor compositions are found
from Eq. (167).
The dew point curve (lower black curve) in Figure 12 is always curved regardless
whether the mixture is ideal or not.

P
2
P
1
T
x
1
,y
1
Figure 12 - Deviations from Raoult's law.

The red curves in Figure 12 indicate deviations from Raoult's law. When the bubble
point curve is above the straight line, we will have positive deviations from Raoult's Law.


When the bubble point curve is below the straight line, we will have negative deviations
from Raoult's Law. This happens for non-ideal mixtures and may lead to azeotropy.

Dew Point Calculation given T, z
i
and
i
P
At the dew point the overall fluid composition coincides with the gas composition. That
is.

i i
y z (170)

Find DP pressure and equilibrium liquid compositions
i i i
i i i
P x P z
P x P y
(171)

P
x
P
z
i
i
i
(172)
1
1
1
]
1

c
N
i
i
i
dp
P
z
P (173)

Once P
dp
is found, the equilibrium liquid compositions are found from Eq. (171).

Bubble Point Temperature Given P, z
i
and
i
P as a function of T
Here we need to find the corresponding bubble point temperature which enters into the
equation indirectly and we must follow an iterative procedure.

i i
x z (174)


Find TB pressure and equilibrium gas compositions
i i i
i i i
P z P y
P x P y
or (175)

The problem is that we do not know yet at what temperature to evaluate the pure
component vapor pressures. See the following diagram

) (
bp i i bp
T P z P (176)

For well-behaved systems (no azeotropes), the searched temperature will be bounded
by the highest and lowest saturation temperature of the components in the mixture at
the selected system pressure, as shown in Figure 13.

T
2
T
1
P
x
1
,y
1
P

Figure 13 - Binding of the searched temperature in a well-behaved system.

For demonstration purposes, lets select a binary and a model such as Antoines for
vapor pressures.


Procedure
1. Evaluate
1
T and
2
T at the given pressure P, which is a saturation pressure.
i i
i
i i
c T
b
a P
+

ln "
i
i i
i
i
c
P a
b
T +
ln
(177)

2. Choose your first guess bubble point temperature as
i
n
i
i
o
bp
T z T
c
1
) (
(178)
using the overall composition of the mixture and the saturation temperatures evaluated
in (177).

3. Define a relative volatility using a reference substance such that all relative volatilities
are either > 0 or < 0 (i.e., monotonically increasing or decreasing).
j
i
ij
P
P
(179)
with the saturation pressures evaluated at the guess temperature evaluated in (178).

4. Expand the volatility as
2
2
1
1
2 1 2 1 12
ln ln ln
c T
b
c T
b
a a P P
+
+
+

(180)
1
]
1
+
+
+

2
2
1
1
2 1 12
exp
c T
b
c T
b
a a (181)
with T from Eq. (177).



Then write the bubble point equation in terms of the volatilities and the reference vapor
pressure for the selected component.

For a binary, you would have only one volatility and this expression would be,

[ ]
2 12 1 2 2
2
1
1 2 2 2 1 1
z z P z
P
P
z P P z P z P +
1
]
1
+ +

(182)

Thus,

[ ]
2 12 1
2
z z
P
P
+
(183)

this is your first guessed saturation pressure for the reference component (here 2) at
the first guessed temperature evaluated in (177).

From the saturation pressure evaluated in (182) use the Antoine equation to find a new
temperature (Eq. (177)).

This new T " new
12
" new
2
P " iterate until two successive temperatures do not
change by a specified tolerance.
The Excel file provided in our WEB site illustrates this procedure for a ternary mixture.
You can modify it and extend it to multicomponents.



Dew Point Temperature given P, z
i
and
i
P as a function of T
You can follow a very similar reasoning as the one developed for the bubble point and
devise the algorithm required to solve this problem using relative volatilities.

Flash Calculations
In this type of calculations, the work-horse of reservoir simulation packages,the
objective is to find fraction of vapor vaporized as well an equilibrium gas and liquid
compositions given the overall mixture composition, P and T.
Start with the same equation as usual

i i i
P x P y (184)

Material balance

( )
v i v i v i l i i
f y f x f y f x z + + 1 (185)

Now replace either liquid or gas compositions using equation (184) in equation (185)

( )
v i v
i
i i
f y f
P
P
y z +
1 (186)

( )
v v
i
i
i
f f
P
P
z
y
+
1
(187)



( )

1
1
1
1
]
1
v v
i
i
i
f f
P
P
z
y
1
(188)

z
i
(T
1
,P
1
)
x
i
(T
1
,P
2
)
y
i
(T
1
,P
2
)
P
1
> P
2
T
1
,P
2

Figure 14 - Separation process.

Objective function (flash function) is

( )
0 1
1
) (
1
1
1
1
]
1
v v
i
i
v
f f
P
P
z
f F (189)

There are several equivalent expressions for the flash function

(a) 0 1
i
y (190)
(b) 0 1
i
x (191)


(c) 0
i i
x y " the best well behaved for numerical solution (Rachford- Rice
function) (192)

Once f
v
is found the equilibrium gas and liquid compositions are evaluated from Eqs.
(189) and (186), respectively.

Systems of Variable Composition - Non-Ideal
Behavior
Here, we will lay the foundation for a general treatment of VLE, LLE, SLE, etc. by
introducing an auxiliary thermodynamic function related to the Gibbss energy, this is
called the fugacity coefficient.
We have seen that

i
n p T
i
i
G
n
nG
i j
,
_
, ,
Partial Molar Gibbs Energy
In general
i j
n p T
i
i
n
nM
M
,
_
, ,
M=U, H, A, G, S, V

Definition of Fugacity and Fugacity Coefficient
For a constant composition fluid at constant T

SdT VdP dG (193)



P
P
P
P
VdP dG
* *
(194)

So, V P , 0
*

P
P
VdP P G P G
*
) ( ) (
*
(195)

Need a function that behaves better and is equivalent.

Recall for an IG @ T=constant
P RTd
P
dP
RT dP V dG
ig ig
ln (196)

In a similar fashion, we define,

f RTd dG ln (197)

Then,

ln ln RTd
P
f
RTd dG dG
ig
with
P
f
(198)

ln RTd dG
R
(199)



Integration provides,

) ( ln T C RTd dG
R
+ (200)

Lets complete the definition of the fugacity by setting the limit of the fugacity of an ideal
gas equal to its pressure,

0 ) ( C and 0 1 T G P f
R ig
(201)

General Expressions
ln
RT
G
R
Applies to a gas mixture and is for mixture. (202)

Similarly,

P
f
RT
G
i
i
i
ln ln pure component (203)

We also saw that,

P R
P
dP
z
RT
G
0
) 1 ( (204)

P
P
dP
z
0
) 1 ( ln (205)


Similarly, for a pure component,

P
i i
P
dP
z
0
) 1 ( ln z
i
= compressiblity factor of component I (206)

Redefining equilibrium for a pure species @
i
P P

v
i
l
i
f
f
i
v
l
i
f RTd dG ln (207)

l
i
v
i l
i
v
i
f
f
RT G G ln (208)

using

i
i
i
P
f
(209)

l
i
v
i l
i
v
i
RT G G
ln (210)

So, at equilibrium,

l
i
v
i
needs models to evaluate this expression (211)



Fugacity for species i in a Mixture
Analogous to previous equations, we can derive a fugacity for mixtures from,

i
ig
i
ig
i
i
ig
i
ig
i
f RTd d G d
Py RTd d G d
ln ) (
ln ) (

(212)

Integration provides,

i i
f RT
ln (213)

For VLE,

l
i
v
i
f f

i = 1, 2,,Nc (214)

where,

P y f
i
v
i
v
i

(215)
P x f
i
l
i
l
i

(216)

Thus, the General VLE problem becomes

i
l
i i
v
i
x P y P

(217)



what results in,

i
l
i i
v
i
x y

(218)

To find the fugacity coefficients there are several models available. For hydrocarbon
fluids we shall use EOS.

Thermodynamics 2

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Thermodynamics 2

Hochgeladen von

Copyright:

Verfügbare Formate

Generalized Phase Equilibria Models, Colombia, Summer 2000 -

Author: Dr. Maria Barrufet - Summer, 2000 Page 1/77

Area PdV (107)

0 dP V and Maxwell equal area

is the pure species saturation pressure at T.

Das könnte Ihnen auch gefallen