Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
PETROLEUM ENGINEERING DEPARTMENT
TEXTBOOK SERIES
VOLUME 2
RESERVOIR FLUIDS
by
Zoltán E. HEINEMANN
Professor for Reservoir Engineering
Brigitte E. WEINHARDT
Assosiate Professor for Reservoir Engineering
Leoben, October 2004
© No part of this publication may be reproduced in any form.
Only students of the University of Leoben may copy for studying purposes.
Edited by DI Claudia Scharf  October 2004
Table of Contents
Chapter 1
Review of Thermodynamic Terminology ........................................................ 1
Chapter 2
Phase Behavior ................................................................................................... 3
2.1 Gibbs’ Phase Rule ........................................................................................................3
2.2 SingleComponent System ..........................................................................................4
2.2.1 Water .............................................................................................................4
2.2.2 nButane .........................................................................................................7
2.3 Critical State and Quantities of Corresponding States ...............................................11
2.4 Binary Systems ..........................................................................................................13
2.5 MultiComponent Systems ........................................................................................20
2.5.1 Ternary Phase Diagrams ..............................................................................20
Chapter 3
Equations of State ............................................................................................ 31
3.1 Change of State at Low Compressibility ...................................................................32
3.2 Equation of State of Perfect and Real Gases .............................................................33
3.3 Cubic Equations of State ............................................................................................35
3.4 Virial Equation of State .............................................................................................47
Chapter 4
Calculation of Phase Equilibria ...................................................................... 49
4.1 Mixtures .....................................................................................................................49
4.1.1 Definitions ...................................................................................................49
4.1.2 Kfactors ......................................................................................................50
4.2 Composition of Phases in Equilibrium ......................................................................54
4.2.1 Definitions ...................................................................................................54
4.2.2 Evaluation of KFactors Using Convergence Pressures .............................62
4.2.3 Evaluation of Convergence Pressure ...........................................................68
4.2.4 Evaluation of by use of PengRobinson equation of state ..........................69
Chapter 5
Phase Properties ............................................................................................... 81
5.1 Natural Gases .............................................................................................................81
5.1.1 Volume ........................................................................................................81
ii Table of Contents
5.1.2 Formation Volume Factor ........................................................................... 83
5.1.3 Compressibility ........................................................................................... 84
5.1.4 Correlation of ZFactor on the Basis of Reduced State Variables .............. 84
5.1.5 Water Content .............................................................................................. 93
5.1.6 Viscosity ...................................................................................................... 94
5.2 Hydrocarbon Liquids ................................................................................................. 98
5.2.1 Volume ........................................................................................................ 98
5.2.2 Formation Volume Factor ......................................................................... 105
5.2.3 Compressibility of Undersaturated Liquids .............................................. 110
5.2.4 Viscosity .................................................................................................... 114
5.3 Brines ....................................................................................................................... 117
5.3.1 Composition of Brines ............................................................................... 118
5.3.2 Formation Volume Factor ......................................................................... 121
5.3.3 Viscosity .................................................................................................... 126
5.3.4 Natural Gas Hydrates ................................................................................ 127
Chapter 6
pVTMeasurements ....................................................................................... 133
6.1 Sampling .................................................................................................................. 133
6.1.1 Objectives .................................................................................................. 133
6.1.2 General Criteria ......................................................................................... 133
6.1.3 Sampling Methods ..................................................................................... 134
6.1.4 Special Problems ....................................................................................... 137
6.2 Experimental Determination of the Volumetric and Phase Behavior ..................... 138
6.2.1 Equipment ................................................................................................. 138
6.2.2 PVTCells .................................................................................................. 139
6.2.3 Volumetric Pumps ..................................................................................... 144
6.2.4 Auxiliary Equipment ................................................................................. 144
6.3 Methods ................................................................................................................... 145
6.3.1 Flash Process ............................................................................................. 145
6.3.2 Differential Process ................................................................................... 146
6.3.3 Reverse Differential Process ..................................................................... 146
List of Figures
Figure 2.1: Water system  schematic (not drawn to scale)4
Figure 2.2: Phase equilibrium surface of a pure substance (from Gyulay, 1967)5
Figure 2.3: Vapor pressure diagram of nbutane (from Gyulay, 1967)7
Figure 2.4: Pressure  volume phase diagram of nbutane (from GYULAY, 1967)8
Figure 2.5: Temperature  density phase diagram of nbutane (from GYULAY, 1967)9
Figure 2.6: Critical pressure as a function of number of Catoms in homologous series (after
Gyulay, 1967)11
Figure 2.7: Critical temperature as a function of numbers of Catoms in homologous series (after
Gyulay, 1967)12
Figure 2.8: Combined reduced pressure  reduced volume phase diagram of paraffins with low
molecular weight (after Gyulay, 1967)14
Figure 2.9: Phase equilibrium surface of the binary system ethane/nheptane (from Gyulay, 1967)14
Figure 2.10: Pressure  temperature phase diagram of the binary system ethane/nheptane (from Kay,
1938)15
Figure 2.11: Pressure  temperature phase diagram of the binary system ethane
(z = 0.9683)/nheptane17
Figure 2.12: Mole fraction(ethane)  temperature diagram of the binary system ethane/nheptane
(from Gyulay, 1967)17
Figure 2.13: Mole fraction (ethane)  Pressure diagram of the binary system ethane/nheptane (from
Gyulay, 1967)18
Figure 2.14: Properties of ternary diagrams20
Figure 2.15: Typical features of a ternary phase diagram21
Figure 2.16: Triangular diagrams for the methane/propane/npentane system at 160
o
F(71
o
C) (after
Dourson et al., 1943)23
Figure 2.17: Critical loci of methane/propane/npentane systems (from Katz et al., 1959)24
Figure 2.18: Phase diagram of a dry gas (from McCain, 1973)26
Figure 2.19: Phase diagram of a wet gas (from McCain, 1973)26
Figure 2.20: Phase diagram of a retrograde gas condensate (from McCain, 1973)26
Figure 2.21: Phase diagram of a highshrinkage crude oil (from McCain, 1973)27
Figure 2.22: Phase diagram of a low shrinkage crude oil (from McCain, 1973)27
Figure 2.23: Phase diagram pairs of gas cap and oil zone28
Figure 2.24: Phase equilibrium surface of oil/natural gas systems (from Gyulay, 1967)29
Figure 3.1: The Van der Waals isotherms near the critical point35
Figure 4.1: Fugacity of natural gases (from Brown, 1945)52
Figure 4.2: Ideal and real Kfactors of nbutane at 60[
o
C]52
Figure 4.3: Flash and differential vaporization59
Figure 4.4: Kfactors for methanepropane at T
c
= 100
o
F (from Sage, Lacey and Schaafsma) (1934)
63
Figure 4.5: Comparison of Kfactors at 100
o
F for 1,000 and 5,000psia convergence pressure (from
NGAA, 1957)64
Figure 4.6: Kfactors for methane, 5,000 psia convergence pressure (from NGAA, 1957)66
Figure 4.7: Kfactors for hexane, 5,000 psia convergence pressure (from NGAA, 1957)67
Figure 4.8: Convergence pressure data  methane for binary hydrocarbon mixtures (from Winn,
1952)70
Figure 5.1: Zfactor of methane, ethane and propane versus pressure at T = 140oF (from Standing,
1977)83
iv List of Figures
Figure 5.2: Zfactor as a function of reduced pressure for a series of reduced temperatures (from
Sage and Lacey, 1949)85
Figure 5.3: Zfactor for natural gases (from Brown et al., 1948) .................................................... 87
Figure 5.4: Pseudocritical temperatures and pressures for heptanes and heavier
(from Matthews et al, 1942)88
Figure 5.5: Pseudocritical properties of Oklahoma City Gases
(from Matthews et al., 1942)88
Figure 5.6: Water content of natural gas in equilibrium with liquid water
(from Katz et al., 1959)92
Figure 5.7: Viscosity of paraffin hydrocarbon gases natural gases at atmospheric pressure (from
Carr et al., 1954)95
Figure 5.8: Viscosity of natural gases at atmospheric pressure (from Carr et al, 1954) ................. 96
Figure 5.9: Correlation of viscosity ratio with pseudoreduced pressure and temperature (from Carr
et al., 1954)96
Figure 5.10: Variation of apparent density of methane and ethane with density of the system (from
Standing and Katz, 1942)99
Figure 5.11: Pseudoliquid density of systems containing methane and ethane
(from Standing, 1952)100
Figure 5.12: Density correction for compressibility of liquids (from Standing, 1952)................... 101
Figure 5.13: Density correction for thermal expansion of liquids
(from Standing, 1952)102
Figure 5.14: Apparent liquid density of natural gases in various API gravity oils
(from Katz, 1952)104
Figure 5.15: Typical graph of formationvolume factor of oil against pressure ............................ 106
Figure 5.16: Pseudoreduced compressibility for undersaturated reservoir fluids
(from Trube, 1957)112
Figure 5.17: Pseudocritical conditions of undersaturated reservoir liquids
(from Trube, 1957)112
Figure 5.18: Viscosity of subsurface samples of crude oil (from Hocott and Buckley, 1941, after Beal,
1946)114
Figure 5.19: Viscosity of gassaturated reservoir crude oils at reservoir conditions
(from Chew and Connally, 1959)115
Figure 5.20: Prediction of crude oil viscosity above bubble point pressure
(from Beal, 1946)116
Figure 5.21: Essential feature of the water pattern analysis system (from Stiff, 1951) .................. 120
Figure 5.22: Course of Arbuckle formation through Kansas shown by water patterns (from Stiff,
1951)120
Figure 5.23: Solubility of natural gas in water (from Dodson and Standing, 1944) ....................... 122
Figure 5.24: Typical graph of formation volume factor of water against pressure......................... 122
Figure 5.25: B
w
for pure water (dashed lines) and pure water saturated with natural gas (solid lines)
as a function of pressure and temperature (from Dodson and Standing, 1944)123
Figure 5.26: Density of brine as a function of total dissolved solids
(from McCain, 1973)124
Figure 5.27: The isothermal coefficient of compressibility of pure water, including effects of gas in
solution (from Dodson and Standing, 1944)125
Figure 5.28: The viscosity of water at oil field temperature and pressure
(from van Wingen, 1950)127
Figure 5.29: Hydrate portion of the phase diagram for a typical mixture of water and a light
List of Figures v
hydrocarbon (from McCain, 1973)129
Figure 5.30: Pressuretemperature curves for predicting hydrate formation
(from Katz, 1945)129
Figure 5.31: Depression of hydrate formation temperature by inhibitors
(from Katz et al., 1959)130
Figure 5.32: Permissible expansion of 0.8 gravity gas without hydrate formation
(from Katz, 1945).131
Figure 6.1: Scheme of PVT equipments......................................................................................... 138
Figure 6.2: Blind PVT cell ............................................................................................................. 140
Figure 6.3: PVT cell (after Burnett) ............................................................................................... 140
Figure 6.4: PVT cell (after DeanPoettman) .................................................................................141
Figure 6.5: Variable volume cell (after Velokivskiy et al.) ...........................................................142
Figure 6.6: PVT cell (after Sloan) .................................................................................................. 142
Figure 6.7: PVT cell (after WellsRoof)......................................................................................... 143
Figure 6.8: Ruska cell .................................................................................................................... 143
Figure 6.9: Ruska volumetric mercury pump ................................................................................144
1
Chapter 1
Review of Thermodynamic Terminology
When considering hydrocarbon reservoirs, terms such as “oil reservoirs” and “gas
reservoirs” are used both in colloquial speech and technical literature. However, these
terms are insufficient. Changes in the state of aggregation during production should
always be taken into account in consequence of changes of the reservoir pressure and
changes of pressure and temperature in the production system (tubing, pipe lines,
separator, tank).
Thermodynamics has evolved to a science of studying changes in the state of a system
with changes in the conditions, i.e. temperature, pressure, composition. A systematic
presentation of basic thermodynamic tools (charts, tables and equations) for sketching the
state of a hydrocarbon system as a function of the state variables is one of the objectives
of this textbook. Therefore, it may be helpful to refurbish the thermodynamic terminology
at the beginning as far as it is indispensable to the understanding of these tools..
Thermodynamic studies are generally focused on arbitrarily chosen systems while the rest
of the universe is assumed as “surroundings”. The surface of the system  real or
imaginary  is called a “boundary”. A system is called a “closed system” if it does not
exchange matter with the surroundings, in opposite to an “open system” which exchanges
matter with the surroundings. Both systems may exchange energy with the surroundings.
The concept of a closed system is of major interest in applied hydrocarbon
thermodynamics. It is called a “homogeneous” closed system if it contains a single phase,
e.g. a natural gas phase or an oil phase. A “heterogeneous” closed system contains more
than one phase.
A “phase” is defined as a physically homogeneous portion of matter. The phases of a
heterogeneous system are separated by interfaces and are optically distinguishable. It is
not obligatory that a phase is chemically homogeneous. It may consists of several
compounds, e.g. of a large number of various hydrocarbons.
The thermodynamic properties of a system are classified into “intensive” and “extensive”
properties. Intensive properties such as temperature and pressure are independent of the
size of the system, i.e. of the amount of the substance in the system. Extensive properties
depend on the amount of the substances, such as volume, enthalpy, entropy etc. However,
2 Review of Thermodynamic Terminology
the extensive properties per unit mass or mole are intensive properties, e.g. the mole
volume.
“State functions” or “state variables” are those properties for which the change in state
only depends on the respective initial and final state. It is this pathindependent
characteristic of the state functions that makes it possible to quantify any change of a
system.
“Equilibrium” has been defined as a “state of rest”. In an equilibrium state, no further
change or  more precisely  no netflux will take place unless one or more properties of
the system are altered. On the other side, a system changes until it reaches its equilibrium
state.. Any change of a system is called a “thermodynamic interest” in the thermodynamic
study of the system:
• adiabatic (no heat added to or removed from the system),
• isothermal (constant temperature),
• isobaric (constant pressure),
• isochoric (constant volume).
A process is called “reversible” if it proceeds through a series of equilibrium states in such
a way that the work done by forward change along the path is identical to the work
attained from the backward change along the same path. However, all real processes are
“irreversible” with varying degrees of departure from a reversible one.
3
Chapter 2
Phase Behavior
Hydrocarbon reservoirs consist of rock and fluids. Water in brine form and a gaseous
and/or liquid hydrocarbon phase are regarded as reservoir fluids. The phase behavior of
the actual hydrocarbon mixture in the reservoir can be described as a function of the state
of the system.
A system in thermodynamic equilibrium posesses an accurately defined relationship
between the state variables. These are united in the socalled “equation of state”:
. (2.1)
By specification of two variables, the third will be stipulated.
2.1 GIBBS’ Phase Rule
When referring to the number of phases coexisting in the thermodynamical equlibrium,
the phase rule introduced by GIBBS (1928) is applied.
(2.2)
where
• P: number of phases,
• C: number of components,
• F: number of degrees of freedom.
C is defined as the smallest number of constituents by which the coexisting phases can be
completely described. F is defined as the number of quantities such as pressure,
temperature, concentrations which can be varied within finite boundaries without
changing the number of phases in the system.
F p V T , , ( ) 0 =
F C P – 2 + =
4 Phase Behavior
Eq. 2.2 describes the system in a qualitative and very general manner. However, no
reference to the state variables (p,T), to the composition of the particular phases or to the
proportions of the phases are given.
To gain a full understanding, it is best to discuss the phase behavior of pure substances
(singlecomponent systems) first. The circumstances in case of 2 or even
multicomponent systems are much more complicated.
2.2 SingleComponent System
2.2.1 Water
Figure 2.1: Water system  schematic (not drawn to scale)
Water is one of the most thoroughly studied chemical compounds. Therefore, it is
discussed as a singlecomponent system in this context. The possible phases are ice (solid
state), water (liquid state) and steam (gaseous state). The phase diagram in Figure 2.1
illustrates at which state of the system  charaterized by p and T  two or all three phases
are in equilibrium:
• The sublimation curve OA signifies the equilibrium between the solid and vapor.
• The melting point curve OB combines the states of equilibrium between the solid
22.09
0.01
0.0006
0.0075 100 374
Temperature [°C]
P
r
e
s
s
u
r
e
[
M
P
a
]
A
B
C
0
Ice
Water
Vapor
Phase Behavior 5
and liquid state.
• The vapor pressure curve OC specifies the states of the system at which the liquid
and vapor coexist. On this curve, the "wet" vapor is in equlibrium with the "saturated"
liquid.
• At the triple point O, all three phases are in eqilibrium. In case of water, the
thermodynamical data at this point are p = 610.6 Pa and T = 273.16 K.
• The end point C of the vapor pressure curve is the critical point and signifies the
highest temperature and pressure at which the liqiud and vapor coexist (p
c
(H
2
O) =
22.09 MPa, T
c
(H
2
O)= 647.15 K.
Example 2.1
The degree(s) of freedom in different states of a singlecomponent
system. by use of Eq. 2.2 (GIBBS’ phase rule) and Figure 2.1:
• F = 0 at the tripel point (P = 3).
• F = 1 on the curves describing the 2phase (P = 2) equlibria.
Either the temperature or the pressure is freely eligible
without counteracting any given phase equlibrium.
• F = 2 in any area of single phase state (P = 1. Both pressure
and temperature (naturally inside finite boundaries) are freely
eligible without transforming the system into a multiphase
system.
Figure 2.2: Phase equilibrium surface of a pure substance (from GYULAY, 1967)
A
A
A
A
B
B
B
B
C
C
C
C
D
D
D
D
E
E
E
E
v
v
v
p
p
p
T
T
T
6 Phase Behavior
The state variables, p, T, V can only assume positive values. Thus, the graphical
illustration of the state of any system is only situated in the positive section of the p,V,
Tcoordinate system An example for an equilibrium surface is giben in Figure 2.2. The
shape of such an equilibrium surface is substance specific.
Assuming that the partial derivatives are steady, it is possible to draw only one single
tangential plane at an optional point of the surface. However, a plane is defined by two
vectors which infers that the differential quotients
are not independent of one another. On the basis of
which describe a change of a system’s state by changing pressure and temperature at
constant specific volume, it is proven that
(2.3)
The three differential quotients describe three essential fundamental properties of the
system: (i) the isothermal compressibility , (ii) the cubic expansion coefficient a, and
(iii) the pressure coefficient :
, (2.4)
(2.5)
(2.6)
Then, according to Eq. 2.3, the following is valid:
(2.7)
Since the specific volume  in contrast to the representation in Figure 2.1  appears now
as a state variable, the 2phase state (e.g. water in equilibrium with steam) is characterized
by any area surrounded by two curves which converge at the critical point:
• On the bubble point curve, an infinitesimal small amount of vapor is in equilibrium
with the “saturated” liquid.
• The dew point curve characterizes states in which a negligible small amount of liquid
∂p
∂T

¸ ,
¸ _
V
∂V
∂T

¸ ,
¸ _
p
∂V
∂p

¸ ,
¸ _
T
, ,
∂p
∂T

¸ ,
¸ _
V
∂V
∂T

¸ ,
¸ _
p
∂V
∂p

¸ ,
¸ _
T

– =
κ
β
κ
1
V

–
¸ ,
¸ _
∂V
∂p

¸ ,
¸ _
T
=
α
1
V

∂V
∂T

¸ ,
¸ _
p
=
β
1
p

∂p
∂T

¸ ,
¸ _
V
=
α pκβ =
Phase Behavior 7
is in equilibrium with “wet” vapor.
It is common to simplify the complex spatial illustration of the equilibrium surface by
applying normal projections. Figure 2.2 displays that
• the projection into the p, Vplane results in isotherms (T = const),
• the projection into the V, Tplane results in isobares (p = const),
• the projection into the p, Tplane results in isochores (V = const).
When regarding the projection which represents the 2phase area (liquidvapor) in the p,
Tplane, the bubble point curve and dew point curve coincide. The resulting single curve
is named vapor pressure curve.
Of course, the vapor pressure curve is not isochoric. However, it is possible to draw
isochores: One upwards into the liquid phase and one downwards into the gas phase. This
aspect will be described in detail by discussing the phase behavior of the simple
hydrocarbon nbutane.
2.2.2 nButane
Projections of the equilibrium surface into two planes of the positive section of the p,V,T
coordinate system are displayed in Figure 2.3 and Figure 2.4.
Figure 2.3: Vapor pressure diagram of nbutane (from GYULAY, 1967)
50 100 150 200
Temperature [°C]
P
r
e
s
s
u
r
e
[
M
P
a
]
0
1
2
3
4
5
6
V
=
0
.
5
m
/M
o
l
3
V
=
0
.
0
5
m
/
M
o
l
3
Liquid
F
A
B D
H
C
E
Vapor
G
8 Phase Behavior
Figure 2.3 illustrates the vapor pressure curve of nbutane, including the critical point C.
In addition, the isochores V = 0,05 m
3
/kmoleinside the liquid phase region and V = 0.5
m
3
/kmole inside the vapor region are shown. In case of the state A, butane is an
undersaturated liquid. When moving to the bubble point B by isothermal expansion,
vaporization commences. Then the continuation of this isothermal expansion includes no
further pressure drop in the system until the last molecule has passed over to the gas phase.
From this moment, further expansion will result in further pressure decrease. At the point
E, nbutane is in the state of a “dry” vapor. An isochoric change of state must be
analogously discussed.
By applying the projection of the equilibrium surface into the p, Vplane (see Figure 2.4),
it is possible to comprise the whole 2phase region. In this region, the isothermal
vaporization or condensation takes place as an isobaric process.
Figure 2.4: Pressure  volume phase diagram of nbutane (from GYULAY, 1967)
Isotherms, which do not intersect the 2phase region, describe those states of the system
without any phase transformation by changing the pressure. The intersection point of all
other isotherms with the bubble point curve (e.g. A) marks the specific volume of the
saturated liquid which is in phase equilibrium with the specific volume of the wet vapor
(e.g. marked by point B). Considering point D within the 2phase region of the system
(specific volume of the system in total), the mass ratio of the liquid and vapor phase being
in equilibrium with one another can be calculated by the principle of the lever:
(2.8)
A D B
130
140
150
C
B
P
D
P Liquid
Vapor
Liquid and Vapor
1
5
5
1
6
0
1
7
0
°
C
1
5
2
.
8
0 0.005 0.010 0.015
V [m/kg]
3
p
[
M
P
a
]
2
3
4
5
m
L
m
v
⁄ DB/AD =
Phase Behavior 9
Example 2.2
100 kg nbutane are filled up in a sealed 10 m
3
container. The
volume of the vapor phase at T = 130°C and p = 2.7 MPa can be
evaluated from Figure 2.4 using Eq. 2.8
From Figure 2.4:
,
m
L
+ m
v
= 100 = 0.3125 m
v
+ m
v
m
v
= 76.19 kg.
The specific volume of the vapor phase is marked by point B in
Figure 2.4:
V = 0.0125 m
3
/kg.
The vapor volume of the system, V
v,
can be now calculated by
multiplying V with m
v
:
V
v
= 0.95 m
3
.
Figure 2.5: Temperature  density phase diagram of nbutane (from GYULAY, 1967)
Figure 2.5 demonstrates the T, diagram of nbutane. The isobare touching the critical
point has an inflection point just as the critical isotherm in Figure 2.4. Inside the 2phase
region, average values of fluid and vapor density are located on a straight line. With the
help of this rule (CAILLETETMATHIAS rule (1886)), the critical density can be calculated
m
L
m
v
⁄ DB AD ⁄ 0.3125 = =
Liquid
D
P
B
P
0 200 400
100
150
200
T
[
°
C
]
ρ [kg/m]
3
Vapor
C
Liquid and Vapor
5
M
P
a
ς
10 Phase Behavior
by extrapolation.
Example 2.3
Use the CAILLETETMATHIAS rule to evaluate the critical density of
methane. The densities of the liquid and the vapor phase being in
equilibrium have been measured at different temperatures (see
table below). The values of averaged densities have already been
calculated.
It is known as the CAILLETETMATHIAS rule that the averaged densities are situated on a
straight line. The slope of a straight line can be evaluated by regression analysis. On the
basis of the averaged densities given above:
.
To evaluate the critical density, , the line must be
extrapolated to the critical temperature of methane, :
,
.
Table 2.1:
Temperature
T
o
C
Liquid Density
kg m
3
Vapor Density
kg m
3
Averaged Density
kg m
3
 158.3
 148.3
 138.3
 128.3
 118.3
 108.3
4.192 E + 2
4.045 E + 2
3.889 E + 2
3.713 E + 2
3.506 E + 2
3.281 E + 2
2.311 E + 0
2.798 E + 0
7.624 E + 0
1.240 E + 1
1.925 E + 1
2.899 E + 1
2.1076 E + 02
2.0365 E + 02
1.9827 E + 02
1.9185 E + 02
1.8493 E + 02
1.7855 E + 02
ρ
liq eq ,
ρ
v eq ,
ρ
eq
tg α 0 654 , kg m
3
⁄ – =
ρ
c
T
c
82 3 , C
o
– =
ρ
c
178.55  26 x 0.654 kg m
3
⁄ =
ρ
c
161.546 kg m
3
⁄ =
Phase Behavior 11
2.3 Critical State and Quantities of Corresponding
States
Figure 2.4 illustrates the inflection point of the critical isotherm at the critical point. At
the point of the inflection, both the first and the second partial derivates of p = p(V) equal
zero that
. (2.9)
The state of the system at this point is characterized by the critical specific volume V
c
, the
critical pressure p
c
, and the critical temperature T
c
.
Figure 2.6: Critical pressure as a function of number of Catoms in homologous series
(after GYULAY, 1967)
∂p
∂V

¸ ,
¸ _
0
∂
2
p
∂V
2

¸ ,
¸ _
= =
Cyclohexane Naphtalene
Toluene
Benzene
iC
4
iC
5
n
o
r
m
a
l

P
a
r
a
f
f
i
n
s
m
o
n
o

O
l
e
f
i
n
e
s
0 2 4 6 8 10 12 14
CAtoms per Mole
p
[
M
P
a
]
c
1
2
4
3
5
6
12 Phase Behavior
Figure 2.7: Critical temperature as a function of numbers of Catoms in homologous
series (after GYULAY, 1967)
Considering the critical data p
c
and T
c
, the homologous series of hydrocarbons show
regularities which can be used for extrapolation. The experimental data in Figure 2.6 and
Figure 2.7 refer to the homologous series of paraffins, C
n
H
2n+2
, and olefines, C
2
H
2n
(n =
1,2,... k)
,
with a margin of error 1 to 2%. Because of thermal decomposition, it is not
possible to obtain experimentally information about the critical data in case of high
molecular weight. However, the critical data of homologous compounds with longer
carbon chains can be extrapolated though an increasing error has to be taken into
consideration.
The “principle of corresponding states” for chemically similar substances  e.g. for
homologous series  results in a close relation between the p, V, Tproperties of pure
hydrocarbons if the state variables are substituted by the so called “reduced quantities”
which are
(2.10)
Figure 2.8 shows a p
r
, V
r
phase diagram which is valid for paraffins from methane (CH
4
)
to hexane (C
6
H
14
).
0 2 4
6
8 10 12 14
CAtoms per Mole
T
[
°
C
]
c
200
100
0
100
200
300
400
500
n
o
r
m
a
l

P
a
r
a
f
f
i
n
s
m
o
n
o

O
l
e
e
f
i
n
s
Toluene
Benzene
Naphtalene
p
r
p
p
c
 ; V
r
V
V
c
 T
r
T
T
c

= ; = =
Phase Behavior 13
2.4 Binary Systems
If a systems consists of more than one component, its state is also a function of
composition. In general, the composition is defined by “mole fractions”.
The mole fraction is defined as the ratio between the number of moles of a certain
component and the sum of moles of all components. A system being composed of k
components is defined by the specification of (k  1) mole fractions because the sum of
the mole fractions always equal 1.Considering a 2component system, every change in
state is described by the equation of state F(p, V, T, z) = 0. z may be the mole fraction of
one (lighter) component.
The phase behavior of the ethane/nheptane system is graphically illustrated by the p, T,
zcoordinate system in Figure 2.9. The volume is equivalent to the mole volume.
In the plane z = 1, the vapor pressure curve of ethane appears, whereas in the plane z = 0
the one of nheptane appears. Covering all other zplanes, an envelope surface encloses
the 2phase state. This is demonstrated by the example of three additional zplanes.
The upper broken line marks the critical points of all compositions which are possible.
This curve divides the envelope surface into two parts: the bubble point surface and the
dew point surface. The region of an undersaturated liquid state is positioned outside the
bubble point surface (low temperature). Outside of the dew point surface (high
temperature), the state of a dry gas is given.
Analogous to the pure substance, the critical state of binary systems is defined as the state
at which the intensive properties of the phases are no more distinguishable. Just as in case
of 1component systems, the critical isotherms have an inflection point according to Eq.
2.9.
14 Phase Behavior
Figure 2.8: Combined reduced pressure  reduced volume phase diagram of paraffins
with low molecular weight (after GYULAY, 1967)
Figure 2.9: Phase equilibrium surface of the binary system ethane/nheptane (from
GYULAY, 1967)
0 1.0 2.0 3.0 4.0
V
r
0.6
0.8
1.0
1.2
1.4
p
r
0.94
0.96
0.98
C
1
.
0
2
1
.0
0
1
.
0
4
1
.
0
6
=
T
r
10
7.5
5.0
2.5
0
0 100 200 300
Temperature [°C]
P
r
e
s
s
u
r
e
[
M
P
a
]
z
1.00
0.75
0.50
0.25
0
C
2
nC
7
Phase Behavior 15
Figure 2.10: Pressure  temperature phase diagram of the binary system
ethane/nheptane (from KAY, 1938)
Figure 2.10 shows the projection of Figure 2.9 into the p, Tplane. At a given pressure, the
bubble point temperature of the mixture is always higher than that of the pure lighter
component. Physically, it can be explained by the fact that the thermal motion of the
lighter molecules is obstructed by the heavier ones which exhibit more inertia.
On the other side, the dew point temperature of the mixture at a given pressure is always
lower than that of the pure heavier component. This is due to the fact that lighter
molecules partially transfer their higher kinetic energy to the heavier ones by collision.
Consequently, the system maintains the state of a gas phase.
Figure 2.10 also shows that T
c
of a mixture lies between the critical temperatures of the
pure substances. In contrast to this, p
c
of the mixture may be obviously higher than the
one of the pure substances.
If the mixture consists of two homologous compounds with quite different volatility (in
consequence of quite different molecular weights), the critical data curve envelopes a very
extensive temperature and pressure region. For example, the maximum of the critical
pressure of a methane/ndecane system equals 37 MPa. The smaller the difference
between the molecular weights and thus between the volatility, the more flat the envelope
curve will be.
Temperature [°F]
100 200 300 400 500
0
200
400
600
800
1000
1200
1400
P
r
e
s
s
u
r
e
[
p
s
i
a
]
Composition
No. [Wt%] Ethane
1 100.00
2 90.22
3 70.22
4 50.25
5 29.91
6 9.78
7 6.14
8 3.27
9 1.25
10 nHeptane
10
9
8
7
6
5
1
3
4
2
16 Phase Behavior
Figure 2.11 illustrates the phase behavior of a certain ethane/nheptane system. Besides
the critical point, the curve enveloping the 2phase region possesses two additional
characteristic points:
• C’: the point of highest pressure on the curve that is called cricondenbare.
• C”: the point of highest temperature on the curve that is called cricondentherm.
As on Figure 2.11, so called "quality lines" are shown on p,Tdiagrams. A quality line
represents a certain mole percentage being liquid or vapor in the state of phase
equilibrium. In Figure 2.11, the quality line "20%" represents the states in which 20% of
the system account for the liquid phase. The bubble point curve and the dew point curve
represent 100% and 0% liquid, respectively. All the quality lines (isochores) converge at
the critical point.
Figure 2.11 also shows an isothermal decrase along the path EF where E defines the
sastem to be a dry gas. If the constant temperature is higher than T
c
but lower than the
cricondentherm  like in case of the path EF , the path surpasses the dew point line twice.
Consequently, a condensate drops out at the dew point D’. At some point between D’ and
D”, the volume of condensate (liquid) will be at its maximum. This maximum is given by
the intersection point of the path EF with the dotted line connecting C and C”. If the
decrease in pressure will be continued, the condensate will be vaporized again. As soon
as the dew point D” has been reached, the condensated phase has been vaporized in total.
This process is called a “retrograde condensation”.
Similar phenomena occur when the temperature is changed by an isobaric process where
the constant pressure is higher than p
c
but lower than the cricondenbar of the system.
In Figure 2.11, the dotted line connecting point C with point C’ marks the states of the
system which exhibit the highest volume percentage of condensate dropout.
Phase Behavior 17
Figure 2.11: Pressure  temperature phase diagram of the binary system ethane
(z = 0.9683)/nheptane
It depends on the composition of the system if the cricondenbar is located on the dew point
curve or on the bubble point curve. As far as the system ethane/nheptane is concerned,
Figure 2.10 elucidates that the cricondenbar is located on the bubble point curve at low
mole fractions of ethane.
Figure 2.12: Mole fraction(ethane)  temperature diagram of the binary system
ethane/nheptane (from GYULAY, 1967)
7
6
5
4
3
2
1
0
0 50 100 150
Temperature [°C]
P
r
e
s
s
u
r
e
[
M
P
a
]
C
C'
A
B
D'
D''
E
F
C''
Vapor
Liquid
and
Vapor
Liquid
1
0
%
2
0
%
0
0.2
0.4
0.6
0.8
1.0
50 0 50 100 150 200 250 300
Temperature [°C]
z
α
β
δ
δ'
1.4
2.0
4.2
5.6
7.0
B
P
L
0.7 [MPa]
D
P
L
β'
18 Phase Behavior
In Figure 2.12, the phase behavior of ethane/nheptane systems is graphically illustrated
in the z,Tplane corresponding to another possible projection of the surface in Figure 2.9.
The mixture achieves bubble point β due to an isobaric (1.4 MPa) heat supply. Point β
symbolizes the composition of the liquid phase which is in equilibrium with an
infinitesimal small vapor phase whose composition is symbolized by the point β’. During
further increase of temperature, dew point state is reached at point . The composition of
the infinitesimal small liquid phase in equilibrium with the vapor phase corresponds with
point ’.
Figure 2.13: Mole fraction (ethane)  Pressure diagram of the binary system
ethane/nheptane (from GYULAY, 1967)
The design of the corresponding p, zdiagram is also possible (see Figure 2.13). An
example may be the composition at the point A(T = 150
o
C). The composition of the liquid
phase is given by point A’, the one of the vapor phase by point A”. Again the relative
masses of both phases can be determined by applying the principle of the lever (see
Example 2.4).
Example 2.4.
Determining the phase composition.
A sealed container (p = 2.86 MPa), T = 150
o
C) is filled up with
100 kg of a ethane(z = 0.47)/nheptane mixture. The mole number of
ethane in the liquid phase and in the vapor phase, respectively,
can be evaluated from Figure 2.13 by using the principle of lever.
α
δ
δ
0 0.2 0.4 0.6 0.8 1.0
A A' A''
1
2
3
4
5
6
7
8
9
z
P
r
e
s
s
u
r
e
[
M
P
a
]
B
P
D
P
2
0
0
°
C
1
5
0
°
C
1
0
0
°
C
Phase Behavior 19
At first, the mole weights (M
C2
= 30 kg/mole, M
C7
= 100 kg/mole)
are inserted into
to evaluate the weight of nheptane in the system, m
C7
. The weight
of ethane, m
C2
, is given by m
C2
= 100  m
C7
.
Now the total mole number of the system, n = n
C2
+ n
C7
, can be
calculated:
,
,
.
From Figure 2.12
where
n
liq
:total mole number in the liquid phase
n
vap
:total mole number in the vapor phase
Thus the total mole number in the vapor phase results in
The composition of the vapor phase is given by point A” in Figure
2.12:
The mole number of ethane in the vapor phase can now be calculated
by
z
m
C2
M
C2
⁄
m
C2
M
C2
m
C7
M
C7
⁄ + ⁄

=
z 0.47
100 m
C7
– ( ) 30 ⁄
100 m
C7
– ( ) 30 m
C7
100 ⁄ + ⁄

= =
m
C7
79.215kg =
m
C2
100 79.215 20.785kg = – =
n
C2
m
C2
M
C2

20.785
30

0.693kmole = = =
n
C7
m
C7
M
C7

79.215
100

0.792kmole = = =
n 0.693 0.792 1.485kmole = + =
n
l i q
n
vap

A A″
A A′
 1.458 = =
n
vap
0.604 kmole =
z 0.82 =
20 Phase Behavior
.
The composition of the liquid phase is given by point A’ in Figure
2.12:
.
The total mole number in the liquid phase results in
.
The mole number of ethane in the liquid phase can now be calculated
by
.
Figure 2.14: Properties of ternary diagrams
2.5 MultiComponent Systems
2.5.1 Ternary Phase Diagrams
It is common to illustrate the phase behavior of 3component systems at constant pressure
and temperature in so called triangular diagrams. Each corner of the triangle represents
one pure component. On the basis of the equilaterality of the triangle, the sum of the
perpendicular distances from any point to each side of the diagram is a constant equal to
length of any of the sides. Thus, the composition  expressed in mole fractions  of a point
n
C2 vap ,
z n
vap
× 0.82 0.672 × 0.495kmole = = =
z 0.23 =
n
liq
n n
vap
1.485 0.604 0.881 kmole = – = – =
n
C2 liq ,
0.23 0.881 0.203kmole = × =
3
3
3
3
2
2
2
2
1
1
1
1
a.
c.
b.
d.
L
2
L
1
L
3
Const. Fraction
Component1
C
o
n
s
t
a
n
t
R
a
t
i
o
o
f
1
t
o
2
A
D
B
Phase Behavior 21
in the interior of the triangle is given by
(2.11)
where
. (2.12)
Several other useful properties of the triangular diagrams are also illustrated by Figure
2.14:
• For mixtures along any line parallel to a side of the diagram, the fraction of the
component of the corner opposite to that side is constant.
• Mixtures lying on any line connecting a corner with the opposite side contain a
constant ratio of the component at the ends of the side.
• Mixtures of any two compositions lie on a straight line connecting the two initial
points on the ternary diagram. The principle of the lever finds application again and
(2.13)
gives the mixing ratio leading to mixture D.
Figure 2.15 shows the 2phase region for chosen p and T. Any mixture with an overall
composition lying inside the binodal curve will split into a liquid and a vapor phase. The
“tie lines” connect compositions of liquid and vapor phases in equilibrium. Any overall
composition on a certain tie line gives the same liquid and vapor composition being in
euqilibrium. Only the amounts of the phases change as the overall composition changes.
Figure 2.15: Typical features of a ternary phase diagram
z
1
L
1
L
T
 , = z
2
L
2
L
T
, = z
3
L
3
L
T
 , =
L
T
L
1
L
2
L
3
+ + =
n
A
n
B

DB
DA

=
Plait
Point
Critical
Region
Liquid
Region
Two Phase
Region
Vapor
Region
Binodal Cu
rv
e
T
i
e
L
i
n
e
3 2
1
22 Phase Behavior
The liquid and vapor portions of the binodal curve meet at the “plait point” which
represents the critical composition. By drawing the tangent in the plait point on the
binodal curve, the singlephase region is splitted into three sections. Mixtures of a
composition being located in the critical region with another one being located in the
liquid or vapor region will, in any case, also result in a singlephase system if the straight
line connecting the two initial compositions does not intersect the 2phase region.
Figure 2.16 illustrates the influence of pressure on the phase behavior of a certain ternary
system at constant temperature. As pressure increases, the 2phase region shrinks.
It is useful to comprise the two heavier components of a ternary system and to reduce this
system to a fictitious binary system, on the basis of a hypothetical component. Figure 2.17
illustrates a corresponding application by the respective p,Tdiagram of the
methane/propane/npentane system. The mole% of methane are specified along the
outermost envelope curve. All envelope curves are characterized by the portion of
propane in the hypothetical component (propane/npentane) which is given by
(2.14)
In accordance to this aspect, the critical state properties, p
c
and T
c
, can be determined for
any mixture of the three components (see Example 2.5).
C
z
3
z
3
z
5
+

=
Phase Behavior 23
Figure 2.16: Triangular diagrams for the methane/propane/npentane system at 160
o
F(71
o
C) (after DOURSON et al., 1943)
Propane
Propane
Propane
nPentane
nPentane
nPentane
Methane
Methane
Methane
0.8
0.8
0.8
0.8
0.8
0.8
0.2
0.2
0.2
0.6
0.6
0.6
0.6
0.6
0.6
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.6
0.6
0.6
0.2
0.2
0.2
0.2
0.2
0.2
0.8
0.8
0.8
p=500 [psia]
p=1000 [psia]
p=1500 [psia]
T=160°F
T=160°F
T=160°F
a.
b.
c.
C
=
1
.
0
C
=
1
.
0
C
=
0
.
0
C
=
0
.
0
C
=
0
.
0
C
=
0
.
2
C
=
0
.
2
C
=
0
.
2
C
=
0
.
4
C
=
0
.
4
C
=
0
.
4
C
=
0
.
6
C
=
0
.
6
C
=
0
.
6
C
=
0
.
8
C
=
0
.
8
1
]
1
¸
+
·
5 3
3
c c
c
x x
x
C
1
]
1
¸
+
·
5 3
3
c c
c
x x
x
C
1
]
1
¸
+
·
5 3
3
c c
c
x x
x
C
24 Phase Behavior
Figure 2.17: Critical loci of methane/propane/npentane systems (from KATZ et al.,
1959)
Example 2.5
The hydrocarbon mixture is composed of 8 [kg] methane (M = 16[kg
kmol
1
], 13,2 [kg] propane (M = 44.1[kg kmol
1
]) and 32.5 [kg]
npentane (M = 72.2[kg kmol
1
]). The critical data of this mixture
can be evaluated by use of Figure 2.17.
At first, the mole numbers and the respective mole fractions must
be calculated.
,
,
,
,
,
.
The portion of propane in the hypothetical component
propane/npentane is given by
0
500
1000
1500
2000
2500
3000
200 100 0 100 200 300 400
Temperature [°F]
C
r
i
t
i
c
a
l
P
r
e
s
s
u
r
e
[
p
s
i
a
]
C
1 C
3
C
5
0
10
20
30
40
50
60
70
80
75
M
o
l
e
[
%
]
C
H
4
C
=1
.0
C
=
0
.
6
C
=
0
.8
C
=
0
.
0
C
=
0
.
2
C
=
0
. 4
1
]
1
¸
+
·
5 3
3
nC C
C
C
n
1
8
16

0.5kmole = =
n
3
13.2
44.1

0.3kmole = =
n
5
32.5
72.2

0.45kmole = =
z
1
0.5
0.5 0.3 0.45 + +

0.4 = =
z
3
0.3
0.5 0.3 0.45 + +

0.24 = =
z
5
0.45
0.5 0.3 0.45 + +

0.36 = =
Phase Behavior 25
.
From Figure 2.17 at C = 0.4 and 40 mole percent methane:
and
.
The application of the triangular diagram is not solely confined to ternary systems. For
example it is possible to partition the paraffinic hydrocarbons into three
pseudocomponents which are
• methane (C
1
) as the light component,
• the lighter pseudocomponent including ethane to hexane (C
2
C
6
),
• the heavier pseudocomponent including heptane and higher hydrocarbons (C
7+
).
Anyway, only poor information of complex natural hydrocarbon systems has been
reported until now. Nevertheless, some generalization makes a description of these
complex systems possible  according to known data. The phase behavior of several
complex and natural hydrocarbon systems are demonstrated in Figure 2.18 to Figure 2.22
by p, Tphase diagrams. For the classification of natural hydrocarbon systems, it is
essential to know
• if the critical temperature is lower or higher than the reservoir temperature,
• which state will be achieved at surface conditions (separator).
Not considered in this classification are changes in composition during production.
Figure 2.18 represents a hydrocarbon system whose critical temperature is significantly
lower than the reservoir temperature. In case of an isothermal pressure decrease (full line
from point 1 to 2), which occurs in the reservoir adjacent to the production well the the
course of production, the system remains in the singlephase (gaseous) state. Even in case
of both pressure and temperature decrease (dotted line), no liquid phase will drop out.
Consequently, the considered hydrocarbon mixture is called a “dry gas”. Dry gases
contain mainly methane, small amounts of ethane, possibly propane and some
hydrocarbons of higher molecular weights.
A so called “wet gas” (see Figure 2.19) remains in a singlephase (gaseous) state in the
reservoir during production (line 12). Anyway, condensate will drop out under separator
conditions.
C
z
3
z
3
z
5
+

0.24
0.24 0.36 +

0.4 = = =
T
c
262.5°F 128°C = =
p
c
1344 psia 9.27MPa = =
26 Phase Behavior
In case of the system shown in Figure 2.20, the reservoir temperature is higher than the
critical one but lower than the cricondentherm. The initial conditions given by Point 1
specifies the hydrocarbon mixture as a dry gas. If the pressure will decrease adjacent to
the production well during production, the dew point of the system is reached at point 2.
Consequently, condensate drops out inside the reservoir. The pressure at point 3
corresponds to the state in which the condensed liquid phase reaches the maximum (in
mole%). In the separator, the amount of condensate is larger than in case of wet gases.
Systems as shown in Figure 2.20 are called “gas condensates”.
Figure 2.18: Phase diagram of a dry gas (from MCCAIN, 1973)
Figure 2.19: Phase diagram of a wet gas (from MCCAIN, 1973)
Figure 2.20: Phase diagram of a retrograde gas condensate (from MCCAIN, 1973)
P
r
e
s
s
u
r
e
Temperature
Gas
Liquid Sep.
1
2
Critical
Point
75
50
25
P
r
e
s
s
u
r
e
Temperature
Gas
Liquid
1
2
Critical
Point
Mole % Liq.
75
50
25
5
0
100
Sep.
P
r
e
s
s
u
r
e
Temperature
Gas
Liquid
1
2
Critical
Point
Mole % Liq.
75
50
25
5
0
10
100
Sep.
3
Phase Behavior 27
Figure 2.21: Phase diagram of a highshrinkage crude oil (from MCCAIN, 1973)
The so called “white oils”  as characterized in Figure 2.21  are referred to as “high
shrinkage oils”. The reservoir temperature is below the critical temperature. Since the
bubble point curve will be reached by the decrease in pressure due to production, from the
initial pressure (point 1) to the pressure 2, a further pressure drop in the reservoir will lead
to point 3 and thus to an increased development of the vapor phase. At separator
conditions, about 65% of the produced hydrocarbon mixture will exist as liquid phase if
the reservoir is produced at bubble point conditions.
Figure 2.22: Phase diagram of a low shrinkage crude oil (from MCCAIN, 1973)
Figure 2.22 shows a “black oil” or “low shrinkage oil”. The initial state is characterized
by point 1 at which the state of the system can be regarded as “undersaturated” liquid. If
the pressure in the neighbourhood of the production well will decrease during production
to point 2, the bubble point curve is reached and the state of the system is now considered
“saturated”. The separator conditions are near the bubble point curve. Consequently,
about 85 mole% of the produced hydrocarbon mixture is in the liquid phase at separator
conditions. In accordance to this fact, the shrinkage of the oil due to gas liberation is less
pronounced than in case "white oils" (see Figure 2.21).
If the hydrocarbon mixture in the reservoir is a 2phase state under initial reservoir
conditions, oil and gas phase can be considered apart from one another (see Figure 2.23).
The equilibrium conditions at the initial state of the system are given by the intersection
point of the dew point curve of the gas cap and the bubble point curve of the oil zone.The
P
r
e
s
s
u
r
e
Temperature
Gas
Liquid
1
2
Critical
Point
Mole % Liq.
Sep.
3
75
50
25
100
P
r
e
s
s
u
r
e
Temperature
Gas
Liquid
Critical Point
Mole % Liq.
Sep.
3
75
50
25
0
100
B
u
b
b
l
e

P
o
i
n
t
L
i
n
e
D
e
w

P
o
i
n
t
L
i
n
e
1 Undersaturated
2 Saturated
28 Phase Behavior
gas cap shows a “retrograde” behavior, if the intersection point is located on the dew point
curve of the gas cap between the critical point and the cricondentherm.
Just as in case of binary systems (see Figure 2.9), the phase behavior of natural
hydrocarbon mixtures can also be illustrated in p, T, zdiagrams.
Figure 2.23: Phase diagram pairs of gas cap and oil zone
In Figure 2.24, composition I represents the separator gas while composition IV
represents the corresponding separator oil of the well stream. Furthermore, the phase
behavior of two representative mixtures of I and IV are given by the compositions II and
III. The system of composition II corresponds to a gascondensate system, the one of
composition III to a white oil. Inside the 2phase region of system II and III, isochores of
the liquid phase and  as dotted lines  the locations of maximum retrograde condensation
are drawn. Again an envelope surface comprises the 2phase region in dependence on the
composition. The spatial curve, which connects the critical points, splits the surface into
two parts which are the dew point surface and the bubble point surface. Outside the
envelope surface, the system is in a singlephase state.
By projecting the phase surface into the p, zplane, information about the composition of
the system will be obtained. If the state of the system is represented by point 1, the
equilibrium composition of the liquid (x
1
), and the one of the vapor phase, (y
1
), is given
by point 4 in the p, zplane.
Temperature Temperature
P
r
e
s
s
u
r
e
P
r
e
s
s
u
r
e
C
C
C
C
Oil Zone
Oil Zone
p
i
p
i
Retrograde Condensating Gas Cap Gas Cap
Phase Behavior 29
Figure 2.24: Phase equilibrium surface of oil/natural gas systems (from GYULAY, 1967)
C''
C'
T
1
T
2
T
3
I
II
III
IV
I
II
III
IV
T
1
x
1
y
1
P
d
P
b
T
2
T
3
C
I
4
C
I
C'
C
I
C''
I
C
1
C
II
C
II
C
II
C'
C''
K
II
C
III
C
III
C'
C
III
C
IV
C
IV
C
IV
C''
L
III
IV
G
GL
GL'
C
0
.
0 0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
0
.
0
0
.
2
0
.
4
z
P
30 Phase Behavior
31
Chapter 3
Equations of State
The preliminary chapter included graphical illustrations of equilibrium surfaces and their
normal projections into the (p,T), (p,V), (p,z) etc. planes. The application of such diagrams
enable the determination of the respective volume for certain states defined by the
corresponding pressure p and temperature T. However, a pure graphical application of the
state functions is not very practical and  on top of that  impossible for multicomponent
systems. This aspect obviously makes a mathematical consideration of these problems
necessary.
On the basis of Eq. 2.1, the following relation regarding the p, V, Tdata is valid in case
of any chemically homogeneous phase:
(3.1)
In this context, V must be defined as the volume of one mole (intensive property). In case
of a 2phase system:
(3.2)
and therefore
(3.3)
where
V
t
: mole volume of the system
V
liq
: mole volume of the liquid phase
V
vap
: mole volume of the vapor phase
n: number of moles in the system
n
liq
: number of moles in the liquid phase
n
vap
: number of moles in the vapor phase.
If phase equilibrium is given, the phases can be regarded as seperate thermodynamic
systems. If a phase  may be the liquid  phase consists of k components, the corresponding
equation of state may be written as follows:
V V p T , ( ) =
n n
l i q
n
vap
+ =
nV
t
n
liq
V
liq
n
vap
V
vap
+ =
32 Equations of State
(3.4)
or
(3.5)
where x
i
is defined as the mole fraction of the component i in the (liquid) phase.
Considering the mole fractions, the so called “constraint equation” is valid:
. (3.6)
In various cases of even practical interests, a multicomponent phase behaves as an ideal
mixture and the volumes are strictly additive. If V
i
is defined as the mole volume of
component i in the phase, the mole volume of the phase will result in
(3.7)
If Eq. 3.7 is valid, the enthalpy of the system must be generally considered additive. This
means that the enthalpy of the system is equal to the sum of the enthalpies of the single
components. In this case, no thermal effect will take place during the mixing procedure.
Put into other words: The mixing energy will be zero.
Regarding Eq. 3.6 in discussing Eq. 3.5, it is obvious that V is a function of 2 + (k  1) =
k + 1 variables. It is impossible to approximate the equilibrium surface for the entire (p,
V, T, x
k
)space by one single equation of state. Therefore some procedure bit by bit is
necessary.
3.1 Change of State at Low Compressibility
The expansion of Eq. 3.1  or just the same of Eq. 3.5 at constant composition  into a
TAYLORSeries leads to
(3.8)
By assuming that the higher derivations are neglectable, Eq. 3.8 may be truncated to
F p, V, T, x
1
x
2
, …x
p
( ) 0 =
V V = p, T, x
1
x
2
, …x
k
( )
x
i
1 =
i 1 =
k
∑
V x
i
V
i
i 1 =
k
∑
=
V p T , ( ) V p
o
T
o
, ( )
V ∂
p ∂

¸ ,
¸ _
T
p p
o
– ( )
1
2

∂
2
V
∂p
2

¸ ,
¸ _
T
p p
o
– ( )
2
x
V p T , ( ) V p
o
T
o
, ( )
V ∂
T ∂

¸ ,
¸ _
p
T T
o
– ( )
1
2

∂
2
V
∂T
2

¸ ,
¸ _
p
T T
o
– ( )
2
...... + + + =
+ + + =
Equations of State 33
(3.9)
where
.
Considering the Eq. 2.4 and Eq. 2.5, Eq. 3.9 may be transformed to
(3.10)
where
: isothermal compressibility
: cubic expansion coefficient
Eq. 3.10 can at best be applied for fluids in a 1phase state. Experience and practice have
shown that the cubic equations of state are most sufficient and beneficial for calculating
the state of gases and of 2phase systems. Of course, there may exist equations which
approximate the measured values more accurately. Anyway, the constants included in
these equations are not always given. Therefore, only the cubic equations of state,
particularly the PENGROBINSON equation, will be discussed in this textbook. In doing so,
the proceedings of generalization in derivating this equation of state will be elucidated.
3.2 Equation of State of Perfect and Real Gases
A gas is defined as perfect, if the intermolecular (VAN DER WAALS) forces are
neglectable.
Then for a molar system:
(3.11)
where R is defined as “universal gas constant” and
R = 8.31434 J/mole K
The compressibility factor Z is defined as
(3.12)
V p T ( , ) V
o
1
1
V
o

∂V
∂p

¸ ,
¸ _
T
p p
o
– ( )
1
V
o

∂V
∂T

¸ ,
¸ _
p
T T
o
– ( ) + + =
V
o
V p
o
T
o
, ( ) =
V p T , ( ) V
o
1 κ p p
o
– ( ) – α T T
o
– ( ) + [ ] =
κ
α
pV RT =
Z
pV
RT

=
34 Equations of State
or
, (3.13)
respectively.
For ideal gases, the factor Z equals 1. For real gases, Z is a state variable and depends on
the pressure, the temperature and the composition. The critical point is defined by the
pressure p
c
, temperature T
c
, and specific volume V
c
and may be determined
experimentally for 1component systems.
The critical compressibility factor, Z
c
, can be evaluated by substituting the critical data p
c
,
T
c
and V
c
into Eq. 3.12:
. (3.14)
Example 3.1
Evaluation of a Zfactor from laboratory data. A cylinder with
volume of 0.075 m
3
has been filled with a gas under atmospheric
pressure and at a temperature of 90.5
o
C. Then the volume of the
cylinder has been isothermally reduced to 0.00035 m
3
(volume
reduction in consequence of mercury insertion). After the volume
reduction, the pressure has been recorded as 13.79 MPa.
If the gas would show an ideal behavior at 90
o
C and 13.79 MPa, the
specific volume could be calculated by
and so
,
.
By use of Eq. 3.13 and considering the measured volume amounting
to 0.00035 m
3
:
Z
V
actual
V
ideal

=
Z
c
p
c
V
c
RT
c

=
P
1
V
1
P
2
V
2
=
V
2
V
1
p
1
p
2
 =
V
2
75 10
3 –
×
101
3 –
×10
1379
2
×10
 55
5 –
×10 m
3
= =
Z
35
5 –
×10
55
5 –
×10
 0.63 = =
Equations of State 35
3.3 Cubic Equations of State
If the pressure of any gaseous system is low, the ideal gas equation remains sufficient to
describe the volumetric behavior. In the year 1873, VAN DER WAALS deducted the first
equation of state which is able  up to a certain degree  to describe the continuity from
gaseous to the liquid state:
. (3.15)
a and b are substance specific constants. b can be interpreted as the inherent volume of the
molecules which is not available for the thermal motion of the molecules. The term a/V
2
regards the pressure reduction in consequence of intermolecular attraction.
Eq. 3.15 may also be written in the following form:
. (3.16)
Figure 3.1 illustrates the VAN DER WAALS isotherms in the vicinity of the critical point.
The dotted section of the isotherms represents the the data which are predicted by using
the VAN DER WAALS equation. Obviously, Eq. 3.16 cannot predict the real behavior of the
system during the vaporization, respectively condensation. The real behavior is shown by
the straight full line BD inside the 2phase region.
The cubic equations of state, which have been formulated by REDLICH and KWONG
(1949), SOAVE (1972), and PENG and ROBINSON (1976), have achieved much better
results.
Figure 3.1: The VAN DER WAALS isotherms near the critical point
p
a
V
2

+
¸ ,
¸ _
V b – ( ) RT =
p
RT
V b – ( )

a
V
2

– =
Critical Point
Calculated from
Van der Waals
Equation
T>T
2 1
T
=
c
o
n
s
t
.
2
Actual
Path
P
r
e
s
s
u
r
e
Specific Volume
T
=
c
o
n
s
t.
1
B
D
36 Equations of State
The REDLICHKWONG equation is given as
(3.17)
where b again is a substance specific constant. Anyway. a is now a function of the
temperature. It is useful to write the parameter a as follows:
(3.18)
where is constant. The original REDLICHKWONG equation included
.
By multiplication of Eq. 3.17 with
and after arrangement of V corresponding to its order of power:
. (3.19)
At the critical point :
. (3.20)
The comparison of Eq. 3.19 and Eq. 3.20 leads to
, (3.21)
, (3.22)
(3.23)
where
. (3.24)
Substitution of Eq. 3.21 and Eq. 3.23 into Eq. 3.22 results in
p
RT
V b – ( )

a
V V b + ( )

– =
a a′f T ( ) =
a′
f T ( ) T
0.5 –
=
V V b + ( ) V b – ( ) p ⁄
V
3
RT
p
 V
2
–
a
p

bRT
p

– b
2
–
¸ ,
¸ _
V
ab
p

0 = – +
V V
c
= ( )
V V
c
– ( )
3
V
3
3V
c
V
2
– 3V
2
c
V V
3
c
0 = – + =
3V
3
RT
c
p
c

=
3V
2
c
a
c
p
c

bRT
c
p
c

– b
2
– =
V
3
c
a
c
b
p
c

=
a
c
a′f T
c
( ) =
Equations of State 37
(3.25)
or after rearrangement, in
, (3.26)
, (3.27)
. (3.28)
Furthermore, Eq. 3.28 and Eq. 3.21 can be combined to
. (3.29)
Inserting Eq. 3.21 and Eq. 3.23 into Eq. 3.29:
(3.30)
The constants and have the following numerical values:
= 0.08664,
= 0.42748.
From Eq. 3.18 and Eq. 3.24:
. (3.31)
It is obvious that , if .
The substitution of into the Eq. 3.19 and the arrangement of Z
corresponding to its order of power results in
. (3.32)
Therefore:
3V
2
c
V
3
c
b

3bV
c
b
2
– – =
b
3
3b
2
V
c
3bV
2
c
V +
3
c
2V
3
c
= + +
b V
c
+ ( )
3
2V
3
c
=
b 2
1 3 ⁄
1 – ( )V
c
=
b
2
1 3 ⁄
1 – RT
c
3p
c

Ω
b
RT
c
p
c
 = =
a
c
RT
c
( )
2
9 2
1 3 ⁄
1 – ( )p
c
 Ω
b
RT
c
( )
2
p
c
 = =
Ω
b
Ω
a
Ω
b
Ω
a
a a
c
f T ( )
f T
c
( )
 a
c
α = =
α 1 = T T
c
=
V ZRT p ⁄ =
ZRT
p

¸ ,
¸ _
3
RT
p

ZRT
p

¸ ,
¸ _
2
–
a
p

bRT
p

– b
2
–
¸ ,
¸ _
ZRT
p

ab
b
 0 = – +
38 Equations of State
(3.33)
or
(3.34)
where
(3.35)
and
. (3.36)
The substitution of Eq. 3.29 and Eq. 3.30 into Eq. 3.35 and Eq. 3.36 leads to
(3.37)
and
. (3.38)
In the original REDLICHKWONG equation:
(3.39)
and so
. (3.40)
Eq. 3.34 includes only two parameters which are p
c
and T
c
. Please note that the ideal gas
equation contains no substancespecific parameters. Since the RedlichKwong cubic
equations of state consider these two substancespecific parameters, it has improved the
calculation of PVTproperties in a fundamental way. Anyway, the increasing yield of
experimental data has more and more indicated that the behavior of many liquids with a
nonspherical molecule structure deviates greatly from the predicted one. This made the
introduction of a third factor necessary.
Z
3
Z
2
–
ap
RT ( )
2

bp
RT

b
2
p
2
RT ( )
2

– – Z
ap
RT ( )
2

bp
RT
 0 = – +
Z
3
Z
2
– A B B
2
– – ( )Z AB 0 = – +
A
ap
RT ( )
2

=
B
bp
RT

=
A 4.2748
p
r
T
2
r

¸ ,
¸ _
α =
B 0.8664
p
r
T
r
 =
f T ( ) T
0.5 –
=
α T
r
0.5 –
=
Equations of State 39
Beginning with the year 1951 (MEISSNER and SEFERIAN), several proposals for a third
parameter have been made. The so called “acentric factor”, , has become the one with
greatest acknowledgement:
(3.41)
where is the reduced boiling point pressure.
The equation of state from SOAVE (1972) only differs from the REDLICHKWONG equation
with respect to the definition of the factor
. (3.42)
The weakness of all these equations ranging from the original REDLICHKWONG equation
to all its modifications (including the SOAVE equation) is the fact of an universal
unrealistic Z
c
factor of . Moreover, the prediction of liquid density is combined with
large errors.
Improved approximation has been achieved with the PENGROBINSON equation:
(3.43)
where a is given by Eq. 3.18 and b is further a substance specific constant.
Just as in case of the REDLICHKWONG equation, the following terms and equations can
be obtained:
, (3.44)
, (3.45)
, (3.46)
, (3.47)
, (3.48)
ω
ω lgp
s
r
1 +
¸ ,
¸ _
at T
r
0.7 = – =
p
s
r
p
s
p
c
⁄ =
α
0.5
1 0.48 1.574 + ω 0.176ω
2
– ( ) + = 1 T
r
0.5
– ( )
1 3 ⁄
p
RT
V b –

a
V V b + ( )+ b V b – ( )

– =
V
3 RT
p

b –
¸ ,
¸ _
V
2
–
a
b

2bRT
p

– 3b
2
–
¸ ,
¸ _
V b
a
p

RT
p
 b – b
2
–
¸ ,
¸ _
0 = – +
V
c
0.307
RT
c
p
c
 =
b 0.07796
RT
c
p
c
 =
a a
c
α =
a
c
0.457235
RT
c
( )
2
p
c
 =
40 Equations of State
(3.49)
, (3.50)
, (3.51)
. (3.52)
Analogies between Eq. 3.44 to Eq. 3.52 on the one side and between Eq. 3.19, Eq. 3.21,
Eq. 3.29 to Eq. 3.31, Eq. 3.34 to Eq. 3.38, and Eq. 3.42 on the other side are obvious. The
universal critical Zfactor of the PENGROBINSON equation results in 0.307 which is much
better than but still far away from reality. Anway, the calculated fluid densities are
much more accurate than the ones calculated by the equations of state previously
discussed.
Eq. 3.16, Eq. 3.17, and Eq. 3.43 have been established for pure substances. The extension
for multicomponent systems requires the calculation of the respective data of the pure
components and mixing rules in order to get the parameters of the mixture.
The mixing rule for the parameter b, which is included in the equations of
REDLICHKWONG, SOAVE and PENGROBINSON, is universally defined as an arithmetic
average by using
. (3.53)
For the temperaturedependent coefficient a , different mixing rules exist and are
presented below.
REDLICHKWONG:
. (3.54)
where a
i
can be calculated on the basis of the critical data of each component by using Eq.
3.30 and Eq. 3.40 into Eq. 3.31.
Z
3
1 B – ( )Z
2
– A 2B – 3B
2
– ( )Z AB B
2
B
3
– – – – + 0, =
A 0.457325
p
r
T
2
r

¸ ,
¸ _
α ⋅ =
B 0.07796
p
r
T
r
 =
α
0.5
1 0.3676 1.54226ω 0.26992ω
4
– +
¸ ,
¸ _
1 T
r
1 2 /
–
¸ ,
¸ _
+ =
1 3 ⁄
b x
i
b
i
i 1 =
k
∑
=
a x
i
a
i
0.5
i 1 =
k
∑
2
=
Equations of State 41
SOAVE:
(3.55)
where a
i
and a
j
, respectively, can be evaluated by Eq. 3.30 and Eq. 3.42.
PENGROBINSON:
(3.56)
where a
i
is defined by Eq. 3.48 and Eq. 3.52.
The mixing rules used by SOAVE and PENGROBINSON consider the binary interaction
between the molecules of the components i and j. In Eq. 3.55 and Eq. 3.56, the terms
are binary interaction coefficients which are assumed to be independent of pressure and
temperature.
Values of the binary interaction coefficients must be obtained by fitting the equation of
state to gasliquid equilibria data for each binary mixture. They have different values for
each binary pair and also take on different values for each equation of state.
Obviously, Eq. 3.55 and Eq. 3.56 reduce to the form of Eq. 3.54 if all binary interaction
coefficients are zero.
Another possibility of obtaining this coefficient  if no data are available  is by mean of
matching the phase behavior of multicomponent systems.
Example 3.2
Calculation of the pressure by use of the PENGROBINSON equation
of state. A laboratory cell at temperature of 100 with volume
of 0.00025 m
3
contains 0.25 mole of gas. The composition of the
system, the critical data and the acentric factors of the
components are tabled below.
a x
i
j 1 =
k
∑
i 1 =
k
∑
x
j
a
i
a
j
( )
0.5
1 k
i j
– ( ) =
a x
i
j 1 =
k
∑
i 1 =
k
∑
x
j
a
i
a
j
( )
0.5
1 k
i j
– ( ) =
k
i j
C
o
42 Equations of State
First, the parameters b and a must be calculated for each component.
For methane from Eq. 3.46:
,
from Eq. 3.48:
,
from Eq. 3.52 by inserting the acentric factor for methane
and the reduced temperature of methane
:
and thus
,
and from Eq. 3.47:
.
Correspondingly to the calculations for methane, the parameters of ethane and nbutane
were evaluated. They are tabled below.
Component Composition
y
i
Critical Data
T
c,i
p
c, i
o
K MPa
Acentric Factor
C
1
0.75 190.6 4.60 0.0115
C
2
0.20 305.4 4.88 0.0908
n  C
4
0.05 425.2 3.80 0.1928
ω
i
b
1
77796 10 ×
6 – 831434 1907 10
6
× ×
463 10
8
×
 2677 10
5 –
× = =
a
c,1
457235 10
6 –
×
831434 1907 10
6
× × ( )
2
463 10
8
×
 0.2494 = =
ω
1
0.0115 =
T
r 1 ,
273.15
190.6

1.958 = =
α
0.5
1
0.8954 =
α
1
0.8018 =
α
1
0.2494 0.8018 0.1999 = × =
Equations of State 43
The parameter b can be calculated by use of Eq. 3.53:
.
To calculate the coefficient a by use of Eq. 3.56, the interactive
coefficients, K
ij
, must be known. They are given below:
a = 0.75 x 0.75(0.1999 x 0.1999)
0.5
(1  0.000000) +
+ 0.75 x 0.20(0.1999 x 0.6045)
0.5
(1  0.002648) +
+ 0.75 x 0.05(0.1999 x 1.8260)
0.5
(1  0.014640) +
+ 0.20 x 0.75(0.6045 x 0.1999)
0.5
(1  0.002648) +
+ 0.20 x 0.20(0.6045 x 0.6045)
0.5
(1  0.000000) +
+ 0.20 x 0.05(0.6045 x 1.8260)
0.5
(1  0.004904) +
+ 0.05 x 0.75(1.8260 x 0.1999)
0.5
(1  0.014640) +
+ 0.05 x 0.20(1.8260 x 0.6045)
0.5
(1  0.004904) +
+ 0.05 x 0.05(1.8260 x 1.8260)
0.5
(1  0.000000) = 0.3108
Now the pressure can be calculated through Eq. 3.43 and by
inserting
,
,
Component b
i
a
c,i
α
i
a
i
C
1
2.677 E5 0.2494 0.8018 0.1999
C
2
4.048 E5 0.6042 1.0010 0.6045
n  C
4
7.245 E5 1.5050 1.2130 1.8260
C
1
C
2
n  C
4
C
1
0.0000 E+0 0.2648 E2 0.1464 E1
C
2
0.2648 E2 0.0000 E+0 0.4904 E2
n  C
4
0.1464 E1 0.4904 E2 0.0000 E+0
b 0.75 2.67E 5 0.20 4. 048E 5 – 0. + × + – × 05 7. × 245E 5 – =
b 3. 180E 5 – =
V
0.00025
0.25

0.001 m
3
mole ⁄ = =
T 373.15 K =
R 8.31434 J K mole × ⁄ =
44 Equations of State
and
.
The calculation procedure results in
Example 3.3
Calculation of the density of methane by use of the PENGROBINSON
equation of state. 2.0 kg methane (MC
1
= 16 kg k/mole is hold at
temperature of 305.12 K and at pressure of 9.26 MPa. The critical
data and the acentric factor of methane are known as
,
,
.
First, the parameters a and b have to be calculated.
From Eq. 3.46:
.
From Eq. 3.48:
.
From Eq. 3.52 by inserting the acentric factor and the reduced
temperature of the system
,
and thus
,
and from Eq. 3.47:
.
Now the mole volume of methane at the given state variables has to
be evaluated through Eq. 3.44. To solve this cubic equation of
b 3.18
5 –
×10 =
a 0.3108 =
p 2.328 = MPa
T
c
190.6 K =
p
c
4.63 MPa =
ω 0.0115 =
b 0.077796
8.31434 190.6 ×
4.63 10
6
×
 2.662
5 –
×10 = =
a
c
0.457235
8.31434 190.6 × ( )
2
463 10
4
×
 0.248 = =
T
r, 1
305. 12
190. 6

1. 601 = =
α
0.5
0.8954 =
α 0.8018 =
a 0.248 0.8018 0.1988 = × =
Equations of State 45
state, the CARDAN equation is applied for
where
,
and
.
The calculation of r, s and t result in
,
,
and
.
By substitution of
,
the formula
is reduced to
where
and
.
The calculation of p and q result in
x
3
rx
2
sx t 0 = + + +
r
RT
p

b –
¸ ,
¸ _
– =
s
a
p

2bRT
p

– 3b
2
–
¸ ,
¸ _
=
t b
a
p

RT
p

– b b
2
–
¸ ,
¸ _
– =
r 0. 2473
3 –
×10 – =
s 0.4755
8 –
×10 =
t 0. 3585
12 –
×10 – =
x y
r
3

– =
x
3
rx
2
sx t 0 = + + +
y
3
py q 0 = + +
p
3s r
2
–
3

=
q
2r
3
27

rs
3

– t + =
p 1. 563
8 –
×10 – =
46 Equations of State
and
.
The discriminant D is defined as
and thus
.
Based on the relationships
,
and
,
the mole volume can be calculated through
:
,
,
and thus
,
and
.
q 1. 089
12 –
×10 – =
D
p
3

¸ ,
¸ _
3
q
2

¸ ,
¸ _
2
+ =
D 1.53776
25 –
×10 =
u
q
2

–
3
D + =
v
q
2

– 3 D – =
y u v + =
x y
r
3

– =
u 9.7800
5 –
×10 =
v 5.3272
5 –
×10 =
y 1.5107
4 –
×10 =
V 2. 3350
4 –
×10 m
3
mole ⁄ =
ρ
m
nV

M
V
 68.522 kg m
3
⁄ = = =
Equations of State 47
3.4 Virial Equation of State
All cubic equations of state mentioned above are more or less empirical or at best
semitheoretical. However, they are obviously qualified for practical application.
In contrast, the virial equation of state has some theoretical background in statistical
mechanics. However, the form of an infinite series concerning the volume is not very
appropriate for practical application.
The pressure is expressed as
. (3.57)
The virial coefficients B, C etc. are solely a function of temperature. Sometimes Eq. 3.57
is applied for the calculation of gaseous states (e.g. natural gas) by considering only the
first two or three terms (depending on the availability of the virial coefficients).
If V is substituted into Eq. 3.57 according to the gas law by Z RT/p, then:
(3.58)
The second virial coefficient B can be shown to be the slope of isotherms on a plot of
. When p approaches zero, the third term in Eq. 3.58 may be
neglected so that
. (3.59)
Since also Z becomes zero as p goes to zero, de l’Hospital’s rule can be applied resulting
at constant T in
. (3.60)
Eq. 3.60 implies that the 2. virial coefficient vanishes at the BOYLEtemperature. The
BOYLE temperature is defined as the temperature above which Zvalues become greater
than 1 for all pressures.
p RT
1
V

B
V
2

C
V
3
 … + +
¸ ,
¸ _
=
Z 1
Bp
ZRT

Cp
2
ZRT ( )
2

… + + + =
Z Z p ( ) at p 0 = =
B ZRT
Z 1 –
p

¸ ,
¸ _
p 0 →
lim =
B RT
∂Z
∂p

¸ ,
¸ _
p 0 →
lim
T
=
48 Equations of State
49
Chapter 4
Calculation of Phase Equilibria
4.1 Mixtures
4.1.1 Definitions
In case of thermodynamical phase equilibrium of a multicomponent system, all phases
are  in physical terms  homogeneous mixtures. Vaporized compounds are always
mixable in all relations, independent of their chemical characteristics. Therefore, only one
vapor phase exists.
Compounds existing in a liquid phase are only mixable in all relations if they are
chemically similar. Since hydrocarbons  especially whose which belong to the same
homologous series  exhibit such chemical conformity, their mixtures will be physically
homogeneous, independent of the composition of the mixture.
The composition of the system and its phases are specified by mole fractions. The mole
fraction of the component i is given by the ratio between its mole number, n
i
, and the total
mole number in the system (in the respective phase), n. The sum of all mole fractions
equals 1. In the following, the different mole fractions are marked by
z
i
: mole fraction of the component i in the total system
x
i
: mole fraction of the component i in the liquid phase
y
i
: mole fraction of the component i in the vapor phase,
so that in the system in total
, (4.1) z
i
n
i
k
i =1
n
i
∑
 =
50 Calculation of Phase Equilibria
, (4.2)
. (4.3)
If a multicomponent system is in the state of thermodynamic phase equilibrium, the
distribution of the component i among the vapor and the liquid phase is characterized by
its distribution coefficient, the so called “Kfactor”:
. (4.4)
The numerical value of K
i
depends on the state of the system. Accordingly, the
composition of the phases will change at any variation in state.
4.1.2 Kfactors
Ideal liquid mixtures have the property of being met by RAOULT’s law so that
(4.5)
where
p
i
: partial pressure of the component i
: tension of component i as a singlecomponent system
x
i
: mole fraction of component i in the liquid phase.
The total pressure of the system adds up to
(4.6)
If a vapor phase exhibits perfect behavior, DALTON’s law can be applied and
(4.7)
The combination of Eq. 4.5 and Eq. 4.7 results in
x
i
n
i
k
i=1
n
i
∑

¸ ,
¸ _
liq
=
y
i
n
i
k
i=1
n
i
∑

¸ ,
¸ _
vap
=
K
i
y
i
x
i

=
x
i
p
i
p
o
i

=
p
o
i
p p
i
i= 1
k
∑
x
i
p
o
i
i= 1
k
∑
= =
y
i
p
i
p

=
Calculation of Phase Equilibria 51
(4.8)
Anyway, certain limitations are placed on the definition of the Kfactor given above. It is
evident that any pure coomponent has a vapor pressure only up to its critical temperature.
In practice, the extrapolation of the vapor pressure curves to higher temperatures than T
c
does not yield satisfactory results in most cases. Since T
c
= 190.37 K for methane, it
becomes evident that some other means must be obtained for the determination of
Kfactors if they are to be used for hydrocarbon mixtures.
On the basis of the thermodynamic concept of fugacity, the vapor pressures in Eq. 4.8 can
be corrected and extrapolated to cover a wider range of conditions. A first approach
originated from DODGE and NEWTON (1937) who have shown that at any given p and T
(4.9)
and
(4.10)
respectively, where
f
i(liq)
: fugacity of component i in a liquid mixture
f
i(vap)
: fugacity of component i in a vapor mixture
: fugacity of pure component i as liquid
: fugacity of pure component i as vapor.
If the phases are in equilibrium at given conditions and thus
, (4.11)
the definition of the Kfactor is given by
. (4.12)
A generalized fugacity chart for natural gases based on reduced temperatures and
pressures is given in Figure 4.1 in terms of f/p, the so called “activity coefficient”. These
fugacities serve to provide ideal equilibrium constants as defined by Eq. 4.12.
K
i
id ( )
y
i
x
i

p
o
i
p

= =
f
i liq ( )
x
i
f
o
i liq ( )
=
f
i vap ( )
y
i
f
o
i vap ( )
=
f
o
i liq ( )
f
o
i vap ( )
f
i vap ( )
f
i l i q ( )
=
K
i id ( )
′
y
i
x
i

f
°
i liq ( )
f
°
i vap ( )

= =
52 Calculation of Phase Equilibria
Figure 4.2 shows that the values are not sufficiently accurate to describe phase
equilibria at high pressures. Anyway, the extrapolation of the fugacities on the basis of
Eq. 4.10  though also limited by the critical state conditions  covers a wider range of state
conditions than the vapor pressure relations.
Figure 4.1: Fugacity of natural gases (from BROWN, 1945)
Figure 4.2: Ideal and real Kfactors of nbutane at 60[
o
C]
In case of multicomponent hydrocarbon systems, attraction forces between the molecules
K
i id ( )
′
1 0.1
0.01
0.1
1.0
2 2 2 3 3 3 4 4 5 5 10
2
2
2
3
3
3
4
4
4
5
5
Pressure [MPa]
l
o
g
K
Temperature=60 °C
K
K'
id
nButane
K
id
Calculation of Phase Equilibria 53
affect the vaporliquid equilibria at higher p and T. As both RAOULT’s and DALTON’s laws
and the fugacity relations given above (see Eq. 4.9 and Eq. 4.10) result in the calculation
of Kfactors without considering the composition, it is necessary to find another solution
for the determination of the Kfactors.
Kfactors can be evaluated in satisfactory accordance with the measured data by the
application of the “successive solution method”. This method  it will be discussed
through Example 5.5  consists in an iterative process. It is based on the definition of the
“fugacity coefficient” of the component i (i = 1, 2,...k) in a multicomponent phase P  in
analogy of the “activity coefficient”, f/p  as
(4.13)
where p is used for the total pressure of the system.
After introducing not only the PENGROBINSON equation of state
in
but also other terms which have been defined by PENGROBINSON (see Chap. 3), NGHIEM,
AZIZ and LI (1983) have calculated fugacity coefficients by use of
(4.14)
where a
ij
is defined by
. (4.15)
If the phases are in equilibrium at given conditions and thus
(see Eq. 4.11), the definition of the Kfactor is given by
(4.16)
φ
i ,P
f
i ,P
x
i P ,
p

=
p
RT
V b –

a
V V b + ( ) b V b – ( ) +

– =
f
p
 Z 1 Z ln
1
RT

RT
V

p –
¸ ,
¸ _
V d
∞
V
∫
+ – – = ln
φ
i
ln
b
i
b
 Z 1 – ( ) Z B – ( )
A
2 2B

–
2
k
i 1 =
∑
x
i
a
ij
a

b
i
b

–
Z 2.414 B +
Z 0.414 B –

¸ ,
¸ _
ln ln – =
a
i j
1 K
ij
– ( ) a
i
0.5
a
j
0.5
( ) =
f
i vap ( )
f
i l i q ( )

1 =
K
i
φ
i l i q ( )
φ
i vap ( )

=
54 Calculation of Phase Equilibria
(see Eq. 4.13),
where
: fugacity coefficient of component i in the liquid phase
: fugacity coefficient of component i in the vapor phase.
4.2 Composition of Phases in Equilibrium
4.2.1 Definitions
If the system at a given pressure, temperature and composition is in a liquidvapor phase
equilibrium, the following must be valid (on the basis of an overall and particular material
balance, respectively):
(4.17)
(4.18)
where
n: total mole number of the system
: mole number of the liquid phase
: mole number of the vapor phase
x
i :
mole fraction of component i in the liquid phase
y
i
: mole fraction of component i in the vapor phase
z
i
: mole fraction of component i in the system
After substitution of Eq. 4.8 into Eq. 4.18 and after rearrangement, the following is valid:
, (4.19)
(4.20)
If the numerical value for n
liq
and has n
vap
have been chosen correctly, the calculations of
x
i
and y
i
(i = 1,...k) (Eq. 4.19 and Eq. 4.20) result in
φ
i liq ( )
φ
i vap ( )
n n
l i q
n
vap
+ =
z
i
n x
i
n
liq
y
i
n
vap
; i 1, …, k = ( ) + =
n
liq
n
vap
x
i
z
i
n
n
liq
n
vap
K
i
+

=
y
i
z
i
n
n
liq
K
i
⁄ n
vap
+

=
Calculation of Phase Equilibria 55
(4.21)
and
. (4.22)
Example 5.1 will demonstrate that it is possible to use the Eq. 4.21 and Eq. 4.22 for
iterative calculation purposes which result in the evaluation of the phase compositions at
given p, T and system composition.
If n = 1, Eq. 4.17 gives
and Eq. 4.21 becomes
. (4.23)
Eq. 4.23 elucidates that
n
liq
must be chosen larger in the following iteration if ,
n
liq
must be chosen smaller in the reverse case of .
Example 4.1
Determination of the gas and liquid phase compositions at p = 14
MPa and T = 92
o
C through an iterative calculation procedure. The
actual composition of the hydrocarbon system is tabled below.
FIRST ITERATION:n
liq
= 0.790
n
vap
= 0.210
x
i
z
i
n
n
l i q
n
vap
K
i
+

1 =
i 1 =
k
∑
=
i 1 =
k
∑
y
i
z
i
n
n
liq
K
i
⁄ n
vap
+

1 =
i 1 =
k
∑
=
i 1 =
k
∑
n
vap
1 n
liq
– =
x
i
i 1 =
k
∑
z
i
n
l i q
K
i
1 n
l i q
– ( ) +

i 1 =
k
∑
1 = =
x
i
1 >
∑
x
i
1 <
∑
56 Calculation of Phase Equilibria
n
liq
must be chosen smaller in the following iteration because of
.
SECOND ITERATION: n
liq
= 0.776
n
vap
= 0.224
THIRD ITERATION: n
liq
= 0.778
n
vap
= 0.222
Component System
Composition
z
i
KFactor
K
i
n
vap
K
i
nl
iq
+n
vap
K
i
C
1
0.4404 2.85 0.598 1.388 0.3173
C
2
0.0432 1.17 0.246 1.036 0.0417
C
3
0.0405 0.68 0.143 0.933 0.0434
C
4
0.0284 0.41 0.086 0.876 0.0324
C
5
0.0174 0.24 0.051 0.841 0.0207
C
6
0.0290 0.15 0.031 0.821 0.0353
C
7+
0.4011 0.01 0.003 0.793 0.5055
Σ 1.0000 0.9963
x
i
z
i
n
liq
n
vap
K
i
+

=
x
i 1 <
∑
Calculation of Phase Equilibria 57
At the bubble point of the system, the amount of vapor is infinitesimal small and thus
.
Eq. 4.22 then becomes
. (4.24)
Example 4.2 illustrates the iterative process of calculating the bubble point pressure, p
b
,
for a given hydrocarbon composition and for a fixed temperature. Naturally, K
i
= K
i(p)
must be known.
In the vicinity of the bubble point, K
i
x
i
are in a linear relation to pressure. Therefore, linear
interpolation can be applied for acceleration purposes.
Example 4.2
Determination of the bubble point pressure of a crude oil. The
composition of the system is given in Example 4.1.
Iteration II Iteration III Iteration III
Component
C1 0.3114 0.3123 0.8896
C2 0.0416 0.0416 0.0487
C3 0.0436 0.0436 0.0296
C4 0.0327 0.0327 0.0134
C5 0.0210 0.0209 0.0051
C6 0.0358 0.0357 0.0054
C7+ 0.5145 0.5132 0.0082
Σ 1.0006 0.9999 1.0000
x
i
z
i
n
liq
n
vap
K
i
+

= x
i
z
i
n
liq
n
vap
K
i
+

= y
i
z
i
n
liq
K
i
+n
vap
⁄

=
n
vap
0, n
l i q
n and x
i
z
i
= = =
z
i
n
n
liq
K
i
n
vap
+ ⁄
 K
i
z
i
K
i
x
i
i 1 =
k
∑
1. = =
i 1 =
k
∑
=
i 1 =
k
∑
58 Calculation of Phase Equilibria
By interpolation: p
b
= 23 MPa
At the dew point of the system, the amount of liquid phase is infinitesimal small and thus
.
From Equation 4.21:
. (4.25)
In analogy of the iterative process demonstrated by Example 4.2 (evaluation of p
b
), it is
theoretically possible to evaluate the dew point pressure by use of Eq. 4.25 by summing
up the values and by variating the pressure as long as .
However, the evaluation of the Kfactor for heavier components is crucial. Therefore, the
calculation of the dew point pressure of hydrocarbon systems by using Eq. 4.25 may be
too inaccurate. This notion is mostly used to check the Kfactors which have been
evaluated for certain state conditions.
Component System
Compos.
z
i
1. Assumption
p
b
= 21 MPa
K
i
y
i
= K
i
z
i
2. Assumption
p
b
= 22.5 MPa
K
i
y
i
= K
i
z
i
3. Assumption
p
b
= 23.2 MPa
K
i
y
i
= K
i
z
i
C
1
0.4404 2.1500 0.9469 2.060 0.9072 2.020 0.8896
C
2
0.0432 1.0300 0.0445 1.025 0.0443 1.020 0.0441
C
3
0.0405 0.6720 0.0272 0.678 0.0274 0.680 0.0275
C
4
0.0284 0.4400 0.0125 0.448 0.0127 0.450 0.0128
C
5
0.0174 0.3000 0.0052 0.316 0.0055 0.323 0.0056
C
6
0.0290 0.2150 0.0062 0.230 0.0067 0.239 0.0069
C
7+
0.4701 0.0235 0.0094 0.250 0.0100 0.026 0.0104
Σ 1.0000 1.0519 1.0138 0.9969
n
liq
0 n
vap
n and y
i
z
i
= = , =
z
i
n
n
liq
n
vap
K
i
+

z
i
K
i

y
i
K
i
 1 =
i 1 =
k
∑
=
i 1 =
k
∑
=
i 1 =
k
∑
y
i
K
i
⁄ Σy
i
K
i
⁄ 1 =
Calculation of Phase Equilibria 59
Figure 4.3: Flash and differential vaporization
In evaluating hydrocarbon systems, two characteristic processes are of importance. These
are illustrated schematically in Figure 4.3:
• Mass and composition of the system remain constant (closed system). The portions of
the phases and their compositions in various states of thermodynamic equilibrium (p
and V are variables) are to be evaluated. This process is specified as the “flash
vaporization (condensation)”.
• Preceding every further change of state, mass and composition of the system are
changed in removing one phase  in most cases the vapor phase  as soon as the
previous phase equilibrium is achieved (open system). Such a process is called
“differential vaporization (condensation)”.
In case of a flash process, it is of importance whether the system is signified as liquid at
bubble point pressure or as a vapor at dew point pressure. In both cases, the volume
increases during an isothermal pressure decrease.
If the temperature of a dry gas system ranges between T
c
and the cricondentherm, a
condensate will drop out below the dew point. The amount of the condensate will increase
up to a certain value as a function of pressure decrease (retrograde condensation). Below
this point, further pressure decrease will now effect the vaporization of the condensate.
In case of an undersaturated liquid system, the vapor phase appears below bubble point
pressure. At the same time, the liquid volume diminishes. Correspondingly, the oil phase
shrinks at decreasing pressure.
V
Hg Hg Hg Hg
L
Hg Hg
Hg Hg
V
L
Hg Hg
V
flash vaporization
at the dew point
flash vaporization at
the bubble point
differential vaporization
at the dew point
differential vaporization
at the bubble point
ntimes ntimes
60 Calculation of Phase Equilibria
The equilibria states appearing at a flash vaporization in series after one another can be
calculated independent of one another by Eq. 4.19 to Eq. 4.21 in regard of their phase
composition. However, the Kfactors must always correspond to the given pressure and
temperature.
In case of a differential vaporization, some material balance has to be applied. The volume
remains constant meanwhile every change in state is connected with a change of the mole
number in the system and its composition (open system).
This would correspond to the following reflections: In case of a saturated liquid, the
amount of component i will decrease by y
i
dn
v
resulting from the liberation and production
of an infinitesimal small amount of gas dn
v
, and
. (4.26)
Since y
i
= K
i
x
i
and dn
vap
= dn
liq
, the following is obtained after neglection of the term
dn
vap
dx
i
:
(4.27)
Eq. 4.27 cannot be integrated, because K
i
is a function of p. Therefore, the method of finite
differences is often applied. The integral is then substituted by summation formulae.
Results have been often proved to correspond with the experimental data in a most
accurate way (see Example 4.3).
Example 4.3
Determination of tank oil composition after 3stage separation on
the basis of the system composition in Example 5.1.
FIRST SEPARATOR: p = 3.33 MPa
T = 48.7
o
C
n
liq
= 0.582
n
vap
= 0.418
yj
i
dn
vap
n
l i q
x
i
n
l i q
dn
vap
– ( ) x
i
dx
i
– ( ) – =
K
i
x
i
dn
l i q
x
i
dn
l i q
n
l i q
dx
i
+ =
Calculation of Phase Equilibria 61
*K
7+
= 0.15 K
C7
SECOND SEPARATOR: p = 0.44 MPa
T = 48.7
o
C
n
liq
= 0.844
n
vap
= 0.156
*K
7+
= 0.15 K
C7
Component Well Stream
Composition
z
i
KFactor
K
i
Fluids
Liquid Vapor
x
i
y
i
C
1
0.4404 8.100 0.1110 0.8991
C
2
0.0432 1.650 0.0339 0.0560
C
3
0.0405 0.590 0.0489 0.0288
C
4
0.0284 0.230 0.0419 0.0096
C
5
0.0174 0.088 0.0281 0.0025
C
6
0.0290 0.039 0.0485 0.0019
C
7+
0.4011 0.003* 0.6877 0.0021
Σ · 1.0000 1.0000 1.0000
Component Feed
Composition
z
i
KFactor
K
i
Fluids
Liquid Vapor
x
i
y
i
C
1
0.1110 60.000 0.0109 0.6530
C
2
0.0339 10.600 0.0135 0.1430
C
3
0.0489 3.400 0.0355 0.1207
C
4
0.0419 1.250 0.0403 0.0504
C
5
0.0281 0.415 0.0309 0.0128
C
6
0.0485 0.170 0.0557 0.0095
C
7+
0.6877 0.013* 0.8132 0.0106
Σ · 1.0000 10000 10000
62 Calculation of Phase Equilibria
TANK: p = 0.1 MPa
T = 48.7
o
C
n
liq
= 0.9486
n
vap
= 0.0514
*K
7+
= 0.15K
C7
One mole reservoir crude oil produced amounts to n
liq1
x n
liq2
x n
liq3
= 0.582 x 0.844 x 0.9486 = 0.466 mol tank oil
4.2.2 Evaluation of KFactors Using Convergence
Pressures
As already mentioned, the K
i
factors in real systems not only depend on temperature and
pressure, but also on composition. KATZ and HACKMUTH have been the first to provide
for appropriate data of gas/oilsystems. Since then the collection of data has been
continuously enlargend.
Component Feed
Composition
z
i
KFactor
K
i
Fluids
Liquid Vapor
x
i
y
i
C
1
0.109 265.0 0.0007 0.1965
C
2
0.0135 46.6 0.0038 0.1885
C
3
0.0355 14.3 0.0210 0.3022
C
4
0.0403 5.35 0.0330 0.1775
C
5
0.0309 1.72 0.0299 0.0513
C
6
0.0557 0.70 0.0566 0.0396
C
7+
0.8132 0.052* 0.8550 0.0444
Σ · 1.0000 1.0000 1.0000
Calculation of Phase Equilibria 63
Figure 4.4: Kfactors for methanepropane at T
c
= 100
o
F (from SAGE, LACEY and
SCHAAFSMA) (1934)
The most simple relations exist in any 2component system. Figure 4.4 is a plot of the
Kfactors for the methanepropane system at the critical temperature of the mixture T
c
=
100
o
F. At any pressure  at which the vapor and the liquid phase are in equilibrium  K >
1 for the highly volatile component methane. If the pressure increases, its Kfactor will
decrease.
On the other side K<1 for the less volatile component propane since its concentration in
the vapor phase is less than in the liquid. At any isothermal change in state, K for the less
volatile component propane will equal 1 at two different pressures. One corresponds  just
as in case of the highly volatile component  with the critical pressure of the mixture and
the other one with the vapor pressure of the pure substance. Above the critical pressure of
the system, Kfactors have no longer any physical meaning.
Generally, the Kfactors increase with increasing T at constant pressure. However,
methane shows some irregular behavior. Its Kfactor increases with temperature up to a
certain limit and decreases beyond that.
According to observations, the two Kisotherms of binary system will also converge to a
certain point at other temperatures than T
c
. This point is also characterized by K = 1(y =
x). However, this convergence is only an “apparent” convergence. In contrast to the
critical isotherms, the Kvalues have no physical meaning up to the point of apparent
convergence because of the existence of only one phase beyond a certain pressure limit.
100
1.0
10.0
1500 200 300 500
0.3
3.0
0.5
5.0
log p [psia]
l
o
g
K
1000
64 Calculation of Phase Equilibria
Anyway, the isotherms may be extrapolated to K = 1 defining the pressure at this point as
the convergence pressure p
k
of the system at the actual temperature.
Figure 4.5: Comparison of Kfactors at 100
o
F for 1,000 and 5,000psia convergence
pressure (from NGAA, 1957)
It has been proved that multicomponent systems have properties very similar to those of
the binary systems with respect to the convergence of all Kisotherms. The (apparent)
convergence pressure p
k
is a function of the composition of the mixture. Figure 4.5
emphasizes the dependence of the Kfactors on the convergence pressure being a function
of the system composition. The differences in Kvalues for the two convergence pressures
at pressures below 0.7 MPa (100 psia) are not significant for the lighter hydrocarbons
ethane through pentane. It then becomes apparent that at low pressures and temperatures,
the Kfactors are nearly independent of composition.
Note:
• If the evaluation of the Kfactors is based on the convergence pressure of the mixture,
the Kfactors depend only on p and T and are independent of composition.
• At a given temperature, the convergence pressure of a mixture coincides with the
critical pressure of another mixing ratio of its components.
In 1957, the NGAA (Natural Gasoline Association of America) published a diagram
series lgK = f(lg p)
T
in its “Equilibrium Ratio Data Book”. The diagrams were based on
average values of already available data from various gas condensates and gas/crude
10 100 1000 2 2 2 3 3 3 4 4 4 5 5 5
Pressure [psia]
0.0001
0.001
0.01
0.1
1
10
100
K
C
1
C
1
C
2
C
3
C
4
C
5
C
6
C
7
C
8
C
9
C
10
Calculation of Phase Equilibria 65
oilsystems. This series includes the paraffinic homologous compounds  methane
through decane  just as ethylene, propylene, nitrogen, carbon dioxide and hydrogen
sulfide within the pressure range from 0.07 to 140 MPa and for temperatures ranging from
 148 to +260 °C.
Figure 4.6 and Figure 4.7 present the respective charts for methane and hexane,
respectively, in case of the convergence pressure p
k
= 34.5 MPa (5000 psia). The values
for this convergence pressure can be used for a large number of hydrocarbon systems.
Today, these data are already available for dataprocessing with computers.
If p
k
is known, it is possible to choose from the measured series those Kfactors , which
are best suited to describe the phase equilibrium vapor/liquid of the given system.
The second parameter, which has to be evaluated to proceed with the prediction of the
phase and volumetric behavior of complex hydrocarbon systems, is the Kfactor of the
heaviest component. In conventional analyses of hydrocarbon fluids, everything heavier
than hexane is grouped together and reported as the pseudocomponent “C
7+
”. This
fraction is a mixture of materials of varying volatility. However, the vapor pressure curves
and the critical properties of the hydrocarbons heavier than hexane are fairly close
together. Thus, it is possible to characterize the mixture “C
7+
” by an average set of
Kvalues.
66 Calculation of Phase Equilibria
Figure 4.6: Kfactors for methane, 5,000 psia convergence pressure (from NGAA,
1957)
10
1
2 5 10
2
2 5 10
3
2 5 10
4
Pressure [PSIA]
10
2
10
1
10
0
10
1
10
2
10
3
2
2
2
2
2
5
5
5
5
5
K
=
y
/
x

1
0
0

1
2
0

1
4
0

1
6
0

1
8
0

2
0
0

2
2
0

2
4
0

2
6
0
0
5
0
0
1
0
0
2
0
0
T
e
m
p
e
r
a
t
u
r
e
[
°
F
]
5
0
Calculation of Phase Equilibria 67
Figure 4.7: Kfactors for hexane, 5,000 psia convergence pressure (from NGAA,
1957)
10
1
2 5 10
2
2 5 10
3
2 5 10
4
Pressure [PSIA]
10
4
10
3
10
2
10
1
10
0
10
1
2
2
2
2
2
5
5
5
5
5
K
=
y
/
x
80
1
0
0

1
0
0
60
40
20
0
20
40
60
1
0
0
8
0
1
2
0
1
4
0
1
6
0
1
8
0
2
0
0
2
2
0
2
4
0
2
6
0
2
8
0
3
0
0
4
0
0
4
5
0
5
0
0
T
e
m
p
e
ra
t
u
re
[°
F
]
68 Calculation of Phase Equilibria
4.2.3 Evaluation of Convergence Pressure
Several iterative methods of estimating p
k
have been proposed. One of the most
convenient methods is based on the assumption of a pseudobinary system and consists of
the following steps:
• A certain convergence pressure is assumed and the Kfactors of components are
evaluated from the NGAAdiagrams (examples are given through Figure 4.6 and
Figure 4.7).
• The respective composition of the liquid phase being in equilibrium are evaluated by
a flash process (see Example 4.1).
• With exception of the lightest component methane, all other components are
summarized in a second fictitious component. The pseudocritical temperature of the
pseudocomponent C
2+
in the liquid phase is evaluated in calculating the averaged
mole weights in the liquid phase, W
i
, through
(4.28)
and the pseudocritical temperature of C
2+
through
. (4.29)
• The critical point of the component C
2+
is plotted on Figure 4.8 which shows the
vapor pressure curves of homologous hydrocarbons.
• The “binary” critical envelope curve of methane with this pseudocomponent is
drawn.
• For the given temperature, the convergence pressure p
k
can be read off because it
coincides with the critical pressure of one of the possible mixing ratios in the “binary”
system.
• If this numerical value of the convergence pressure does not coincide with the value
initially assumed, the whole procedure has to be repeated starting with the assumption
of another convergence pressure. The iterative process must go on until
correspondence of the assumed convergence pressure and the evaluated one will be
achieved.
This iterative method will be demonstrated by Example 4.4.
The Kfactors obtained through this method exhibit only satisfactory results, if the
operating pressure is considerably lower than p
k
. As the operating pressure approaches p
k
,
the Kfactors become very sensitive to the convergence pressure used and care must be
W
i
x
i
M
i
=
T
pc
k
i 2 =
∑
W
i
T
ci
k
i 2 =
∑
W
i

=
Calculation of Phase Equilibria 69
taken in the selection of the correct value of p
k
. However, charts exist only for some
distinct convergence pressure (800, 1000, 1500, 2000, 3000, 5000 and 10,000 psia). Thus
the interpolation  inevitable in estimating the Kvalues  may lead to great errors.
4.2.4 Evaluation of by use of PENGROBINSON
equation of state
With help of the “successive solution method”, the Kfactors of the components of the
system can be evaluated by an iterative process like the one presented through Example
4.5:
• The composition of the system, the pressure and temperature as well as the critical
data of the components, their acentric factors and the interactive coefficients must be
initially known.
• On the basis of assumed Kfactors, the phase compositions are calculated through an
iterative calculation procedure (see Example 4.1).
• On the basis of the calculated phase compositions, the EOSparameters a, b, A, and B
after PENGROBINSON are calculated for both phases.
• The Zfactors both for the liquid and the vapor phase can be evaluated solving the
equation of state (Eq. 3.49) through use of the method of CARDAN (see Example 3.3).
• The fugacity coefficients for both phases are evaluated by use of Eq. 4.14 and Eq.
4.15:
and
.
• New Kvalues can be evaluated by inserting these values in Eq. 4.16:
• On the basis of the evaluated Kvalues, the calculation procedure is repeated. The
steps described above will be repeated until the Kfactors converge within a certain
limit.
• The last flash iteration results in the evaluation of the mole number in the liquid
φ
i
ln
b
i
b
 Z 1 – ( ) Z B – ( ) ln –
A
2 2B

2
k
i 1 =
x
i
a
ij
∑
a

b
i
b

–
Z 2 414B ⋅ +
Z 0.414B –

¸ ,
¸ _
ln – =
a
i j
1 K
i j
– ( ) a
i
0.5
a
j
0.5
( ) =
K
i
φ
i l i q ,
φ
i vap ,

=
70 Calculation of Phase Equilibria
phase, n
liq
, the mole number in the vapor phase n
vap
and in the evaluation of the
composition of the phases.
• Using Z
liq
and Z
vap
, the volume and the density of the two phases can be calculated,
too.
Figure 4.8: Convergence pressure data  methane for binary hydrocarbon mixtures
(from WINN, 1952)
200 300 400 1000 2000 4000 6000 600
Convergence Pressure [psia]
T
e
m
p
e
r
a
t
u
r
e
[
°
F
]
100
0
100
200
300
400
500
600
700
800
900
C
3
C
4
C
7
C C
7 23
CC
5 24
C
16
C
10
C
8
iC
5
C
5
C
6
iC
4
C
4
C*
C
3
C*
C
2
C*
C
2
C
1
Calculation of Phase Equilibria 71
Example 4.4
Evaluation of the convergence pressure by the iterative NGAA
process described above. The composition of the system is tabled
below, and the temperature and the pressure are fixed as T = 49
o
C
and p = 15.5 MPa.
FIRST ITERATION: p
k
= 26.7 MPa (4000 psia)
n
liq
= 0.5876
n
vap
= 0.4124
*assumption:
On Figure 4.8 T
pc
= 514.06
o
F is close to T
c
of pure nheptane. For
the given temperature of the system (T = 49
o
C and 120
o
F,
respectively), the convergence pressure p
k
can be read off on the
envelope curve between methane and nheptane at
p
k
= 3150 psia = 21.72 MPa.
This value does not coincide with the initially assumed value. Thus
a second iteration must be done.
SECOND ITERATION: p
k
= 20.4 MPa (3000 psia)
Component z
i
M
i
kg/kmole
T
c,i
K
K
i
x
i
W
i
=x
i
M
i
W
i
T
ci
C
1
0.6973   1.6500 0.5499  
C
2
0.0318 30.07 305.4 0.9000 0.0332 0.998 304.79
C
3
0.0194 44.09 369.8 0.6300 0.0229 1.010 373.50
nC
4
0.0223 58.12 425.2 0.4300 0.0292 1.697 721.56
nC
5
0.0221 72.15 469.7 0.2250 0.0325 2.345 1101.45
nC
6*
0.0309 86.17 507.4 0.1700 0.0470 4.050 2054.97
C
7*
0.1762 114.22 568.8 0.0720 0.2854 32.598 18541.74
Σ 1.0000 42.698 23098.01
C
7+
nC
8
≈
T
pc
k
i 2 =
∑
W
i
T
c i ,
k
i 2 =
∑
W
i

23098.01
42.698

540.96 K 514.06 F
o
= = = =
72 Calculation of Phase Equilibria
n
liq
= 0.5604
n
vap
= 0.4496
*assumption:
From Fig. 4.8:
and thus
.
The agreement with the assumption is now satisfactory. Further
refinement is impossible (Kvalues only evaluated for 1000 psia
steps in p
k
).
Example 4.5
Evaluation of the Kfactors of the three components of a system at
p = 1.38 MPa and T = 338.7 K. The successive solution method is
used on the basis of the composition of the system, the critical
and reduced data as well as the acentric factors which are given
below in Table I. Table II presents the interaction coefficients
K
ij
which must be known, too.
Component K
i
x
i
W
i
= x
i
M
i
W
i
Tc
i
C
1
1.700 0.5332  
C
2
0.8800 0.0336 1.010 308.45
C
3
0.6200 0.0233 1.027 379.78
nC
4
0.4200 0.0299 1.738 739.00
nC5 0.2200 0.0336 2.424 1138.55
nC
6*
0.1800 0.0483 4.162 2111.80
C
7*
0.0700 0.2980 34.038 19360.81
Σ 1.0000 44.399 24038.39
C
7+
nC
8
≈
k
i=2
W
i
T
c i ,
∑
k
i=2
W
i
∑

24038.39
44.399

541.42K 514.9 F
o
= = =
pk 3150psia =
p
k
21.72MPa =
Calculation of Phase Equilibria 73
Table 4.1: Composition of the System, Critical Data and Acentric Factors
For the first iteration of a flash process,Kfactors as well as
the mole numbers in the phases, n
liq
and n
vap
, are assumed. The gas
and the liquid phase compositions are calculated through an
iterative calculation procedure (see Example 4.1). The assumed
Kfactors as well as the evaluated phase compositions are given in
Table 4.3.
Then the flash iteration starts with the evaluation of a and b. At
first the parameters b and a for each component have to be
evaluated through
Component Composition
z
i
Critical Data
T
c
K p
c
MPa
Reduced Data
T
r
p
r
Acentric Factor
ω
i
C
3
0.61 369.8 4.245 0.916 0.325 0.152
n  C
4
0.28 425.2 3.800 0.797 0.363 0.193
n  C
5
0.11 469.6 3.374 0.721 0.409 0.251
Table 4.2: Interaction Coefficients of the Components
C
3
n  C
4
n  C
5
C
3
0.000 0.001 0.003
n  C
4
0.001 0.000 0.001
n  C
5
0.003 0.001 0.000
Table 4.3:
FIRST ITERATION: n
liq
= 0.52
n
vap
= 0.48
Component KFactor
C
3
1.745 0.440 0.767
n  C
4
0.537 0.369 0.198
n  C
5
0.182 0.191 0.035
Σ 1.000 1.000
x
i
z
i
n
v
K
i
n
L
+

= y
i
z
i
n
L
K
i
⁄ n
v
+

=
74 Calculation of Phase Equilibria
,
(see Eq. 3.46),
,
(see Eq. 3.48),
,
(see Eq. 3.52) and through
.
(see Eq. 3.47).
The results of these calculations are tabled below (see Table 4.4).
By use of Eq. 3.53
and Eq. 3.56
,
a and b for the liquid phase as well as for the vapor phase are
calculated on the basis of phase compositions evaluated through
the first evaluation (see Table 4.3.) The calculation procedure is
presented through Example 3.2.
Then the parameters A and B are calculated after PENGROBINSON
using Eq. 3.35
Table 4.4: First Iteration
Component b
i
a
i
C
3
5.635 E5 1.072
n  C
4
7.247 E5 1.729
n  C
5
9.003 E5 2.557
b
i
0.0778
RT
c
p
c
 =
a
c i ,
0.45724
R
2
T
2
p
c
 =
α
i
0.5
1 0.37646 1.54226ω 1.54226ω
2
– + ( ) 1 T
r
0.5
– ( ) + =
a
i
a
c
α ( )
i
=
b x
i
b
i
i 1 =
k
∑
=
a x
i
x
j
a
i
a
j
( )
0.5
1 K
i j
– ( )
j 1 =
k
∑
i 1 =
k
∑
=
Calculation of Phase Equilibria 75
and Eq. 3.36
.
The evaluated numerical values of the parameters a, b, A and B for
both phases are summarized in Table 4.5.
Now the Zfactors both for the liquid and the vapor phase can be
evaluated solving the equation of state given by Eq. 3.49:
.
To solve this cubic equation of state, the CARDAN equation is
applied for
where
,
and
.
The factors r, s and t have to be evaluated both for the liquid
and the vapor phase.
By substitution of
,
the formula
is reduced to
Table 4.5: First Iteration
a b A B
Liquid Phase 1.551 6.873 E5 0.2699 3.368 E2
Vapor Phase 1.232 6.072 E5 0.2144 2.975 E2
A
ap
R
2
T
2

=
B
b p
R T

=
Z
3
1 B – ( ) Z
2
– A 2B – 3B
2
– ( )Z AB B
2
– B
3
– ( ) 0 = – +
x
3
rx
2
sx t 0 = + + +
r 1 B – ( ) – =
s A 2B 3B
2
– – ( ) =
t AB B
2
B
3
– – ( ) – =
x y
r
3

– =
x
3
rx
2
sx t 0 = + + +
76 Calculation of Phase Equilibria
where
and
.
The discriminant D is defined as
Contrary to Example 3.3, D < 0 both in case of the liquid phase and in case of the vapor
phase. Thus the cubic equation of state has three real solutions, respectively. In defining
and
,
the three solutions for the reduced form
are given by
and
.
Introducing y
1
, y
2
, and y
3
into
,
the Zfactors(=x) of the phases can be evaluated. The solutions
are summarized in Table 4.6.
y
3
py q 0 = + +
p
3s r
2
–
3

=
q
2r
3
27

rs
3

– t + =
D
p
3

¸ ,
¸ _
3
q
2

¸ ,
¸ _
2
+ =
ς
p
3
27

–
2
=
ϕ cos
q
2ς

– =
y
3
py q 0 = + +
y
1
2 ς
3
ϕ 3 ⁄ ( ) cos =
y
2
2 ς
3
ϕ 3 2π 3 ⁄ + ⁄ ( ) cos =
y
3
2 ς
3
ϕ 3 4π 3 ⁄ + ⁄ ( ) cos =
x y
r
3

– =
Calculation of Phase Equilibria 77
Table 4.6: First Iteration Solution f or the ZFactors after CARDAN
Obviously, the actual Zfactor of the liquid phase must be
identical with the smaller one of the three solutions for the
liquid phase, meanwhile the actual Zfactor for the vapor phase is
equal to the greatest one of the three solutions for the vapor
phase. Thus the first iteration results in
and
.
The Zfactors are introduced into Equation 4.15
,
respectively, to evaluate the fugacity coefficients (of
propane), , of nbutane) and (of npentane) both for the
liquid and the vapor phase. The results of this calculation
procedure are summarized in Table 4.7.
As a result of the first iteration flash, Kfactors are evaluated
through Eq. 4.16
and summarized in Table 4.8
Z
1
Z
2
Z
3
Liquid Phase 0.0523 0.2172 0.6968
Vapor Phase 0.0526 0.1325 0.7852
Table 4.7: First Iteration Fugacity Coefficients
φ
3
φ
4
φ
5
Liquid Phase 1.2580 0.4707 0.1831
Vapor Phase 0.8473 0.7519 0.6672
Z
l i q
0.0523 =
Z
vap
0.7852 =
φ
i
ln
b
i
b
 Z 1 – ( ) Z B – ( ) ln –
A
2 2B

2 x
i
a
ij
i 1 =
k
∑
a

b
i
b

–
Z 2 414B ⋅ +
Z 0.414B –

¸ ,
¸ _
ln – =
φ
3
φ
4
φ
5
K
i
φ
i liq ,
φ
i vap ,

=
78 Calculation of Phase Equilibria
Table 4.8: First Iteration
These Kfactors are the basis of the second iteration which starts
again with the iterative calculation procedure presented through
Example 4.1 (see Table 4.9)
On the basis of the data in Table 4.4 and in Table 4.9 the numerical
values of the parameters a, b, A and B for both phases are
evaluated again and summarized in Table 4.10.
Now the Zfactors both for the liquid and the vapor phase can be
evaluated solving the equation of state given by Eq. 3.49
through the CARDAN equation. Again D < 0 both in case of the liquid
phase and in case of the vapor phase. Correspondingly, the cubic
equation of state has three real solutions. The solutions are
summarized in Table 4.11.
Table 4.9:
SECOND ITERATION: n
liq
= 0.556
n
vap
= 0.444
Components KFactor
C
3
1.4847
n  C
4
0.6260
n  C
5
0.2744
Component KFactor
C
3
1.4847 0.5020 0.7453
n  C
4
0.6260 0.3358 0.2102
n  C
5
0.2742 0.1623 0.0445
Σ 1.0001 1.0000
Z
3
1 B – ( )Z
2
– A 2B 3B
2
– – ( )Z AB B
2
B
3
– – ( ) 0 = – +
x
i
z
i
n
v
K
i
n
L
+

= y
i
z
i
n
L
K
i
n
v
+ ⁄

=
Calculation of Phase Equilibria 79
Table 4.10: Second Iteration
Once more again, the actual Zfactor of the liquid phase must be
identical with the smallest one of the three solutions for the
liquid phase, meanwhile the actual Zfactor for the vapor phase is
equal to the greatest one of the three solutions for the vapor
phase. Thus the second iteration results in
Z
liq
= 0.0521
and
Z
vap
= 0.7812.
The Zfactors are introduced into Eq. 4.15
,??????
respectively, to evaluate the fugacity coefficients (of
nbutane)and (of npentane) both for the liquid and the vapor
phase. The results of this calculation procedure are summarized in
Table 4.12.
Table 4.12: Second Iteration
Fugacity Coefficients
a b A B
Liquid Phase 1.490 6.725 E5 0.2592 3.296 E2
Vapor Phase 1.249 6.116 E5 0.2173 2.997 E2
Table 4.11: Second Iteration Solution for the ZFactors, after CARDAN
Z
1
Z
2
Z
3
Liquid Phase 0.0521 0.1988 0.7161
Vapor Phase 0.0525 0.1364 0.7812
φ
3
φ
4
φ
5
Liquid Phase 1.2520 0.4717 0.1846
Vapor Phase 0.8477 0.7511 0.6655
φ
i
ln
b
i
b
 Z 1 – ( ) Z B – ( ) ln –
A
2 2B

2
k
i 1 =
x
i
a
i j
∑
a

b
i
b

–
Z 2 414B ⋅ +
Z 0.414B –

¸ ,
¸ _
ln – =
φ
4
φ
5
80 Calculation of Phase Equilibria
The Kfactors are evaluated through Eq. 4.16
and given in Table 4.13
Table 4.13: Second Iteration
KFactors
The Kfactors from the two flash iterations exhibit satisfactory
convergence. On the basis of the Kfactors given in Table 4.13, a
final iterative calculation procedure analogous to Example 4.1
results in the mole numbers and in the composition of the phases.
Table 4.14:
Components KFactor
C
3
1.4769
n  C
4
0.6280
n  C
5
0.2773
FINAL RESULTS: n
liq
= 0.562
n
vap
= 0.438
Component KFactor
C
3
1.4769 0.5046 0.7452
n  C
4
0.6280 0.3345 0.2101
n  C
5
0.2773 0.1609 0.0446
Σ 1.0000 0.9999
K
i
φ
i liq ,
φ
i vap ,

=
x
i
z
i
n
v
K
i
n
L
+

= y
i
z
i
n
L
K
i
n
v
+ ⁄

=
81
Chapter 5
Phase Properties
5.1 Natural Gases
5.1.1 Volume
As previously discussed in Chap. 3, the state of ideal gases is defined by the following
equation of state:
(5.1)
where R is the universal gas constant and n is the number of moles. The value R depends
on the system of units. In the SIsystem:
R = 8.31432 J / K mole.
In case of ideal gases, no intermolecular forces are acting. At any isothermal change of
state, the decrease in volume is directly proportional to the increase in pressure. It has
been observed that in many cases  e.g. in case of natural gases  the volume becomes less
than half if the pressure is doubled. This behavior of a gas in called a real one.
Below the BOYLE temperature, the real gases are more compressible than an ideal one,
because of intermolecular forces of attraction and repulsion. If the pressure is low (the
forces of attraction can be neglected because of the great distance of molecules) or in case
of high temperatures (dominance of the kinetic energy of the molecules), real gases
behave similarly to perfect gases.
In case of real gases, the change in state can be well described by the equations of state
which are also presented in Chap. 3 (VAN DER WAALS, REDLICHKWONG,
PENGROBINSON). In the petroleum industry, the basic relation of Eq. 3.12
pV = ZnRT
pV nRT =
82 Phase Properties
is commonly used because of its easy applicability.
The Zfactor is the so called “deviation factor” or “compressibility factor”. It is defined as
the ratio of the volume of a real gas to the volume of a perfect one at identical temperature
and pressure.
By introducing
where
m: mass
M: molecular weight,
pV = ZnRT can be transformed into
(5.2)
if m = 1[g] gas
The reciprocal of the specific volume V is the density and so
. (5.3)
The atmospheric density of the gas is often related to the atmospheric density of air at the
same standard temperature. In doing so, the “specific gravity” of the gas is defined by
(5.4)
where the subscript “a” indicates the values for air and the superscript “o” those at
standard conditions (atmospheric pressure).
If Eq. 5.4 is combined with Eq. 5.2, the gas density results in
(5.5)
As expected, the Zfactor depends both on pressure and temperature. Zfactors of different
gases, which can be determined experimentally, vary considerably. Figure 5.1 shows the
Zfactor of methane, ethane and propane as a function of the pressure.
n
m
M

=
pV
ZRT
M

=
ρ
g
1
V

pM
ZRT

= =
γ
g
γ
g
ρ
g
0
ρ
g a ,
0

M
M
a

M
28.97

= = =
ρ
g
1
V

M
a
R
γ
g
p
ZT
 0.003473γ
g
p
ZT
 = = =
Phase Properties 83
Figure 5.1: Zfactor of methane, ethane and propane versus pressure at T = 140
o
F
(from STANDING, 1977)
5.1.2 Formation Volume Factor
In petroleum industry, the term “formation volume factor” is widely used. It expresses
which volume is taken up by one cubic meter (at standard state conditions, p
o
and V
o
) at
a pressure p and a temperature T.
The formations volume factor for gases, B
g
, is defined by
(5.6)
where the standard conditions p
o
and T
o
are generally defined in
Europe: p
o
: 1.00 bar
T
o
: 273.15 K
and in
USA: p
o
: 14.7 psia (1.01325 bar)
T
o
: 60.0
o
F (15
o
C).
In case of one mole, Eq. 5.3 results for standard conditions in
(5.7)
If Eq. 5.3 and Eq. 5.7 are substituted in Eq. 5.6, then:
Pressure [MPa]
0 4 8 12 16 20
0
0.2
0.4
0.6
0.8
1.0
z
p
V
n
R
T
Methane
P
r
o
p
a
n
e
E
t
h
a
n
e
B
g
B
g
p T , ( )
V
V °

= =
p°V ° Z
°
RT
°
=
84 Phase Properties
(5.8)
where
.
5.1.3 Compressibility
On the basis of the gas law, the isothermal gas compressibility results in
. (5.9)
For ideal gases: Z = 1, i.e. .
Therefore, the gas compressibility is only a function of pressure and is inversely
proportional to the pressure.
5.1.4 Correlation of ZFactor on the Basis of Reduced
State Variables
In general, the components of natural gases are homologous paraffins with similar
molecular structures and the theorem of corresponding states can be applied. Figure 5.2
shows the excellent correspondence of the Zfactors of methane, propane and pentane as
a function of reduced pressure for a series of reduced temperatures.
B
g
p,T ( )
ZT
p
 C =
C
p
o
Z
o
T
o

=
c
g
1
V

∂V
∂p

¸ ,
¸ _
T
1
B
g

∂B
g
∂p

¸ ,
¸ _
T
1
p

1
Z

∂Z
∂p

¸ ,
¸ _
– = – = – =
∂Z
∂p
 0 =
Phase Properties 85
Figure 5.2: Zfactor as a function of reduced pressure for a series of reduced
temperatures (from SAGE and LACEY, 1949)
The theorem of corresponding states can also be applied to gas mixtures through the
evaluation of “pseudocritical” data, T
pc
and p
pc
, which are given by
(5.10)
and
(5.11)
where
p
ci
: critical pressure of component i
T
ci
: critical temperature of component i
y
i
: mole fraction of component i in the gas phase.
Then the “pseudoreduced” data, T
pr
and p
pr
, can be evaluated:
, (5.12)
(5.13)
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
Reduced Pressure, p
r
z
p
V
n
R
T
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
CH
4
C H
3 8
C H
5 12
T=0.9
r
CH
4
C H
3 8 C H
5 12
T=1.0
r CH
4
CH
4
CH
4
CH
4
C H
3 8
T=1.1
r
1
1
3
3
5
5
T=1.2
r
T=1.3
r
T=1.5
r
C H
3 8
C H
5 12
C H
5 12
C H
5 12
C H
3 8
C H
3 8
C H
5 12
T
pc
y
i
T
c,i
i 1 =
k
∑
=
p
pc
y
i
p
ci
i 1 =
k
∑
=
T
pr
T
T
pc

=
p
pr
p
p
pc

=
86 Phase Properties
where
T: operating temperature
p: operating pressure
Note that both T and T
pc
in Eq. 5.12 must be absolute temperatures, either given in K
(Kelvin, SI unit) or in R (Rankin, Field Unit).
The definition of corresponding states on the basis of the pseudoreduced data results in
a generalized plot of Zfactors for natural gases (see Figure 5.3). This plot can be used to
evaluate the Zfactor of a natural gas on the basis of its composition and the critical data
(see Example 5.1).
Several natural gases contain considerable amounts of heavy components which will be
combined  depending on the feasibility of chemical analysis  to the pseudo components
C
7+
or C
9+
. The compositions of these mixtures cannot exactly be defined. In the
laboratory practice, the C
7+
fraction is obtained as liquid residue from a fractional
destillation and two properties  averaged molecular weight and specific gravity  are
determined. These properties have been correlated with the pseudocritical data of this
fraction (see Figure 5.4). On the basis of these correlations, the engineer is able to
calculate pseudoreduced properties of natural gases for which conventional analyses are
available.
In many cases, the data of fractional distillation are not available. Anyway, several
diagrams have been published which correlate the pseudocritical functions of state with
the gas gravity. By means of these diagrams, the Zfactor can be determined even if the
gas composition is not known. These diagrams are only valid for a given quality or only
for one area, formation, reservoir, basin, etc. The respective diagram for Oklahoma City
gases is given in Figure 5.5 (see also Example 5.2.)
Phase Properties 87
Figure 5.3: Zfactor for natural gases (from BROWN et al., 1948)
7
0.9
1.0
1.1
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
8 9 10 11 12
13 14 15
Pseudo Reduced Pressure
C
o
m
p
r
e
s
s
i
b
i
l
i
t
y
F
a
c
t
o
r
Z
0 1 2 3 4 5 6 7 8
Pseudo Reduced Pressure
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
0.95
1.0
1.1
C
o
m
p
r
e
s
s
i
b
i
l
i
t
y
F
a
c
t
o
r
Z
0.25
3.0
2.8
2.6
2.4
2.2
2.0
1.9
1.8
1.7
1.6
1.5
1.45
1.4
1.35
1.3
1.25
1
.2
1
.
1
5
1
.1
1
.
0
5
Pseudo Reduced
Temperature
88 Phase Properties
Figure 5.4: Pseudocritical temperatures and pressures for heptanes and heavier
(from MATTHEWS et al, 1942)
Figure 5.5: Pseudocritical properties of Oklahoma City Gases
(from MATTHEWS et al., 1942)
300
400
500
1.000
1.100
1.200
1.300
1.400
200
120 140 160 180 200 220 100 240
Molecular Weight
P
s
e
u
d
o
C
r
i
t
i
c
a
l
P
r
e
s
s
.
[
p
s
i
a
]
P
s
e
u
d
o
C
r
i
t
i
c
a
l
T
e
m
p
.
[
°
R
]
Sp. gr.
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Gas Gravity
P
s
e
u
d
o
C
r
i
t
i
c
a
l
T
e
m
p
e
r
a
t
u
r
e
[
°
R
]
P
s
e
u
d
o
C
r
i
t
i
c
a
l
P
r
e
s
s
u
r
e
[
i
b
/
s
q
i
n
.
]
a
b
s
o
l
u
t
e
700
675
650
625
600
575
550
525
500
475
450
425
400
375
350
325
Phase Properties 89
Example 5.1
Evaluation of the density of a natural gas at T = 105
o
C and at p
= 22.55 MPa. The Zfactor is determined by using Figure 5.3 The
table below summarizes the composition of the gas, the critical
data as well as the calculated reduced data of the components.
assumption:
From Figure 5.3
Z = 0.925
which is introduced in
where
R = 8.31432[J°K
1
mol
1
].
Component M
i
kg/kmole
y
i
y
i
M
i
kg/kmole
T
c,i
K
y
i
x T
ci
K
p
ci
MPa
y
i
x p
ci
MPa
C
1
16.0 0.9443 15.109 190.6 179.98 4.60 4.344
C
2
30.1 0.0158 0.476 305.4 4.83 4.88 0.077
C
3
44.1 0.0132 0.582 369.8 4.88 4.25 0.056
iC
4
58.1 0.0051 0.296 408.1 2.08 3.65 0.019
nC
4
58.1 0.0053 0.308 425.2 2.25 3.79 0.020
iC
5
72.2 0.0023 0.166 461.4 1.06 3.38 0.008
nC
5
72.2 0.0021 0.152 469.7 0.99 3.37 0.007
C
6
86.2 0.0033 0.284 507.4 1.67 3.01 0.010
114.23 0.0086 0.982 568.8 4.89 2.48 0.021
Σ 1.000 18.355 202.63 4.562
C
*
7+
C
7+
C
8
≈
T
pr
T
T
pc

105 + 273.15
202.63

1.87 = = =
p
pr
p
p
pc

22.5
4.562

4.93 =
¸ ,
¸ _
= =
ρ
g
p
k
i 1 =
y
i
M
i
∑
ZRT

=
90 Phase Properties
Example 5.2
Evaluation of the density and the formation volume factor of an
Oklahoma City Gas of the specific gravity at
temperature T = 90.5
o
C(655
R)and pressure p = 12.5 MPa(1813 psia).
At first the pseudo critical data are read off from Figure 5.5 as
and
.
Now the pseudo reduced data are calculated as
and
.
On the basis of the pseudoreduced data,
results from Figure 5.3.
Because
,
where
M
a
= 28.97 E3 kg/mole,
and
R = 8.31432J/K mole,
the density can be calculated through
so that
ρ
g
225.5 E 5 18.355 E  3 × +
0.925 8.31432 378.15 × ×
 kg m
3
⁄ =
ρ
g
142.32kg m
3
⁄ =
γ
g
0.7018 =
T
pc
390R =
p
pc
670psia =
T
pr
T
T
pc

655
390
 1.68 = = =
p
pr
p
p
pc

1813
670
 2.71 = = =
Z 0.86 =
ρ
g
M
a
R
 γ
g
p
ZT
 =
ρ
g
0.02897 0.7018 125 × × E 5 +
8.31432 0.86 363.65 × ×
 kg m
3
⁄ =
Phase Properties 91
= 97.774 kg/m
3
.
The formation volume factor can be evaluated through Eq. 5.8
where
.
Assuming that the gas is a perfect one at European standard
conditions
and
so that Z = 1.0, its formation volume factor results in
,
.
ρ
g
B
g
p,T ( )
ZT
p
 C =
C
p
o
Z
o
T
o

=
p
o
1bar =
T
o
273.15 K =
B
g
0.86 (90.5 + 273.15) ×
12.5

0.1
273.15
 =
B
g
0.00916 =
92 Phase Properties
Figure 5.6: Water content of natural gas in equilibrium with liquid water
(from KATZ et al., 1959)
70 50 30 10 10 30 50 70 90 110 130 150 170 190 210 230 250 270 290 320 340 360 380400 440 480 520 560 600 640 680 720
10
0
10
1
2
5
10
2
2
5
10
3
2
5
10
4
2
5
2
5
W
a
t
e
r
c
o
n
t
e
n
t
I
b
H
O
p
e
r
m
i
l
l
i
o
n
c
u
f
t
t
o
t
a
l
g
a
s
2
Temperature, °F
Saturation 100% HO, 47,489 Ib per mm cu ft
2
1
4
.
7
p
s
i
a
3
0
5
0
1
0
0
2
0
0
4
0
0
5
0
0
6
0
0
8
0
0
1
0
0
0
1
2
5
0
1
5
0
0
2
0
0
0
2
5
0
0
3
0
0
0
3500
4000
5000
6000
8000
10,000
Phase Properties 93
5.1.5 Water Content
The conventional gas analysis does not indicate any water content. However, the natural
gas in a reservoir is in thermodynamic equilibrium with the interstitial water. Therefore
the gas must be assumed to be saturated with steam. The mole fraction of the steam, y
w
,
can be calculated by the DALTON law (see Eq. 5.7).
(5.14)
where
p
w
: vapor pressure of water at operating temperature
p: operating pressure of the system.
From numerous measurements it could be deduced that the steam content can be well
correlated with pressure and temperature (see Figure 5.6 and Example 5.3). On the other
hand, the influence of the gas composition has been found to be insignificant.
Natural gas and liquid water will combine to form solids resembling wet snow at
temperatures somewhat above the crystallization temperature of pure water. These solids
are called “gas hydrates”. Because these solids can form at temperatures and pressures
normally encountered in naturalgas pipelines, this phenomenon is of particular interest
to the petroleum industry. The conditions for the gas hydrate formation will be
summarized later on.
Example 5.3
The evaluation of the water content of a natural gas as a function
of operating pressure and temperature. This can be done by
correlation through Figure 5.6
FIRST STATE: p = 13.8 MPa(2000 psia)
T = 93.3
o
C (200
o
F)
The water content at this state is evaluated through Figure 5.6 as
.
SECOND STATE: p = 3.45 MPa (500 psia)
T = 60
o
C (140
o
F)
The water content at this state was evaluated through Figure 5.6 as
.
THIRD STATE: p = 2.8 MPa (400 psia)
y
w
p
w
p

=
3.8 E 4lb cu ft ⁄ 6.1 E 3kg m
3
⁄ – = –
3.25 E 4 lb cu ft ⁄ 5.2 E 3kg m
3
⁄ – = –
94 Phase Properties
T = 60
o
C (140
o
F)
The water content at this state was evaluated through Figure 5.6 as
5.1.6 Viscosity
Like all intensive physical properties, the viscosity is completely described by
(5.15)
in case of natural gases and
(5.16)
in case of hydrocarbon liquids.
Eq. 5.15 and Eq. 5.16 simply state that the viscosity is a function of pressure, temperature
and composition. LOHRENZ et.al. (1964) presented respective calculations which are
useful for
• determining viscosities of reservoir fluids in compositional material balance
computations and
• predicting the decrease in viscosity which occurs when gases dissolve in reservoir
oils.
The correlations presented below may be viewed as modifications of Eq. 5.15. The
assumptions are practical because the composition is frequently not known. Anyway, the
assumptions are sufficiently valid so that these correlations are frequently used for
reservoir engineering computations.
The gas viscosity µ
g
increases with the temperature. Figure 5.7 shows the dependence of
the viscosity of pure gases on the temperature at atmospheric pressure (0.1 MPa and 14.7
psi, respectively).
Figure 5.8 presents the viscosity of natural gases at atmospheric pressure as a function of
the averaged molecular weight. This diagram elucidates that the viscosity decreases along
with increasing molecular weight. It also contains correction functions for CO
2
, N
2
and
H
2
S.
The theorem of corresponding states has been proved to be valid once more again. The
viscosity at any state of the gaseous system can be predicted through the reduced
viscosity which is defined by
3.8 E 4 lb cu ft ⁄ 6.1 E 3 lb cu ft ⁄ – = –
µ F p T y
1
,y
2
, y
3
…y
k 1 –
, , ( ) =
µ F p T x
1
,x
2
, x
3
…x
k 1 –
, , ( ) =
Phase Properties 95
. (5.17)
The procedure to evaluate µ at certain p and T is the following (see Example 5.4):
• From Figure 5.8, the atmospheric viscosity µ
g,1
has to be evaluated for the given
gravity (or averaged molecular weight) of the gas and the operating temperature.
• According toFigure 4.4 and Figure 5.5, the pseudocritical pressure, p
pc
, and the
pseudocritical temperature, T
pc
, are determined.
• After the calculation of the pseudoreduced state quantities p
pr
and T
pr
, the reduced
viscosity µ
g,r
, which is given by the ratio between the actual viscosity and the one at
atmospheric pressure, is evaluated by use of Figure 5.9.
• The actual gas viscosity can be calculated through Eq. 5.17 where
The gas viscosity determined through this method deviates from laboratory
measurements by 3% at the maximum. This is a satisfactory solution for practical
purposes.
Figure 5.7: Viscosity of paraffin hydrocarbon gases natural gases at atmospheric
pressure (from CARR et al., 1954)
µ
g r ,
p
pr
T
pr
, ( )
µ
g
p T ( , )
µ
g 1 ,
p 1 T , = ( )

=
µ
g
µ
g r ,
µ
g 1 ,
=
50 100 150 200 250 300 350 400
Temperature [°F]
V
i
s
c
o
s
i
t
y
[
c
P
]
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.018
0.020
0.022
0.024
H
e
l
i
u
m
A
i
r
N
i
t
r
o
g
e
n
C
a
r
b
o
n
D
i
o
x
i
d
e
H
y
d
r
o
g
e
n
S
u
l
f
i
d
e
M
e
t
h
a
n
e
E
t
h
a
n
e
P
r
o
p
a
n
e
n
O
c
ta
n
e
n
D
ec
a
n
e
n

N
o
n
a
n
e
i
B
u
ta
n
e
n
B
u
ta
n
e
96 Phase Properties
Figure 5.8: Viscosity of natural gases at atmospheric pressure (from CARR et al, 1954)
Figure 5.9: Correlation of viscosity ratio with pseudoreduced pressure and
temperature (from CARR et al., 1954)
0.0020
0.0010
0.0015
0
0
C
o
r
r
e
c
t
i
o
n
o
d
d
e
d
t
o
V
i
s
c
o
s
i
t
y
[
c
P
]
5 10
Mol e % HS
2
15
H S
2
γg
=
2
.0
1.0
1
.5
0.6
0.0020
0.0010
0.0015
0
0
C
o
r
r
e
c
t
i
o
n
o
d
d
e
d
t
o
V
i
s
c
o
s
i
t
y
[
c
P
]
5 10
Mole % CO
2
CO
2
γ g
=
2
.
0
1
.0
0
.6
1
.
5
0.0020
0.0010
0.0015
0
0
C
o
r
r
e
c
t
i
o
n
o
d
d
e
d
t
o
V
i
s
c
o
s
i
t
y
[
c
P
]
5 10
Mole % N
2
15
N
2 γ g
=
2
.
0
1
.
5
0
.6
1
.
0
15
0
0.004
0.005
0.006
0.007
0.008
0.009
0.010
0.011
0.012
0.013
0.014
0.015
0.016
0.5 1.0 1.5 2.0 2.5 3.0 3.5
10 20 30 40 50 60 70 80 90
Molecular Weight
Gas Gravity (Air = 1.000)
V
i
s
c
o
s
i
t
y
a
t
1
[
a
t
m
]
[
c
P
]
µ
1
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
V
i
s
c
o
s
i
t
y
R
a
t
i
o
,
µ
/
µ
1
0.1 0.2 0.4 0.6 1.0 2 4 3 6 8 10 20 0.3
Pseudo Reduced Pressure, p
r
P
s
e
u
d
o
R
e
d
u
c
e
d
T
e
m
p
e
r
a
t
u
r
e
,
T
r
0
.8
0
.
9
1
.
0
0
1
.
0
5
1
.
1
0
1
.
1
5
1
.
2
0
1
.
3
0
1
.
4
0
1
.
5
0
1
.
7
5
2
.
0
0
2
.
2
5
2
.
5
0
3
.
0
0
Phase Properties 97
Example 5.4
Determination of the gas viscosity at temperature T = 90.5
o
C
(194.9
o
F, 654.7
R) and at pressure p = 12.5 MPa (1800 psia). The
specific gas gravity was found as .
From Figure 5.8, the viscosity at atmospheric pressure and
operating temperature can be read off as
.
Then the pseudocritical data of the system are evaluated from
Figure 5.5 where
p
pc
= 670 psia
and
T
pc
= 385
o
R.
The calculation of the pseudoreduced data of the system result in
and
.
The ratio between the actual gas viscosity, µ
g,,
and its viscosity
at atmospheric pressure, µ
g,1
, can now be read off from Figure 5.9:
so that
and
,
respectively.
γ
g
0.7018 =
µ
g 1 ,
0.0122cP =
p
pr
p
p
pc

1800
670
 2.69 = = =
T
T
pc

654.6
385
 1.70 = =
µ
g r ,
1.35 =
µ
g
µ
g r ,
µ
g 1 ,
× 1.35 0.0122 0.0165 = × cP = =
µ
g
0.0165 mPa s =
98 Phase Properties
5.2 Hydrocarbon Liquids
5.2.1 Volume
For practical purposes, it is sufficient to consider hydrocarbon liquids as ideal mixtures
whose volumes can be calculated by adding up the volumes of the components.
The interesting properties (density, viscosity, etc.), which depend on the oil composition,
can be easily determined at the standard (normal) state (atmospheric pressure and T =
60
o
F (15
o
C)). The theorem of corresponding states can be applied again to determine the
properties at any other state of the system (e.g. at reservoir conditions).
However, a crude oil does not remain in the singlephase state at “normal” conditions if
it contains any high mole fraction of light components which will be liberated at
atmospheric pressure and surface temperature. This inconvenience can be avoided by
attributing “apparent densities” or “apparent specific volumes” at standard conditions to
the dissolved and easily volatilized components. Then the respective “pseudoliquid” data
of any hydrocarbon liquid at standard conditions may be evaluated from the real one at
standard conditions and from the apparent data of the liberated light components.
STANDING and KATZ (1942) examined the properties of many crude oils. They established
that the apparent densities of the two light components methane and ethane depend on the
density of the system (see Figure 5.10). Based on these data and assuming that propane
and heavier components follow the rule of additive volumes, they developed a method for
computing the density of hydrocarbon mixtures. This calculation procedure is presented
through Example 5.5. It is based on Figure 5.11 presenting the correlation of the density
of a system containing methane and ethane with the density of the propanesplus fraction,
the weight % ethane in the ethanesplus and the weight% methane in the system. Then,
the resulting pseudoliquid density of the system containing methane and ethane has to be
corrected with regard to the operating temperature and pressure. Figure 5.12 and Figure
5.13 serve this purpose.
The correlation method after STANDING and KATZ is satisfactory for stocktank liquids
which have low concentrations of methane and ethane.
Phase Properties 99
Figure 5.10: Variation of apparent density of methane and ethane with density of the
system (from STANDING and KATZ, 1942)
Ethane  NButane
Ethane  Heptane
Ethane  Crystal Oil
Methane  Cyclo Hexane
Methane  Benzene
Methane  Pentane
Methane  Hexane
Methane  Heptane
Methane  Propane
Methane  Crystal Oil
Methane  Crude Oil
0.3
0.3
0.4
0.4
0.5
0.5
0.6
0.6
0.7 0.8 0.9
Density of System [g/cc], 60°F & atm. Pressure
0.1
0.2
0.3
0.4
A
p
p
a
r
r
e
n
t
D
e
n
s
i
t
y
o
f
M
e
t
h
a
n
e
[
g
/
c
c
]
100 Phase Properties
Figure 5.11: Pseudoliquid density of systems containing methane and ethane
(from STANDING, 1952)
50 40 30 20 10 0
0
10
20
30
10
20
30
40
50
60
70
30
40
50
60
70
D
e
n
s
i
t
y
o
f
P
r
o
p
a
n
e
p
l
u
s
[
i
b
/
c
u
f
t
]
D
e
n
s
i
t
y
o
f
S
y
s
t
e
m
i
n
c
l
u
d
i
n
g
M
e
t
h
a
n
e
a
n
d
E
t
h
a
n
e
[
i
b
/
c
u
f
t
]
W
t
%
E
t
h
a
n
e
i
n
E
t
h
a
n
e
p
l
u
s
M
a
t
e
r
i
a
l
W
t
%
M
e
t
h
a
n
e
i
n
e
n
t
i
r
e
S
y
s
t
e
m
Phase Properties 101
Figure 5.12: Density correction for compressibility of liquids (from STANDING, 1952)
25 30 35 40 45 50 55 60
65
Density at 60 [°F] and 14.7 [psia], [ib/cu ft]
D
e
n
s
i
t
y
a
t
P
r
e
s
s
u
r
e
m
i
n
u
s
D
e
n
s
i
t
y
a
t
6
0
[
°
F
]
a
n
d
1
4
.
7
[
p
s
i
a
]
,
[
i
b
/
c
u
f
t
]
0
1
2
3
4
5
6
7
8
9
10
1
0
0
0
2
0
0
0
3
0
0
0
4
0
0
0
5
0
0
0
6
0
0
0
8
0
0
0
1
0
0
0
0
1
5
0
0
0
P
r
e
s
s
u
r
e
[
p
s
i
a
]
102 Phase Properties
Figure 5.13: Density correction for thermal expansion of liquids
(from STANDING, 1952)
Example 5.5
Determination of the volume of 1 mole oil at bubble point through
STANDINGKATZ correlations. The bubble point is defined by
and
.
The composition of the system is tabled below.
25 30 35 40 45 50 55 60
65
Density at 60 [°F] and Pressure p, [ib/cu ft]
D
e
n
s
i
t
y
a
t
6
0
[
°
F
]
m
i
n
u
s
D
e
n
s
i
t
y
a
t
T
e
m
p
e
r
a
t
u
r
e
,
[
i
b
/
c
u
f
t
]
0
1
2
3
4
5
6
7
8
9
10
2
4
0
2
2
0
2
0
0
1
8
0
1
6
0
1
4
0
1
2
0
100
60
T
e
m
p
e
r
a
t
u
r
e
[
°
F
]
p
ob
3300psia 22.55MPa ( ) =
T
ob
220
o
F 105°C ( ) =
Phase Properties 103
*) at standard conditions: 0.1 MPa, 15
o
C
The density of the propaneplus can be calculated through
Then the weight % ethane in the ethaneplus material
and
the weight % methane in the entire system
are calculated.
From Figure 5.11, the pseudoliquid density of the hydrocarbon
mixture at standard conditions is read off as 51.5 lb/cu ft.
From Figure 5.12, the pressure correction for p
ob
(3300 psia) is
read off as
Composition Mole
Fraction
z
i
= x
i
Mole
Weight
M
i
kg/kmole
z
i
M
i
Density
kg/m
3
Volume
V
i
*
m
3
C
1
0.4404 16.0 7.046
C
2
0.0432 30.1 1.300
C
3
0.0405 44.1 1.786 507.5 0.0035
C
4
0.0284 58.1 1.650 575.0 0.0029
C
5
0.0174 72.2 1.256 626.0 0.0020
C
6
0.0290 86.2 2.500 662.5 0.0038
C
7+
0.4001 287.0 114.829 907.0 0.1271
Σ 1.0000 130.367 0.1393
ρ
i
k
i 3 =
z
i
M
i
∑
k
i 3 =
V
i
∑

122.021
0.1393
 876kg m
3
⁄ 54.69 lb cuft ⁄ = = =
z
2
M
2
k
i 2 =
z
i
M
i
 100
1.300
123.321

= × 100 1.05 % [ ] = ×
z
1
M
1
k
i 1 =
z
i
M
i
∑
 100
7.046
130.367

= × 100 5.4% = ×
104 Phase Properties
0.9 lb/cu ft
so that the liquid density at 3300 psia and 60
o
F results in
51.5 + 0.9 = 52.4 lb/cu ft.
Then the temperature correction for 220
o
F is read off from Figure
6.13 as
3.7 lb/cu ft
so that the oil density at 3300 psia and 220
o
F results in
and
,
respectively.
Figure 5.14: Apparent liquid density of natural gases in various API gravity oils
(from KATZ, 1952)
ρ
ob
52.4 3.7 48.7 lb cu ft ⁄ = – =
ρ
ob
778.5kg m
3
⁄ =
15
20
25
30
35
40
45
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
Gas Gravity  Air = 1.0
A
p
p
a
r
r
e
n
t
D
e
n
s
i
t
y
i
n
P
o
u
n
d
s
p
e
r
C
u
b
i
c
F
o
o
t
o
f
d
i
s
s
o
l
v
e
d
G
a
s
a
t
6
0
[
°
F
]
a
n
d
1
A
t
m
o
s
p
h
e
r
e
3
0
°
4
0
°
5
0
°
6
0
°
2
0
°
A
.
P
.
I
.
C
r
u
d
e
G
r
a
v
it
y
Phase Properties 105
Natural gases (dissolved gas) predominantly consist of methane. Therefore, KATZ (1942)
has applied the method of the apparent liquid volume also to crude oil  natural gas
systems. The easily volatilized component is marked by its gravity, , and the scarcely
volatilized fraction is the crude oil with the density and its gravity (in °API) at
standard conditions. The apparent liquid density of the respective natural gas can be
determined according to Figure 5.14.
The amount of the dissolved gas is always indicated as the number of normal cubic meter
per actual liquid volume at standard conditions, nm
3
/m
3
. This ratio is designated as R
s
(solution ratio). In terms of the production, the ratio is also often designated as GOR
(producing gas  oil ratio).
In other words: R
s
expresses how much gas will be dissolved in one cubic meter oil
(volume at standard conditions) at the operating conditions. Therefore, R
s
depends on
pressure and temperature, but also on the oil quality.
5.2.2 Formation Volume Factor
The oil volume inside the tank at surface conditions differs from the volume of the oil
leaving the reservoir. This change in volume has three main sources:
• The most important factor is the gas liberation below bubble point pressure p
ob.
• The difference between reservoir pressure and atmospheric pressure causes slight
expansion of the remaining oil phase.
• The increase in volume because of expansion is more or less compensated by the
shrinkage of the oil resulting from the transfer from reservoir to surface temperature.
γ
g
ρ
o
γ
o
106 Phase Properties
Figure 5.15: Typical graph of formationvolume factor of oil against pressure
The formation volume factor of the oil phase, which must be taken into account, is defined
as
Correspondingly, B
o
can be defined as the volume of one cubic meter tank oil (oil at
standard conditions) at certain operating pressure and temperature.
Figure 5.15 illustrates typical characteristics the relationship between B
o
and p. As long
as the pressure does not fall below the bubble point pressure p
ob
during a pressure
decrease from initial reservoir conditions, change in B
o
will be ruled by the expansion of
the oil. As soon as the reservoir pressure falls below p
ob
, gas will be liberated. This gas
liberation results in a shrinkage of the oil volume. Thus, B
o
decreases at decreasing
pressure as soon as p
ob
has been crossed.
The reciprocal value of B
o
is called “shrinkage factor”.
By multiplying the tank oil volume with the actual B
o
, the volume of the oil inside the
reservoir is obtained. On the other side, the multiplication of the shrinkage factor with the
volume inside the reservoir will result in the stock oil volume.
On the basis of many investigations, STANDING (1947) amplified the correlation to permit
the calculation of bubble point pressures and volume factors, B
o
. The respective
STANDING charts are appended at the end of this textbook:
• Working with Chart 1, it is possible to take the solution ratio, gas gravity, tank oil
gravity and reservoir temperature to determine the pressure at which the given amount
of gas would be in solution. This pressure is the bubble point pressure of the
0
p
b
Reservoir Pressure
B
o,b
1.0
B
o
oil volume and volume of dissolved gas at reservoir conditions
oil volume at standard conditions

=
Phase Properties 107
respective hydrocarbon system, p
ob
.
• Charts 2 permits the evaluation of B
o
at bubble point conditions if GOR represents the
solution ratio.
• The total formation volume factor, B
t
, can be estimated by use of Chart 3. B
t
is
defined by
(5.18)
where
Bo: volume factor of the actual oil phase
Bg: volume factor of the actual gas phase
Rs,i: solution ratio at initial conditions
Rs: solution ratio at operating conditions.
B
o
can be evaluated
• by use of the KATZ correlation (see Figure 5.14) and the STANDING diagrams (see
Figure 5.12 and Figure 5.13)
• or by use of the Charts from STANDING (see appendix).
In most cases, these empirical correlations (see Example 6.6) yield values which are
comparable within 3 or 4% to measured laboratory data.
Example 5.6
Evaluation of the oil volume factor at bubble point conditions from
operation data tabled below both in SIunits and Fieldunits. The
thermodynamic data of the system were given in Example 5.5. The
calculation is based on one cubic meter tank oil.
B
t
B
o
R
si
R
s
– ( )B
g
+ =
108 Phase Properties
In summing up the producing gas/oil ratios GOR, the initial solution ratio R
si
results in
and
respectively.
The computation of the average specific gas gravity results in
Now the amount of gas can be calculated through
where is the air density:
m
g
= 94 x 1.225 x 0.722 = 83.1 kg
Because the specific gravities are given as
γ
g
= 0.722
and
γ
o
= 30
o
API,
the apparent liquid density of the gas can be evaluated through
Figure 5.14 as
and
,
respectively.
GOR
nm
3
/m
3
]
GOR
cu ft/bbl
kg/m
3 o
API
Separator 1 74 415.5 0.640
Separator 2 16 89.8 0.897
Tank 4 22.5 1.540 876.2 30
Σ 94 527.8
γ
g
ρ
o
γ
o
R
si
94nm
3
m
3
⁄ =
R
si
527.8 cu ft bbl ⁄ =
γ
g
74 0. 640 16 0. 897 4 1. 540 × + × + ×
74 16 4 + +

67.9
94
 0.722. = = =
m
g
R
si
ς
a
γ
g
=
ρ
a
ρ
g a ,
24.3 lb/cu ft =
ρ
g a ,
389.2 kg/m
3
=
Phase Properties 109
Now the apparent liquid volume of the liberated gas at standard
conditions (T =15
o
C and atmospheric pressure) can be calculated
through
and so
.
In summarizing 1 m
3
tank oil and 0.214 m
3
apparent liquid volume
of the liberated gas, as well as 876.2 kg tank oil and 83.1kg
liberated gas,the pseudoliquid density of the given hydrocarbon
system at standard conditions results in
and so
.
From Figure 5.12, the pressure correction of this density for p =
3300 psia results in
.
From Figure 5.13, the final temperature correction for T = 220
o
F
results in
or, in SIunits, respectively, in
.
On the basis of the calculated oil density under the conditions of
the reservoir, the volume of one cubic meter oil under reservoir
conditions results in
and so
.
In other words: 1.288 m
3
saturated reservoir liquid under
reservoir conditions shrinks to 1 m
3
tank oil. The shrinkage
results in
V
g app ,
m
g
ρ
g a ,

=
V
g app ,
83.1
389.2
 0.214 m
3
= =
ρ
o app ,
876.2 83.1 +
1000 0.214 +
 790.2kg m
3
⁄ = =
ρ
o app ,
49.3 lb/ cu ft =
ρ
o
49.3 0.9 50.2 lb/ cu ft = + =
ρ
o
50.2 3.8 – 46.4 lb/ cu ft = =
ρ
o
743.7kg m
3
⁄ =
V
ob
959.3
743.7

1.29m
3
= =
B
ob
1.29 =
110 Phase Properties
5.2.3 Compressibility of Undersaturated Liquids
An undersaturated oil is a compressed liquid in the pressure range above the bubble point
pressure p
b
. Its isothermal compressibility is to be calculated in the same way as for the
gases (see Eq. 5.9) so that
(5.19)
and after integration
. (5.20)
Because
where C is a constant, the density of an undersaturated oil phase can be calculated through
(5.21)
In general, the bubble point pressure, p
b,
is chosen as reference. Eq. 5.20 and Eq. 5.21 will
develop in a Taylor series. Because c
o
is very small, the terms of higher order can be
disregarded so that
(5.22)
and
. (5.23)
Similar formulae are also valid for water.
The compressibility of undersaturated liquids is constant within wide pressure intervals.
Therefore, it can be calculated between any two states of the system, (V
1
, p
1
) and (V
2
, p
2
)
by
1.29 1.000 –
1.29
 100 × 22.48% =
c
o
1
V

∂V
∂p

¸ ,
¸ _
T
1
B
o

∂B
o
∂p

¸ ,
¸ _
T
= – =
B
o
B
ob
e
c
o
–
p p
b
– ( )
=
ρ
o
C
B
o

=
ρ
o
ρ
ob
e
c
o
p p
b
– ( )
=
B
o
~B
ob
1 c
o
p p
b
– ( ) – [ ]
ρ
o
ρ
o b ,
1 c
o
p p
b
– ( ) + [ ] =
Phase Properties 111
(5.24)
where V is averaged volume between p
1
and p
2
.
TRUBE (1957) has reported a method of calculating the compressibility of undersaturated
liquids. This method is based on the determination of the pseudocritical data of the
system and on the evaluation of the pseudoreduced compressibility of the undersaturated
liquid, c
r
as a function the pseudoreduced pressure, p
pr
, and pseudoreduced temperature,
T
pr
(see Figure 5.16). Then the oil compressibility can be calculated through
(5.25)
If the composition of the undersaturated system is not known, the pseudoreduced
pressure and temperature can be approximately evaluated through Figure 5.17 on the basis
of
• the calculation of the pseudoliquid density of the hydrocarbon mixture at standard
conditions (see procedure of Example 5.6),
• the pressure correction of this pseudoliquid density from Figure 5.12.
Then, the evaluated oil density is related to the density of water at the standard
temperature, .
c
o
V
1
V
2
–
V p
1
p
2
– ( )

– =
c
o
c
r
p
pc
⁄ =
ρ
w
1000kg m
3
⁄ =
112 Phase Properties
Figure 5.16: Pseudoreduced compressibility for undersaturated reservoir fluids
(from TRUBE, 1957)
Figure 5.17: Pseudocritical conditions of undersaturated reservoir liquids
(from TRUBE, 1957)
1 2 3 5 7 10 20 30 50 70100
p
r
c
r
10
1
10
2
10
3
10
4
0.60
0.50
0.40
1.10
1.00
0.95
0.90
0.80
0.70
Pseudo Reduced
Temperature
0.62 0.64 0.66 0.68 0.70 0.72 0.74 0.76 0.78 0.80 0.82 0.84 0.86 0.88
Specific Gravity of Undersaturated Reservoir Liquid
at Reservoir Pressure. Corrected to 60 [°F]
200
300
400
500
600
700
800
900
1000
1100
1200
1300
P
s
e
u
d
o
C
r
i
t
i
c
a
l
P
r
e
s
s
u
r
e
[
p
s
i
a
]
P
s
e
u
d
o
C
r
i
t
i
c
a
l
T
e
m
p
e
r
a
t
u
r
e
[
°
R
]
P
se
ud
o
C
ritic
a
l T
e
m
p
e
ra
tu
re
Pseudo C
ritic
al P
re
ss
u
r
e
Phase Properties 113
Example 5.7
Evaluation of B
oi
for an under saturated oil at reservoir
temperature T = 105
o
C and under the initial pressure p
i
= 25 MPa
through the reduced compressibility from TRUBE. The composition of
the system was given in Example 5.5 and B
ob
was calculated through
Example 5.6. The critical data are tabled below.
The computation of the pseudoreduced temperature and pressure
results in
and
.
From Figure 5.16, the reduced compressibility can be read off as
.
The compressibility of the undersaturated oil phase can be
calculated through
Because
(see Example 5.6), B
oi
can be calculated through Eq. 5.23
and results in
B
oi
= 1.29[1  0.00292(25  22.5)] = 1.281.
T
pr
T
T
pc

105 273.15 +
374.70
 1.01 = = =
p
pr
p
p
pc

25
3.655
 6.84 = = =
c
r
0.02 =
c
o
c
r
p
pc

0.02
6.84

2.92 E 3MPa
1 –
– = = =
B
o b ,
1.29 =
B
oi
B
ob
1 c
o
p
i
p
b
– ( ) – [ ] =
114 Phase Properties
5.2.4 Viscosity
Figure 5.18: Viscosity of subsurface samples of crude oil (from HOCOTT and
BUCKLEY, 1941, after BEAL, 1946)
In general, liquids decrease in viscosity with increasing temperature and increase in
viscosity with increasing pressure. HOCOTT and BUCKLEY (1941) presented
measurements of the viscosity of crude oils containing dissolved gases. Figure 5.18
demonstrates that with increasing pressure, the viscosity decreases up to the bubble point
because of the increasing fraction of light components. The dependence on pressure is
more significant for high viscous oils than for low viscous oils. Anyway, above the bubble
point pressure, the viscosity increases with increasing pressure.
The viscosity of the crude oil can be determined with sufficient accurancy on the basis of
the known viscosity of the tank oil µ
1
at the reservoir temperature T and atmospherical
pressure, the solution ratio R
si
and the undersaturated pressure in the reservoir, ∆p = p
i

p
b
. Example 5.8 elucidates the respective correlation procedure.
0 500 1000 1500 2000 2500 3000 3500
Reservoir Pressure [psia]
A
b
s
o
l
u
t
e
V
i
s
c
o
s
i
t
y
[
c
P
]
0
1
2
3
4
5
6
7
8
9
A Permian, West Texas
B Frio Sand, Gulf Coast
C Palo Pinto Lime, North Texas
D Frio Sand, Southwest Texas
B
u
b
b
l
e
P
o
i
n
t
B
u
b
b
l
e
P
o
i
n
t
B
u
b
b
l
e
P
o
i
n
t
Bubble Point
158°F 800ft /
3
bbl GOR 45.7°API G=0.84
1
6
4
°
F
3
8
5
f
t
/
3
b
b
l
G
O
R
2
6
.1
°
A
P
I G
=0
.65
A
B
C
D
100°F Reservoir Temperature
191 ft/bbl Gas in Solution at
Bubble Point Pressure
29.7°API Oil Gravity
1.12 Gas Gravity
3
Legend
Phase Properties 115
Figure 5.19: Viscosity of gassaturated reservoir crude oils at reservoir conditions
(from CHEW and CONNALLY, 1959)
Example:
Problem: Find the gassaturated viscosity of
a crude oil having a solution gas/oil ratio of
600 ft/bbl and dead oil viscosity of 1.50 cP
all at the same temperature.
Procedure: Locate 1.5 cP on the dead oil
viscosity scale (abscissa) and go up vertically
to the 600 gas/oil ratio line. Then go left
horizontally to read the answer 0.58 on the
gassaturated oil viscosity scale (ordinate).
3
0.3
0.3
0.5
0.5
0.7
0.7
1.0
1.0
2
2
3
3
5
5
7
7
10
10
20
20
30
30
50
50
70
70
100
100
Viscosity of Dead Oil [cP]
(at Reservoir Temperature and Atmospheric Pressure)
V
i
s
c
o
s
i
t
y
o
f
G
a
s

S
a
t
u
r
a
t
e
d
O
i
l
[
c
P
]
(
a
t
R
e
s
e
r
v
o
i
r
T
e
m
p
e
r
a
t
u
r
e
a
n
d
S
a
t
u
r
a
t
i
o
n
P
r
e
s
s
u
r
e
)
0.1
0.2
Example
1
6
0
0
1
4
0
0
1
2
0
0
1
0
0
0
8
0
0
6
0
0
4
0
0
3
0
0
2
0
0
0
Solution Gas/Oil Ratio
[ft /bbl]
3
1
0
0
5
0
116 Phase Properties
Figure 5.20: Prediction of crude oil viscosity above bubble point pressure
(from BEAL, 1946)
7
6
5
4
3
2
1
Detail
0
0
1
1000
2
2000
3
3000
4
4000
5
5000
6
7
8
9
Pressure [psi]
A
b
s
o
l
u
t
e
V
i
s
c
o
s
i
t
y
[
c
P
]
0
100
10
20
30
40
50
60
70
80
90
A
b
s
o
l
u
t
e
V
i
s
c
o
s
i
t
y
o
f
C
r
u
d
e
O
i
l
a
b
o
v
e
B
u
b
b
l
e
P
o
i
n
t
P
r
e
s
s
u
r
e
[
c
P
]
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Undersaturated Pressure [psi]
(Pressure above Bubble Point less Pressure at Bubble Point)
8
5
8
0
7
5
7
0
6
5
6
0
5
5
5
0
4
5
4
0
3
0
25
20
15
10
5
A
b
s
o
lu
te
V
isc
o
s
ity
o
f G
a
s S
a
tu
ra
te
d
C
ru
de
a
t B
u
b
b
l e
P
o
in
t P
res
s
u
re
[c
P
]
3
5
Crude Oil Prediction above
Bubble Point Pressure
Average Deviation 2.7%
Extrapolated
1
.
0
0
.
9
0
.8
0
.
7
0
.6
0
.5
0
.4
0
.3
0.2
0.1
0 1 2 3 4 5
Pressure 10 [psi]
3
0
10
1
11
2
12
3
13
4
5
6
7
8
9
V
i
s
c
o
s
i
t
y
[
c
P
]
Detail
Phase Properties 117
Exampl e 6. 8
Eval uat i on of t he oi l vi scosi t y at r eser voi r condi t i ons whi ch ar e gi ven
bel ow as wel l as t he i nt er est i ng pr oduct i on dat a.
.
Fr om Fi gur e 5. 19, t he oi l vi scosi t y at bubbl e poi nt condi t i ons can be
r ead of f as
Fr om Fi gur e 5. 20, t he vi scosi t y of t he under sat ur at ed cr ude oi l at p =
3300 psi a can be eval uat ed:
or
5.3 Brines
It has been recognized relatively late that water is  without any exception  one of the
reservoir fluids. GARRISON (1935) and SCHILTHUIS (1938) have been the first who
presented detailed information on the distribution of oil and water in porous rock and on
the origin and occurrence of “connate water” (formation water, capillary water, interstitial
water). Furthermore, they established a relationship between the water saturation and the
rock permeability.
Nowadays, knowledge of the chemical and physiochemical data of the respective
reservoir water is of great importance for the petroleum engineer. Water analyses
• give valuable information about the area of prospecting,
• help to answer questions about the origin of the produced water,
SIUnits Field Units
Reservoir Temperature T
105
o
C
220
o
F
Bubble Point Pressure pb 22.55 MPa 3300 psia
Reservoir Pressure p 25.00 MPa 3600 psia
Initial Solution Ratio Rsi
94 nm
3
/m
3
] 527.8 cu ft/bbl
Standard Oil Viscosity µ
1
0.9 mPas 0.9 cP
µ
ob
0.45cP =
µ
ob
0.45cP =
µ
oi
0.48cP =
µ
oi
0.48 mPas [ ] =
118 Phase Properties
• make sure that injected water is chemically compatible with the reservoir water or
with added water (plugging by precipitations must be avoided),
• are indispensable for feasibility studies in case of chemical EOR processes.
Last but not least, data such as density, compressibility, formation volume factor and
viscosity are required for calculating the displacement process and for establishing a
material balance.
Additionally, this section presents an unusual phenomenon, the socalled “gas hydrate
formation”, during which water and natural gas form a solid phase at temperatures above
the freezing point of water.
5.3.1 Composition of Brines
Reservoir water always contains dissolved inorganic salts, above all sodium chloride,
NaCl. Therefore, these waters are sometimes called “brines” although they cannot be
compared with sea water because both the total salt concentration and the content of
distinctions are different.
The cations usually dissolved in reservoir brine are
Na
+
, K
+
, Ca
2+
, Mg
2+
In certain cases,
Ba
2+
, Li
+
, Fe
2+
,
and sometimes Sn
2+
can be found.
The generally present anions are
and, in some cases, also
.
The elements are present in reservoir water within the following concentration ranges:
Na, Cl in%
Ca, SO
4
in% or ppm
K, Sr in 100 ppm
Cl

, SO
2
4
, HCO

3
CO
2
3
, NO

3
, Br

, J

, S
2
Phase Properties 119
Al, B, Ba, Fe, Li in 1  100 ppm
Cr, Cu, Mn, Ni, Sn, Ti, Zr in ppb (in most cases)
Be, Co, Ga, Ge, Pb, V, W, Zn in ppb (in some cases)
where
% : g substance per 100 g brine
ppm: mg substance per 1 kg brine
ppb: mg substance per 10
3
kg brine
Usually microorganisms of different species are also present in oil field brines. The
origin of these organisms is not yet totally clear. They contribute to corrosion in the bore
hole and to permeability reduction during water flooding.
There is a wide variability of concentrations and chemical characteristics resulting from
many factors. Such factors may be
• the nonmarine origin of a few sediments,
• dilution with ground water,
• enrichment because of vaporization by migrating gas,
• adsorption and cation exchange with clay minerals,
• dissolution of mineral salts in the migrating formation water,
• the chemical reaction with components of adjacent sediments.
Therefore, it is important that water from a certain horizont is characteristically distinct
from all the other brines, even those from the immediate neighbourhood. Thus, water
analyses are mainly used for identifying the produced water with regard to the geological
formation it originates from. In doing so, a simple and easily readable representation of
the great data set from a water sample is desired. The graphical method proposed by STIFF
(1951) may be the simplest graphical method, has maximum utility and, hence, is the most
popular.
This graphical method is shown in Figure 5.21. Horizontal lines extend right and left from
a centrally located vertical line. The positive ions are plotted on the the left of the vertical
line, while the anions are plotted on the right. Characteristic positions are designated for
the ions which are most frequently found in reservoir brines.
120 Phase Properties
Figure 5.21: Essential feature of the water pattern analysis system (from STIFF, 1951)
The analysis data for all ions in question are indicated in ppm (milligram per liter). To
make these values comparable with each other, the contents have to be converted to meq
(milliequivalents) by dividing the respective value with the milliequivalent weight (ionic
weight in mg per valence). If there is a difference between Σ meq of the positive and Σ
meq of the negative ions, the difference is presented as sodium.
Although various scales can be employed, most reservoir waters may be plotted with
sodium and chloride on a scale of 100 meq and a scale of 10 meq for the others. When the
data of the water analysis are plotted on the graph and the adjacent points are connected
by straight lines, a closed “pattern” is formed as a “fingerprint” of the respective brine.
One of many practical applications in correlating producing formations is illustrated by
Figure 5.22. The characteristic pattern of the formations is evident.
Figure 5.22: Course of Arbuckle formation through Kansas shown by water patterns
(from STIFF, 1951)
Scale: meq/liter
Na
100
Ca
10
Mg
10
Fe
10
Cl
100
HCO
10
3
CO
10
3
SO
10
4
Scale:
meq/liter
Na
100
Ca
10
Mg
10
Fe
10
Cl
100
HCO
10
3
CO
10
3
SO
10
4
St. John Field Drach Field
Bloomer Field
Stoltenberg Field
Barton
Stafford
Rice
Ellsworth
Phase Properties 121
5.3.2 Formation Volume Factor
The definition of the formation volume factor of water B
w
corresponds to the definition
of B
o
. Similarly to the hydrocarbon liquids, the change in volume may have three reasons:
• liberation of dissolved gas due to pressure decrease,
• expansion of the brine because of pressure decrease,
• contraction of the brine due to temperature reduction.
In contrast to the solubility of natural gases in oil, their solubility in water is low (see
Figure 5.23a). Additionally, the solubility decrease with increasing salinity of the brine
(see Figure 5.23b). Consequently, the contribution of the gas liberation to B
w
is low.
Since the expansion due to the decrease in pressure and the contraction because of
temperature decrease are also insignificant, B
w
is always numerically small, rarely above
1.06.
Figure 5.24 shows a typical relationship between B
w
and p. In the present case, the initial
reservoir pressure is higher than the bubble point pressure. If the reservoir pressure is
reduced from initial pressure to the bubble point pressure, B
w
increases because of the
water expansion. A further pressure reduction leads to gas liberation. This loss of liquid
volume only balances partly the continuing expansion so that B
w
further increases with
further pressure drop. Along with the reduction of the pressure to atmospheric pressure,
B
w
will reach its maximum value. The difference between this value and 1.0 represents
the thermal contraction resulting from the cooling of the brine from the reservoir
temperature to the standard temperature of 15
o
C. However, B
w
will be smaller than 1.0,
if the increase in volume by dissolved gas is too low to compensate the volume reduction
resulting from high pressure.
122 Phase Properties
Figure 5.23: Solubility of natural gas in water (from DODSON and STANDING, 1944)
Figure 5.24: Typical graph of formation volume factor of water against pressure
The assumption is customary that the bubble point pressure of a gassaturated brine is
equal to the bubble point pressure of the crude oil inside the corresponding reservoir.
24
22
20
18
16
14
12
10
8
6
4
2
0
60 100 140 180 220 260
Temperature [°F]
a)
S
o
l
u
b
i
l
i
t
y
o
f
N
a
t
u
r
a
l
G
a
s
i
n
W
a
t
e
r
[
f
t
/
b
b
l
]
3
4
0
00
35
0
0
2
0
5
0
200
0
1500
500
1000
4
5
0
0
5
0
0
0
3
00
0
Pressure [psi]
0 10 20 30 40
Total Solid in Brine 10 [ppm]
b)
3
Correction of Brine Salinity
R
a
t
i
o
:
S
o
l
u
b
i
l
i
t
y
i
n
B
r
i
n
e
S
o
l
u
b
i
l
i
t
y
i
n
W
a
t
e
r
0.8
0.9
1.0
150°F
100°F
250°F
200°F
0
p
b
Reservoir Pressure
B
w,b
1.0
0
Phase Properties 123
Figure 5.25 presents data of B
w
for pure water saturated with natural gas (solid lines).
Naturally, gas saturated pure water has a higher volume factor than pure water at a given
pressure and temperature (dashed lines). B
w
for a reservoir water can be computed with
aid of Figure 5.25 and Figure 5.23 as follows:
• The volume factors for pure water and pure water saturated with natural gas are read
from Figure 5.25 for the given reservoir data.
• The gas solution ratio, R
sw
, for pure water is read from Figure 5.23a. and corrected for
salinity using Figure 5.23b.
• Assuming the effect of gas solubility on B
w
to be linear, the volume factor is
computed by interpolation.
Figure 5.25: B
w
for pure water (dashed lines) and pure water saturated with natural gas
(solid lines) as a function of pressure and temperature (from DODSON and STANDING,
1944)
Example 5.9
Formation volume factor of reservoir water. The reservoir
conditions as well as the salinity of the brine are given below
both in SI and Field units.
0 1000 2000 3000 4000 5000
Pressure [psia]
W
a
t
e
r
F
o
r
m
a
t
i
o
n
V
o
l
u
m
e
F
a
c
t
o
r
[
b
b
l
/
b
b
l
]
0.98
0.99
1.00
1.01
1.02
1.03
1.04
1.05
1.06
1.07
200°F
150°F
250°F
100°F
124 Phase Properties
The evaluation of B
w
for pure water from Figure 5.25 results in
B
w
= 1.0275
B
w
for pure water saturated with natural brine can be read off
from Figure 5.25 as
B
w
= 1.034.
The gas solution ratio in pure water from Figure 5.23a results in
R
sw
= 15.3 cu ft/bbl
After correction of the gas solubility from Figure 5.23b, R
sw
results in
R
sw
= 15.3 x 0.88 = 13.46 cu ft/bbl
Then the formation volume factor of the reservoir water results in
B
w
= 1.0275 + (1.034  1.0275) .
Figure 5.26: Density of brine as a function of total dissolved solids
(from MCCAIN, 1973)
In Figure 5.26, the density of reservoir water at standard conditions is plotted as a function
of the total salt concentration. By dividing the respective density with B
w
, the
SI Units Field Units
Reservoir Temperature T
93.3
o
C 200
o
F
Reservoir Pressure p 20.56 MPa 3000 psia
Salinity 30,000 ppm 30,000 ppm
13.46
15.3
 1.033 =
0 5 10 15 20 25 30
Total dissolved Solids [%]
B
r
i
n
e
D
e
n
s
i
t
y
a
t
1
4
.
7
p
s
i
a
a
n
d
6
0
°
F
[
i
b
/
c
u
f
t
]
76
60
62
64
66
68
70
72
74
Phase Properties 125
corresponding density under reservoir conditions can be received.
From Figure 5.27a the coefficient of the isothermal water compressibility, c
w
, can be
evaluated without considering the gas in solution. At a given pressure and temperature,
the effect of gas in solution is to increase c
w
(see Figure 5.27b).
Figure 5.27: The isothermal coefficient of compressibility of pure water, including
effects of gas in solution (from DODSON and STANDING, 1944)
Example 5.10
Evaluation of the compressibility of reservoir water at the
reservoir data of Example 5.9.
From Figure 5.27a at 3000 psia and 200
o
F, c
w
can be read off as
.
60 100 140 180 220 260
Temperature [°F]
a)
2.4
2.8
3.2
3.6
4.0
2.6
3.0
3.4
3.8
0 5 10 15 20 25
Gas  Water Ratio [ft /bbl]
b)
3
1.0
1.1
1.2
1.3
R
a
t
i
o
:
S
o
l
u
t
i
o
n
C
o
m
p
r
e
s
s
i
b
i
l
i
t
y
W
a
t
e
r
C
o
m
p
r
e
s
s
i
b
i
l
i
t
y
Correction for Gas in Solution
C
o
m
p
r
e
s
s
i
b
i
l
i
t
y
o
f
W
a
t
e
r
c
1
0
[
b
b
l
/
b
b
l
p
s
i
]
w
6
1
0
0
0
[p
s
i
a
]
4
0
0
0
5
0
0
0
6
0
0
0
2
0
0
0
3
0
0
0
T
w
P
V
V
c
,
_
¸
¸
∂
∂
,
_
¸
¸
− ·
1
c
w
3.1 E  6 psi
1 –
=
126 Phase Properties
From Figure 5.23a, the gas solubility in pure water results in
.
Correction of gas solubility from Figure 5.23b:
.
After evaluating the effect of the dissolved gas from Figure 5.23b,
the compressibility of the reservoir brine results in
,
or
,
respectively.
5.3.3 Viscosity
Regarding the viscosity of oil field brines, only few data are available. Figure 5.28 shows
the water viscosity as a function of the temperature. The viscosity should rise with
pressure and increase significantly with increasing salt concentration. These
dependencies can be observed in Figure 5.28. It is recommended to use this graph for
estimating the viscosity of reservoir waters without conisdering both the pressure and the
salinity of the brine.
The fact that low molecular paraffins are dissolved in water under reservoir conditions
should furthermore lead to a considerable reduction of the brine viscosity. Unfortunately,
no corresponding data have been published until now.
R
sw
15.3 cu ft bbl
1 –
=
R
w
15.3 0.88 × 13.46 cu ft / bbl = =
c
w
1.12 3.1 × E 6 3.47 E  6 psi
1 –
= – =
c
w
5.03 E  10 Pa
1 –
=
Phase Properties 127
Figure 5.28: The viscosity of water at oil field temperature and pressure
(from VAN WINGEN, 1950)
5.3.4 Natural Gas Hydrates
Natural gas and liquid water can form solid material which remind of wet snow. This
process takes place at temperatures slightly higher than the freezing point of pure water.
These solid substances are called “gas hydrates”.
The phenomenon is of special interest for petroleum industry. For example, the pressure
decreases in the production line, such as across a choke or in a separator. This can cause
a reduction in the temperature of the gas (adiabatic process) and, therefore, the
condensation of water from the gas. The conditions necessary for hydrate formation could
be established.
Gas hydrates behave rather like dissolutions of gases in ice than like chemical
compounds. The lattice of the hydrate crystal is mainly formed by water molecules while
the hydrocarbon molecules occupy vacancies within the lattice. Therefore, the hydrate
formation is rather of physical than of chemical nature. The water lattice seems to be
similar to ice, because the formation enthalpy is comparable.
Saline Water (60000 [ppm]) at
14.2 [psia]
14.7 [psia]
7100 [psia]
Vapor Pressure
0 50 100 150 200 250 300 350
Temperature [°F]
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
D
y
n
a
m
i
c
V
i
s
c
o
s
i
t
y
[
c
P
]
128 Phase Properties
Although gas hydrates can be compared rather to solid solutions than to chemical
compounds, a certain number of water molecules is associated to each gas molecule. This
may be caused by the crystal structure. The ratio of hydrocarbon to water depends mainly
on the size of gas molecules.
The presence of liquid water is the most important prerequisite for hydrate formation.
However, even if liquid water is present, a metastable equilibrium between water and
hydrocarbon gas may occur (under conditions at which generally hydrate formation takes
place). Seed crystals can immediately induce the hydrate formation.
Figure 5.29 shows a part of the phase diagram of a mixture of water and a light
hydrocarbon. The point Q
2
is a “quadruple point” at which four phases are in equilibrium.
The quadruple point Q
1
is at approximately 0
o
C (ice, gas hydrate, water and natural gas
are in equilibrium). The connection Q
1
 Q
2
represents the equilibrium natural
gaswaterhydrate. This line is of special interest.
Figure 5.30 presents the hydrate formation conditions for natural gases of different
gravities. The curves can be used to estimate the conditions of hydrate formation.
Anyway, the results should be applied with caution because of the large discrepancies of
the few published experimental data and the correlation present, especially in case of high
gas gravity and temperature.
Additionally, it has to be taken into account that electrolytes present in the water will lead
to decreasing temperatures at which hydrate formation will occur. Thus, watersoluble
substances are often added to the system to inhibit or suppress the hydrate formation. The
effect of the concentration of various inhibitors on the hydrateformation temperatures of
a natural gas is plotted in Figure 5.31.
Hydrate portion of the phase diagram for a typical mixture of water and a light
hydrocarbon (from MCCAIN, 1973)
Phase Properties 129
Figure 5.29: Hydrate portion of the phase diagram for a typical mixture of water and a
light hydrocarbon (from MCCAIN, 1973)
Figure 5.30: Pressuretemperature curves for predicting hydrate formation
(from KATZ, 1945)
P
r
e
s
s
u
r
e
Temperature
Q
1
Q
2
C
Hydrate
+
Ice
Hydrate
+
Water
Hydrocarbon Liquid
+
Water
Water
+
Hydrocarbon Gas
Ice
+
Hydrocarbon Gas
30 40 50 60 70 80 90
Temperature [°F]
40
60
80
100
150
200
300
400
600
800
1000
1500
3000
4000
6000
30
500
2000
5000
50
P
r
e
s
s
u
r
e
f
o
r
H
y
d
r
a
t
e
F
o
r
m
a
t
i
o
n
[
p
s
i
a
]
M
e
t
h
a
n
e
0
.
6
G
r
a
v
i
t
y
G
a
s
0
.
7
0
.
8
0
.
9
1
.
0
130 Phase Properties
Figure 5.31: Depression of hydrate formation temperature by inhibitors
(from KATZ et al., 1959)
Temperature calculations have been carried out for typical natural gases, and the results
have been combined with the hydrate formation conditions given in Figure 5.30. This has
been done to evaluate diagrams by which the maximum adiabatic pressure reduction
without gas hydrate formation can be estimated. One of these diagrams is given by Figure
5.32.
Example 5.11
Evaluation of the maximum of adiabatic expansion of a natural gas
without any danger of hydrate formation.
Problem: Up to which pressure can this gas be expanded
0 4 8 12 16 20 24
Weight per cent Inhibitor
0
2
4
6
8
10
12
14
16
18
20
D
e
p
r
e
s
s
i
o
n
o
f
H
y
d
r
a
t
e
F
o
r
m
a
t
i
o
n
T
e
m
p
e
r
a
t
u
r
e
[
°
F
]
1
2
3
4
1 Methanol on 0.63 Gravity
Gas (521)
2 Sodium Chloride on 0.68
Gravity Gas (543)
3 Calcium Chloride on 0.63
Gravity Gas (521)
4 Sodium Chloride on
Methane on 0.555 Gas
Gravity (543)
Phase Properties 131
adiabatically without hydrate formation and what will the
temperature be?
Solution steps:
 The ordinate of Figure 5.32 was entered at p
i
= 3000 psia.
 It follows a horizontal movement to the 160
o
F isotherm.
 The next step consists of the vertical movement to the abscissa.
800 psia is the lowest final pressure which will preclude hydrate
formation.
The intersection of the 800 psia line with the dashed line results
in 65
o
F (18.3
o
C) after expansion to 800 psia (5.52 MPa).
Figure 5.32: Permissible expansion of 0.8 gravity gas without hydrate formation
(from KATZ, 1945).
SIUnits FieldUnits
Gas Gravity 0.80 0.80
Initial Pressure
p
i 20.68 [MPa] 3000 psia
Initial Temperature
T
i 71[
o
C] 160
o
F
γ
g
100
100
150
150
200
200
300
300
600
600
400
400
1000
1000
2000
2000
3000
3000
Final Pressure [psia]
I
n
i
t
a
i
l
P
r
e
s
s
u
r
e
[
p
s
i
a
]
4000
10000
8000
6000
5000
1500
500
800
40
200
12
0
11
0
10
0
9
0
8
0
70
60
50
197.5
19
5
1
9
0
180
170
160
15
0
14
0
1
3
0
Initial Temp. [°F]
132 Phase Properties
133
Chapter 6
pVTMeasurements
Accurate crude oil data are necessary to control the production of a hydrocarbon reservoir
efficiently. They are of importance for further calculations of the processes which take
place inside a reservoir.
6.1 Sampling
6.1.1 Objectives
The objectives of sampling are to receive samples from a suitable place in the production
wells or surface facilities. The samples should represent the system in the reservoir under
its initial conditions. Only then, the determination of its type, volumetric and phase
behavior, and its composition is assured to supply data for
• geological and reservoir engineering evaluation and forecasting,
• laboratory studies concerning enhanced oil recovery (EOR) methods modifying PVT
properties (and the viscosity) of the system.
6.1.2 General Criteria
Based on the general remarks about the types and behavior of hydrocarbon systems,
conclusions can be drawn about general sampling criteria. It is obvious that  except in
case of dry and wet gases  characteristic samples can be received only in the early life of
the reservoir. Therefore, it can be generally stated that the best time of sampling is at the
beginning of production.
134 pVTMeasurements
To meet all requirements, the careful selection of wells for sampling is very important.
Representative samples cannot be received from wells which are perforated at the
water/oil or gas/oil contact and are producing water and/or nonequilibrium gas. The wells
should be allocated at characteristic sections of the reservoir. Number of samples will
depend on the dip of the structure as well as on the pay thickness. To determine the fluid
properties along with the depth of the pay zone (where monophase system exists), a
minimum of one sample per 30 m can be recommended.
There are restrictions which must always be considered:
• Considering the gravitational effect, it is clear that the sampling in closed wells could
easily lead to misleading data.
• No characteristic fluid can be received within the flowing wells, where a multiphase
system exists or is developing.
• Reliable sampling conditions in the flowing wells and at the wellhead only exist, if
the total composition of the system entering the well and leaving it at the wellhead
remains constant.
Favorable conditions at steady state flow are indicated by stabilization of the pressure and
temperature at the wellhead and in the separator. Moreover, the production gas/liquid
ratio, GOR, should remain constant during a long period of time.
6.1.3 Sampling Methods
There are three methods of sampling. Samples can be taken from the flowing wells
(subsurface sampling) or from the equilibrium phases of the separators (separator
samples). The third possibility is to get representative samples from the wellhead
(wellhead sampling). In every case, steady state conditions must be maintained. The
method used depends on the type of system to be sampled.
There are certain approximations for selecting sampling methods which consider some of
the properties of the produced fluids. Values summarized in Table 6.1 represent average
characteristics which can be used as some rule of thumb. It is clear, however, that the
behavior of any system also depends on reservoir conditions which are missing on the
tabulated data. Therefore, preliminary screening must be used cautiously.
pVTMeasurements 135
On the basis of the general remarks on the phase behavior of different systems in their
initial state and on the changes in their behavior in the depression zone and in the wells,
the following guidelines can be applied in screening sampling methods.
Sampling of equilibrium phases from separators can be used for all systems, but steady
state conditions must exist. Note, that no separator sampling is needed for dry gases and
for saturated and undersaturated oils, if no free water enters the production well and if
there is a considerable length along which a1phase oil.
For sampling, only those separators can be used where construction ensures highly
efficient separation of the entering phases. In a wide region of flow rates, no liquid
droplets should depart the system with the gas.
Separators must contain pressure gauges and thermometers as well as devices for the
accurate measurement of the gas volumes (liberated and/or delivered), liquid
hydrocarbons, and of water. Each phase can be sampled through sampling valves built in
the separator at that part where pure gas, oil, and water exist.
To determine the phase boundary during the time of sampling, separators often contain
high pressure windows for visual control. When highly paraffinic systems are produced
exhibiting a cloud point in the vicinity of the separator temperature, separators should be
heated. The continuous monitoring of the concentration of some key components in the
gas phase is recommended. This procedure should be considered to control the
development and maintenance of steadystate flow.
Samples of the equilibrium phases should be take using specially designed bottles with
valves at their extreme ends. These bottles are in vertical position during sampling. The
upper valve is connected with the sampling valve of the separator by a flexible tubing
which is resistant to pressure and corrosion by phases sampled.
Sometimes, the bottles and the connecting tube are filled up with an indifferent liquid
(mercury or water from the separator or from other sources) to maintain pressure and to
Table 6.1: Screening sampling methods
Type Density of the
produced liquid
[kg/m
3
]
Gas/liquid ratio
[m
3
/m
3
]
Dry gas 
Wet gas < 740 11000  20000
Gas/condensate
system
740  780 1000  13000
High shrinkage
oils
780  800 200  1500
Low shrinkage
oils
> 800 < 200
136 pVTMeasurements
exclude air. It is then displaced by the sampled phase (gas or liquid). Whenever the
temperature in the separator is lower than the ambient temperature, the use of mercury (or
brine) is recommended in sampling the liquid phase. In doing so, about 90% of mercury
(or brine) has to be displaced by the separator liquid. After closing the valves into he
container and in the separator and after dismounting the container, additional 10% of
mercury (or brine) should be removed by opening the lower valve. The created gas cap
will prevent uncontrolable pressure increase during shipping. The valves in the contains
must be protected by steel caps during shipping. Data for identification must be delivered
with the sample.
Bottom hole sampling
Bottom hole sampling can be used when only the phase to be sampled is present at the
perforated zone. If water is also produced with the hydrocarbons but remains at the bottom
under steadystate conditions, samples can be taken above the water level. However, the
flow of a 1phase hydrocarbon fluid must be assured. The method can be controlled by
sampling at different depths and by analyzing the water content of samples. No sampling
within the wells is needed for dry gases or  in general  for systems which exist in a
1phase state up to the wellhead.
Bottom hole samplers are specially designed devices containing one valve or two valves
at their extreme ends. They are run into the well through a sealed tube mounted to the
christmas tree by a mechanically driven cable. After reaching the desired depth (usually
with open valves to allow the flushing of the sampler by the flowing fluids), the valves
are closed. The valves are operated mechanically or electrically from the surface or within
the sampler at the predetermined depth or time.
At the surface, the sample should immediately be transferred to containers through a
flexible steel tube which  together with the shipping bottle  is filled up with an inert
liquid (e.g. mercury). The sample is displaced by injection of the inert liquid into the
sampler using a volumetric pump or  in a simple manner  by gravitational effects. The
protection of the samples in the shipping bottles and their correct marking for
identification are important as it has already been mentioned.
Sampling at the wellhead
Sampling at the wellhead is a suitable process in case of the production of dry gas or
undersaturated oil at wellhead conditions.
The container is connected with a valve at the wellhead and flushed with the fluid
produced. The previous remarks referring to the protection and identification of the
shipping bottles are also valid for wellhead samples.
pVTMeasurements 137
6.1.4 Special Problems
There are sometimes problems arising from low permeability oil reservoirs (wells being
produced with high depression). As a result, the ratio of the phases entering the wells does
not characterize the initiallyinplace fluids (bubble point oil system), especially if water
is also present. In this case, it is only possible to take samples from the equilibrium phases
of the separator and to recombine them under separator conditions corresponding to their
volume ratio. Afterwards, pressure and temperature are increased up to the reservoir
conditions.
The basic concept of this method is that  if the liquid is in equilibrium with other phase(s)
 the composition of the oil phase is not influenced by the contacted volume of the other
phase(s) being in equilibrium with the oil. However, no processes in the reservoir/well
system should take place which may result in a difference between the overall
composition of the reservoir fluid and the produced fluid.
A similar approach can also be used for gascondensate systems.
Whenever water production cannot be avoided and the volume of the produced water is
high, dissolved gases may be liberated from the water having a composition which could
largely differ from that of the equilibrium hydrocarbon system. In this case, a similar
procedure as above should be used to restore the composition of the hydrocarbon phase.
However, the samples must be preserved in their original state. As for liquids, it is
important
• to prevent any loss  especially in their lighter fractions ,
• to avoid the contact with air because oxidation processes could drastically modify the
composition of hydrocarbon liquids, primarily those of oils.
Specially designed sampling is needed to determine the microbiological contamination of
the produced fluid(s). This procedure is often necessary in case of reservoirs containing
only water from which water will be produced for industrial or public supply. Sterile
samplers and shipping bottles but also anaerobic conditions (in the course of sample
transfer) must be guaranteed.
138 pVTMeasurements
6.2 Experimental Determination of the Volumetric
and Phase Behavior
To determine the volumetric and phase behavior of naturally occurring systems under
reservoir/well/surface conditions, equipments and methods have been developed for the
petroleum industry. A general scheme of the set of equipments used for PVT studies
shown in Figure 6.1.
6.2.1 Equipment
Figure 6.1: Scheme of PVT equipments
Samples to be studied are transferred to the P V T cell. Some known volume of an inert
liquid is injected into the sampling bottle(s), while the same liquid is withdrawn from the
cell initially filled with it.
The selection criteria for the inert liquid are:
• It must exhibit low volatility under (i) ambient conditions and also (ii) under
temperatures and pressures which represent the working conditions.
• It should not react with the fluids contacted.
Control
unit
p
V
T
C
e
l
l
p
Gauge
S
a
m
p
l
e
S
a
m
p
l
e
Hg
Volumetric Pump
Separator
Gasometer
Analysis
V
g
L
G
pVTMeasurements 139
• Its physical and chemical properties must be well known.
Although mercury is widely used, sometimes low melting point alloys are used. No
restriction is needed, however, if the liquid transported by the volumetric pump(s) is
separated from the samples. In this case, the samples will be transferred and measured by
a piston built into the containers and the PVT cell.
Doubleacting or two singleacting volumetric pump(s) are needed for this process.
Sometimes only one pump may be used for this purpose, if the inert liquid leaves the cell
through a suitable operated valve at the bottom of the cell. This valve controls the rate and
volume of the outflowing inert liquid.
The sample being transferred into the cell must be in a 1phase state. Therefore, liberated
gas or condensate in the shipping bottle has to be dissolved or evaporated in advance.
Referring to separator samples, the proportions of gas and liquid to be injected into the
PVT cell must be equal to those measured into the field separator at the time of sampling.
6.2.2 PVTCells
Evaluation of the PVT behavior is carried out on the system transferred into the PVT cell.
The cell has to be a heat and pressureresistant vessel which is equipped with valves
made of noncorrosive materials. The cell is mounted into a thermostatic bath which
ensures from detrimental temperature inconstancy in the cell. The electrically heated (or
cooled) medium circulating around the cell could be air or liquid. To obtain the desired
rapid establishment of the equilibrium conditions at any given pressure and/or
temperature, the cell is agitated or stirred inside.
However, the accuracy of measurements depends on the resolution of the measuring
instruments. Nevertheless, the relative error can be decreased by a reasonable increase of
the sample volume studied. Therefore, “micro PVT cells” are not used in practice.
Moreover, considering the change in volume of the system within the pressure range
applied in the PVT studies, larger cell volumes are used to study (i) gases, (ii)
gascondensate systems, (iii) critical oils, and (iv) water. On the other hand, relatively
smaller ones are used to investigate characteristic oil systems.
Although many different PVT cells are specified, two basic types can be distinguished:
• blind PVT cells (without any possibility of visual observation of the system to be
measured),
• windowed PVT cells, where the volumetric and phase behavior can directly be
observed through one or more pressure and temperature resistant windows built into
the cell body. The inside of the cell is lighted from outside by transmitted or reflected
(from the inner wall of the cell) light.
140 pVTMeasurements
Figure 6.2: Blind PVT cell
Blind PVT cells are usually used to study systems in which no phase transition takes place
or phase transition can easily be determined using the isothermal p  V relationship (dry
gases, oil systems, water). A characteristic blind cell is shown in Figure 6.2. Its volume is
about 0.001[m
3
] which is large enough to be used for gas and oil studies as well.
Figure 6.3: PVT cell (after BURNETT)
A special PVT cell for gases was developed by BURNETT (see Figure 6.3). It contains two
or more chambers with different volumes which are separated by valves. The chamber
possessing the smaller volume contains the gas under given pressure. By opening
Valve
Valve
Valve
Hg
Heating Wire
Jacket
Heat Isolator
Bath
Thermocouple
Steel Body
Thermometer
Cell
V
1
V
2
In
Out
Pressure Gauge
pVTMeasurements 141
successively the valves of the evacuate other chambers with known volume, the change
in pressure can be measured. No volumetric pumps are needed in this case.
Windowed P V T cells
Windowed PVT cells are usually used to evaluate systems where the knowledge of the
volumes of coexisting phases is necessary (e.g. wet gases under well or separator
conditions, gas condensate systems). It is also required whenever phase transition cannot
exactly be determined by isothermal p  V relationships (volatile oils or partially miscible
liquid/liquid systems with or without free gas phase).
Figure 6.4: PVT cell (after DEANPOETTMAN)
A typical onewindowed cell is shown in Figure 6.4. The volumes of the liquid phase can
be determined by the angle of rotating the cell until liquid phases appear in the window.
A PVT cell with variable volume is presented in Figure 6.5. The condensed phase can be
moved to the window by simultaneous lifting (or sinking) the two builtin plungers. The
volume corresponding to that motion can be read on the calibrated stem of the plungers.
Gas
Hg
Liquid
Gas
Hg
Liquid
Windows
142 pVTMeasurements
Figure 6.5: Variable volume cell (after VELOKIVSKIY et al.)
As far as SLOAN’s windowed cell (see Figure 6.6) is concerned, there are three sections.
In the upper part, a plunger is moved by the fluid injected or withdrawn by a volumetric
pump. In the lower part, the condensed liquid is collected and can be lifted up to the
windows which are built into the middle part. This can be done by injecting inert fluid.
Meanwhile, pressure can be kept at the desired constant value by the simultaneous lifting
of the plunger.
Another type contains one capillary glass tube within the cell through which the inner
volume can be lighted and observed using an optical system. It serves also for the
measurement of the volume of phases. The scheme of this cell is shown in Figure 6.7.
Three windows are built into the body of the most generally used cell as shown in Figure
6.8.
Figure 6.6: PVT cell (after SLOAN)
Window
1 Orifice for Introduction of Mercury through
the Upper Cap, for Operation of the Piston
2 Piston with joist Packing
3 Hole and Valve for Introduction of Gasoline
4 Hole and Valve for Introduction and Sampling
of Gas
5 Spiral for stirring Gas. The Liquid is stirred by
a Spiral Coil Wound around the axis
6 Sapphire Viewing Window
7 Viewing Window
8 Lighting Window
9 Liquid Volume Measurement Chamber
10 Magnetic Agitator
1
2
3 4
5
6
7 8
9
10
pVTMeasurements 143
Figure 6.7: PVT cell (after WELLSROOF)
Figure 6.8: RUSKA cell
Capillary
(Glass)
Windows
144 pVTMeasurements
Figure 6.9: RUSKA volumetric mercury pump
6.2.3 Volumetric Pumps
Volumetric pumps  the so called positive displacement pumps  are used to measure the
volume of the inert liquid injected to and withdrawn from the cell respectively. The
pressure can be increased or decreased. The transported volume is proportional to the
change in volume of the system studied.
In the generally used types, the inert liquid is displaced by a plunger from a pressure
resistant and thermostatically controlled cylinder. The movement of the connecting
screwspindle can be read on a calibrated rod and on a dial to increase the accuracy.
In practice, hand and mechanically operated pumps are used, the latter sometimes with
variable speed. A special type of volumetric pumps is the doubleacting one: Two
plungers are working simultaneously, but in opposite directions.
A scheme of a handoperated volumetric pump is shown in Figure 6.9. Its general
construction does not differ from the pump which is mechanically driven.
6.2.4 Auxiliary Equipment
The PVT cell, the volumetric pump(s), and other parts of the PVT system are
interconnected by pressureresistant stainless steel tubes. In order to separate single parts
of the equipment, valves are mounted in the network of tubes.
Pressure in the system is measured by gauges containing BOURDONtubes. However, also
dead weight gauges can be used for this purpose and for the calibration of the
BOURDONtubegauges as well. There are also gauges in which an electric signal is
proportional to the pressure measured. These gauges are also built into the tube network.
They are filled with the inert liquid used in the volumetric pumps.
In order to store the inert liquid (mercury), a container is used which is also interconnected
pVTMeasurements 145
through valves with the different parts of the PVT system.
The temperature of the volumetric pump(s), in the PVT cells, and in the thermostate
around the cell is usually measured by thermocouples (may be controlled by
thermometers). Whenever pressure, temperature, and the volume in the pumps are
measured electrically, the process can be automated and computerized.
If the gas phase liberated in the PVT cell will be displaced, its volume must be measured.
For this purpose gasometers are used. However, gas could condense under ambient
conditions or at preselected pressures. Therefore, a laboratory separator must be inserted
between the PVT cell and the gasometer.
6.3 Methods
Basically, there are three methods generally used in PVT studies. They are distinguishable
by considering the change and the direction of change in the composition of the system.
6.3.1 Flash Process
In the course of flash processes, pressure is gradually increased or decreased at constant
temperature under equilibrium conditions. At each pressure step, the volumes of the
system and that of the existing phases are measured and the development of phases is
observed. Measurements are usually repeated at other temperatures, too, covering some
range between reservoir and surface temperatures (in the separator or in the stock tanks).
Measurements start at or above the reservoir pressure. The lowest pressure applied is
determined by the system studied. For systems in which no phase transition takes place,
the lowest pressure selected could be the atmospheric one. In this case the flash process
simulates the whole course of production.
Whenever phase transition occurs, measurements should be carried out in a pressure
region which correctly characterizes the phase transition and the 2phase (or multiphase)
behavior of the system. In this case the flash process  carried out at reservoir temperature
 simulates the behavior of the system far inside the reservoir where its composition is not
influenced by the production. A scheme of the flash process has been presented in Figure
4.3.
146 pVTMeasurements
6.3.2 Differential Process
The differential process describes the change in the volumetric and phase behavior of the
system at reservoir temperature and under equilibrium conditions, where any decrease in
the reservoir pressure results from the stepwise production of one of the equilibrium
phases (usually liberated or free gas).
In contrast to the flash process during which the composition is maintained and the
volume is varied, the differential process is carried out at constant volume and by
variation of the system composition.
In the course of the differential study  also explained in Figure 4.3 , the volumes of the
system and those of the existing phases are measured. The development or disappearance
of phases is observed. To follow the change in composition of the system and to
characterize the displaced phase (gas), the composition of the latter is also determined. By
use of a laboratory separator in which pressure and temperature correspond to those
during the field separation process, further information on the volume (and composition)
of phases separated can be given.
The differential process is repeatedly carried out, usually down to the lowest possible
reservoir (separator) pressure. As a result of the gradually changing composition, the
values of characteristic parameters will differ from those determined by the flash process.
except the starting point representing the intact system under initial reservoir conditions.
6.3.3 Reverse Differential Process
The reverse differential process is used to study the effect of any gas injection process on
the volumetric and phase behavior of systems. This can be done under initial conditions
(gas cycling) or by starting from a given depleted state (gas injection with pressure
increase). Both are constant volume processes. In the course of gas injection (using a
preselected gas), the phase volumes are measured and phase transition(s) are observed at
reservoir temperature. There are two subgroups of the reverse differential process.
If the effect of pressure maintenance by gas injection is being studied after each step of
the differential measurement, p is reestablished by injecting a measured volume of the
preselected gas. This method is used, when
• in the initial state of a gas condensate system, the liquid phase exists beyond the
restriction of retrograde condensation,
• partial or total evaporation of the immobile liquid condensate is needed.
In depleted (gas condensate or oil) reservoirs, the effect of increasing the pressure by
preselected gas can also be studied. In the course of a stepwise process, an isothermal
pressure increase is realized under constant volume conditions. The highest pressure to be
pVTMeasurements 147
reached corresponds to the fracturing pressure of the formation, or  if this information is
not available  it is about 10% higher than the initial reservoir pressure. If some 1phase
system develops below this pressure, it is obviously no need to continue the process in a
higher pressure region. The effect of gas injection is often the combination of the methods
discussed above.
© No part of this publication may be reproduced in any form. Only students of the University of Leoben may copy for studying purposes. Edited by DI Claudia Scharf  October 2004
Table of Contents
Chapter 1 Review of Thermodynamic Terminology ........................................................ 1 Chapter 2 Phase Behavior ................................................................................................... 3
2.1 2.2 Gibbs’ Phase Rule ........................................................................................................3 SingleComponent System ..........................................................................................4 2.2.1 Water .............................................................................................................4 2.2.2 nButane .........................................................................................................7 Critical State and Quantities of Corresponding States ...............................................11 Binary Systems ..........................................................................................................13 MultiComponent Systems ........................................................................................20 2.5.1 Ternary Phase Diagrams ..............................................................................20
2.3 2.4 2.5
Chapter 3 Equations of State ............................................................................................ 31
3.1 3.2 3.3 3.4 Change of State at Low Compressibility ...................................................................32 Equation of State of Perfect and Real Gases .............................................................33 Cubic Equations of State ............................................................................................35 Virial Equation of State .............................................................................................47
Chapter 4 Calculation of Phase Equilibria ...................................................................... 49
4.1 Mixtures .....................................................................................................................49 4.1.1 Definitions ...................................................................................................49 4.1.2 Kfactors ......................................................................................................50 Composition of Phases in Equilibrium ......................................................................54 4.2.1 Definitions ...................................................................................................54 4.2.2 Evaluation of KFactors Using Convergence Pressures .............................62 4.2.3 Evaluation of Convergence Pressure ...........................................................68 4.2.4 Evaluation of by use of PengRobinson equation of state ..........................69
4.2
Chapter 5 Phase Properties ............................................................................................... 81
5.1 Natural Gases .............................................................................................................81 5.1.1 Volume ........................................................................................................81
...........................................1 Flash Process ...................................................................... 137 Experimental Determination of the Volumetric and Phase Behavior ....................2................................ 117 5............2 Differential Process ......... 98 5...................... 118 5..........2 General Criteria ...........................................2............................................................1.................................................................................................................................1...... 146 6......3 Compressibility of Undersaturated Liquids ................................................................3............ 84 5.............2...2....................4 Auxiliary Equipment ...........................................ii Table of Contents 5..... 145 6.... 105 5..3................................... 114 Brines ...................... 98 5.......1.............................................................................................................................1 Objectives ..............3 5....2 Formation Volume Factor ............................................3..1 Composition of Brines ....................5 Water Content ............................. 133 6......................................................... 133 6....2..........................................................................................................110 5.................................. 144 Methods ..............................................1.....1................................ 144 6..........2.........3 Reverse Differential Process .....................................2 Formation Volume Factor .................. 146 6..................................................................2................................ 134 6.........................................................................4 Viscosity ....................3 Sampling Methods ...................2....3 .... 127 Chapter 6 pVTMeasurements ............................6 Viscosity ................................ 145 6........................................................................................................................2 5.....1.....................3...............................................3............ 138 6............................1 Sampling ...............................................................................................................................................................................3. 121 5. 126 5.................1....... 133 6........................................................... 139 6.......................................................................................4 Natural Gas Hydrates ...........3..........................................................2 Formation Volume Factor ...................................... 83 5.....................4 Correlation of ZFactor on the Basis of Reduced State Variables .................. 93 5...........3 Viscosity .1 Equipment ......3 Volumetric Pumps ...3 Compressibility .2 6..................................1 Volume ...... 94 Hydrocarbon Liquids .... 133 6..................................................................................84 5.....................................................2 PVTCells .1.......................1..........4 Special Problems .................................................................138 6...........
Lacey and Schaafsma) (1934) 63 Comparison of Kfactors at 100 oF for 1.3: Figure 4.volume phase diagram of nbutane (from GYULAY.5: Figure 4. 1957)66 Kfactors for hexane.000 psia convergence pressure (from NGAA.reduced volume phase diagram of paraffins with low molecular weight (after Gyulay.000psia convergence pressure (from NGAA.temperature phase diagram of the binary system ethane/nheptane (from Kay. 1967)18 Properties of ternary diagrams20 Typical features of a ternary phase diagram21 Triangular diagrams for the methane/propane/npentane system at 160 oF(71 oC) (after Dourson et al. ethane and propane versus pressure at T = 140oF (from Standing.3: Figure 2.7: Figure 2.10: Figure 2.7: Figure 4. 5. 1967)29 The Van der Waals isotherms near the critical point35 Fugacity of natural gases (from Brown.4: Figure 4.temperature phase diagram of the binary system ethane (z = 0.List of Figures Figure 2.9: Figure 2.2: Figure 4.. 1973)26 Phase diagram of a highshrinkage crude oil (from McCain.15: Figure 2.schematic (not drawn to scale)4 Phase equilibrium surface of a pure substance (from Gyulay. 1973)27 Phase diagram of a low shrinkage crude oil (from McCain. 1967)9 Critical pressure as a function of number of Catoms in homologous series (after Gyulay. 1973)26 Phase diagram of a retrograde gas condensate (from McCain.16: Figure 2. 1967)14 Pressure .density phase diagram of nbutane (from GYULAY. 5.22: Figure 2. 1977)83 .temperature diagram of the binary system ethane/nheptane (from Gyulay.Pressure diagram of the binary system ethane/nheptane (from Gyulay.methane for binary hydrocarbon mixtures (from Winn.8: Figure 5. 1938)15 Pressure . 1957)64 Kfactors for methane.11: Figure 2.19: Figure 2.5: Figure 2.24: Figure 3.13: Figure 2.2: Figure 2.8: Figure 2. 1967)14 Phase equilibrium surface of the binary system ethane/nheptane (from Gyulay.18: Figure 2.4: Figure 2. 1967)5 Vapor pressure diagram of nbutane (from Gyulay.1: Figure 2. 1952)70 Zfactor of methane.20: Figure 2.1: Water system .21: Figure 2.14: Figure 2. 1959)24 Phase diagram of a dry gas (from McCain.000 psia convergence pressure (from NGAA. 1967)11 Critical temperature as a function of numbers of Catoms in homologous series (after Gyulay. 1967)17 Mole fraction (ethane) . 1943)23 Critical loci of methane/propane/npentane systems (from Katz et al.1: Figure 4. 1967)8 Temperature .6: Figure 4. 1957)67 Convergence pressure data . 1967)7 Pressure .12: Figure 2.23: Figure 2.17: Figure 2. 1973)27 Phase diagram pairs of gas cap and oil zone28 Phase equilibrium surface of oil/natural gas systems (from Gyulay. 1945)52 Ideal and real Kfactors of nbutane at 60[oC]52 Flash and differential vaporization59 Kfactors for methanepropane at Tc = 100 oF (from Sage..000 and 5.6: Figure 2. 1973)26 Phase diagram of a wet gas (from McCain.9683)/nheptane17 Mole fraction(ethane) .1: Figure 4. 1967)12 Combined reduced pressure .
....29: Hydrate portion of the phase diagram for a typical mixture of water and a light .. 122 Figure 5.......4: Pseudocritical temperatures and pressures for heptanes and heavier (from Matthews et al.6: Water content of natural gas in equilibrium with liquid water (from Katz et al.. 1950)127 Figure 5...15: Typical graph of formationvolume factor of oil against pressure .. 1942)88 Figure 5........14: Apparent liquid density of natural gases in various API gravity oils (from Katz... 1944) ....26: Density of brine as a function of total dissolved solids (from McCain. 1949)85 Figure 5..... 1944)123 Figure 5.........21: Essential feature of the water pattern analysis system (from Stiff. 96 Figure 5... after Beal.. 1959)92 Figure 5....7: Viscosity of paraffin hydrocarbon gases natural gases at atmospheric pressure (from Carr et al. 101 Figure 5. 106 Figure 5. 1946)114 Figure 5.. 1951)120 Figure 5......8: Viscosity of natural gases at atmospheric pressure (from Carr et al.27: The isothermal coefficient of compressibility of pure water.13: Density correction for thermal expansion of liquids (from Standing.....iv List of Figures Figure 5. 1952)102 Figure 5........3: Zfactor for natural gases (from Brown et al...19: Viscosity of gassaturated reservoir crude oils at reservoir conditions (from Chew and Connally...20: Prediction of crude oil viscosity above bubble point pressure (from Beal...24: Typical graph of formation volume factor of water against pressure.. 122 Figure 5.22: Course of Arbuckle formation through Kansas shown by water patterns (from Stiff..11: Pseudoliquid density of systems containing methane and ethane (from Standing.....18: Viscosity of subsurface samples of crude oil (from Hocott and Buckley......... 1941..16: Pseudoreduced compressibility for undersaturated reservoir fluids (from Trube..... 1952)104 Figure 5......... 1957)112 Figure 5.............28: The viscosity of water at oil field temperature and pressure (from van Wingen.....9: Correlation of viscosity ratio with pseudoreduced pressure and temperature (from Carr et al.....25: Bw for pure water (dashed lines) and pure water saturated with natural gas (solid lines) as a function of pressure and temperature (from Dodson and Standing.... 1957)112 Figure 5.. 1944)125 Figure 5. 1952)100 Figure 5... 1954)96 Figure 5......23: Solubility of natural gas in water (from Dodson and Standing.17: Pseudocritical conditions of undersaturated reservoir liquids (from Trube..5: Pseudocritical properties of Oklahoma City Gases (from Matthews et al... 120 Figure 5.... 1942)99 Figure 5......... 1946)116 Figure 5. including effects of gas in solution (from Dodson and Standing..... 87 Figure 5... 1948) .... 1942)88 Figure 5..12: Density correction for compressibility of liquids (from Standing............ 1954)95 Figure 5... 1954) .2: Zfactor as a function of reduced pressure for a series of reduced temperatures (from Sage and Lacey..... 1973)124 Figure 5....... 1952)...10: Variation of apparent density of methane and ethane with density of the system (from Standing and Katz.... 1959)115 Figure 5.. 1951) ..
..........30: Pressuretemperature curves for predicting hydrate formation (from Katz..........31: Depression of hydrate formation temperature by inhibitors (from Katz et al...........138 Figure 6.. 1945)..........................................................3: PVT cell (after Burnett) ..............................143 Figure 6.......................................8 gravity gas without hydrate formation (from Katz........................................................................2: Blind PVT cell ............... 1945)129 Figure 5..........................................................131 Figure 6...............7: PVT cell (after WellsRoof)............................6: PVT cell (after Sloan) .....................140 Figure 6..................................................................32: Permissible expansion of 0................................142 Figure 6.. 1973)129 Figure 5...................1: Scheme of PVT equipments..................8: Ruska cell .........................List of Figures v hydrocarbon (from McCain................................... 1959)130 Figure 5..........................142 Figure 6..............................................................................141 Figure 6.4: PVT cell (after DeanPoettman) ..9: Ruska volumetric mercury pump ...................................140 Figure 6.......................144 .......................) ......................143 Figure 6...............................................5: Variable volume cell (after Velokivskiy et al..............
.
It is called a “homogeneous” closed system if it contains a single phase. e. such as volume. of a large number of various hydrocarbons. The thermodynamic properties of a system are classified into “intensive” and “extensive” properties. A “phase” is defined as a physically homogeneous portion of matter. in opposite to an “open system” which exchanges matter with the surroundings. entropy etc. composition. a natural gas phase or an oil phase. these terms are insufficient. Both systems may exchange energy with the surroundings. A “heterogeneous” closed system contains more than one phase. i. Therefore.g.e. However. pipe lines. tables and equations) for sketching the state of a hydrocarbon system as a function of the state variables is one of the objectives of this textbook. enthalpy. The concept of a closed system is of major interest in applied hydrocarbon thermodynamics. 1 . pressure.real or imaginary . e. However. tank). of the amount of the substance in the system.g.is called a “boundary”. temperature. i. Intensive properties such as temperature and pressure are independent of the size of the system. A system is called a “closed system” if it does not exchange matter with the surroundings. The surface of the system . A systematic presentation of basic thermodynamic tools (charts.e. Changes in the state of aggregation during production should always be taken into account in consequence of changes of the reservoir pressure and changes of pressure and temperature in the production system (tubing. it may be helpful to refurbish the thermodynamic terminology at the beginning as far as it is indispensable to the understanding of these tools. It is not obligatory that a phase is chemically homogeneous. Extensive properties depend on the amount of the substances. Thermodynamic studies are generally focused on arbitrarily chosen systems while the rest of the universe is assumed as “surroundings”. separator. terms such as “oil reservoirs” and “gas reservoirs” are used both in colloquial speech and technical literature. Thermodynamics has evolved to a science of studying changes in the state of a system with changes in the conditions. It may consists of several compounds. The phases of a heterogeneous system are separated by interfaces and are optically distinguishable.Chapter 1 Review of Thermodynamic Terminology When considering hydrocarbon reservoirs..
no further change or . It is this pathindependent characteristic of the state functions that makes it possible to quantify any change of a system. “State functions” or “state variables” are those properties for which the change in state only depends on the respective initial and final state. “Equilibrium” has been defined as a “state of rest”. However. In an equilibrium state. • isothermal (constant temperature).no netflux will take place unless one or more properties of the system are altered. • isochoric (constant volume). Any change of a system is called a “thermodynamic interest” in the thermodynamic study of the system: • adiabatic (no heat added to or removed from the system). a system changes until it reaches its equilibrium state. • isobaric (constant pressure).g. the mole volume.more precisely .2 Review of Thermodynamic Terminology the extensive properties per unit mass or mole are intensive properties.. . all real processes are “irreversible” with varying degrees of departure from a reversible one. e. On the other side. A process is called “reversible” if it proceeds through a series of equilibrium states in such a way that the work done by forward change along the path is identical to the work attained from the backward change along the same path.
C is defined as the smallest number of constituents by which the coexisting phases can be completely described.2) . A system in thermodynamic equilibrium posesses an accurately defined relationship between the state variables. V. temperature. • C: number of components. F is defined as the number of quantities such as pressure. • F: number of degrees of freedom. the phase rule introduced by GIBBS (1928) is applied.Chapter 2 Phase Behavior Hydrocarbon reservoirs consist of rock and fluids. F = C–P+2 where • P: number of phases. the third will be stipulated.1) 2. concentrations which can be varied within finite boundaries without changing the number of phases in the system. The phase behavior of the actual hydrocarbon mixture in the reservoir can be described as a function of the state of the system. 3 (2. Water in brine form and a gaseous and/or liquid hydrocarbon phase are regarded as reservoir fluids. T ) = 0 . By specification of two variables. (2. These are united in the socalled “equation of state”: F ( p.1 GIBBS’ Phase Rule When referring to the number of phases coexisting in the thermodynamical equlibrium.
2.schematic (not drawn to scale) Water is one of the most thoroughly studied chemical compounds. To gain a full understanding.01 0.1 illustrates at which state of the system .or even multicomponent systems are much more complicated.1: Water system .2 describes the system in a qualitative and very general manner. Therefore. 2. water (liquid state) and steam (gaseous state).T). • The melting point curve OB combines the states of equilibrium between the solid . The circumstances in case of 2.2. However. The possible phases are ice (solid state).charaterized by p and T . The phase diagram in Figure 2.0006 0 Vapor A 0.two or all three phases are in equilibrium: • The sublimation curve OA signifies the equilibrium between the solid and vapor. it is best to discuss the phase behavior of pure substances (singlecomponent systems) first. to the composition of the particular phases or to the proportions of the phases are given.09 C Pressure [MPa] Water Ice 0. it is discussed as a singlecomponent system in this context.4 Phase Behavior Eq.0075 100 374 Temperature [°C] Figure 2.2 SingleComponent System 2. no reference to the state variables (p.1 Water B 22.
Example 2.6 Pa and T = 273. • The vapor pressure curve OC specifies the states of the system at which the liquid and vapor coexist. • The end point C of the vapor pressure curve is the critical point and signifies the highest temperature and pressure at which the liqiud and vapor coexist (pc (H2O) = 22. • F = 2 in any area of single phase state (P = 1. • F = 1 on the curves describing the 2phase (P = 2) equlibria. the thermodynamical data at this point are p = 610. • At the triple point O. Tc(H2O)= 647. all three phases are in eqilibrium. 2. by use of Eq. C D E A B p D A C C p T D E B v E B A p T v T v C A D B E Figure 2.15 K. the "wet" vapor is in equlibrium with the "saturated" liquid.16 K.Phase Behavior 5 and liquid state. On this curve. 1967) .2 (GIBBS’ phase rule) and Figure 2.2: Phase equilibrium surface of a pure substance (from GYULAY.1: • F = 0 at the tripel point (P = 3).09 MPa. In case of water. Either the temperature or the pressure is freely eligible without counteracting any given phase equlibrium. Both pressure and temperature (naturally inside finite boundaries) are freely eligible without transforming the system into a multiphase system.1 The degree(s) of freedom in different states of a singlecomponent system.
 p ∂T V Then. However.6 Phase Behavior The state variables. Thus. it is proven that ∂V  ∂T p ∂p .  . the 2phase state (e. it is possible to draw only one single tangential plane at an optional point of the surface. On the basis of which describe a change of a system’s state by changing pressure and temperature at constant specific volume.3) The three differential quotients describe three essential fundamental properties of the system: (i) the isothermal compressibility κ . 2.6) Since the specific volume . according to Eq. . Tcoordinate system An example for an equilibrium surface is giben in Figure 2. (ii) the cubic expansion coefficient a. V can only assume positive values.1 . p.V. ∂V . water in equilibrium with steam) is characterized by any area surrounded by two curves which converge at the critical point: • On the bubble point curve. ∂V V ∂T p 1 ∂p β = . and (iii) the pressure coefficient β : 1 ∂V κ = – . a plane is defined by two vectors which infers that the differential quotients ∂p . The shape of such an equilibrium surface is substance specific.4) (2. V ∂p T 1 α = . = –  ∂T V ∂V  ∂p T (2. T.g.7) (2. Assuming that the partial derivatives are steady. the graphical illustration of the state of any system is only situated in the positive section of the p. • The dew point curve characterizes states in which a negligible small amount of liquid . ∂V ∂T V  p  T ∂T ∂p are not independent of one another.in contrast to the representation in Figure 2. the following is valid: α = pκ β (2.appears now as a state variable. an infinitesimal small amount of vapor is in equilibrium with the “saturated” liquid.2.3.5) (2.
the bubble point curve and dew point curve coincide.4.2. Tplane results in isochores (V = const). • the projection into the V.2 nButane Projections of the equilibrium surface into two planes of the positive section of the p.V. • the projection into the p.5 V= l /Mo 2 H E Vapor 1 0 50 100 150 200 Temperature [°C] Figure 2. When regarding the projection which represents the 2phase area (liquidvapor) in the p.3 and Figure 2. Vplane results in isotherms (T = const). This aspect will be described in detail by discussing the phase behavior of the simple hydrocarbon nbutane. Tplane.2 displays that • the projection into the p.05m3/Mol A C 3 3 F Liquid B D m 0 . Tplane results in isobares (p = const). The resulting single curve is named vapor pressure curve.Phase Behavior 7 is in equilibrium with “wet” vapor. However. It is common to simplify the complex spatial illustration of the equilibrium surface by applying normal projections. 6 5 G Pressure [MPa] 4 V=0. Of course. Figure 2. it is possible to draw isochores: One upwards into the liquid phase and one downwards into the gas phase. the vapor pressure curve is not isochoric. 2. 1967) .T coordinate system are displayed in Figure 2.3: Vapor pressure diagram of nbutane (from GYULAY.
8) . describe those states of the system without any phase transformation by changing the pressure. the mass ratio of the liquid and vapor phase being in equilibrium with one another can be calculated by the principle of the lever: mL ⁄ m v = DB/AD (2. the isochores V = 0. vaporization commences. which do not intersect the 2phase region. it is possible to comprise the whole 2phase region. From this moment.4: Pressure . the isothermal vaporization or condensation takes place as an isobaric process.8 140 3 130 BP A D B Liquid and Vapor 2 0 0.015 V [m/kg] 3 Figure 2. including the critical point C.g.010 0. By applying the projection of the equilibrium surface into the p. At the point E. The intersection point of all other isotherms with the bubble point curve (e. An isochoric change of state must be analogously discussed. A) marks the specific volume of the saturated liquid which is in phase equilibrium with the specific volume of the wet vapor (e.005 0. marked by point B).05 m3/kmoleinside the liquid phase region and V = 0. When moving to the bubble point B by isothermal expansion. 1967) Isotherms.4).3 illustrates the vapor pressure curve of nbutane. 5 Vapor 4 C 150 Liquid DP 17 0 °C p [MPa] 16 15 0 15 5 2 .5 m3/kmole inside the vapor region are shown.g. Then the continuation of this isothermal expansion includes no further pressure drop in the system until the last molecule has passed over to the gas phase.8 Phase Behavior Figure 2. further expansion will result in further pressure decrease. In addition. Considering point D within the 2phase region of the system (specific volume of the system in total). In case of the state A. nbutane is in the state of a “dry” vapor.volume phase diagram of nbutane (from GYULAY. In this region. Vplane (see Figure 2. butane is an undersaturated liquid.
The isobare touching the critical point has an inflection point just as the critical isotherm in Figure 2.4. Inside the 2phase region. Vv.8 From Figure 2. 200 Liquid T [°C] 5 M C 150 Vapor Pa DP Liquid and Vapor 200 400 3 BP 100 0 ρ [kg/m ] Figure 2.2 100 kg nbutane are filled up in a sealed 10 m3 container.95 m3 . can be now calculated by multiplying V with mv: Vv = 0.3125 .3125 m v + mv mv = 76.0125 m3/kg.19 kg.Phase Behavior 9 Example 2. the critical density can be calculated . The volume of the vapor phase at T = 130°C and p = 2. The vapor volume of the system.4: V = 0.density phase diagram of nbutane (from GYULAY. ς diagram of nbutane.5: Temperature . 2.5 demonstrates the T.4 using Eq.7 MPa can be evaluated from Figure 2. The specific volume of the vapor phase is marked by point B in Figure 2. 1967) Figure 2. mL + mv = 100 = 0.4: mL ⁄ mv = DB ⁄ AD = 0. average values of fluid and vapor density are located on a straight line. With the help of this rule (C AILLETETMATHIAS rule (1886)).
3 .192 E + 2 4. the line must o be extrapolated to the critical temperature of methane. Example 2.9185 E + 02 1. 3 C : ρc = 178.148.240 E + 1 1.281 E + 2 2.3 4.118.506 E + 2 3.10 Phase Behavior by extrapolation.3 Use the CAILLETETMATHIAS rule to evaluate the critical density of methane.9827 E + 02 1.3 .eq kg m3 Vapor Density ρ v.0365 E + 02 1.26 x 0. 3 3 . Table 2.1076 E + 02 2.045 E + 2 3. The slope of a straight line can be evaluated by regression analysis.138.654 kg ⁄ m . On the basis of the averaged densities given above: tg α = –0.713 E + 2 3.1: Temperature T oC Liquid Density ρ liq . 654 kg ⁄ m . 3 ρc . eq kg m3 Averaged Density ρe q kg m3 . The densities of the liquid and the vapor phase being in equilibrium have been measured at different temperatures (see table below).3 .899 E + 1 2.925 E + 1 2.128. T c = – 82.8493 E + 02 1.546 kg ⁄ m .3 . ρc = 161. To evaluate the critical density.158.798 E + 0 7.889 E + 2 3.311 E + 0 2.7855 E + 02 It is known as the C AILLETETMATHIAS rule that the averaged densities are situated on a straight line.624 E + 0 1.3 . The values of averaged densities have already been calculated.108.55 .
the critical pressure pc.Phase Behavior 11 2. ∂V ∂ V2 (2. At the point of the inflection.4 illustrates the inflection point of the critical isotherm at the critical point. both the first and the second partial derivates of p = p(V) equal zero that 2 ∂p ∂ p  = 0 =  . 1967) .9) The state of the system at this point is characterized by the critical specific volume Vc.3 Critical State and Quantities of Corresponding States Figure 2. and the critical temperature Tc. 6 e Ol no mo 5 Benzene Toluene Cyclohexane Naphtalene es f in pc [MPa] 4 iC4 iC5 3 no m r al Pa ra ffin s 2 1 0 2 4 6 8 10 12 14 CAtoms per Mole Figure 2.6: Critical pressure as a function of number of Catoms in homologous series (after GYULAY.
the critical data of homologous compounds with longer carbon chains can be extrapolated though an increasing error has to be taken into consideration. with a margin of error 1 to 2%. k). Because of thermal decomposition. the homologous series of hydrocarbons show regularities which can be used for extrapolation.e..12 Phase Behavior 500 Naphtalene 400 300 Toluene Benzene 200 100 0 100 200 0 2 4 6 8 10 12 14 no rm almo Pa no ra O ffi lef n ine s s Tc [°C] CAtoms per Mole Figure 2.results in a close relation between the p. Vc T T r = Tc (2. Vr phase diagram which is valid for paraffins from methane (CH4) to hexane (C6H14).. Tproperties of pure hydrocarbons if the state variables are substituted by the so called “reduced quantities” which are p pr = . for homologous series . . 1967) Considering the critical data pc and Tc..7: Critical temperature as a function of numbers of Catoms in homologous series (after GYULAY.2. it is not possible to obtain experimentally information about the critical data in case of high molecular weight. The experimental data in Figure 2. pc V r = V . C 2H2n(n = 1.g.7 refer to the homologous series of paraffins.10) Figure 2. V. The “principle of corresponding states” for chemically similar substances .8 shows a pr.6 and Figure 2.. However. and olefines. CnH2n+2.
its state is also a function of composition. V. In the plane z = 1. Analogous to the pure substance. This curve divides the envelope surface into two parts: the bubble point surface and the dew point surface. The volume is equivalent to the mole volume. an envelope surface encloses the 2phase state. The phase behavior of the ethane/nheptane system is graphically illustrated by the p. Outside of the dew point surface (high temperature). Just as in case of 1component systems.Phase Behavior 13 2. . Covering all other zplanes. z may be the mole fraction of one (lighter) component. every change in state is described by the equation of state F(p. T. the critical isotherms have an inflection point according to Eq. the state of a dry gas is given. In general. T. The region of an undersaturated liquid state is positioned outside the bubble point surface (low temperature).9. The mole fraction is defined as the ratio between the number of moles of a certain component and the sum of moles of all components. the critical state of binary systems is defined as the state at which the intensive properties of the phases are no more distinguishable. This is demonstrated by the example of three additional zplanes.1) mole fractions because the sum of the mole fractions always equal 1. 2. zcoordinate system in Figure 2. z) = 0.Considering a 2component system. A system being composed of k components is defined by the specification of (k .9. The upper broken line marks the critical points of all compositions which are possible.4 Binary Systems If a systems consists of more than one component. whereas in the plane z = 0 the one of nheptane appears. the vapor pressure curve of ethane appears. the composition is defined by “mole fractions”.
0 Vr Figure 2.0 0 0.5 0.0 2.98 0.8: Combined reduced pressure .reduced volume phase diagram of paraffins with low molecular weight (after GYULAY.2 Tr 6= 1 .0 .00 300 Temperature [°C] Figure 2. 1967) 10 nC7 Pressure [MPa] 7.0 4.0 2 pr 1.9: Phase equilibrium surface of the binary system ethane/nheptane (from GYULAY.0 4 1 C 1.14 Phase Behavior 1.0 1. 1967) .96 0.4 1.50 5.6 0 1.94 0.75 0 0 100 200 1.8 0.0 3.5 0 0.0 z 2.25 C2 0.
For example. This is due to the fact that lighter molecules partially transfer their higher kinetic energy to the heavier ones by collision.00 2 90. the more flat the envelope curve will be. 1938) Figure 2.10 also shows that Tc of a mixture lies between the critical temperatures of the pure substances. pc of the mixture may be obviously higher than the one of the pure substances. Physically.78 7 6. Figure 2. the system maintains the state of a gas phase.14 8 3. At a given pressure.27 9 1.22 3 70. If the mixture consists of two homologous compounds with quite different volatility (in consequence of quite different molecular weights). The smaller the difference between the molecular weights and thus between the volatility. In contrast to this. .25 5 29.temperature phase diagram of the binary system ethane/nheptane (from KAY.10: Pressure .10 shows the projection of Figure 2. the dew point temperature of the mixture at a given pressure is always lower than that of the pure heavier component.91 6 9. On the other side.9 into the p. the maximum of the critical pressure of a methane/ndecane system equals 37 MPa. Tplane. it can be explained by the fact that the thermal motion of the lighter molecules is obstructed by the heavier ones which exhibit more inertia. the critical data curve envelopes a very extensive temperature and pressure region. the bubble point temperature of the mixture is always higher than that of the pure lighter component. [Wt%] Ethane 1 100.22 4 50.25 10 nHeptane 6 600 7 400 8 9 10 200 0 100 200 300 400 500 Temperature [°F] Figure 2.Phase Behavior 15 1400 4 1200 3 5 1000 Pressure [psia] 2 800 1 Composition No. Consequently.
11 illustrates the phase behavior of a certain ethane/nheptane system. The bubble point curve and the dew point curve represent 100% and 0% liquid. Similar phenomena occur when the temperature is changed by an isobaric process where the constant pressure is higher than pc but lower than the cricondenbar of the system. the path surpasses the dew point line twice. the dotted line connecting point C with point C’ marks the states of the system which exhibit the highest volume percentage of condensate dropout. the condensate will be vaporized again. Consequently. If the constant temperature is higher than Tc but lower than the cricondentherm . Figure 2. As on Figure 2.like in case of the path EF . a condensate drops out at the dew point D’. In Figure 2. A quality line represents a certain mole percentage being liquid or vapor in the state of phase equilibrium.11.Tdiagrams. respectively. If the decrease in pressure will be continued. so called "quality lines" are shown on p. In Figure 2. Besides the critical point.11. As soon as the dew point D” has been reached.11 also shows an isothermal decrase along the path EF where E defines the sastem to be a dry gas.16 Phase Behavior Figure 2. This maximum is given by the intersection point of the path EF with the dotted line connecting C and C”. the volume of condensate (liquid) will be at its maximum. the curve enveloping the 2phase region possesses two additional characteristic points: • C’: the point of highest pressure on the curve that is called cricondenbare. At some point between D’ and D”. All the quality lines (isochores) converge at the critical point.11. the condensated phase has been vaporized in total. . This process is called a “retrograde condensation”. • C”: the point of highest temperature on the curve that is called cricondentherm. the quality line "20%" represents the states in which 20% of the system account for the liquid phase.
Phase Behavior 17 7 C' E D' 6 C 5 Liquid C'' Pressure [MPa] 4 Liquid and Vapor Vapor 3 20% 10% D'' 2 A 1 F B 0 0 50 100 150 Temperature [°C] Figure 2.6 0.temperature phase diagram of the binary system ethane (z = 0.4 BPL 0.0 4. As far as the system ethane/nheptane is concerned.8 β' 1.11: Pressure .0 50 0 50 100 150 200 250 300 Temperature [°C] Figure 2.6 7.10 elucidates that the cricondenbar is located on the bubble point curve at low mole fractions of ethane. 0 0.2 0.7 [MPa] δ' δ α β 1. 1967) .9683)/nheptane It depends on the composition of the system if the cricondenbar is located on the dew point curve or on the bubble point curve.0 D PL z 0.4 2.temperature diagram of the binary system ethane/nheptane (from GYULAY.12: Mole fraction(ethane) .2 5. Figure 2.
Tplane corresponding to another possible projection of the surface in Figure 2. 9 8 7 BP Pressure [MPa] 6 5 100 °C 3 A' 2 D P 200 °C 4 15 0 °C A A'' 1 0 0.86 MPa).4 0.13 by using the principle of lever. the phase behavior of ethane/nheptane systems is graphically illustrated in the z. T = 150oC) is filled up with 100 kg of a ethane(z = 0.4. The composition of the liquid phase is given by point A’.13). The composition of the infinitesimal small liquid phase in equilibrium with the vapor phase corresponds with point δ ’.9.6 0. . Determining the phase composition.12.8 1. An example may be the composition at the point A(T = 150 oC). can be evaluated from Figure 2.0 z Figure 2. The mixture α achieves bubble point β due to an isobaric (1. Example 2. Point β symbolizes the composition of the liquid phase which is in equilibrium with an infinitesimal small vapor phase whose composition is symbolized by the point β’. dew point state is reached at point δ . 1967) The design of the corresponding p. A sealed container (p = 2.Pressure diagram of the binary system ethane/nheptane (from GYULAY. respectively.2 0.18 Phase Behavior In Figure 2.13: Mole fraction (ethane) .4). Again the relative masses of both phases can be determined by applying the principle of the lever (see Example 2.4 MPa) heat supply. The mole number of ethane in the liquid phase and in the vapor phase. zdiagram is also possible (see Figure 2. During further increase of temperature. the one of the vapor phase by point A”.47)/nheptane mixture.
The weight of ethane. mC7.792 = 1.604 kmole The composition of the vapor phase is given by point A” in Figure 2. n = nC2 + nC7. From Figure 2.693 + 0.693 kmole .82 The mole number of ethane in the vapor phase can now be calculated by .12 n liq .458 n vap A A′ where nliq:total mole number in the liquid phase nvap:total mole number in the vapor phase Thus the total mole number in the vapor phase results in n vap = 0. is given by mC2 = 100 .12: z = 0.= .= 0.Phase Behavior 19 At first.785 n C2 = .47 = ( 100 – mC7 ) ⁄ 30 + mC7 ⁄ 100 mC7 = 79.215kg mC2 = 100 – 79. MC7 = 100 kg/mole) are inserted into mC 2 ⁄ M C2 z = mC 2 ⁄ M C2 + mC 7 ⁄ M C7 to evaluate the weight of nheptane in the system.= .mC7.785kg Now the total mole number of the system.792 kmole .= 0.485kmole .215 = 20. mC2. the mole weights (MC2 = 30 kg/mole.215 n C7 = . ( 100 – mC7 ) ⁄ 30 z = 0. MC2 30 mC7 79. MC7 100 n = 0. can be calculated: mC2 20.= A A″ = 1.
881 kmole . the composition .23 .82 × 0.12: z = 0.203 kmole . v a p = z × n vap = 0.5. The total mole number in the liquid phase results in nliq = n – n vap = 1.expressed in mole fractions . the sum of the perpendicular distances from any point to each side of the diagram is a constant equal to length of any of the sides. Figure 2. On the basis of the equilaterality of the triangle.485 – 0.881 = 0.14: Properties of ternary diagrams 2.495kmole .of a point . Each corner of the triangle represents one pure component. 2 3 d.604 = 0. The mole number of ethane in the liquid phase can now be calculated by nC2 .5 MultiComponent Systems 2. Fraction Component1 2 3 2 a. 1 b. liq = 0.23 × 0.672 = 0. The composition of the liquid phase is given by point A’ in Figure 2. Thus. 1 1 L2 L1 3 L3 Const.1 Ternary Phase Diagrams It is common to illustrate the phase behavior of 3component systems at constant pressure and temperature in so called triangular diagrams.20 Phase Behavior nC2 . 1 C io at tR 2 n s ta t o on 1 of A 3 D B 2 c.
15: Typical features of a ternary phase diagram . (2.14: • For mixtures along any line parallel to a side of the diagram. The “tie lines” connect compositions of liquid and vapor phases in equilibrium. Only the amounts of the phases change as the overall composition changes. • Mixtures lying on any line connecting a corner with the opposite side contain a constant ratio of the component at the ends of the side. Any overall composition on a certain tie line gives the same liquid and vapor composition being in euqilibrium.15 shows the 2phase region for chosen p and T. Figure 2.13) Vapor Region Two Phase Region Li n e ve Binod al C ur Plait Point Critical Region Liquid Region 3 Ti e 2 Figure 2. LT where LT = L1 + L2 + L3 .. the fraction of the component of the corner opposite to that side is constant. LT L3 z 3 = . The principle of the lever finds application again and nA . LT (2. Any mixture with an overall composition lying inside the binodal curve will split into a liquid and a vapor phase.12) L2 z 2 = . 1 (2.Phase Behavior 21 in the interior of the triangle is given by L z1 = 1 .11) Several other useful properties of the triangular diagrams are also illustrated by Figure 2.= DB nB DA gives the mixing ratio leading to mixture D. • Mixtures of any two compositions lie on a straight line connecting the two initial points on the ternary diagram.
16 illustrates the influence of pressure on the phase behavior of a certain ternary system at constant temperature. pc and Tc.Tdiagram of the methane/propane/npentane system.14) In accordance to this aspect. By drawing the tangent in the plait point on the binodal curve.17 illustrates a corresponding application by the respective p. It is useful to comprise the two heavier components of a ternary system and to reduce this system to a fictitious binary system. the 2phase region shrinks. can be determined for any mixture of the three components (see Example 2. All envelope curves are characterized by the portion of propane in the hypothetical component (propane/npentane) which is given by z3 C = z3 + z5 (2. on the basis of a hypothetical component. The mole% of methane are specified along the outermost envelope curve. Mixtures of a composition being located in the critical region with another one being located in the liquid or vapor region will. also result in a singlephase system if the straight line connecting the two initial compositions does not intersect the 2phase region. the singlephase region is splitted into three sections. in any case. . As pressure increases.22 Phase Behavior The liquid and vapor portions of the binodal curve meet at the “plait point” which represents the critical composition. Figure 2. Figure 2. the critical state properties.5).
2 nPentane 0.0 C= 0.8 Propane 0.0 C= x c3 C = x c3 + x c5 0.0 C= 0.6 0. 6 C= 0 .2 0.8 0 1.6 0.4 0.2 nPentane x c3 C = x c3 + x c5 c.6 C= 0.8 Propane 0.4 0.4 0.6 T=160°F 0.4 C=0.2 b.2 0.6 0. 8 1.8 0.6 0.8 0. Methane p=1500 [psia] 0.4 C= 0. 1943) .4 0.4 C=0.6 C= 0.6 0. Methane p=1000 [psia] 0.4 0.2 C=0 .Phase Behavior 23 Methane p=500 [psia] 0.2 0.4 0.2 a.6 0.8 0.8 0. Figure 2.8 0.2 T=160°F 0.8 0.4 0.0 C= 0.16: Triangular diagrams for the methane/propane/npentane system at 160 oF(71 oC) (after D OURSON et al.2 T=160°F 0.2 nPentane 0.6 0..6 C=0.4 0. C= 0.2 Propane C= 0.4 x c3 C = x c3 + x c5 C= 0.8 C=0.
.= 0.= 0. At first. 16 13.24 Phase Behavior 3000 2500 C3 C= C 3 + nC 5 80 75 70 [ le Mo C= C Critical Pressure [psia] 0 C= 2000 C= 0. 6 C= 0.36 .4 0 0. 0. the mole numbers and the respective mole fractions must be calculated.2[kg kmol1 ]).3 + 0. The critical data of this mixture can be evaluated by use of Figure 2.5 The hydrocarbon mixture is composed of 8 [kg] methane (M = 16[kg kmol1 ]. 72.5 + 0.= 0. 0 50 C %] 40 30 20 10 0 H4 1000 500 C1 0 200 C3 C5 100 0 100 200 300 400 Temperature [°F] Figure 2. 8 n1 = .2 [kg] propane (M = 44.1 32.5 n5 = .45 The portion of propane in propane/npentane is given by the hypothetical component . 0.2 60 1500 C= 1.= 0.5 [kg] npentane (M = 72.17: Critical loci of methane/propane/npentane systems (from KATZ et al.3 + 0.5kmole .3 z3 = .4 .2 n3 = ..17.8 =0 .= 0. 0.45 0.= 0. 44.24 .3kmole .45 z5 = .45kmole .5 + 0.3 + 0. 13.45 0.5 + 0. 1959) Example 2.2 0.5 z1 = .1[kg kmol1 ]) and 32.
A so called “wet gas” (see Figure 2. Consequently. • the lighter pseudocomponent including ethane to hexane (C2C6). • the heavier pseudocomponent including heptane and higher hydrocarbons (C7+). the system remains in the singlephase (gaseous) state. Nevertheless. some generalization makes a description of these complex systems possible . Figure 2. condensate will drop out under separator conditions. Anyway. In case of an isothermal pressure decrease (full line from point 1 to 2).17 at C = 0. Tphase diagrams.22 by p. Even in case of both pressure and temperature decrease (dotted line).18 to Figure 2. The application of the triangular diagram is not solely confined to ternary systems.= 0.4 .24 C = . it is essential to know • if the critical temperature is lower or higher than the reservoir temperature.18 represents a hydrocarbon system whose critical temperature is significantly lower than the reservoir temperature. z3 + z5 0. For the classification of natural hydrocarbon systems.Phase Behavior 25 z3 0.36 From Figure 2.19) remains in a singlephase (gaseous) state in the reservoir during production (line 12).4 and 40 mole percent methane: T c = 262. which occurs in the reservoir adjacent to the production well the the course of production. For example it is possible to partition the paraffinic hydrocarbons into three pseudocomponents which are • methane (C1) as the light component. possibly propane and some hydrocarbons of higher molecular weights. small amounts of ethane. The phase behavior of several complex and natural hydrocarbon systems are demonstrated in Figure 2.= . Anyway. Not considered in this classification are changes in composition during production. only poor information of complex natural hydrocarbon systems has been reported until now. no liquid phase will drop out. Dry gases contain mainly methane.according to known data. the considered hydrocarbon mixture is called a “dry gas”.24 + 0.5°F = 128°C and p c = 1344 psia = 9. . • which state will be achieved at surface conditions (separator).27MPa .
The initial conditions given by Point 1 specifies the hydrocarbon mixture as a dry gas. Consequently. 1 Pressure Critical Point Liquid 75 50 25 Sep. If the pressure will decrease adjacent to the production well during production. Systems as shown in Figure 2. 2 Gas Temperature Figure 2.18: Phase diagram of a dry gas (from MCC AIN. The pressure at point 3 corresponds to the state in which the condensed liquid phase reaches the maximum (in mole%). 1973) Liquid Critical Point 1 2 3 Sep.26 Phase Behavior In case of the system shown in Figure 2. the dew point of the system is reached at point 2. condensate drops out inside the reservoir.20: Phase diagram of a retrograde gas condensate (from MCC AIN.20. 1973) . 100 75 50 25 5 0 2 Sep.20 are called “gas condensates”. 1973) Liquid Critical Point 1 Pressure Mole % Liq. the amount of condensate is larger than in case of wet gases. 100 75 50 25 10 5 0 Temperature Figure 2. Gas Temperature Figure 2. In the separator. the reservoir temperature is higher than the critical one but lower than the cricondentherm.19: Phase diagram of a wet gas (from MCCAIN. Gas Pressure Mole % Liq.
Liquid Mole % Liq. from the initial pressure (point 1) to the pressure 2.21).are referred to as “high shrinkage oils”. about 85 mole% of the produced hydrocarbon mixture is in the liquid phase at separator conditions. the shrinkage of the oil due to gas liberation is less pronounced than in case "white oils" (see Figure 2.22 shows a “black oil” or “low shrinkage oil”.The D ew  Gas .22: Phase diagram of a low shrinkage crude oil (from MCCAIN. If the pressure in the neighbourhood of the production well will decrease during production to point 2. about 65% of the produced hydrocarbon mixture will exist as liquid phase if the reservoir is produced at bubble point conditions. Since the bubble point curve will be reached by the decrease in pressure due to production. At separator conditions. The initial state is characterized by point 1 at which the state of the system can be regarded as “undersaturated” liquid. If the hydrocarbon mixture in the reservoir is a 2phase state under initial reservoir conditions. 1973) The so called “white oils” . The separator conditions are near the bubble point curve. 1973) Figure 2. The reservoir temperature is below the critical temperature. In accordance to this fact.21: Phase diagram of a highshrinkage crude oil (from MCCAIN. Temperature Figure 2. 100 1 Undersaturated 2 Saturated Critical Point Pressure oin tL i ne 3 75 Bu bbl e P 50 25 0 Po in t L in e Sep. a further pressure drop in the reservoir will lead to point 3 and thus to an increased development of the vapor phase. 100 Critical Point Pressure 75 3 Sep.23). the bubble point curve is reached and the state of the system is now considered “saturated”. 50 Gas 25 Temperature Figure 2.21 . Consequently.as characterized in Figure 2. oil and gas phase can be considered apart from one another (see Figure 2. The equilibrium conditions at the initial state of the system are given by the intersection point of the dew point curve of the gas cap and the bubble point curve of the oil zone.Phase Behavior 27 Liquid 1 2 Mole % Liq.
The system of composition II corresponds to a gascondensate system.28 Phase Behavior gas cap shows a “retrograde” behavior.23: Phase diagram pairs of gas cap and oil zone In Figure 2. . the phase behavior of two representative mixtures of I and IV are given by the compositions II and III. the phase behavior of natural hydrocarbon mixtures can also be illustrated in p. Just as in case of binary systems (see Figure 2.as dotted lines . zdiagrams. the one of composition III to a white oil. Again an envelope surface comprises the 2phase region in dependence on the composition. isochores of the liquid phase and .9). The spatial curve. Outside the envelope surface. (y1). is given by point 4 in the p. the system is in a singlephase state. Furthermore. information about the composition of the system will be obtained.the locations of maximum retrograde condensation are drawn. which connects the critical points. zplane. zplane. By projecting the phase surface into the p. If the state of the system is represented by point 1. composition I represents the separator gas while composition IV represents the corresponding separator oil of the well stream. Inside the 2phase region of system II and III. splits the surface into two parts which are the dew point surface and the bubble point surface. and the one of the vapor phase. T. Gas Cap Retrograde Condensating Gas Cap Pressure C Oil Zone Pressure C pi C pi C Oil Zone Temperature Temperature Figure 2. the equilibrium composition of the liquid (x1). if the intersection point is located on the dew point curve of the gas cap between the critical point and the cricondentherm.24.
0 III GL C1 P IV G II T1 T2 T3 I III z Figure 2. 1967) .2 CIV IV C IV C'' II GL' C'' CI I 0 .4 0 .24: Phase equilibrium surface of oil/natural gas systems (from GYULAY.0 C'' CII C' CI CI 0.Phase Behavior 29 IV III II I Pd C Pb CII T1 T2 T3 y1 0.2 0 .0 4 x1 0 .8 0.4 C' CIII CIII C' CII K 1.6 L C'IV C C'' CIII 0 .
30 Phase Behavior .
a pure graphical application of the state functions is not very practical and . In case of a 2phase system: n = nl i q + nv a p and therefore nVt = n liq V liq + n vap V vap where Vt: Vliq: Vvap: n: nliq: nvap: mole volume of the system mole volume of the liquid phase mole volume of the vapor phase number of moles in the system number of moles in the liquid phase number of moles in the vapor phase.T).3) (3.phase consists of k components. planes.1) In this context.1. T ) (3. V must be defined as the volume of one mole (intensive property).on top of that . 2. (p. If a phase . (p. V.V). The application of such diagrams enable the determination of the respective volume for certain states defined by the corresponding pressure p and temperature T. the following relation regarding the p.Chapter 3 Equations of State The preliminary chapter included graphical illustrations of equilibrium surfaces and their normal projections into the (p.z) etc.impossible for multicomponent systems. (3. the phases can be regarded as seperate thermodynamic systems.2) If phase equilibrium is given. However. On the basis of Eq. the corresponding equation of state may be written as follows: 31 .may be the liquid . This aspect obviously makes a mathematical consideration of these problems necessary. Tdata is valid in case of any chemically homogeneous phase: V = V ( p.
xk)space by one single equation of state. 3. …xk ) (3. the so called “constraint equation” is valid: ∑ xi = 1 . Eq.6) In various cases of even practical interests. To ) + ∂V ( T – To ) + 1 ∂ V ( T – T o ) + .. x1 x2 . T. 3. or x1 x2 .or just the same of Eq. T o ) +  . the mole volume of the phase will result in V = ∑ xi Vi i=1 k (3. In this case. This means that the enthalpy of the system is equal to the sum of the enthalpies of the single components.4) V = V ( p. T ) = V ( p o . i=1 k (3.7 is valid. It is impossible to approximate the equilibrium surface for the entire (p.32 Equations of State F ( p. a multicomponent phase behaves as an ideal mixture and the volumes are strictly additive. T.8 may be truncated to . 3. 3. T.  o o ∂p T 2 ∂p 2 T 2 2 V ( p..7) If Eq. Put into other words: The mixing energy will be zero. 3. Therefore some procedure bit by bit is necessary. T) = V ( p o. V. no thermal effect will take place during the mixing procedure. Considering the mole fractions.  .8) By assuming that the higher derivations are neglectable.5 at constant composition .6 in discussing Eq. If Vi is defined as the mole volume of component i in the phase.into a TAYLORSeries leads to 2 ∂V ( p – p ) + 1 ∂ V ( p – p )2 + x V ( p.5) where xi is defined as the mole fraction of the component i in the (liquid) phase. V. 3. ∂T p 2 ∂T 2 p (3.1) = k + 1 variables.. it is obvious that V is a function of 2 + (k . the enthalpy of the system must be generally considered additive..5. …x p ) = 0 (3.1 . Regarding Eq..1 Change of State at Low Compressibility The expansion of Eq. 3.
particularly the PENGROBINSON equation. Then for a molar system: pV = RT where R is defined as “universal gas constant” and R = 8. Therefore. 2.10 can at best be applied for fluids in a 1phase state.9 may be transformed to V ( p. Eq. 3. if the intermolecular (VAN neglectable.31434 J/mole K The compressibility factor Z is defined as pV Z = RT (3.12) (3.11) DER WAALS) forces are . the constants included in these equations are not always given. ∂V ( p – po ) + . In doing so.4 and Eq. will be discussed in this textbook. T o ) . ∂V ( T – To ) V o ∂p T V o ∂T p where V o = V (p o. the proceedings of generalization in derivating this equation of state will be elucidated. 2.V (p . 3.T ) = V o 1 + .Equations of State 33 1 1. there may exist equations which approximate the measured values more accurately. only the cubic equations of state. Considering the Eq. Anyway.2 Equation of State of Perfect and Real Gases A gas is defined as perfect.5.10) Eq.9) (3. Experience and practice have shown that the cubic equations of state are most sufficient and beneficial for calculating the state of gases and of 2phase systems. Of course. T) = V o [ 1 – κ ( p – p o ) + α ( T – T o )] where κ : isothermal compressibility α : cubic expansion coefficient (3. 3.
= 55 ×10 m . the pressure has been recorded as 13.13) Example 3. A cylinder with volume of 0..5o C.12: pc V c Zc = . the factor Z equals 1. Zc. For ideal gases. After the volume reduction.79 MPa. 3.13 and considering the measured volume amounting to 0. 3. the temperature and the composition. and specific volume Vc and may be determined experimentally for 1component systems. Z is a state variable and depends on the pressure. Tc and V c into Eq.79 MPa. Then the volume of the cylinder has been isothermally reduced to 0.63 –5 55 ×10 –5 . 2 1379 ×10 –3 By use of Eq. If the gas would show an ideal behavior at 90oC and 13. V ideal respectively.34 Equations of State or V actual Z = . The critical compressibility factor. RT c (3.00035 m3 (volume reduction in consequence of mercury insertion). For real gases.14) (3. The critical point is defined by the pressure pc..= 0. can be evaluated by substituting the critical data pc..00035 m3 : 35 ×10 Z = .075 m3 has been filled with a gas under atmospheric pressure and at a temperature of 90.1 Evaluation of a Zfactor from laboratory data. temperature Tc. p2 V 2 = 75 × 10 –3 –5 3 101 ×10 . the specific volume could be calculated by P1V1 = P2V2 and so p1 V 2 = V 1 .
T1 = co ns t. In the year 1873. b can be interpreted as the inherent volume of the molecules which is not available for the thermal motion of the molecules. The real behavior is shown by the straight full line BD inside the 2phase region.16 cannot predict the real behavior of the system during the vaporization. 3. SOAVE (1972). Eq.1: The VAN DER WAALS isotherms near the critical point .– . VAN DER WAALS deducted the first equation of state which is able . The dotted section of the isotherms represents the the data which are predicted by using the VAN DER WAALS equation. and P ENG and R OBINSON (1976). 2 V (3.15 may also be written in the following form: RT a p = . ( V – b) V2 (3.to describe the continuity from gaseous to the liquid state: a p +  ( V – b ) = RT ..1 illustrates the VAN DER WAALS isotherms in the vicinity of the critical point. Obviously. 3. which have been formulated by REDLICH and KWONG (1949).3 Cubic Equations of State If the pressure of any gaseous system is low. Pressure Critical Point Calculated from Van der Waals Equation T>T 1 2 T 2= co ns t. the ideal gas equation remains sufficient to describe the volumetric behavior. The cubic equations of state. Eq.16) Figure 3. have achieved much better results.up to a certain degree . respectively condensation. B Actual Path D Specific Volume Figure 3. The term a/V2 regards the pressure reduction in consequence of intermolecular attraction.15) a and b are substance specific constants.Equations of State 35 3.
a is now a function of the temperature. 3.19) – 0. c c The comparison of Eq.17) where b again is a substance specific constant. p p p p At the critical point ( V = V c ) : 2 3 3 3 2 ( V – Vc ) = V – 3V c V + 3V V – V = 0 .21 and Eq.36 Equations of State The R EDLICHKWONG equation is given as RT a p = .20) (3. 3. 3.23) (3. It is useful to write the parameter a as follows: a = a′f ( T ) where a′ is constant.. 3. Substitution of Eq.18) (3.21) (3.– .5 (3.24) .22) ac b 3 V c = pc where ac = a′f ( T c ) .– . 3.22 results in (3. Anyway.– b . 3.– b V – . The original R EDLICHKWONG equation included . pc 3V a c bRT c 2 2 = .20 leads to RT c 3V 3 = . pc pc c (3. f( T ) = T By multiplication of Eq.23 into Eq.19 and Eq.– (V – b ) V (V + b ) (3.= 0 .17 with V( V + b) ( V – b) ⁄ p and after arrangement of V corresponding to its order of power: a bRT ab 3 2 2 V – RT V + .
– 3bV c – b c b or after rearrangement.= Ω b . if T = T c .18 and Eq.21 can be combined to 2 – 1 RTc RT c b = .25) (3. in 2 3 3 3 2 b + 3b V c + 3bV c + V c = 2V c .19 and the arrangement of Z corresponding to its order of power results in 2 ZRT 3 – RT ZRT + a – bRT – b2 ZRT – ab = 0 . 3.24: f(T) a = ac . 3.31) (3.30) (3. f ( Tc ) It is obvious that α = 1 . Furthermore. p p p p p p b 2 2 1⁄ 3 (3. 3.23 into Eq. 3 3 ( b + Vc ) = 2V c . From Eq.= Ω b 1⁄ 3 pc 9( 2 – 1 )pc The constants Ω b and Ω a have the following numerical values: Ω b = 0. Eq.08664. Ω a = 0.21 and Eq. 3. 3.Equations of State 37 3 Vc 2 2 3V = . 3. The substitution of V = ZRT ⁄ p into the Eq. 3..42748.29) (3..26) (3. b = (2 1⁄3 (3. 3pc pc Inserting Eq. .29: ( RT c ) ( RT c ) a c = ..= ac α .28 and Eq.27) (3. 3.28) – 1 )V c .32) Therefore: .
35) 3 2 2 (3.30 into Eq.8664 .36 leads to (3.2748  α T 2 r p (3.38 Equations of State bp b p ap ap bp Z – Z + . Please note that the ideal gas equation contains no substancespecific parameters. 3. 3.34 includes only two parameters which are pc and Tc.5 – 0. 3.37) and pr B = 0. RT The substitution of Eq..39) . 3.– .33) or Z – Z + ( A – B – B )Z – AB = 0 where ap A = 2 ( RT) and bp B = .– ..35 and Eq. Anyway. the increasing yield of experimental data has more and more indicated that the behavior of many liquids with a nonspherical molecule structure deviates greatly from the predicted one. This made the introduction of a third factor necessary..Z – . Tr In the original REDLICHKWONG equation: f( T ) = T and so α = Tr – 0.34) r A = 4.40) Eq. 3. it has improved the calculation of PVTproperties in a fundamental way. .5 (3.= 0 2 2 RT 2 ( RT ) ( RT ) ( RT ) RT 3 2 2 2 (3. (3.29 and Eq. Since the RedlichKwong cubic equations of state consider these two substancespecific parameters.38) (3.36) (3.
pc RTc b = 0.– 3b V – b .43) (3. ω . Just as in case of the R EDLICHKWONG equation.– RT b – b = 0 ... Moreover.44) RT c V c = 0.5 (3.45) (3.574ω – 0. the prediction of liquid density is combined with large errors. several proposals for a third parameter have been made. the following terms and equations can be obtained: 3 RT 2 a 2bRT 2 a 2 .307 . pc 2 (3. Improved approximation has been achieved with the PENGROBINSON equation: RT a p = .42) The weakness of all these equations ranging from the original R EDLICHKWONG equation to all its modifications (including the SOAVE equation) is the fact of an universal unrealistic Zc factor of 1 ⁄ 3 .176ω )( 1 – T r ) .7 r s s where p r = p ⁄ p c is the reduced boiling point pressure.07796 .457235 .V – . pc a = ac α . The equation of state from S OAVE (1972) only differs from the REDLICHKWONG equation with respect to the definition of the factor α 0.5 (3.48) .47) (3.Equations of State 39 Beginning with the year 1951 (MEISSNER and SEFERIAN).– V – b V ( V + b )+ b( V – b ) where a is given by Eq. 2 0.  p b p p p (3. ( RTc ) a c = 0.41) = 1 + ( 0. has become the one with greatest acknowledgement: s ω = – lgp + 1 at T r = 0.48 + 1. 3..– .18 and b is further a substance specific constant.46) (3.– b V + . The so called “acentric factor”.
49) (3.43 have been established for pure substances. The extension for multicomponent systems requires the calculation of the respective data of the pure components and mixing rules in order to get the parameters of the mixture.34 to Eq. Anway. S OAVE and PENGROBINSON.5 2 . Eq. the calculated fluid densities are much more accurate than the ones calculated by the equations of state previously discussed.29 to Eq. T 2 r 3 2 2 2 3 (3.457325 ⋅ pr  α . + 1.5 = 1 + 0. and Eq.31. 3.52 on the one side and between Eq. 3. 3.51) 4 1 T r / 2 .38.07796 . . (3.17.. Eq.54) where ai can be calculated on the basis of the critical data of each component by using Eq. different mixing rules exist and are presented below. Tr α 0. 3. is universally defined as an arithmetic average by using b = i=1 ∑ xi bi . 3. Eq.16. A = 0. 3.31.53) For the temperaturedependent coefficient a .19.52) Analogies between Eq. R EDLICHKWONG: a = i=1 ∑ xi a i k 0.44 to Eq.3676 (3.50) pr B = 0. Eq. 3. 3.26992ω 1 – (3. 3.21. The universal critical Zfactor of the PENGROBINSON equation results in 0. 3.54226ω – 0. 3. 3.42 on the other side are obvious.40 into Eq.307 which is much better than 1 ⁄ 3 but still far away from reality. 3. and Eq. The mixing rule for the parameter b. Eq.40 Equations of State Z – ( 1 – B )Z + ( A – 2B – 3B )Z – – AB – B – B = 0.30 and Eq. 3. 3. which is included in the equations of R EDLICHKWONG. k (3.
55 and Eq.56.54 if all binary interaction coefficients are zero.55 and Eq. Example 3.30 and Eq. .5 ( 1 – kij ) (3. Eq.is by mean of matching the phase behavior of multicomponent systems. 3. In Eq. The mixing rules used by SOAVE and PENGROBINSON consider the binary interaction between the molecules of the components i and j.00025 m3 contains 0. can be evaluated by Eq. A laboratory cell at temperature of 100 o C with volume of 0. 3. Obviously. The composition of the system. 3.56 reduce to the form of Eq.25 mole of gas.2 Calculation of the pressure by use of the PENGROBINSON equation of state.55) where ai and aj. 3. 3.52.Equations of State 41 SOAVE: a = i=1 ∑ j ∑1 xix j ( a i aj ) = k k 0.56) where ai is defined by Eq. 3.42.if no data are available . Values of the binary interaction coefficients must be obtained by fitting the equation of state to gasliquid equilibria data for each binary mixture.5 ( 1 – kij ) (3. the critical data and the acentric factors of the components are tabled below. respectively.48 and Eq. PENGROBINSON: a = ∑ ∑ xix j ( a i aj ) i = 1j = 1 k k 0. 3. 3. the terms k ij are binary interaction coefficients which are assumed to be independent of pressure and temperature. Another possibility of obtaining this coefficient . They have different values for each binary pair and also take on different values for each equation of state. 3.
the parameters of ethane and nbutane were evaluated.4 4. They are tabled below.8018 . 8 463 × 10 – 6 831434 6 from Eq.0115 and the reduced temperature of methane 273. i oK Acentric Factor ωi 0. 3.1 × 1907 × 10 ) = 457235 × 10 .= 0.1928 MPa C1 C2 n .15 Tr .i pc. 3.6 α 0. 3.C4 190.75 0.0908 0.47: α1 = 0. .46: –5 × 1907 × 10 b1 = 77796 × 10 .= 2677 × 10 .= 1. and from Eq.1999 .48: ac .2 3.8018 = 0. 8 463 × 10 – 6 ( 831434 6 2 from Eq.1 = .80 First.958 : 190.52 by inserting the acentric factor for methane ω1 = 0.2494 × 0.5 = 0.0115 0.42 Equations of State Component Composition yi 0. Correspondingly to the calculations for methane.6 4. For methane from Eq.2494 .88 425.20 0.05 Critical Data Tc.60 305. 3.8954 1 and thus α1 = 0. the parameters b and a must be calculated for each component.
5(1 . 048E – 5 + 0. R = 8.000000) = 0.20 × 4.1999 x 0.56.20 x 0.8260 x 0.000000) + + 0. 180E – 5 .6045)0. 3.2494 0.5(1 .8260 x 1.6045 x 0.C4 bi 2.0000 E+0 0.6045)0.05 x 0.20(0.75(0.8260 The parameter b can be calculated by use of Eq.677 E5 4.00025 3 V = .0.20(1.0.0.= 0.0000 E+0 0. 3.6045 1.5(1 .8260)0.75(1.1999 x 1.20(0. 245E – 5 b = 3.6045)0.5(1 .001 m ⁄ mole .25 T = 373.5(1 .0.53: b = 0.2648 E2 0. To calculate the coefficient a by use of Eq.8260 x 0.002648) + + 0.0010 1. the interactive coefficients.0.0. 3.014640) + + 0.5(1 .245 E5 ac.1999)0.5(1 .000000) + + 0.2648 E2 0.4904 E2 n .75 × 2. They are given below: C1 C1 C2 n .75 x 0.2130 ai 0.1999)0.0000 E+0 a = 0.8260)0.4904 E2 0.05(0.05(1.75(0.004904) + + 0.3108 Now the pressure inserting can be calculated through Eq.5050 αi 0.8018 1.1999 0.6045 x 0.014640) + + 0.67E – 5 + 0.05(0.43 and by 0.C4 0.20 x 0.15 K .i 0.1999 x 0.C4 0.5(1 . must be known. Kij.31434 J ⁄ K × mole .75 x 0.1464 E1 C2 0.05 x 0.75 x 0.004904) + + 0.8260)0.05 x 0.5(1 . 0.6045 x 1.0.0.20 x 0.1464 E1 0.Equations of State 43 Component C1 C2 n .002648) + + 0.6042 1.1999)0. 05 × 7.048 E5 7.0.
8018 = 0. ω = 0.= 1. 3.328 MPa Example 3.6 K . 4 463 × 10 From Eq.5 = 0.48: ( 8.662 ×10 . 190. pc = 4. and from Eq.63 MPa . The critical data and the acentric factor of methane are known as Tc = 190.12 K and at pressure of 9.26 MPa. 601 . First. 12 Tr .= 0. 6 4.0 kg methane (MC1 = 16 kg k/mole is hold at temperature of 305.0115 .31434 × 190.3108 .63 × 10 From Eq.077796 8. 6 α 0.6 ) ac = 0.8954 and thus α = 0.8018 .18 ×10 and –5 a = 0.1988 . 3. The calculation procedure results in p = 2.3 Calculation of the density of methane by use of the PENGROBINSON equation of state.457235 .248 .6 = 2. From Eq.46: –5 b = 0.52 by inserting the acentric factor and the reduced temperature of the system 2 305. 3. Now the mole volume of methane at the given state variables has to be evaluated through Eq.44. To solve this cubic equation of . 3. 3.31434 × 190. 2.248 × 0.44 Equations of State b = 3. the parameters a and b have to be calculated. 1 = .47: a = 0.
s and t result in r = – 0.b – b .– b . 3 the formula x + rx + sx + t = 0 is reduced to 3 2 y + py + q = 0 where 3 3s – r p = 3 and 2 2r rs q = .p p The calculation of r. the CARDAN equation is applied for x + rx + sx + t = 0 where 3 2 RT r = – . p a 2bRT 2 s = .+ t . –8 .– .– . t = –0.Equations of State 45 state. 2473 ×10 s = 0.. 563×10 –8 . By substitution of r x = y–.– 3b p p and RT 2 t = – b a – .4755 ×10 and –3 . . 3585 ×10 – 12 . 27 3 The calculation of p and q result in 3 p = – 1.
M ρ = .53776 ×10 – 25 .+ D . The discriminant D is defined as 3 q 2 D = p + .46 Equations of State and q = – 1. –5 –4 .3272 ×10 y = 1.7800 ×10 v = 5. the mole volume can be calculated through r x = y–.  2 3 and thus D = 1.5107 ×10 and thus –5 . V = 2.– D 2 v = and 3 y = u + v. nV V . 2 q –. 089×10 – 12 . 3350 ×10 and –4 m ⁄ mole .= . 3 3 m .522 kg ⁄ m .: 3 u = 9. Based on the relationships u = 3 q –.= 68.
However.59) Since also Z becomes zero as p goes to zero. p → 0 p (3. they are obviously qualified for practical application.57 according to the gas law by Z RT/p. de l’Hospital’s rule can be applied resulting at constant T in ∂Z B = RT lim  . are solely a function of temperature. p → 0 ∂ p T (3.+ . then: Bp Cp Z = 1 + . 3. Sometimes Eq.+ … ZRT ( ZRT )2 2 (3. natural gas) by considering only the first two or three terms (depending on the availability of the virial coefficients).57 is applied for the calculation of gaseous states (e.58) The second virial coefficient B can be shown to be the slope of isotherms on a plot of Z = Z ( p ) at p = 0 .4 Virial Equation of State All cubic equations of state mentioned above are more or less empirical or at best semitheoretical.60 implies that the 2.58 may be neglected so that B = ZRT lim Z – 1 . C etc. 2 V 3 V V (3. virial coefficient vanishes at the BOYLEtemperature.+ . the virial equation of state has some theoretical background in statistical mechanics.60) Eq. . In contrast. 3.… . The BOYLE temperature is defined as the temperature above which Zvalues become greater than 1 for all pressures.57) The virial coefficients B.g. However. If V is substituted into Eq.+ . 3. When p approaches zero. the form of an infinite series concerning the volume is not very appropriate for practical application. 3. The pressure is expressed as 1 B C p = RT . the third term in Eq.Equations of State 47 3.
48
Equations of State
Chapter 4
Calculation of Phase Equilibria
4.1 Mixtures
4.1.1 Definitions
In case of thermodynamical phase equilibrium of a multicomponent system, all phases are  in physical terms  homogeneous mixtures. Vaporized compounds are always mixable in all relations, independent of their chemical characteristics. Therefore, only one vapor phase exists. Compounds existing in a liquid phase are only mixable in all relations if they are chemically similar. Since hydrocarbons  especially whose which belong to the same homologous series  exhibit such chemical conformity, their mixtures will be physically homogeneous, independent of the composition of the mixture. The composition of the system and its phases are specified by mole fractions. The mole fraction of the component i is given by the ratio between its mole number, ni, and the total mole number in the system (in the respective phase), n. The sum of all mole fractions equals 1. In the following, the different mole fractions are marked by zi: mole fraction of the component i in the total system xi: mole fraction of the component i in the liquid phase yi: mole fraction of the component i in the vapor phase, so that in the system in total
n i z =  , i k n i =1 i
∑
(4.1)
49
50
Calculation of Phase Equilibria
n i x =  , i k n i i=1 liq
∑
(4.2)
yi =
n i .  k n i i=1 vap
∑
(4.3)
If a multicomponent system is in the state of thermodynamic phase equilibrium, the distribution of the component i among the vapor and the liquid phase is characterized by its distribution coefficient, the so called “Kfactor”: yi K i =  . xi (4.4)
The numerical value of Ki depends on the state of the system. Accordingly, the composition of the phases will change at any variation in state.
4.1.2
Kfactors
Ideal liquid mixtures have the property of being met by RAOULT’s law so that xi =
p i o p i
(4.5)
where pi: partial pressure of the component i p o : tension of component i as a singlecomponent system i xi : mole fraction of component i in the liquid phase.
The total pressure of the system adds up to p =
i= 1
∑ pi
k
=
i= 1
∑ xipoi
k
(4.6)
If a vapor phase exhibits perfect behavior, DALTON’s law can be applied and pi yi = p The combination of Eq. 4.5 and Eq. 4.7 results in (4.7)
10) respectively.. the vapor pressures in Eq. the so called “activity coefficient”. the extrapolation of the vapor pressure curves to higher temperatures than Tc does not yield satisfactory results in most cases.12) A generalized fugacity chart for natural gases based on reduced temperatures and pressures is given in Figure 4.8 can be corrected and extrapolated to cover a wider range of conditions.12.11) (4.= xi o p i p (4. Since Tc = 190. certain limitations are placed on the definition of the Kfactor given above. .8) Anyway.9) (4. It is evident that any pure coomponent has a vapor pressure only up to its critical temperature. On the basis of the thermodynamic concept of fugacity.37 K for methane. it becomes evident that some other means must be obtained for the determination of Kfactors if they are to be used for hydrocarbon mixtures. In practice.1 in terms of f/p. the definition of the Kfactor is given by f° yi i (liq ) ′ K i ( id ) = . 4. 4. If the phases are in equilibrium at given conditions and thus fi ( v a p ) = f i( liq) . A first approach originated from DODGE and NEWTON (1937) who have shown that at any given p and T fi ( liq ) = x i f o and f i ( v a p ) = yi f o i ( vap) i ( liq ) (4. where fi(liq) : fugacity of component i in a liquid mixture fi(vap) : fugacity of component i in a vapor mixture f f o i ( liq ) o i ( vap ) : fugacity of pure component i as liquid : fugacity of pure component i as vapor.= . xi f ° i (vap ) (4. These fugacities serve to provide ideal equilibrium constants as defined by Eq.Calculation of Phase Equilibria 51 K i( id ) yi = .
0 5 4 3 log K K 2 Temperature=60 °C 0.1 Kid 5 4 3 2 K'id 0. the extrapolation of the fugacities on the basis of Eq.1: Fugacity of natural gases (from BROWN. Anyway.though also limited by the critical state conditions .2: Ideal and real Kfactors of nbutane at 60[oC] In case of multicomponent hydrocarbon systems. 4.10 .covers a wider range of state conditions than the vapor pressure relations. attraction forces between the molecules .2 shows that the K i( id ) values are not sufficiently accurate to describe phase equilibria at high pressures. ′ Figure 4. 1945) 4 3 2 nButane 1.01 0.52 Calculation of Phase Equilibria Figure 4.1 2 3 4 5 1 2 3 4 5 10 2 3 Pressure [MPa] Figure 4.
Kfactors can be evaluated in satisfactory accordance with the measured data by the application of the “successive solution method”. As both RAOULT’s and DALTON’s laws and the fugacity relations given above (see Eq. NGHIEM. AZIZ and LI (1983) have calculated fugacity coefficients by use of b A ln φi = i ( Z – 1 ) – ln ( Z – B ) – b 2 2B where a ij is defined by a ij = ( 1 – K ij ) ( a i aj ) .in analogy of the “activity coefficient”.5 k xa 2 b i = 1 i ij .P φi.– V – b V ( V + b ) + b ( V – b) in f 1 V RT ln . 4. it is necessary to find another solution for the determination of the Kfactors.= 1 f i( liq) (see Eq. 2.5 0.= Z – 1 – ln Z + .consists in an iterative process...k) in a multicomponent phase P . 4.11). P p where p is used for the total pressure of the system.13) ∑ Z + 2.414 B ln .10) result in the calculation of Kfactors without considering the composition.Calculation of Phase Equilibria 53 affect the vaporliquid equilibria at higher p and T.∫ . 4. 3).. If the phases are in equilibrium at given conditions and thus f i( vap ) . the definition of the Kfactor is given by φ i( liq) K i = φi ( vap ) (4.16) 0.as fi.– i a b (4. f/p .15) .14) (4.414 B (4.it will be discussed through Example 5. Z – 0.P = xi .5 . It is based on the definition of the “fugacity coefficient” of the component i (i = 1.9 and Eq. After introducing not only the PENGROBINSON equation of state RT a p = .– p dV p RT ∞ V but also other terms which have been defined by PENGROBINSON (see Chap. This method .
4. the calculations of xi and yi(i = 1. 4. the following must be valid (on the basis of an overall and particular material balance. temperature and composition is in a liquidvapor phase equilibrium. where φi ( liq ) : fugacity coefficient of component i in the liquid phase φi ( v a p ) : fugacity coefficient of component i in the vapor phase.19 and Eq. 4. where n: n liq : n vap : xi : yi : zi : ( i = 1. 4. k ) (4..2. the following is valid: zin xi = ..k) (Eq.. nliq + n vap K i zi n yi = nliq ⁄ K i + n vap (4.20) If the numerical value for nliq and has nvap have been chosen correctly. 4.54 Calculation of Phase Equilibria (see Eq.17) (4.2 Composition of Phases in Equilibrium 4.18 and after rearrangement.13). 4.18) total mole number of the system mole number of the liquid phase mole number of the vapor phase mole fraction of component i in the liquid phase mole fraction of component i in the vapor phase mole fraction of component i in the system After substitution of Eq.20) result in . respectively): n = nl i q + n v a p zi n = x i nliq + y i nvap .1 Definitions If the system at a given pressure..8 into Eq. ….19) (4.
4.790 nvap = 0. 4.23 elucidates that nliq must be chosen larger in the following iteration if nliq must be chosen smaller in the reverse case of ∑ xi > 1 .1 will demonstrate that it is possible to use the Eq.1 Determination of the gas and liquid phase compositions at p = 14 MPa and T = 92 o C through an iterative calculation procedure.22) Example 5. FIRST ITERATION:nliq = 0.23) Eq. Example 4. If n = 1.21 becomes ∑ xi i=1 k = i = 1 liq i ∑ ) n + K (1 – n i liq k z = 1.17 gives n vap = 1 – n liq and Eq. T and system composition. n liq ⁄ K i + n v a i=1 k k zi n (4. The actual composition of the hydrocarbon system is tabled below. 4.Calculation of Phase Equilibria 55 i=1 ∑ x i= ∑ i = 1 n liq + n vap K i=1 k k zin (4.21 and Eq.21) and i=1 ∑ y i = ∑ p = 1 .22 for iterative calculation purposes which result in the evaluation of the phase compositions at given p. 4. ∑ xi < 1 . (4. 4. Eq.210 .
56 Calculation of Phase Equilibria nliq must be chosen smaller in the following iteration because of Component KFactor System zi x i = Composition n liq + nvap K i zi Ki nvapKi nliq +nvapKi 0.17 0.876 0.598 0.4011 1.24 0.933 0.143 0.0000 2.4404 0.086 0.0405 0.051 0.0353 0.793 0.3173 0.031 0.841 0.388 1.0432 0.0434 0.0174 0. SECOND ITERATION: nliq = nvap = 0.778 0.036 0.41 0.821 0.85 1.003 1.0207 0.776 0.5055 0.68 0.224 THIRD ITERATION: nliq = nvap = 0.222 .01 0.9963 C1 C2 C3 C4 C5 C6 C7+ Σ ∑ xi < 1 .15 0.246 0.0417 0.0290 0.0284 0.0324 0.
24) Example 4.0487 0.0416 0. Example 4.3123 0. .0327 0.0436 0.1. n liq = n and x i= zi .9999 zi yi = nliq ⁄ K i +nv a p 0.0134 0.0082 1.0051 0. .0209 0.8896 0.0296 0. The composition of the system is given in Example 4. Ki = Ki(p) must be known.0210 0. Naturally. Eq. In the vicinity of the bubble point.0054 0. (4.0436 0. Kixi are in a linear relation to pressure.0006 zi x i = n liq + n vap K i 0. linear interpolation can be applied for acceleration purposes.Calculation of Phase Equilibria 57 Iteration II Component Iteration III Iteration III zi x i = n liq + nvap K i 0.0416 0.2 illustrates the iterative process of calculating the bubble point pressure.5132 0.0358 0.0000 C1 C2 C3 C4 C5 C6 C7+ Σ At the bubble point of the system.5145 1.0357 0. the amount of vapor is infinitesimal small and thus n vap = 0.2 Determination of the bubble point pressure of a crude oil. 4.3114 0.= n liq ⁄ K i + n vap ∑ i= 1 k Ki zi = ∑ Ki x i = i=1 k 1. pb.0327 0. for a given hydrocarbon composition and for a fixed temperature. Therefore.22 then becomes ∑ i=1 k zi n .
0275 0.0235 By interpolation: pb = 23 MPa At the dew point of the system.0055 0.0441 0. zi 0.026 1.060 0. However. 4.4400 0. it is theoretically possible to evaluate the dew point pressure by use of Eq.58 Calculation of Phase Equilibria Component System Compos.448 0.020 0.0056 0. 4.323 0.0094 0.9969 C1 C2 C3 C4 C5 C6 C7+ Σ 2.9469 2. From Equation 4.0125 0.0128 0. This notion is mostly used to check the Kfactors which have been evaluated for certain state conditions.= Ki i=1 ∑ i = K i k y 1.0052 0.316 0.025 0.0138 3.9072 2. the calculation of the dew point pressure of hydrocarbon systems by using Eq.= n liq + n vap K i ∑ i=1 k zi .020 0. Assumption pb = 23.0290 0. .21: ∑ i=1 k zin .0445 1.1500 1.0104 0.680 0.25 may be too inaccurate.0443 1.0519 2.2 MPa yi = Kizi 0.4701 1.0067 0.25) In analogy of the iterative process demonstrated by Example 4. n vap = n and yi = z i .0174 0.4404 0.0127 0.678 0.0405 0.0432 0.0284 0.2150 0.239 0.25 by summing up the y i ⁄ K i values and by variating the pressure as long as Σyi ⁄ K i = 1 .0274 0.3000 0. the evaluation of the Kfactor for heavier components is crucial.0100 0.450 0.0000 Ki 1.5 MPa yi = Ki zi Ki 0.0062 0.0300 0.230 0.250 1. Assumption pb = 21 MPa yi = Ki zi Ki 0.0272 0. (4. Therefore.6720 0.0069 0. the amount of liquid phase is infinitesimal small and thus nliq = 0.8896 0. Assumption pb = 22.2 (evaluation of pb).
it is of importance whether the system is signified as liquid at bubble point pressure or as a vapor at dew point pressure. . This process is specified as the “flash vaporization (condensation)”. the vapor phase appears below bubble point pressure. a condensate will drop out below the dew point.as soon as the previous phase equilibrium is achieved (open system).in most cases the vapor phase . the oil phase shrinks at decreasing pressure. further pressure decrease will now effect the vaporization of the condensate. the volume increases during an isothermal pressure decrease.3: Flash and differential vaporization In evaluating hydrocarbon systems. In case of a flash process. Correspondingly. the liquid volume diminishes. These are illustrated schematically in Figure 4. mass and composition of the system are changed in removing one phase . If the temperature of a dry gas system ranges between Tc and the cricondentherm. In both cases. At the same time. two characteristic processes are of importance. The amount of the condensate will increase up to a certain value as a function of pressure decrease (retrograde condensation). Below this point.3: • Mass and composition of the system remain constant (closed system). In case of an undersaturated liquid system. • Preceding every further change of state. Such a process is called “differential vaporization (condensation)”. The portions of the phases and their compositions in various states of thermodynamic equilibrium (p and V are variables) are to be evaluated.Calculation of Phase Equilibria 59 V L Hg Hg Hg Hg Hg flash vaporization at the bubble point Hg flash vaporization at the dew point V V L Hg Hg Hg Hg ntimes differential vaporization at the dew point ntimes differential vaporization at the bubble point Figure 4.
Example 4. (4.27 cannot be integrated. Therefore.21 in regard of their phase composition. the following is obtained after neglection of the term dnvapdxi: K ix i dn liq = xi dn liq + n l i qdx i (4. The volume remains constant meanwhile every change in state is connected with a change of the mole number in the system and its composition (open system).26) Since yi = Kixi and dnvap = dnliq. 4.3 Determination of tank oil composition after 3stage separation on the basis of the system composition in Example 5.19 to Eq.60 Calculation of Phase Equilibria The equilibria states appearing at a flash vaporization in series after one another can be calculated independent of one another by Eq. the method of finite differences is often applied. because Ki is a function of p.27) Eq.582 nvap = 0.33 MPa T = 48.7 oC nliq = 0. In case of a differential vaporization.3). Results have been often proved to correspond with the experimental data in a most accurate way (see Example 4. FIRST SEPARATOR: p = 3. The integral is then substituted by summation formulae.1. This would correspond to the following reflections: In case of a saturated liquid. 4. However. 4. some material balance has to be applied. the amount of component i will decrease by yidnv resulting from the liberation and production of an infinitesimal small amount of gas dnv.418 . and yji dn vap = n liqx i – ( n l i q – dn v a p ) ( x i – d xi ) . the Kfactors must always correspond to the given pressure and temperature.
0019 0.0109 0.44 MPa T = 48.1430 0.0128 0.844 nvap = 0.0095 0.0096 0.4404 0.1110 0.0135 0.0000 KFactor Ki 60.088 0.0560 0.7 oC nliq = 0.1207 0.0489 0.0309 0.0557 0.8991 0.415 0.250 0.0504 0.0485 0.15 KC7 .6877 1.170 0.0000 KFactor Ki 8.0000 1.6530 0.0339 0.0284 0.0419 0.0355 0.0489 0.0485 0.1110 0.0290 0.0281 0.003* Fluids Liquid Vapor xi yi 0.0419 0.0174 0.230 0.4011 1.400 1.013* Fluids Liquid Vapor xi yi 0.0281 0.000 10.0021 1.0106 10000 C1 C2 C3 C4 C5 C6 C7+ Σ= *K7+ = 0.590 0.0405 0.0432 0.0403 0.8132 10000 0.0000 C1 C2 C3 C4 C5 C6 C7+ Σ= *K7+ = 0.600 3.039 0.Calculation of Phase Equilibria 61 Component Well Stream Composition zi 0.0025 0.100 1.0339 0.15 KC7 SECOND SEPARATOR: p = 0.156 Component Feed Composition zi 0.0288 0.6877 0.650 0.
0299 0.0566 0.15KC7 Component Feed Composition zi 0.0514 *K 7+ = 0.0000 KFactor Ki 265.0330 0.0396 0.70 0.052* Fluids Liquid Vapor xi yi 0. the Kifactors in real systems not only depend on temperature and pressure.8132 1.0000 0.0038 0.0444 1.0355 0.72 0.466 mol tank oil 4.0309 0.844 x 0.1885 0.582 x 0. KATZ and HACKMUTH have been the first to provide for appropriate data of gas/oilsystems.0513 0.35 1.0135 0.109 0. .2. Since then the collection of data has been continuously enlargend. but also on composition.0000 C1 C2 C3 C4 C5 C6 C7+ Σ= One mole reservoir crude oil produced amounts to nliq1 x nliq2 x nliq3 = 0.9486 nvap = 0.1 = 48.2 Evaluation of KFactors Using Convergence Pressures As already mentioned.6 14.0557 0.9486 = 0.62 Calculation of Phase Equilibria TANK: p T = 0.8550 1.3 5.1775 0.0007 0.3022 0.7 MPa o C nliq = 0.0 46.0403 0.0210 0.1965 0.
Its Kfactor increases with temperature up to a certain limit and decreases beyond that. However. Above the critical pressure of the system. According to observations. However. Figure 4. In contrast to the critical isotherms.0 3.K > 1 for the highly volatile component methane.4: Kfactors for methanepropane at T c = 100 oF (from SAGE.0 log K 1. the two Kisotherms of binary system will also converge to a certain point at other temperatures than Tc. LACEY and SCHAAFSMA) (1934) The most simple relations exist in any 2component system. One corresponds . this convergence is only an “apparent” convergence. At any isothermal change in state.0 5. the Kfactors increase with increasing T at constant pressure. This point is also characterized by K = 1(y = x). the Kvalues have no physical meaning up to the point of apparent convergence because of the existence of only one phase beyond a certain pressure limit.4 is a plot of the Kfactors for the methanepropane system at the critical temperature of the mixture T c = 100 oF. Generally. Kfactors have no longer any physical meaning. At any pressure .just as in case of the highly volatile component .5 0.with the critical pressure of the mixture and the other one with the vapor pressure of the pure substance.at which the vapor and the liquid phase are in equilibrium . methane shows some irregular behavior. K for the less volatile component propane will equal 1 at two different pressures. its Kfactor will decrease. If the pressure increases.Calculation of Phase Equilibria 63 10. . On the other side K<1 for the less volatile component propane since its concentration in the vapor phase is less than in the liquid.0 0.3 100 200 300 500 1000 1500 log p [psia] Figure 4.
the Kfactors depend only on p and T and are independent of composition. Note: • If the evaluation of the Kfactors is based on the convergence pressure of the mixture.5 emphasizes the dependence of the Kfactors on the convergence pressure being a function of the system composition.1 C7 C8 C9 0.64 Calculation of Phase Equilibria Anyway. the isotherms may be extrapolated to K = 1 defining the pressure at this point as the convergence pressure pk of the system at the actual temperature. the convergence pressure of a mixture coincides with the critical pressure of another mixing ratio of its components. The diagrams were based on average values of already available data from various gas condensates and gas/crude . Figure 4. the Kfactors are nearly independent of composition. The (apparent) convergence pressure pk is a function of the composition of the mixture. 1957) It has been proved that multicomponent systems have properties very similar to those of the binary systems with respect to the convergence of all Kisotherms. the NGAA (Natural Gasoline Association of America) published a diagram series lgK = f(lg p)T in its “Equilibrium Ratio Data Book”.000 and 5. 100 C2 C3 C4 C1 C1 10 1 C5 C6 K 0. In 1957. It then becomes apparent that at low pressures and temperatures. • At a given temperature.0001 10 2 3 45 100 2 3 45 1000 2 3 45 Pressure [psia] Figure 4.000psia convergence pressure (from NGAA.5: Comparison of Kfactors at 100 oF for 1.001 0.7 MPa (100 psia) are not significant for the lighter hydrocarbons ethane through pentane.01 C10 0. The differences in Kvalues for the two convergence pressures at pressures below 0.
Calculation of Phase Equilibria
65
oilsystems. This series includes the paraffinic homologous compounds  methane through decane  just as ethylene, propylene, nitrogen, carbon dioxide and hydrogen sulfide within the pressure range from 0.07 to 140 MPa and for temperatures ranging from  148 to +260 °C. Figure 4.6 and Figure 4.7 present the respective charts for methane and hexane, respectively, in case of the convergence pressure pk = 34.5 MPa (5000 psia). The values for this convergence pressure can be used for a large number of hydrocarbon systems. Today, these data are already available for dataprocessing with computers. If pk is known, it is possible to choose from the measured series those Kfactors , which are best suited to describe the phase equilibrium vapor/liquid of the given system. The second parameter, which has to be evaluated to proceed with the prediction of the phase and volumetric behavior of complex hydrocarbon systems, is the Kfactor of the heaviest component. In conventional analyses of hydrocarbon fluids, everything heavier than hexane is grouped together and reported as the pseudocomponent “C7+”. This fraction is a mixture of materials of varying volatility. However, the vapor pressure curves and the critical properties of the hydrocarbons heavier than hexane are fairly close together. Thus, it is possible to characterize the mixture “C7+” by an average set of Kvalues.
66
Calculation of Phase Equilibria
103 5
2 102 5
Te m pe ra tu re 50 [ °F ] 200 100 0 50 0
2 101
00 20 40 160  1 1 1 
K = y/x
5
2 10
0
80 1
2 0 0
5
2 20
2 40
2 101 5
2 60
2 10
2
101
2
5
102
2
5
103
2
5
104
Pressure [PSIA]
Figure 4.6: Kfactors for methane, 5,000 psia convergence pressure (from NGAA, 1957)
Calculation of Phase Equilibria
67
101 5
Te m pe ra t
2 100 5
ure
50
0
[° F ]
45 0 40 0
2 101
30 0 28 0 26 0 24 0 22 0 20 0
18 0
16 0 14 0
K = y/x
5
1 20
10 0
2 10
2
80
60
40
5
20
0
2 103 5
 20
40
60
2
1 0 0
10
4
101
2
5
102
2
5
103
10 0
80
2
5
104
Pressure [PSIA]
Figure 4.7: Kfactors for hexane, 5,000 psia convergence pressure (from NGAA, 1957)
∑ i =k 2Wi (4. One of the most convenient methods is based on the assumption of a pseudobinary system and consists of the following steps: • A certain convergence pressure is assumed and the Kfactors of components are evaluated from the NGAAdiagrams (examples are given through Figure 4. This iterative method will be demonstrated by Example 4. • With exception of the lightest component methane.29) • The critical point of the component C2+ is plotted on Figure 4. the whole procedure has to be repeated starting with the assumption of another convergence pressure. The Kfactors obtained through this method exhibit only satisfactory results.2.68 Calculation of Phase Equilibria 4.4. Wi.28) ∑ i =k 2 WiTc i Tpc = .1). the Kfactors become very sensitive to the convergence pressure used and care must be . • For the given temperature. The pseudocritical temperature of the pseudocomponent C2+ in the liquid phase is evaluated in calculating the averaged mole weights in the liquid phase. if the operating pressure is considerably lower than pk.8 which shows the vapor pressure curves of homologous hydrocarbons.6 and Figure 4.. • The “binary” critical envelope curve of methane with this pseudocomponent is drawn. • The respective composition of the liquid phase being in equilibrium are evaluated by a flash process (see Example 4.7). The iterative process must go on until correspondence of the assumed convergence pressure and the evaluated one will be achieved. As the operating pressure approaches pk.3 Evaluation of Convergence Pressure Several iterative methods of estimating pk have been proposed. through Wi = x i M i and the pseudocritical temperature of C2+ through (4. all other components are summarized in a second fictitious component. • If this numerical value of the convergence pressure does not coincide with the value initially assumed. the convergence pressure pk can be read off because it coincides with the critical pressure of one of the possible mixing ratios in the “binary” system.
Calculation of Phase Equilibria 69 taken in the selection of the correct value of pk. 3000. However. 4. • New Kvalues can be evaluated by inserting these values in Eq. Thus the interpolation . 2000. 4.4 Evaluation of by use of PENGROBINSON equation of state With help of the “successive solution method”. b. • On the basis of assumed Kfactors.. the pressure and temperature as well as the critical data of the components.16: φi.000 psia). 4. and B after PENGROBINSON are calculated for both phases. the calculation procedure is repeated. • The Zfactors both for the liquid and the vapor phase can be evaluated solving the equation of state (Eq.5: • The composition of the system.may lead to great errors. the phase compositions are calculated through an iterative calculation procedure (see Example 4.2.inevitable in estimating the Kvalues . 5000 and 10.15: 2 ∑ k xi aij b bi i = 1 i A Z + 2 ⋅ 414B ln φi = . 1000. A.14 and Eq. l i q K i = φ i.( Z – 1 ) – ln ( Z – B ) – .ln . Z – 0. • On the basis of the calculated phase compositions. The steps described above will be repeated until the Kfactors converge within a certain limit. charts exist only for some distinct convergence pressure (800. 4.5 ) . 3.1). • The fugacity coefficients for both phases are evaluated by use of Eq.414B a b b 2 2B and a ij= ( 1 – K ij ) ( a i0.3).5 aj0. the Kfactors of the components of the system can be evaluated by an iterative process like the one presented through Example 4. • The last flash iteration results in the evaluation of the mole number in the liquid . vap • On the basis of the evaluated Kvalues.– . 1500.49) through use of the method of C ARDAN (see Example 3. their acentric factors and the interactive coefficients must be initially known. the EOSparameters a.
too. the mole number in the vapor phase nvap and in the evaluation of the composition of the phases. nliq. • Using Zliq and Zvap . 1952) . 900 C16 800 C7C23 CC 24 5 700 C10 600 C8 500 C7 C6 400 C5 iC5 C4 C* C4 iC4 C3 C* C3 Temperature [°F] 300 200 100 C2 C* C2 0 100 200 300 400 C1 600 1000 2000 4000 6000 Convergence Pressure [psia] Figure 4. the volume and the density of the two phases can be calculated.8: Convergence pressure data .70 Calculation of Phase Equilibria phase.methane for binary hydrocarbon mixtures (from WINN.
and the temperature and the pressure are fixed as T = 49 and p = 15.5499 0.22 Tc.697 2.0318 0.09 58.5876 nvap = 0.56 1101. respectively).15 86.6973 0.010 1. This value does not coincide with the initially assumed value.8 0.1762 1.345 4.12 72.698 23098.0221 0.17 114. For the given temperature of the system (T = 49 o C and 120 o F.050 304.9000 0.0470 568.4 0.07 44. SECOND ITERATION: pk = 20.79 373.5 MPa.4 0.72 MPa.698 On Figure 4.01 *assumption: C 7+ ≈ nC 8 Tp c = ∑ i =k 2 Wi Tc. The composition of the system is tabled below.50 721.0292 469.Calculation of Phase Equilibria 71 Example 4.1700 0.0325 507.01 o = .4 MPa (3000 psia) . i k ∑ i = 2W i 23098.0194 0.45 2054.6300 0.998 1.74 1. the convergence pressure pk can be read off on the envelope curve between methane and nheptane at pk = 3150 psia = 21.7 MPa (4000 psia) nliq = 0.97 305.8 0.7 0.0223 0.96 K = 514.0332 369.0000 42.2250 0.0309 0.= 540.i K Ki xi Wi=xi Mi WiTci C1 C2 C3 nC4 nC5 nC6* C7* Σ 0. o C FIRST ITERATION: pk = 26.598 18541.2 0.0229 425.6500 0. Thus a second iteration must be done.8 T pc = 514.06 F 42.4 Evaluation of the convergence pressure by the iterative NGAA process described above.4300 0.06o F is close to T c of pure nheptane.0720 0.2854 32.4124 Component zi Mi kg/kmole 30.
Example 4.5 Evaluation of the Kfactors of the three components of a system at p = 1.2200 0.162 34.2980 1.5604 nvap = 0.0336 0.027 1.038 44.0299 0.78 739.72 MPa .424 4.39 o = .010 1. i k ∑ i=2Wi 24038.4496 Component C1 C2 C3 nC4 nC5 nC6* C7* Σ Ki 1. The agreement with the assumption is now satisfactory.80 19360. 4. Table II presents the interaction coefficients Kij which must be known.39 *assumption: C 7+ ≈ nC 8 k ∑ i=2WiTc.8800 0.6200 0.9 F 44.45 379.0000 Wi = xi Mi 1.738 2.00 1138.5332 0.38 MPa and T = 338.81 24038. .55 2111. too.0233 0.1800 0.8: p k = 3150psia and thus pk = 21. The successive solution method is used on the basis of the composition of the system.4200 0.7 K. Further refinement is impossible (Kvalues only evaluated for 1000 psia steps in pk).0700 xi 0. the critical and reduced data as well as the acentric factors which are given below in Table I.0336 0.700 0.72 Calculation of Phase Equilibria nliq = 0.42 K = 514.= 541.399 From Fig.0483 0.399 W i T ci 308.
745 0.3: FIRST ITERATION: nliq = 0.797 0.409 Table 4. At first the parameters b and a for each component have to be evaluated through .245 0.2: Interaction Coefficients of the Components C3 C3 n . The gas and the liquid phase compositions are calculated through an iterative calculation procedure (see Example 4.363 3.8 425.1: Composition of the System.001 0.C5 0.035 1. are assumed.325 3.1).000 zi y i = n L ⁄ Ki + n v 0.3. Critical Data and Acentric Factors Component Composition Critical Data Reduced Data Acentric Factor zi C3 n .003 0.182 Then the flash iteration starts with the evaluation of a and b.C4 n .251 4.2 469.000 0.916 0.800 0.52 nvap = 0.191 1.Calculation of Phase Equilibria 73 Table 4.C5 0.152 0.440 0.537 0.61 0.193 0. The assumed Kfactors as well as the evaluated phase compositions are given in Table 4.6 pc MPa Tr pr ωi 0.001 0.28 0.767 0.000 For the first iteration of a flash process.369 0.000 0.11 Tc K 369.001 0.001 n .721 0.C4 n . Table 4.C4 n .000 C3 n .C4 0.C5 0.C5 Σ 1.Kfactors as well as the mole numbers in the phases.374 0.48 Component KFactor zi xi = nv K i + n L 0.003 n .198 0. nliq and nvap.
3.247 E5 9. (see Eq. 3.072 1.003 E5 ai 1.557 By use of Eq.635 E5 7.3.74 Calculation of Phase Equilibria RT c bi = 0. 2 2 αi = 1 + ( 0.54226ω ) ( 1 – T r ) . 3. 3.53 b = ∑ xi bi i=1 k k k and Eq.5 ( 1 – K ij) .C4 n . The results of these calculations are tabled below (see Table 4.C5 bi 5. (see Eq.4).5 ai = ( a c α )i . Then the parameters A and B are calculated after PENGROBINSON using Eq.5 2 0. i = 0. 3. Table 4. pc (see Eq.47).37646 + 1.54226ω – 1.35 . 3.2..56 a = ∑ ∑ i = 1j = 1 xi x j ( ai a j ) 0. 3. a and b for the liquid phase as well as for the vapor phase are calculated on the basis of phase compositions evaluated through the first evaluation (see Table 4. pc (see Eq.729 2.46).48).) The calculation procedure is presented through Example 3.52) and through 0. ac .45724 R T .0778 .4: First Iteration Component C3 n .
Calculation of Phase Equilibria 75 ap A = 2 2 R T and Eq. 3 the formula x + rx + sx + t = 0 is reduced to 3 2 .5: First Iteration a Liquid Phase Vapor Phase 1.2144 B 3.. By substitution of 2 3 r x = y–. Table 4. the CARDAN equation is applied for 3 2 2 2 3 x + rx + sx + t = 0 where 3 2 r = –( 1 – B ) . 3.2699 0. The factors r.975 E2 Now the Zfactors both for the liquid and the vapor phase can be evaluated solving the equation of state given by Eq. b. To solve this cubic equation of state. A and B for both phases are summarized in Table 4.49: Z – ( 1 – B ) Z + ( A – 2B – 3B )Z – ( AB – B – B ) = 0 . s = ( A – 2B – 3B ) and 2 t = –( AB – B – B ) .36 bp B = . 3. RT The evaluated numerical values of the parameters a.232 b 6.551 1.. s and t have to be evaluated both for the liquid and the vapor phase.368 E2 2.5.873 E5 6.072 E5 A 0.
respectively. Thus the cubic equation of state has three real solutions. 27 3 The discriminant D is defined as 3 q 2 D = p + .+ t . 2ς the three solutions for the reduced form y + py + q = 0 are given by y1 = 2 3 ς cos ( ϕ ⁄ 3 ) y2 = 2 3 ς cos ( ϕ ⁄ 3 + 2π ⁄ 3 ) and 3 y3 = 2 3 ς cos ( ϕ ⁄ 3 + 4π ⁄ 3 ) . The solutions are summarized in Table 4.. y2. D < 0 both in case of the liquid phase and in case of the vapor phase. Introducing y1.3. and y3 into r x = y–. 3 the Zfactors(=x) of the phases can be evaluated.– . .76 Calculation of Phase Equilibria y + py + q = 0 where 3s – r p = 3 and 2r rs q = . In defining ς = and 2 p – 27 3 q cos ϕ = – .  2 3 3 2 3 Contrary to Example 3..6.
to evaluate the fugacity coefficients φ3 (of propane).– .414B b a b 2 2B k respectively. φ 4 .2172 0. The results of this calculation procedure are summarized in Table 4. The Zfactors are introduced into Equation 4. meanwhile the actual Zfactor for the vapor phase is equal to the greatest one of the three solutions for the vapor phase. liq K i = φi. .6968 0.15 2 ∑ xi aij bi bi Z + 2 ⋅ 414B A i=1 ln φi = .( Z – 1 ) – ln ( Z – B ) – .6672 As a result of the first iteration flash. Thus the first iteration results in Z l i q = 0.1325 Z3 0.0526 Z2 0.7852 .0523 0. v a p and summarized in Table 4.7852 Obviously.1831 0. 4.6: First Iteration Solution f or the ZFactors after CARDAN Z1 Liquid Phase Vapor Phase 0.ln .16 φ i.8473 φ4 0.7519 φ5 0.7: First Iteration Fugacity Coefficients φ3 Liquid Phase Vapor Phase 1.7.Calculation of Phase Equilibria 77 Table 4. the actual Zfactor of the liquid phase must be identical with the smaller one of the three solutions for the liquid phase.8 .4707 0. Table 4. of nbutane) and φ 5 (of npentane) both for the liquid and the vapor phase..  Z – 0. Kfactors are evaluated through Eq.0523 and Z vap = 0.2580 0.
1 (see Table 4.10.2102 0. Again D < 0 both in case of the liquid phase and in case of the vapor phase.8: First Iteration Components C3 n .0445 1.2744 These Kfactors are the basis of the second iteration which starts again with the iterative calculation procedure presented through Example 4.9) On the basis of the data in Table 4. Correspondingly. 3.C4 n .C5 KFactor 1. the cubic equation of state has three real solutions.C5 Σ KFactor 1. The solutions are summarized in Table 4.9: SECOND ITERATION: nliq= 0.4847 0.0001 zi y i = n L ⁄ K i + nv 0.11.3358 0. A and B for both phases are evaluated again and summarized in Table 4.2742 zi x i = n v Ki + n L 0.4 and in Table 4.6260 0.78 Calculation of Phase Equilibria Table 4.6260 0. b.4847 0.0000 .444 Component C3 n .5020 0.C4 n .1623 1.49 Z – ( 1 – B )Z + ( A – 2B – 3B )Z – ( AB – B – B ) = 0 through the CARDAN equation.556 nvap= 0. Now the Zfactors both for the liquid and the vapor phase can be evaluated solving the equation of state given by Eq.7453 0. 3 2 2 2 3 Table 4.9 the numerical values of the parameters a.
Table 4.1846 0.7161 0. meanwhile the actual Zfactor for the vapor phase is equal to the greatest one of the three solutions for the vapor phase. 4.414B coefficients the fugacity φ4 (of nbutane)and φ 5 (of npentane) both for the liquid and the vapor phase.725 E5 6.12.249 b 6. to evaluate k 2 xa i = 1 i ij b i . the actual Zfactor of the liquid phase must be identical with the smallest one of the three solutions for the liquid phase.0521 and Zvap = 0.0525 Z2 0.12: Second Iteration Fugacity Coefficients φ3 Liquid Phase Vapor Phase 1.7812.490 1.4717 0.1988 0. after CARDAN Z1 Liquid Phase Vapor Phase 0.15 b A ln φi = i ( Z – 1 ) – ln ( Z – B ) – b 2 2B respectively.?????? Z – 0.2520 0.7812 Once more again.– b a ∑ ln Z + 2 ⋅ 414B .2173 B 3. The results of this calculation procedure are summarized in Table 4. The Zfactors are introduced into Eq.1364 Z3 0.11: Second Iteration Solution for the ZFactors.116 E5 A 0. Thus the second iteration results in Zliq = 0.6655 .7511 φ5 0.Calculation of Phase Equilibria 79 Table 4.0521 0.2592 0.997 E2 Table 4.296 E2 2.10: Second Iteration a Liquid Phase Vapor Phase 1.8477 φ4 0.
14: FINAL RESULTS: nliq = 0.1 results in the mole numbers and in the composition of the phases.13 Table 4.2101 0.0000 zi y i = n L ⁄ K i + nv 0.13. Table 4.438 Component KFactor zi x i = n v Ki + n L 0.C5 KFactor 1.0446 0.80 Calculation of Phase Equilibria The Kfactors are evaluated through Eq.C4 n .562 nvap= 0.7452 0.2773 . On the basis of the Kfactors given in Table 4.2773 The Kfactors from the two flash iterations exhibit satisfactory convergence.13: Second Iteration KFactors Components C3 n .C5 Σ 1.6280 0.16 φi.4769 0.1609 1. v a p and given in Table 4.6280 0.C4 n . liq K i = φi.4769 0. 4.9999 C3 n .5046 0. a final iterative calculation procedure analogous to Example 4.3345 0.
3 (VAN DER WAALS. PENGROBINSON).1 Natural Gases 5.1.1 Volume As previously discussed in Chap. no intermolecular forces are acting. In the SIsystem: R = 8. the state of ideal gases is defined by the following equation of state: pV = nRT (5. in case of natural gases . R EDLICHKWONG. In case of real gases. At any isothermal change of state. The value R depends on the system of units. the real gases are more compressible than an ideal one. In case of ideal gases. the decrease in volume is directly proportional to the increase in pressure. It has been observed that in many cases .31432 J / K mole. This behavior of a gas in called a real one.12 pV = ZnRT 81 .the volume becomes less than half if the pressure is doubled.g. the basic relation of Eq.Chapter 5 Phase Properties 5. Below the BOYLE temperature.1) where R is the universal gas constant and n is the number of moles. because of intermolecular forces of attraction and repulsion. real gases behave similarly to perfect gases. 3. If the pressure is low (the forces of attraction can be neglected because of the great distance of molecules) or in case of high temperatures (dominance of the kinetic energy of the molecules). the change in state can be well described by the equations of state which are also presented in Chap. In the petroleum industry.e. 3.
= 0 M a 28.1 shows the Zfactor of methane. pV = ZnRT can be transformed into ZRT pV = M if m = 1[g] gas The reciprocal of the specific volume V is the density and so 1 pM ρg = . 5.5) As expected.= . If Eq.3) (5.4 is combined with Eq. a 0 (5.4) where the subscript “a” indicates the values for air and the superscript “o” those at standard conditions (atmospheric pressure).. Zfactors of different gases. the Zfactor depends both on pressure and temperature.= . the “specific gravity” of the gas γ g is defined by ρg M M γ g = .82 Phase Properties is commonly used because of its easy applicability. The Zfactor is the so called “deviation factor” or “compressibility factor”. In doing so. the gas density results in Ma p 1 p ρg = . V ZRT (5. By introducing m n = M where m: mass M: molecular weight. Figure 5. .003473γ g V R ZT ZT (5.= γ g . vary considerably. 5.= 0. which can be determined experimentally.97 ρg .2) The atmospheric density of the gas is often related to the atmospheric density of air at the same standard temperature.2. It is defined as the ratio of the volume of a real gas to the volume of a perfect one at identical temperature and pressure. ethane and propane as a function of the pressure.
Eq. T ) = V° where the standard conditions po and T o are generally defined in Europe: po: 1.2 0 0 4 8 12 16 20 Pressure [MPa] Figure 5.6.01325 bar) T o: 60.7 are substituted in Eq.8 pV nRT z 0. 5. then: (5.7) (5.3 results for standard conditions in p°V ° = Z ° RT° If Eq. 5.4 0.0 M etha ne e pan Pro ne ha Et 0. 1977) 5.7 psia (1. The formations volume factor for gases. In case of one mole.6 0.2 Formation Volume Factor In petroleum industry. B g. is defined by V B g = B g ( p.1: Zfactor of methane. po and V o) at a pressure p and a temperature T. It expresses which volume is taken up by one cubic meter (at standard state conditions.3 and Eq. 5.0 oF (15 oC).Phase Properties 83 1.15 K and in USA: po: 14.00 bar T o: 273.6) . 5. the term “formation volume factor” is widely used. ethane and propane versus pressure at T = 140oF (from STANDING.1.
cg = – .. .= 0 . (5.C p where p C = . o o Z T o (5..1.4 Correlation of ZFactor on the Basis of Reduced State Variables In general. propane and pentane as a function of reduced pressure for a series of reduced temperatures.e. the isothermal gas compressibility results in 1 1 1 ∂B g .  V ∂p T B g ∂p T p Z ∂p ∂Z For ideal gases: Z = 1.8) 5.– 1 ∂Z .3 Compressibility On the basis of the gas law. .2 shows the excellent correspondence of the Zfactors of methane. Figure 5. ∂p Therefore.1. ∂V = – . i. the components of natural gases are homologous paraffins with similar molecular structures and the theorem of corresponding states can be applied. the gas compressibility is only a function of pressure and is inversely proportional to the pressure.84 Phase Properties ZT B g ( p.T ) = .9) 5.  = .
10) and p pc = i=1 ∑ y ip c i k (5.12) (5. T pc p p pr = pp c (5. pr Figure 5.3 0. 1949) The theorem of corresponding states can also be applied to gas mixtures through the evaluation of “pseudocritical” data.0 1.7 CH4 C3H8 C5H12 5 T=0.. Tpc and ppc.11) where pci: critical pressure of component i Tci: critical temperature of component i yi: mole fraction of component i in the gas phase.2 0.0 T=1.2 r CH4 T=1.9 r 3 T=1.0 2.8 1.4 2.1 0 0 0.i k (5.5 r 0.13) .6 1.8 0. Tpr and ppr.8 2.8 3.2 0.6 0.Phase Properties 85 1.5 0.2 2.0 r CH4 C5H12 C3H8 1 3 CH4 T=1.2 1.2: Zfactor as a function of reduced pressure for a series of reduced temperatures (from SAGE and LACEY.4 0.9 0.3 r C5H12 CH4 C5H12 C3H8 C3H8 pV nRT z 0.1 r 1 CH4 C3H8 C 5H 12 C H 3 8 5 C 5H 12 T=1.6 2.0 Reduced Pressure. which are given by Tp c = i=1 ∑ y i Tc .4 0. can be evaluated: T T p r = .6 0.4 1. Then the “pseudoreduced” data.
are determined.4). These diagrams are only valid for a given quality or only for one area.86 Phase Properties where T: operating temperature p: operating pressure Note that both T and Tpc in Eq.) . several diagrams have been published which correlate the pseudocritical functions of state with the gas gravity. the data of fractional distillation are not available. By means of these diagrams. formation. SI unit) or in R (Rankin. The definition of corresponding states on the basis of the pseudoreduced data results in a generalized plot of Zfactors for natural gases (see Figure 5. Field Unit). These properties have been correlated with the pseudocritical data of this fraction (see Figure 5. 5.5 (see also Example 5. The respective diagram for Oklahoma City gases is given in Figure 5.1). The compositions of these mixtures cannot exactly be defined. either given in K (Kelvin. Several natural gases contain considerable amounts of heavy components which will be combined .2. the C7+ fraction is obtained as liquid residue from a fractional destillation and two properties . etc. On the basis of these correlations.depending on the feasibility of chemical analysis .averaged molecular weight and specific gravity .12 must be absolute temperatures. reservoir. In the laboratory practice. the Zfactor can be determined even if the gas composition is not known. This plot can be used to evaluate the Zfactor of a natural gas on the basis of its composition and the critical data (see Example 5. Anyway.3). basin. In many cases. the engineer is able to calculate pseudoreduced properties of natural gases for which conventional analyses are available.to the pseudo components C 7+ or C9+.
0 1.4 0. 1.9 1 .7 1 . 1948) Compressibility Factor Z .0 0.5 15 1.6 0.6 2.0 2. 1.8 1.5 1.1 Pseudo Reduced Temperature 3 . 2 1.0 1.35 1.3 0.8 1.0 0.45 Compressibility Factor Z 0.2 1.3: Zfactor for natural gases (from BROWN et al.9 2.7 1.Phase Properties 87 Pseudo Reduced Pressure 1.3 0.6 1 .0 1.1 0 1 2 3 4 5 6 7 8 1. 05 1 1.95 0.4 1 .5 0.1 1.9 7 8 9 10 11 12 13 14 0.8 2.4 1.2 1.4 2 .25 1.6 1 ..2 5 1.3 0.7 1.1 1.9 15 Pseudo Reduced Pressure Figure 5.
5: Pseudocritical properties of Oklahoma City Gases (from MATTHEWS et al.400 Pseudo Critical Temp.8 0..1 1.000 Pseudo Critical Press.2 Gas Gravity Figure 5.100 1. [°R] 1. gr. [psia] 500 400 300 200 100 120 140 160 180 200 220 240 Molecular Weight Figure 5.7 0. 1942) 700 Pseudo Critical Temperature [°R] Pseudo Critical Pressure [ib/sq in.9 1.200 1.6 0.] absolute 675 650 625 600 575 550 525 500 475 450 425 400 375 350 325 0.5 0. 1. 1942) .300 Sp.0 1.4: Pseudocritical temperatures and pressures for heptanes and heavier (from MATTHEWS et al.88 Phase Properties 1.
1 44.355 305.37 0. The Zfactor is determined by using Figure 5.1 58.1 58.2 461.582 0.98 4.166 0.000 0.9443 15.1 72.1 425.8 408.= .8 4.55 MPa.93 p p c 4.65 0.296 0.2 114.= 1.08 2.6 179.63 22.7 507.5 p pr = p.0023 0.0051 0.0158 0.284 0.562 From Figure 5.109 190.077 4.89 202. Component C1 C2 C3 iC4 nC4 iC5 nC5 C6 C * 7+ Mi yi yiMi Tc.152 0.87 T pc 202.0 0.63 4.48 0.4 568.2 72.3 Z = 0.982 18.67 4.83 4.0021 0.056 3.99 1.60 4. .23 0.021 4.562 Σ assumption: C 7+ ≈ C 8 T 105 + 273.0033 0.2 86.0086 1.01 0.25 0.4 469.31432[J°K 1mol1].= .15 T p r = .019 3.476 0.008 3. the critical data as well as the calculated reduced data of the components.79 0.007 3.0132 0.308 0.344 30.88 2.= 4.925 which is introduced in p∑ yM i = 1 i i ρ g = ZRT k where R = 8.1 Evaluation of the density of a natural gas at T = 105 o C and at p = 22.Phase Properties 89 Example 5.38 0.0053 0.010 2.4 369.88 0.i yi x Tci pci yi x pci kg/kmole kg/kmole K K MPa MPa 16.25 1.020 3.3 The table below summarizes the composition of the gas.06 0.
Because Ma p ρg = .65 so that . pp c 670 On the basis of the pseudoreduced data.2 Evaluation of the density and the formation volume factor of an Oklahoma City Gas of the specific gravity γ g = 0.31432J/K mole.90 Phase Properties 3 225.32kg ⁄ m 3 Example 5.86 results from Figure 5.31432 × 378.kg ⁄ m 0. Z = 0.31432 × 0.15 ρg = 142.7018 at temperature T = 90.355 E .= 1813 = 2.3. Now the pseudo reduced data are calculated as T Tp r = . the density can be calculated through 3 0.5 oC(655 R)and pressure p = 12.5 as Tp c = 390R and ppc = 670psia .925 × 8. and R = 8.97 E3 kg/mole.kg ⁄ m 8.5 MPa(1813 psia).= 655 = 1.5 E + 5 × 18. R ZT where Ma = 28.02897 × 0.3 ρg = .71 .86 × 363. At first the pseudo critical data are read off from Figure 5..7018 × 125 E + 5 ρg = .68 T pc 390 and ppr = p.γ g .
1 B g = . The formation volume factor can be evaluated through Eq.. .15 B g = 0.15 K so that Z = 1.00916 . o o Z T Assuming that the gas is a perfect one at European standard conditions o p = 1bar and T o o = 273.Phase Properties 91 ρ g = 97.0..8 ZT B g ( p.5 + 273.C p where p C = .86 × (90.T ) = . 12.15) 0.5 273. its formation volume factor results in 0. 5.774 kg/m3..
92
Phase Properties
ps ia
5
30
2 10
4
50
Saturation 100% HO, 47,489 Ib per mm cu ft 2
Water content Ib H2O per million cu ft total gas
5
2 103 5
2 10
2
5
2 101 5
2 100
70 50 30 10 10 30 50 70 90 110 130 150 170 190 210 230 250 270 290 320 340 360 380 400 440 480 520 560 600 640 680 720
Temperature, °F
Figure 5.6: Water content of natural gas in equilibrium with liquid water (from KATZ et al., 1959)
80 0 10 00 12 50 15 00 20 00 25 00 30 00
14 .7
10 0
20 0
40 0 50 0 60 0
3500 4000 5000 6000 8000 10,000
Phase Properties
93
5.1.5
Water Content
The conventional gas analysis does not indicate any water content. However, the natural gas in a reservoir is in thermodynamic equilibrium with the interstitial water. Therefore the gas must be assumed to be saturated with steam. The mole fraction of the steam, yw, can be calculated by the DALTON law (see Eq. 5.7). pw yw = p where pw: vapor pressure of water at operating temperature p: operating pressure of the system. From numerous measurements it could be deduced that the steam content can be well correlated with pressure and temperature (see Figure 5.6 and Example 5.3). On the other hand, the influence of the gas composition has been found to be insignificant. Natural gas and liquid water will combine to form solids resembling wet snow at temperatures somewhat above the crystallization temperature of pure water. These solids are called “gas hydrates”. Because these solids can form at temperatures and pressures normally encountered in naturalgas pipelines, this phenomenon is of particular interest to the petroleum industry. The conditions for the gas hydrate formation will be summarized later on. (5.14)
Example 5.3 The evaluation of the water content of a natural gas as a function of operating pressure and temperature. This can be done by correlation through Figure 5.6 FIRST STATE: p = 13.8 MPa(2000 psia) T = 93.3 oC (200
o F)
The water content at this state is evaluated through Figure 5.6 as
3.8 E – 4lb ⁄ cu ft = 6.1 E – 3kg ⁄ m .
SECOND STATE: p = 3.45 MPa (500 psia) T = 60
oC
3
(140
o F)
The water content at this state was evaluated through Figure 5.6 as
3.25 E – 4 lb ⁄ cu ft= 5.2 E – 3kg ⁄ m .
THIRD STATE: p = 2.8 MPa (400 psia)
3
94
Phase Properties
T = 60
o
C (140
o
F)
The water content at this state was evaluated through Figure 5.6 as
3.8 E – 4 lb ⁄ cu ft = 6.1 E – 3 lb ⁄ cu ft
5.1.6
Viscosity
Like all intensive physical properties, the viscosity is completely described by µ = F ( p, T, y 1 ,y 2 ,y3 …y k – 1 ) in case of natural gases and µ = F ( p, T, x 1 ,x 2 ,x3 …x k – 1 ) in case of hydrocarbon liquids. Eq. 5.15 and Eq. 5.16 simply state that the viscosity is a function of pressure, temperature and composition. LOHRENZ et.al. (1964) presented respective calculations which are useful for • determining viscosities of reservoir fluids in compositional material balance computations and • predicting the decrease in viscosity which occurs when gases dissolve in reservoir oils. The correlations presented below may be viewed as modifications of Eq. 5.15. The assumptions are practical because the composition is frequently not known. Anyway, the assumptions are sufficiently valid so that these correlations are frequently used for reservoir engineering computations. The gas viscosity µg increases with the temperature. Figure 5.7 shows the dependence of the viscosity of pure gases on the temperature at atmospheric pressure (0.1 MPa and 14.7 psi, respectively). Figure 5.8 presents the viscosity of natural gases at atmospheric pressure as a function of the averaged molecular weight. This diagram elucidates that the viscosity decreases along with increasing molecular weight. It also contains correction functions for CO 2, N2 and H2S. The theorem of corresponding states has been proved to be valid once more again. The viscosity at any state of the gaseous system can be predicted through the reduced viscosity which is defined by (5.16) (5.15)
0. 1 ( p = 1.014 0.T) µg. and the pseudocritical temperature.020 0.010 0. the reduced viscosity µg. r µ g.1 has to be evaluated for the given gravity (or averaged molecular weight) of the gas and the operating temperature. r ( p pr . T p r ) = . 1954) . the atmospheric viscosity µg.Phase Properties 95 µg (p.016 0.9.8.022 0.012 0.7: Viscosity of paraffin hydrocarbon gases natural gases at atmospheric pressure (from C ARR et al.018 lium He r Ai 0.4): (5. This is a satisfactory solution for practical purposes. • According toFigure 4.024 0.008 0.006 0.. are determined. is evaluated by use of Figure 5. • After the calculation of the pseudoreduced state quantities ppr and Tpr .5. T pc.17 where µg = µg. which is given by the ratio between the actual viscosity and the one at atmospheric pressure. µg. the pseudocritical pressure. • The actual gas viscosity can be calculated through Eq. T ) The procedure to evaluate µ at certain p and T is the following (see Example 5.17) • From Figure 5. ppc. 1 The gas viscosity determined through this method deviates from laboratory measurements by 3% at the maximum.4 and Figure 5. 5.r .004 50 n ge itro N ide iox nD e rbo l fid Ca Su n ge dro Hy ne tha Me tane iBu ne a Eth e pan Pro nane tane nNo nBu ane nOct ane nDec Viscosity [cP] 100 150 200 250 300 350 400 Temperature [°F] Figure 5..
3 0 1.5 2.9 0.0 0.5 2.0 1.0015 . Viscosity at 1 [atm] µ1 [cP] 0.9: Correlation of viscosity ratio with pseudoreduced pressure and temperature (from C ARR et al.0 CO2 Correction odded to Vis cosity [cP] 0.016 Correction odded to Vis cosity [cP] 0. 1954) 6.05 20 Pseudo Reduced Pressure.4 0.0 2 3 4 6 8 10 2.0 0.96 Phase Properties Gas Gravity (Air = 1.5 N2 γg = 2.5 4.5 2.5 1 .0010 0 0 5 10 15 0 0 5 10 15 Mole % N2 Mole % CO2 0. 1.8: Viscosity of natural gases at atmospheric pressure (from CARR et al.5 0.0020 3.010 0. 0 0.004 0 H2S 0.0 =2 γg 5 1.0 0 0 5 10 15 Mole % HS 2 10 20 30 40 50 60 70 80 90 Molecular Weight Figure 5.005 0. 5 3.2 0 5 00 3.0015 0.6 1.0 1 .0 5. 1 .0010 0. 1 . 1.008 0.0020 6 0.20 duc ed 1. 8 Temp erat u re.014 0 1.006 0.5 0.007 0.00 T r Pseudo R e Viscosity Ratio.012 0.013 0.75 0 0 1.5 5.0 3. 1.10 1.0 1.5 0..0015 γ g= 2.6 0.0010 0.015 0.6 1 .5 40 2.0 2.2 0.0 2.5 3.011 0. pr Figure 5.000) 0.009 Correction odded to Viscosity [cP] 0 .0020 0.1 0.0 4. 1954) .15 1. 0 1.3 0. µ/µ 1 1.
9 F. 654.6 = 1. the viscosity at atmospheric operating temperature can be read off as pressure and µg.70 . The calculation of the pseudoreduced data of the system result in p 1800 p pr = . . From Figure 5. r × µg .8.5 o o C (194. µg.0165 mPa s .0122cP . The specific gas gravity was found as γ g = 0.0165cP and µg = 0.5 MPa (1800 psia). 1 = 1.35 × 0.= . 1 = 0. respectively. can now be read off from Figure 5.. Then the pseudocritical data of the system are evaluated from Figure 5.= 654.35 so that µg = µg.7018 .5 where ppc = 670 psia and Tpc = 385oR. and its viscosity at atmospheric pressure. µ g.7 R) and at pressure p = 12.0122 = 0.9: µg.69 pp c 670 and T .= 2.1. r = 1.Phase Properties 97 Example 5. T pc 385 The ratio between the actual gas viscosity.4 Determination of the gas viscosity at temperature T = 90.
Then the respective “pseudoliquid” data of any hydrocarbon liquid at standard conditions may be evaluated from the real one at standard conditions and from the apparent data of the liberated light components. The interesting properties (density. Based on these data and assuming that propane and heavier components follow the rule of additive volumes. a crude oil does not remain in the singlephase state at “normal” conditions if it contains any high mole fraction of light components which will be liberated at atmospheric pressure and surface temperature.5. The theorem of corresponding states can be applied again to determine the properties at any other state of the system (e. the resulting pseudoliquid density of the system containing methane and ethane has to be corrected with regard to the operating temperature and pressure. The correlation method after STANDING and KATZ is satisfactory for stocktank liquids which have low concentrations of methane and ethane.g.11 presenting the correlation of the density of a system containing methane and ethane with the density of the propanesplus fraction. etc. This calculation procedure is presented through Example 5.). the weight % ethane in the ethanesplus and the weight% methane in the system. It is based on Figure 5. which depend on the oil composition. Figure 5.2.13 serve this purpose. can be easily determined at the standard (normal) state (atmospheric pressure and T = 60oF (15 oC)). they developed a method for computing the density of hydrocarbon mixtures.98 Phase Properties 5.10).2 Hydrocarbon Liquids 5.1 Volume For practical purposes.12 and Figure 5. STANDING and KATZ (1942) examined the properties of many crude oils. This inconvenience can be avoided by attributing “apparent densities” or “apparent specific volumes” at standard conditions to the dissolved and easily volatilized components. it is sufficient to consider hydrocarbon liquids as ideal mixtures whose volumes can be calculated by adding up the volumes of the components. at reservoir conditions). Then. . They established that the apparent densities of the two light components methane and ethane depend on the density of the system (see Figure 5. However. viscosity.
9 Density of System [g/cc].6 0.5 0.7 0.3 0.Crude Oil 0. 60°F & atm.Benzene Methane .Cyclo Hexane Methane .Heptane Ethane .Propane Methane .Heptane Methane .6 0.Phase Properties 99 0. Pressure Figure 5. 1942) .2 Ethane .10: Variation of apparent density of methane and ethane with density of the system (from STANDING and KATZ.Pentane Methane .1 0.5 Apparrent Density of Methane [g/cc] 0.3 0.4 0.4 0.Crystal Oil Methane .3 0.NButane Ethane .Crystal Oil Methane .Hexane Methane .4 0.8 0.
100 Phase Properties Density of Propane plus [ib/cu ft] 70 50 40 30 20 10 0 50 70 0 60 10 50 40 30 W in t % en M t ir et e h Sy an st e em 20 40 30 30 20 10 Figure 5.11: Pseudoliquid density of systems containing methane and ethane (from STANDING. 1952) Density of System including Methane and Ethane [ib/cu ft] 60 Wt Eth % Etha ane n plus e in Mat er ia l .
1952) . [ib/cu ft] 10 9 8 0 00 00 10 80 0 00 15 7 6 5 4 3 00 00 00 000 60 50 40 3 00 20 10 0 0 re su es Pr a si [p ] 2 1 0 25 30 35 40 45 50 55 60 65 Density at 60 [°F] and 14.7 [psia].7 [psia].12: Density correction for compressibility of liquids (from STANDING.Phase Properties 101 Density at Pressure minus Density at 60 [°F] and 14. [ib/cu ft] Figure 5.
102 Phase Properties 10 Density at 60 [°F] minus Density at Temperature.55MPa ) and To b = 220 F ( 105°C ) . The composition of the system is tabled below.5 Determination of the volume of 1 mole oil at bubble point through STANDINGKATZ correlations.13: Density correction for thermal expansion of liquids (from STANDING. o . 1952) Example 5. [ib/cu ft] 9 8 7 6 5 4 3 24 0 22 0 20 0 18 0 160 140 12 0 Tem per atur e [°F] 2 1 0 25 100 60 30 35 40 45 50 55 60 65 Density at 60 [°F] and Pressure p. The bubble point is defined by pob = 3300psia ( 22. [ib/cu ft] Figure 5.
= k ∑ i = 3V i k 122.× 100 = 1.046 1.× 100 = .× 100 = .2 287.021 = 876kg ⁄ m 3 = 54.1 MPa.0290 0.× 100 = 5.0 626. the pseudoliquid density of the hydrocarbon mixture at standard conditions is read off as 51.2 86.4404 0.0020 0.0284 0.5 lb/cu ft.1 58.321 z iM i i = 2 and the weight % methane in the entire system z1M 1 7.0174 0. 15o C The density of the propaneplus can be calculated through ∑ i = 3 zi M i .0 662.05 [% ] k 123.0 0.256 2.367 507.500 114. From Figure 5. From Figure 5.69 lb ⁄ cuft 0.0038 0.0035 0.4001 1.12.829 130.5 907.1393 *) at standard conditions: 0.1393 Then the weight % ethane in the ethaneplus material z 2 M2 1.Phase Properties 103 Composition Mole Fraction zi = xi Mole Weight Mi kg/kmole 16.4% k 130.300 .1 72.0 Density ziMi ρi kg/m3 Volume Vi * m3 C1 C2 C3 C4 C5 C6 C7+ Σ 0.0 30.11.0432 0.5 575.1 44.1271 0.0405 0.046 .0029 0. the pressure correction for pob (3300 psia) is read off as .650 1.786 1.0000 7.300 1.367 ∑ zi M i i = 1 are calculated.
1 1. 3 Apparrent Density in Pounds per Cubic Foot of dissolved Gas at 60 [°F] and 1 Atmosphere 45 ity rav eG ud .2 1.14: Apparent liquid density of natural gases in various API gravity oils (from KATZ.I A 20° 30° 40 35 40° 30 50° 60° 25 20 15 0.4 – 3.5 + 0.0 1.7 0.4 lb/cu ft.9 1.0 Figure 5.8 0.7 lb ⁄ cu ft and ρob = 778. Then the temperature correction for 220o F is read off from Figure 6.4 Gas Gravity .9 = 52.9 lb/cu ft so that the liquid density at 3300 psia and 60 oF results in 51. 1952) .3 1.7 = 48. Cr . respectively.13 as 3.6 0.104 Phase Properties 0.Air = 1.7 lb/cu ft so that the oil density at 3300 psia and 220oF results in ρob = 52.P.5kg ⁄ m .
natural gas systems. In terms of the production. The apparent liquid density of the respective natural gas can be determined according to Figure 5.14.oil ratio). . Therefore. • The difference between reservoir pressure and atmospheric pressure causes slight expansion of the remaining oil phase.2.Phase Properties 105 Natural gases (dissolved gas) predominantly consist of methane. The easily volatilized component is marked by its gravity. This ratio is designated as Rs (solution ratio). but also on the oil quality. Rs depends on pressure and temperature. This change in volume has three main sources: • The most important factor is the gas liberation below bubble point pressure pob. and the scarcely volatilized fraction is the crude oil with the density ρ o and its gravity γ o (in °API) at standard conditions. Therefore. • The increase in volume because of expansion is more or less compensated by the shrinkage of the oil resulting from the transfer from reservoir to surface temperature. The amount of the dissolved gas is always indicated as the number of normal cubic meter per actual liquid volume at standard conditions.2 Formation Volume Factor The oil volume inside the tank at surface conditions differs from the volume of the oil leaving the reservoir. the ratio is also often designated as GOR (producing gas . 5. KATZ (1942) has applied the method of the apparent liquid volume also to crude oil . γ g . nm3/m3. In other words: Rs expresses how much gas will be dissolved in one cubic meter oil (volume at standard conditions) at the operating conditions.
As soon as the reservoir pressure falls below pob.15 illustrates typical characteristics the relationship between Bo and p. This pressure is the bubble point pressure of the . Bo can be defined as the volume of one cubic meter tank oil (oil at standard conditions) at certain operating pressure and temperature. On the other side. Bo. This gas liberation results in a shrinkage of the oil volume.0 0 pb Reservoir Pressure Figure 5. The reciprocal value of Bo is called “shrinkage factor”. the volume of the oil inside the reservoir is obtained. is defined as oil volume and volume of dissolved gas at reservoir conditions B o = oil volume at standard conditions Correspondingly. Figure 5. change in Bo will be ruled by the expansion of the oil.106 Phase Properties Bo.b 1. tank oil gravity and reservoir temperature to determine the pressure at which the given amount of gas would be in solution. gas will be liberated.15: Typical graph of formationvolume factor of oil against pressure The formation volume factor of the oil phase. the multiplication of the shrinkage factor with the volume inside the reservoir will result in the stock oil volume. Bo decreases at decreasing pressure as soon as pob has been crossed. On the basis of many investigations. which must be taken into account. STANDING (1947) amplified the correlation to permit the calculation of bubble point pressures and volume factors. gas gravity. Thus. it is possible to take the solution ratio. By multiplying the tank oil volume with the actual Bo. The respective STANDING charts are appended at the end of this textbook: • Working with Chart 1. As long as the pressure does not fall below the bubble point pressure pob during a pressure decrease from initial reservoir conditions.
Bo can be evaluated • by use of the KATZ correlation (see Figure 5. The calculation is based on one cubic meter tank oil. The thermodynamic data of the system were given in Example 5.12 and Figure 5.6 Evaluation of the oil volume factor at bubble point conditions from operation data tabled below both in SIunits and Fieldunits.Phase Properties 107 respective hydrocarbon system.6) yield values which are comparable within 3 or 4% to measured laboratory data.5. Bt. Bt is defined by B t = B o + ( R si – R s )B g where Bo: volume factor of the actual oil phase Bg: volume factor of the actual gas phase Rs. pob. can be estimated by use of Chart 3. • Charts 2 permits the evaluation of Bo at bubble point conditions if GOR represents the solution ratio.18) Example 5. these empirical correlations (see Example 6.i: solution ratio at initial conditions Rs: solution ratio at operating conditions. .14) and the STANDING diagrams (see Figure 5. (5.13) • or by use of the Charts from STANDING (see appendix). In most cases. • The total formation volume factor.
108 Phase Properties GOR nm 3/m3] Separator 1 Separator 2 Tank Σ 74 16 4 94 GOR cu ft/bbl 415.2 kg/m . 640 + 16 × 0. 3 .5 527.5 89. respectively. the apparent liquid density of the gas can be evaluated through Figure 5.640 0.8 22.14 as ρg . .897 1. 540 = 67. 897 + 4 × 1.1 kg Because the specific gravities are given as γg = 0.2 30 In summing up the producing gas/oil ratios GOR.722 = 83.74 + 16 + 4 94 Now the amount of gas can be calculated through m g = R s i ςa γ g where ρ a is the air density: mg = 94 x 1.3 lb/cu ft and ρg .722.540 876. the initial solution ratio Rsi results in R s i = 94nm ⁄ m and 3 3 R s i = 527.722 and γ o = 30o API. a = 389. The computation of the average specific gas gravity results in γ g = 74 × 0. a = 24.225 x 0.8 γg ρo kg/m 3 γo o API 0.8 cu ft ⁄ bbl respectively.9 = 0.
in SIunits. In other words: 1. a and so 3 83.7 and so B o b = 1.2 In summarizing 1 m3 tank oil and 0.3 lb/ cu ft .2 kg tank oil and 83.3 3 V o b = . From Figure 5.9 = 50. respectively.8 = 46. 389.= 0.13.214 m .1kg liberated gas. a p p = ρ g. a p p = . From Figure 5.3 + 0.7kg ⁄ m .214 and so ρ o.2kg ⁄ m 1000 + 0.1 ρ o.214 m3 apparent liquid volume of the liberated gas. app = 49.29m 743. app = . the pressure correction of this density for p = 3300 psia results in ρ o = 49.the pseudoliquid density of the given hydrocarbon system at standard conditions results in 3 876. On the basis of the calculated oil density under the conditions of the reservoir.4 lb/ cu ft or. in ρ o = 743.Phase Properties 109 Now the apparent liquid volume of the liberated gas at standard conditions (T =15 o C and atmospheric pressure) can be calculated through mg V g. as well as 876.= 790.= 1.2 –3.1 V g.29 .288 m3 saturated reservoir liquid under reservoir conditions shrinks to 1 m3 tank oil.2 + 83. The shrinkage results in . the final temperature correction for T = 220 o F results in ρ o = 50. the volume of one cubic meter oil under reservoir conditions results in 3 959.2 lb/ cu ft .12.
Therefore.21) In general.× 100 = 22.9) so that ∂B 1 co = – . p1) and (V2. the bubble point pressure. 5.29 – 1.22) . (V1. 5.23) (5. pb.110 Phase Properties 1. Its isothermal compressibility is to be calculated in the same way as for the gases (see Eq. Similar formulae are also valid for water. 5. the density of an undersaturated oil phase can be calculated through ρo = ρ o be c o (p – p b ) – c o( p – p b) (5. b [ 1 + c o ( p – p b ) ] .20 and Eq.   1. the terms of higher order can be disregarded so that B o ~B ob [ 1 – c o ( p – p b ) ] and ρo = ρ o. p2) by (5.000 .48% 1.29 5. Eq.21 will develop in a Taylor series.19) .3 Compressibility of Undersaturated Liquids An undersaturated oil is a compressed liquid in the pressure range above the bubble point pressure pb. The compressibility of undersaturated liquids is constant within wide pressure intervals. Because co is very small. o V ∂p T B o ∂p T and after integration B o = B ob e Because C ρo = Bo where C is a constant. ∂V = . (5.20) (5.2. it can be calculated between any two states of the system. is chosen as reference.
3 .Phase Properties 111 V1 – V2 co = – V (p 1 – p2 ) where V is averaged volume between p1 and p2.17 on the basis of • the calculation of the pseudoliquid density of the hydrocarbon mixture at standard conditions (see procedure of Example 5. the pseudoreduced pressure and temperature can be approximately evaluated through Figure 5. Then. ppr.16). Tpr (see Figure 5.12. • the pressure correction of this pseudoliquid density from Figure 5. Then the oil compressibility can be calculated through co = c r ⁄ p pc (5. (5. and pseudoreduced temperature.25) If the composition of the undersaturated system is not known. cr as a function the pseudoreduced pressure. the evaluated oil density is related to the density of water at the standard temperature. This method is based on the determination of the pseudocritical data of the system and on the evaluation of the pseudoreduced compressibility of the undersaturated liquid.24) TRUBE (1957) has reported a method of calculating the compressibility of undersaturated liquids.6). ρw = 1000kg ⁄ m .
74 0.68 0.78 0.86 0.70 0.76 0.62 0.80 0. 1957) .66 0.40 10 4 1 2 3 5 7 10 20 30 50 70 100 pr Figure 5.82 0.88 Pseudo Critica l Press u re Specific Gravity of Undersaturated Reservoir Liquid at Reservoir Pressure.10 1.84 0.95 0.16: Pseudoreduced compressibility for undersaturated reservoir fluids (from TRUBE.70 0.17: Pseudocritical conditions of undersaturated reservoir liquids (from TRUBE.90 103 0.50 0.00 0.112 Phase Properties 101 Pseudo Reduced Temperature 102 cr 1.60 0. 1957) 1300 Pseudo Critical Temperature [°R] 1200 1100 1000 900 800 700 Pseud re ratu mp e Te i cal o Crit Pseudo Critical Pressure [psia] 600 500 400 300 200 0. Corrected to 60 [°F] Figure 5.80 0.64 0.72 0.
00292(25 .16. Boi can be calculated through Eq.5)] = 1.84 pp c 3. .02 .22.= .281. b = 1.5 and Bob was calculated through Example 5. The computation of the pseudoreduced temperature and pressure results in T T p r = .= .= 6.84 Because B o. p pr = . the reduced compressibility can be read off as cr = 0.0.02 –1 co = .70 and p 25 .6).7 Evaluation of Boi for an under saturated oil at reservoir temperature T = 105 o C and under the initial pressure pi = 25 MPa through the reduced compressibility from TRUBE.= 2.15 = 1.= 105 + 273. The critical data are tabled below. The compressibility calculated through of the undersaturated oil phase can be cr 0.29[1 .01 T pc 374.6.92 E – 3MPa p pc 6. 5. The composition of the system was given in Example 5.23 B o i = B ob [ 1 – co ( p i – p b ) ] and results in Boi = 1.655 From Figure 5.Phase Properties 113 Example 5.29 (see Example 5.
18 demonstrates that with increasing pressure.12 Gas Gravity bb 16 4° F lG OR 26. Example 5.8 elucidates the respective correlation procedure. ∆p = pi pb. North Texas Frio Sand. Figure 5.2. . 1941. above the bubble point pressure. after BEAL. The dependence on pressure is more significant for high viscous oils than for low viscous oils.84 D 0 500 1000 1500 2000 2500 3000 3500 Reservoir Pressure [psia] Figure 5. Gulf Coast Palo Pinto Lime.4 Viscosity 9 8 7 A B C D Permian. West Texas Frio Sand.7°API Oil Gravity 1. HOCOTT and BUCKLEY (1941) presented measurements of the viscosity of crude oils containing dissolved gases. 1946) In general.6 5 Bubble Point 38 5ft 3 / B C Bubble Point 1 58°F 8 0 3 0ft / bbl GOR 45 . Anyway. the viscosity decreases up to the bubble point because of the increasing fraction of light components. the solution ratio Rsi and the undersaturated pressure in the reservoir. The viscosity of the crude oil can be determined with sufficient accurancy on the basis of the known viscosity of the tank oil µ1 at the reservoir temperature T and atmospherical pressure. the viscosity increases with increasing pressure.114 Phase Properties 5.1 ° AP I G=0 Bubble Point . Southwest Texas Absolute Viscosity [cP] Bubble Point A 6 5 4 3 2 1 0 Legend 100°F Reservoir Temperature 3 191 ft/bbl Gas in Solution at Bubble Point Pressure 29.7°API G=0.18: Viscosity of subsurface samples of crude oil (from HOCOTT and BUCKLEY. liquids decrease in viscosity with increasing temperature and increase in viscosity with increasing pressure.
3 0. 0 30 Viscosity of GasSaturated Oil [cP] (at Reservoir Temperature and Saturation Pressure) 30 20 0 20 10 7 5 3 2 0 60 0 80 00 10 00 120 0 14 0 40 Solution Gas/Oil Ratio [ft3/bbl] 1.3 0.0 0.Phase Properties 115 100 70 50 Example: Problem: Find the gassaturated viscosity of a crude oil having a solution gas/oil ratio of 3 600 ft/bbl and dead oil viscosity of 1.7 0.58 on the gassaturated oil viscosity scale (ordinate). 1959) 0 50 0 10 .2 0 160 Example 0.5 0. Procedure: Locate 1. Then go left horizontally to read the answer 0.1 0.5 0.0 2 3 5 7 10 20 30 50 70 100 Viscosity of Dead Oil [cP] (at Reservoir Temperature and Atmospheric Pressure) Figure 5.19: Viscosity of gassaturated reservoir crude oils at reservoir conditions (from CHEW and CONNALLY.5 cP on the dead oil viscosity scale (abscissa) and go up vertically to the 600 gas/oil ratio line.50 cP all at the same temperature.7 1.
50 45 P] re [c s su Pre 12 11 10 9 8 7 6 5 4 3 2 1 0 9 0.4 0.20: Prediction of crude oil viscosity above bubble point pressure (from BEAL. 0.8 Viscosity [cP] e o lut A bs 30 0.116 Phase Properties Absolute Viscosity [cP] Crude Oil Prediction above Bubble Point Pressure Average Deviation 2. 1946) .7 0.5 0.1 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 1 2 3 4 5 Undersaturated Pressure [psi] (Pressure above Bubble Point less Pressure at Bubble Point) Pressure 10 [psi] 3 Figure 5.6 25 20 0 .3 0.7% Extrapolated 100 8 75 80 5 6 5 70 9 8 7 6 5 4 3 2 1 0 0 Detail 7 6 5 4 3 2 1 90 Absolute Viscosity of Crude Oil above Bubble Point Pressure [cP] 80 70 60 50 40 30 20 10 0 0 1000 2000 3000 4000 5000 Pressure [psi] 55 60 13 nt Poi bble t Bu de a 40 Cr u t ed tur a 5 of Gas Sa 3 ity os Vis c Detail 0 1.2 15 10 5 0.
Phase Properties
117
Example 6.8 Evaluation of the oil viscosity at reservoir conditions which are given below as well as the interesting production data. µob = 0.45cP .
SIUnits Reservoir Temperature Bubble Point Pressure Reservoir Pressure Initial Solution Ratio Standard Oil Viscosity T pb p Rsi µ1 105 oC 22.55 MPa 25.00 MPa 94 nm3/m3] 0.9 mPas Field Units 220
o
F
3300 psia 3600 psia 527.8 cu ft/bbl 0.9 cP
From Figure 5.19, the oil viscosity at bubble point conditions can be read off as µob = 0.45cP From Figure 5.20, the viscosity of the under saturated crude oil at p = 3300 psia can be evaluated: µoi = 0.48cP or µoi = 0.48 [ mPas ]
5.3
Brines
It has been recognized relatively late that water is  without any exception  one of the reservoir fluids. GARRISON (1935) and SCHILTHUIS (1938) have been the first who presented detailed information on the distribution of oil and water in porous rock and on the origin and occurrence of “connate water” (formation water, capillary water, interstitial water). Furthermore, they established a relationship between the water saturation and the rock permeability. Nowadays, knowledge of the chemical and physiochemical data of the respective reservoir water is of great importance for the petroleum engineer. Water analyses • give valuable information about the area of prospecting, • help to answer questions about the origin of the produced water,
118
Phase Properties
• make sure that injected water is chemically compatible with the reservoir water or with added water (plugging by precipitations must be avoided), • are indispensable for feasibility studies in case of chemical EOR processes. Last but not least, data such as density, compressibility, formation volume factor and viscosity are required for calculating the displacement process and for establishing a material balance. Additionally, this section presents an unusual phenomenon, the socalled “gas hydrate formation”, during which water and natural gas form a solid phase at temperatures above the freezing point of water.
5.3.1
Composition of Brines
Reservoir water always contains dissolved inorganic salts, above all sodium chloride, NaCl. Therefore, these waters are sometimes called “brines” although they cannot be compared with sea water because both the total salt concentration and the content of distinctions are different. The cations usually dissolved in reservoir brine are Na+, K +, Ca2+, Mg 2+ In certain cases, Ba2+, Li +, Fe2+, and sometimes Sn2+ can be found. The generally present anions are
2Cl , SO , HCO 4 3
and, in some cases, also CO 2 , NO  ,B r , J , S .
23 3
The elements are present in reservoir water within the following concentration ranges: Na, Cl Ca, SO 4 K, Sr in% in% or ppm in 100 ppm
Phase Properties
119
Al, B, Ba, Fe, Li Cr, Cu, Mn, Ni, Sn, Ti, Zr Be, Co, Ga, Ge, Pb, V, W, Zn where %: ppm: ppb:
in 1  100 ppm in ppb (in most cases) in ppb (in some cases)
g substance per 100 g brine mg substance per 1 kg brine mg substance per 103 kg brine
Usually microorganisms of different species are also present in oil field brines. The origin of these organisms is not yet totally clear. They contribute to corrosion in the bore hole and to permeability reduction during water flooding. There is a wide variability of concentrations and chemical characteristics resulting from many factors. Such factors may be • the nonmarine origin of a few sediments, • dilution with ground water, • enrichment because of vaporization by migrating gas, • adsorption and cation exchange with clay minerals, • dissolution of mineral salts in the migrating formation water, • the chemical reaction with components of adjacent sediments. Therefore, it is important that water from a certain horizont is characteristically distinct from all the other brines, even those from the immediate neighbourhood. Thus, water analyses are mainly used for identifying the produced water with regard to the geological formation it originates from. In doing so, a simple and easily readable representation of the great data set from a water sample is desired. The graphical method proposed by STIFF (1951) may be the simplest graphical method, has maximum utility and, hence, is the most popular. This graphical method is shown in Figure 5.21. Horizontal lines extend right and left from a centrally located vertical line. The positive ions are plotted on the the left of the vertical line, while the anions are plotted on the right. Characteristic positions are designated for the ions which are most frequently found in reservoir brines.
Barton Ellsworth Stoltenberg Field Stafford Rice Bloomer Field Na 100 Ca 10 Mg 10 Fe 10 Cl 100 HCO3 10 SO4 10 CO3 10 St. a closed “pattern” is formed as a “fingerprint” of the respective brine. If there is a difference between Σ meq of the positive and Σ meq of the negative ions. The characteristic pattern of the formations is evident.22. Although various scales can be employed. the difference is presented as sodium.21: Essential feature of the water pattern analysis system (from S TIFF. most reservoir waters may be plotted with sodium and chloride on a scale of 100 meq and a scale of 10 meq for the others. 1951) . When the data of the water analysis are plotted on the graph and the adjacent points are connected by straight lines.22: Course of Arbuckle formation through Kansas shown by water patterns (from STIFF. One of many practical applications in correlating producing formations is illustrated by Figure 5.120 Phase Properties Na 100 Ca 10 Mg 10 Fe 10 Cl 100 HCO3 10 SO4 10 CO3 10 Scale: meq/liter Figure 5. the contents have to be converted to meq (milliequivalents) by dividing the respective value with the milliequivalent weight (ionic weight in mg per valence). John Field Drach Field Scale: meq/liter Figure 5. To make these values comparable with each other. 1951) The analysis data for all ions in question are indicated in ppm (milligram per liter).
the solubility decrease with increasing salinity of the brine (see Figure 5. Since the expansion due to the decrease in pressure and the contraction because of temperature decrease are also insignificant. Consequently. In contrast to the solubility of natural gases in oil.06. Figure 5.0 represents the thermal contraction resulting from the cooling of the brine from the reservoir temperature to the standard temperature of 15 oC. Bw increases because of the water expansion.23b). Bw is always numerically small. • expansion of the brine because of pressure decrease.23a).3. If the reservoir pressure is reduced from initial pressure to the bubble point pressure. Similarly to the hydrocarbon liquids. the contribution of the gas liberation to B w is low.24 shows a typical relationship between Bw and p. Bw will reach its maximum value. However. A further pressure reduction leads to gas liberation. Along with the reduction of the pressure to atmospheric pressure. if the increase in volume by dissolved gas is too low to compensate the volume reduction resulting from high pressure. the initial reservoir pressure is higher than the bubble point pressure. • contraction of the brine due to temperature reduction. Additionally. rarely above 1. . their solubility in water is low (see Figure 5. In the present case.Phase Properties 121 5. The difference between this value and 1. Bw will be smaller than 1. the change in volume may have three reasons: • liberation of dissolved gas due to pressure decrease.2 Formation Volume Factor The definition of the formation volume factor of water Bw corresponds to the definition of B o.0. This loss of liquid volume only balances partly the continuing expansion so that Bw further increases with further pressure drop.
23: Solubility of natural gas in water (from DODSON and STANDING.8 0 150°F 100°F 10 20 30 3 250°F 200°F 40 Total Solid in Brine 10 [ppm] b) Figure 5. 1944) Bw.9 0.0 0. .122 Phase Properties 24 Solubility of Natural Gas in Water [ft3/bbl] 22 20 18 16 14 12 10 8 6 4 2 0 60 100 140 Pressure [psi] 5000 4500 4000 3500 3000 2500 20 00 1500 1000 50 0 180 220 260 Temperature [°F] a) Ratio: Solubility in Brine Solubility in Water Correction of Brine Salinity 1.0 0 0 pb Reservoir Pressure Figure 5.24: Typical graph of formation volume factor of water against pressure The assumption is customary that the bubble point pressure of a gassaturated brine is equal to the bubble point pressure of the crude oil inside the corresponding reservoir.b 1.
R sw. gas saturated pure water has a higher volume factor than pure water at a given pressure and temperature (dashed lines).02 1.9 Formation volume factor of reservoir water. • Assuming the effect of gas solubility on Bw to be linear. 1.25 for the given reservoir data.25 and Figure 5. .06 1. the volume factor is computed by interpolation.25 presents data of Bw for pure water saturated with natural gas (solid lines). Naturally.01 1.23a.04 1. for pure water is read from Figure 5. and corrected for salinity using Figure 5.25: Bw for pure water (dashed lines) and pure water saturated with natural gas (solid lines) as a function of pressure and temperature (from DODSON and STANDING.23 as follows: • The volume factors for pure water and pure water saturated with natural gas are read from Figure 5.07 Water Formation Volume Factor [bbl/bbl] 1.23b.03 200°F 1. • The gas solution ratio. B w for a reservoir water can be computed with aid of Figure 5.00 100°F 0.99 0.Phase Properties 123 Figure 5. 1944) Example 5.05 1.98 150°F 250°F 0 1000 2000 3000 4000 5000 Pressure [psia] Figure 5. The reservoir conditions as well as the salinity of the brine are given below both in SI.and Field units.
3 cu ft/bbl After correction of the gas solubility from Figure 5. By dividing the respective density with Bw. the .034 . The gas solution ratio in pure water from Figure 5.23a results in Rsw = 15.46 cu ft/bbl Then the formation volume factor of the reservoir water results in Bw = 1.25 results in Bw = 1.000 ppm 3000 psia 30.0275) 13. the density of reservoir water at standard conditions is plotted as a function of the total salt concentration.56 MPa 30.1.3 o Field Units 200 o C F 20.034.25 as Bw = 1. 15.7 psia and 60°F [ib/cu ft] 76 74 72 70 68 66 64 62 60 0 5 10 15 20 25 30 Total dissolved Solids [%] Figure 5.26.3 x 0.124 Phase Properties SI Units Reservoir Temperature Reservoir Pressure Salinity T p 93.033 .0275 Bw for pure water saturated with natural brine can be read off from Figure 5. Rsw results in Rsw = 15.88 = 13.23b.0275 + (1. 1973) In Figure 5.3 Brine Density at 14.46 = 1.26: Density of brine as a function of total dissolved solids (from MCCAIN.000 ppm The evaluation of Bw for pure water from Figure 5.
27b). the effect of gas in solution is to increase cw (see Figure 5. cw can be read off as cw = 3.6 3. can be evaluated without considering the gas in solution. 4.8 3.8 2. including effects of gas in solution (from DODSON and STANDING.10 Evaluation of the compressibility reservoir data of Example 5. of reservoir water at the From Figure 5.9.1 1.0 0 5 10 15 3 20 25 Gas .1 E .6 2.2 1.27: The isothermal coefficient of compressibility of pure water. From Figure 5. .4 60 100 140 180 220 260 1 ∂V c w = − V ∂P T ] sia 0 [p 100 00 20 0 0 3000 40 0 0 50 0 600 Temperature [°F] a) Ratio: Solution Compressibility Water Compressibility Correction for Gas in Solution 1.0 Compressibility of Water 6 cw 10 [bbl/bbl psi] 3.2 3.Water Ratio [ft /bbl] b) Figure 5.27a at 3000 psia and 200 oF.6 psi –1 .0 2.27a the coefficient of the isothermal water compressibility.Phase Properties 125 corresponding density under reservoir conditions can be received.3 1. cw. At a given pressure and temperature. 1944) Example 5.4 3.
3 cu ft bbl –1 . Unfortunately.23a. It is recommended to use this graph for estimating the viscosity of reservoir waters without conisdering both the pressure and the salinity of the brine.28. Correction of gas solubility from Figure 5. only few data are available.6 psi or –1 .12 × 3. .3.1 E – 6 = 3.23b: R w = 15. The viscosity should rise with pressure and increase significantly with increasing salt concentration. the gas solubility in pure water results in R sw = 15.3 × 0. The fact that low molecular paraffins are dissolved in water under reservoir conditions should furthermore lead to a considerable reduction of the brine viscosity.3 Viscosity Regarding the viscosity of oil field brines.03 E .126 Phase Properties From Figure 5.88 = 13. cw = 5. 5. –1 .47 E .23b. After evaluating the effect of the dissolved gas from Figure 5. These dependencies can be observed in Figure 5.46 cu ft / bbl . Figure 5. no corresponding data have been published until now. the compressibility of the reservoir brine results in cw = 1.28 shows the water viscosity as a function of the temperature.10 Pa respectively.
3. Gas hydrates behave rather like dissolutions of gases in ice than like chemical compounds. such as across a choke or in a separator. because the formation enthalpy is comparable. the hydrate formation is rather of physical than of chemical nature. The conditions necessary for hydrate formation could be established.2 [psia] 14.7 [psia] 7100 [psia] Vapor Pressure 1. the pressure decreases in the production line.6 0. This can cause a reduction in the temperature of the gas (adiabatic process) and. therefore.4 Natural Gas Hydrates Natural gas and liquid water can form solid material which remind of wet snow. .4 Dynamic Viscosity [cP] 1. This process takes place at temperatures slightly higher than the freezing point of pure water. Therefore.0 1. the condensation of water from the gas. The phenomenon is of special interest for petroleum industry.4 0.2 0 0 50 100 150 200 250 300 350 Temperature [°F] Figure 5.28: The viscosity of water at oil field temperature and pressure (from VAN WINGEN.Phase Properties 127 2.6 1. These solid substances are called “gas hydrates”. The water lattice seems to be similar to ice.8 Saline Water (60000 [ppm]) at 14.8 0. For example.0 0. The lattice of the hydrate crystal is mainly formed by water molecules while the hydrocarbon molecules occupy vacancies within the lattice. 1950) 5.2 1.
This may be caused by the crystal structure. water and natural gas are in equilibrium). watersoluble substances are often added to the system to inhibit or suppress the hydrate formation. a metastable equilibrium between water and hydrocarbon gas may occur (under conditions at which generally hydrate formation takes place).29 shows a part of the phase diagram of a mixture of water and a light hydrocarbon. 1973) . gas hydrate. Figure 5.Q2 represents the equilibrium natural gaswaterhydrate. a certain number of water molecules is associated to each gas molecule. the results should be applied with caution because of the large discrepancies of the few published experimental data and the correlation present.30 presents the hydrate formation conditions for natural gases of different gravities. However. The ratio of hydrocarbon to water depends mainly on the size of gas molecules.128 Phase Properties Although gas hydrates can be compared rather to solid solutions than to chemical compounds. especially in case of high gas gravity and temperature. Seed crystals can immediately induce the hydrate formation. Figure 5. The connection Q1 . Additionally. The curves can be used to estimate the conditions of hydrate formation. even if liquid water is present. Thus.31. The quadruple point Q1 is at approximately 0 oC (ice. The point Q2 is a “quadruple point” at which four phases are in equilibrium. it has to be taken into account that electrolytes present in the water will lead to decreasing temperatures at which hydrate formation will occur. The effect of the concentration of various inhibitors on the hydrateformation temperatures of a natural gas is plotted in Figure 5. The presence of liquid water is the most important prerequisite for hydrate formation. This line is of special interest. Anyway. Hydrate portion of the phase diagram for a typical mixture of water and a light hydrocarbon (from MCC AIN.
0 .Phase Properties 129 Hydrocarbon Liquid + Water Hydrate + Water Q2 C Hydrate + Ice Pressure Water + Hydrocarbon Gas Q1 Ice + Hydrocarbon Gas Temperature Figure 5. 1. 7 0.30: Pressuretemperature curves for predicting hydrate formation (from KATZ. 1973) 6000 5000 4000 3000 Pressure for Hydrate Formation [psia] 2000 1500 1000 800 600 500 400 300 200 150 100 80 50 60 30 6 0. 1945) .29: Hydrate portion of the phase diagram for a typical mixture of water and a light hydrocarbon (from MCCAIN. 0 9 ty vi ra G M ne ha et as G 40 30 40 50 60 70 80 90 Temperature [°F] Figure 5. 8 0.
555 Gas Gravity (543) 18 Depression of Hydrate Formation Temperature [°F] 16 14 12 10 1 2 3 4 8 6 4 2 0 0 4 8 12 16 20 24 Weight per cent Inhibitor Figure 5.11 Evaluation of the maximum of adiabatic expansion of a natural gas without any danger of hydrate formation. Example 5. This has been done to evaluate diagrams by which the maximum adiabatic pressure reduction without gas hydrate formation can be estimated.63 Gravity Gas (521) 4 Sodium Chloride on Methane on 0.130 Phase Properties 20 1 Methanol on 0. and the results have been combined with the hydrate formation conditions given in Figure 5.32.68 Gravity Gas (543) 3 Calcium Chloride on 0.31: Depression of hydrate formation temperature by inhibitors (from KATZ et al. 1959) Temperature calculations have been carried out for typical natural gases..63 Gravity Gas (521) 2 Sodium Chloride on 0. One of these diagrams is given by Figure 5. Problem: Up to which pressure can this gas be expanded .30.
The ordinate of Figure 5. 800 psia is the lowest final pressure which will preclude hydrate formation.80 0.32: Permissible expansion of 0.Phase Properties 131 SIUnits Gas Gravity Initial Pressure Initial Temperature adiabatically without temperature be? Solution steps: FieldUnits γ g 0. . . 1945). .5 195 190 180 170 160 150 140 13 0 1 20 110 100 Initail Pressure [psia] 2000 1500 1000 800 600 500 400 300 50 90 80 70 60 200 150 40 100 100 150 200 300 400 600 1000 2000 3000 Final Pressure [psia] Figure 5.52 MPa).3 o C) after expansion to 800 psia (5. [°F] 6000 200 5000 4000 3000 1 97.68 [MPa] 3000 psia Ti 71[oC] 160 oF formation and what will the hydrate .80 pi 20.8 gravity gas without hydrate formation (from KATZ.32 was entered at pi = 3000 psia. 10000 8000 Initial Temp.The next step consists of the vertical movement to the abscissa. The intersection of the 800 psia line with the dashed line results in 65 o F (18.It follows a horizontal movement to the 160 o F isotherm.
132 Phase Properties .
conclusions can be drawn about general sampling criteria. it can be generally stated that the best time of sampling is at the beginning of production.Chapter 6 pVTMeasurements Accurate crude oil data are necessary to control the production of a hydrocarbon reservoir efficiently. The samples should represent the system in the reservoir under its initial conditions.1 Objectives The objectives of sampling are to receive samples from a suitable place in the production wells or surface facilities.except in case of dry and wet gases . It is obvious that . They are of importance for further calculations of the processes which take place inside a reservoir. volumetric and phase behavior. and its composition is assured to supply data for • geological and reservoir engineering evaluation and forecasting. 6.1 Sampling 6. the determination of its type. • laboratory studies concerning enhanced oil recovery (EOR) methods modifying PVT properties (and the viscosity) of the system. 133 . 6.1.1.2 General Criteria Based on the general remarks about the types and behavior of hydrocarbon systems.characteristic samples can be received only in the early life of the reservoir. Therefore. Only then.
The method used depends on the type of system to be sampled. • Reliable sampling conditions in the flowing wells and at the wellhead only exist. the production gas/liquid ratio. Therefore. 6.1 represent average characteristics which can be used as some rule of thumb. There are restrictions which must always be considered: • Considering the gravitational effect.134 pVTMeasurements To meet all requirements. the careful selection of wells for sampling is very important. Values summarized in Table 6. Favorable conditions at steady state flow are indicated by stabilization of the pressure and temperature at the wellhead and in the separator. steady state conditions must be maintained. It is clear. The wells should be allocated at characteristic sections of the reservoir. it is clear that the sampling in closed wells could easily lead to misleading data. To determine the fluid properties along with the depth of the pay zone (where monophase system exists). Representative samples cannot be received from wells which are perforated at the water/oil or gas/oil contact and are producing water and/or nonequilibrium gas. Number of samples will depend on the dip of the structure as well as on the pay thickness.1. should remain constant during a long period of time. The third possibility is to get representative samples from the wellhead (wellhead sampling).3 Sampling Methods There are three methods of sampling. preliminary screening must be used cautiously. • No characteristic fluid can be received within the flowing wells. GOR. Moreover. if the total composition of the system entering the well and leaving it at the wellhead remains constant. where a multiphase system exists or is developing. Samples can be taken from the flowing wells (subsurface sampling) or from the equilibrium phases of the separators (separator samples). that the behavior of any system also depends on reservoir conditions which are missing on the tabulated data. There are certain approximations for selecting sampling methods which consider some of the properties of the produced fluids. . however. a minimum of one sample per 30 m can be recommended. In every case.
Sampling of equilibrium phases from separators can be used for all systems. The continuous monitoring of the concentration of some key components in the gas phase is recommended. separators often contain high pressure windows for visual control. that no separator sampling is needed for dry gases and for saturated and undersaturated oils. These bottles are in vertical position during sampling. This procedure should be considered to control the development and maintenance of steadystate flow. Separators must contain pressure gauges and thermometers as well as devices for the accurate measurement of the gas volumes (liberated and/or delivered). oil. When highly paraffinic systems are produced exhibiting a cloud point in the vicinity of the separator temperature. Sometimes. Note. if no free water enters the production well and if there is a considerable length along which a1phase oil.1500 < 200 Gas/liquid ratio [m3 /m3 ] Dry gas Wet gas Gas/condensate system High shrinkage oils Low shrinkage oils initial state and on the changes in their behavior in the depression zone and in the wells.20000 1000 . the following guidelines can be applied in screening sampling methods. but steady state conditions must exist. Samples of the equilibrium phases should be take using specially designed bottles with valves at their extreme ends. Each phase can be sampled through sampling valves built in the separator at that part where pure gas. For sampling. no liquid droplets should depart the system with the gas. only those separators can be used where construction ensures highly efficient separation of the entering phases.1: Screening sampling methods Type Density of the produced liquid [kg/m3 ] < 740 740 .800 > 800 11000 . liquid hydrocarbons. separators should be heated.780 780 . and water exist.pVTMeasurements 135 On the basis of the general remarks on the phase behavior of different systems in their Table 6. In a wide region of flow rates. and of water. The upper valve is connected with the sampling valve of the separator by a flexible tubing which is resistant to pressure and corrosion by phases sampled.13000 200 . the bottles and the connecting tube are filled up with an indifferent liquid (mercury or water from the separator or from other sources) to maintain pressure and to . To determine the phase boundary during the time of sampling.
At the surface.g.by gravitational effects. However. Bottom hole samplers are specially designed devices containing one valve or two valves at their extreme ends.is filled up with an inert liquid (e. Data for identification must be delivered with the sample.for systems which exist in a 1phase state up to the wellhead. The valves are operated mechanically or electrically from the surface or within the sampler at the predetermined depth or time.136 pVTMeasurements exclude air. Bottom hole sampling Bottom hole sampling can be used when only the phase to be sampled is present at the perforated zone. mercury). .in a simple manner . The created gas cap will prevent uncontrolable pressure increase during shipping. about 90% of mercury (or brine) has to be displaced by the separator liquid. The method can be controlled by sampling at different depths and by analyzing the water content of samples. In doing so. the flow of a 1phase hydrocarbon fluid must be assured. The protection of the samples in the shipping bottles and their correct marking for identification are important as it has already been mentioned.together with the shipping bottle . They are run into the well through a sealed tube mounted to the christmas tree by a mechanically driven cable. The container is connected with a valve at the wellhead and flushed with the fluid produced. The previous remarks referring to the protection and identification of the shipping bottles are also valid for wellhead samples. If water is also produced with the hydrocarbons but remains at the bottom under steadystate conditions. additional 10% of mercury (or brine) should be removed by opening the lower valve. No sampling within the wells is needed for dry gases or . the use of mercury (or brine) is recommended in sampling the liquid phase.in general . It is then displaced by the sampled phase (gas or liquid). After reaching the desired depth (usually with open valves to allow the flushing of the sampler by the flowing fluids). The sample is displaced by injection of the inert liquid into the sampler using a volumetric pump or . Sampling at the wellhead Sampling at the wellhead is a suitable process in case of the production of dry gas or undersaturated oil at wellhead conditions. The valves in the contains must be protected by steel caps during shipping. Whenever the temperature in the separator is lower than the ambient temperature. the sample should immediately be transferred to containers through a flexible steel tube which . samples can be taken above the water level. the valves are closed. After closing the valves into he container and in the separator and after dismounting the container.
it is important • to prevent any loss . Specially designed sampling is needed to determine the microbiological contamination of the produced fluid(s).pVTMeasurements 137 6. Afterwards.especially in their lighter fractions .4 Special Problems There are sometimes problems arising from low permeability oil reservoirs (wells being produced with high depression). However. it is only possible to take samples from the equilibrium phases of the separator and to recombine them under separator conditions corresponding to their volume ratio. As for liquids. the ratio of the phases entering the wells does not characterize the initiallyinplace fluids (bubble point oil system).1. A similar approach can also be used for gascondensate systems. Sterile samplers and shipping bottles but also anaerobic conditions (in the course of sample transfer) must be guaranteed. • to avoid the contact with air because oxidation processes could drastically modify the composition of hydrocarbon liquids. primarily those of oils. In this case. In this case. especially if water is also present. This procedure is often necessary in case of reservoirs containing only water from which water will be produced for industrial or public supply.if the liquid is in equilibrium with other phase(s) . the samples must be preserved in their original state. a similar procedure as above should be used to restore the composition of the hydrocarbon phase. As a result. pressure and temperature are increased up to the reservoir conditions. However. The basic concept of this method is that .the composition of the oil phase is not influenced by the contacted volume of the other phase(s) being in equilibrium with the oil. Whenever water production cannot be avoided and the volume of the produced water is high. no processes in the reservoir/well system should take place which may result in a difference between the overall composition of the reservoir fluid and the produced fluid. . dissolved gases may be liberated from the water having a composition which could largely differ from that of the equilibrium hydrocarbon system.
1: Scheme of PVT equipments Samples to be studied are transferred to the P V T cell. equipments and methods have been developed for the petroleum industry. Some known volume of an inert liquid is injected into the sampling bottle(s).1.138 pVTMeasurements 6.1 Equipment Gasometer Vg G Control unit Separator L Analysis pVT Cell Sample Hg p Gauge Volumetric Pump Figure 6. 6. • It should not react with the fluids contacted. The selection criteria for the inert liquid are: • It must exhibit low volatility under (i) ambient conditions and also (ii) under temperatures and pressures which represent the working conditions. while the same liquid is withdrawn from the cell initially filled with it.2 Experimental Determination of the Volumetric and Phase Behavior To determine the volumetric and phase behavior of naturally occurring systems under reservoir/well/surface conditions. Sample .2. A general scheme of the set of equipments used for PVT studies shown in Figure 6.
relatively smaller ones are used to investigate characteristic oil systems.2. the samples will be transferred and measured by a piston built into the containers and the PVT cell. the accuracy of measurements depends on the resolution of the measuring instruments. . The cell has to be a heat. Therefore. however.and pressureresistant vessel which is equipped with valves made of noncorrosive materials. The sample being transferred into the cell must be in a 1phase state. 6.2 PVTCells Evaluation of the PVT behavior is carried out on the system transferred into the PVT cell. Although many different PVT cells are specified. (ii) gascondensate systems. and (iv) water. Although mercury is widely used. On the other hand. However. To obtain the desired rapid establishment of the equilibrium conditions at any given pressure and/or temperature. if the liquid transported by the volumetric pump(s) is separated from the samples. Referring to separator samples. This valve controls the rate and volume of the outflowing inert liquid. The inside of the cell is lighted from outside by transmitted or reflected (from the inner wall of the cell) light. In this case. (iii) critical oils. two basic types can be distinguished: • blind PVT cells (without any possibility of visual observation of the system to be measured).pVTMeasurements 139 • Its physical and chemical properties must be well known. the cell is agitated or stirred inside.and temperature. Sometimes only one pump may be used for this purpose. “micro PVT cells” are not used in practice. Moreover. The cell is mounted into a thermostatic bath which ensures from detrimental temperature inconstancy in the cell. Nevertheless. larger cell volumes are used to study (i) gases. where the volumetric and phase behavior can directly be observed through one or more pressure. the relative error can be decreased by a reasonable increase of the sample volume studied. No restriction is needed. sometimes low melting point alloys are used. considering the change in volume of the system within the pressure range applied in the PVT studies. the proportions of gas and liquid to be injected into the PVT cell must be equal to those measured into the field separator at the time of sampling. • windowed PVT cells. liberated gas or condensate in the shipping bottle has to be dissolved or evaporated in advance. if the inert liquid leaves the cell through a suitable operated valve at the bottom of the cell. Doubleacting or two singleacting volumetric pump(s) are needed for this process. Therefore.resistant windows built into the cell body. The electrically heated (or cooled) medium circulating around the cell could be air or liquid.
3: PVT cell (after BURNETT) A special PVT cell for gases was developed by B URNETT (see Figure 6. Its volume is about 0. oil systems. It contains two or more chambers with different volumes which are separated by valves. The chamber possessing the smaller volume contains the gas under given pressure.2: Blind PVT cell Blind PVT cells are usually used to study systems in which no phase transition takes place or phase transition can easily be determined using the isothermal p . In V1 Out Pressure Gauge V2 Figure 6.001[m3] which is large enough to be used for gas and oil studies as well.2.V relationship (dry gases.140 pVTMeasurements Valve Thermometer Valve Steel Body Thermocouple Cell Bath Heat Isolator Jacket Heating Wire Hg Valve Figure 6. A characteristic blind cell is shown in Figure 6. water).3). By opening .
It is also required whenever phase transition cannot exactly be determined by isothermal p .V relationships (volatile oils or partially miscible liquid/liquid systems with or without free gas phase). Windowed P V T cells Windowed PVT cells are usually used to evaluate systems where the knowledge of the volumes of coexisting phases is necessary (e.5.g. the change in pressure can be measured. A PVT cell with variable volume is presented in Figure 6.4: PVT cell (after DEANPOETTMAN) A typical onewindowed cell is shown in Figure 6. The volume corresponding to that motion can be read on the calibrated stem of the plungers.pVTMeasurements 141 successively the valves of the evacuate other chambers with known volume. gas condensate systems). Gas Liquid Hg Windows Gas Liquid Hg Figure 6. . wet gases under well or separator conditions. The condensed phase can be moved to the window by simultaneous lifting (or sinking) the two builtin plungers.4. The volumes of the liquid phase can be determined by the angle of rotating the cell until liquid phases appear in the window. No volumetric pumps are needed in this case.
5: Variable volume cell (after VELOKIVSKIY et al.6) is concerned.8. It serves also for the measurement of the volume of phases. Meanwhile. This can be done by injecting inert fluid.7. The scheme of this cell is shown in Figure 6.142 pVTMeasurements Window Figure 6. for Operation of the Piston 2 Piston with joist Packing 3 Hole and Valve for Introduction of Gasoline 4 Hole and Valve for Introduction and Sampling of Gas 5 Spiral for stirring Gas. the condensed liquid is collected and can be lifted up to the windows which are built into the middle part. pressure can be kept at the desired constant value by the simultaneous lifting of the plunger. In the upper part. Another type contains one capillary glass tube within the cell through which the inner volume can be lighted and observed using an optical system.6: PVT cell (after S LOAN) . 1 2 3 5 4 6 7 8 9 1 Orifice for Introduction of Mercury through the Upper Cap. The Liquid is stirred by a Spiral Coil Wound around the axis 6 Sapphire Viewing Window 7 Viewing Window 8 Lighting Window 9 Liquid Volume Measurement Chamber 10 Magnetic Agitator 10 Figure 6. In the lower part. a plunger is moved by the fluid injected or withdrawn by a volumetric pump. Three windows are built into the body of the most generally used cell as shown in Figure 6.) As far as S LOAN’s windowed cell (see Figure 6. there are three sections.
8: RUSKA cell .pVTMeasurements 143 Capillary (Glass) Figure 6.7: PVT cell (after WELLSROOF) Windows Figure 6.
9: RUSKA volumetric mercury pump 6. In order to store the inert liquid (mercury). the volumetric pump(s). also dead weight gauges can be used for this purpose and for the calibration of the B OURDONtubegauges as well. hand. In practice. These gauges are also built into the tube network.are used to measure the volume of the inert liquid injected to and withdrawn from the cell respectively.and mechanically operated pumps are used. a container is used which is also interconnected . the latter sometimes with variable speed. but in opposite directions.2.9. However. Its general construction does not differ from the pump which is mechanically driven.2.4 Auxiliary Equipment The PVT cell. valves are mounted in the network of tubes. Pressure in the system is measured by gauges containing BOURDONtubes. A scheme of a handoperated volumetric pump is shown in Figure 6. In the generally used types. There are also gauges in which an electric signal is proportional to the pressure measured. and other parts of the PVT system are interconnected by pressureresistant stainless steel tubes. the inert liquid is displaced by a plunger from a pressure resistant and thermostatically controlled cylinder.3 Volumetric Pumps Volumetric pumps . The pressure can be increased or decreased. 6. A special type of volumetric pumps is the doubleacting one: Two plungers are working simultaneously.the so called positive displacement pumps .144 pVTMeasurements Figure 6. The movement of the connecting screwspindle can be read on a calibrated rod and on a dial to increase the accuracy. In order to separate single parts of the equipment. They are filled with the inert liquid used in the volumetric pumps. The transported volume is proportional to the change in volume of the system studied.
3. They are distinguishable by considering the change and the direction of change in the composition of the system. and in the thermostate around the cell is usually measured by thermocouples (may be controlled by thermometers). gas could condense under ambient conditions or at preselected pressures. measurements should be carried out in a pressure region which correctly characterizes the phase transition and the 2phase (or multiphase) behavior of the system.carried out at reservoir temperature . Measurements start at or above the reservoir pressure. its volume must be measured. If the gas phase liberated in the PVT cell will be displaced.3. and the volume in the pumps are measured electrically. A scheme of the flash process has been presented in Figure 4.1 Flash Process In the course of flash processes. Measurements are usually repeated at other temperatures. a laboratory separator must be inserted between the PVT cell and the gasometer. in the PVT cells. For this purpose gasometers are used. temperature. The lowest pressure applied is determined by the system studied.3 Methods Basically. In this case the flash process . For systems in which no phase transition takes place. 6. At each pressure step. In this case the flash process simulates the whole course of production. the lowest pressure selected could be the atmospheric one. Therefore. pressure is gradually increased or decreased at constant temperature under equilibrium conditions. Whenever phase transition occurs. the process can be automated and computerized. 6.pVTMeasurements 145 through valves with the different parts of the PVT system. However. The temperature of the volumetric pump(s). too. . covering some range between reservoir and surface temperatures (in the separator or in the stock tanks).simulates the behavior of the system far inside the reservoir where its composition is not influenced by the production. there are three methods generally used in PVT studies. the volumes of the system and that of the existing phases are measured and the development of phases is observed. Whenever pressure.
In the course of a stepwise process. To follow the change in composition of the system and to characterize the displaced phase (gas). an isothermal pressure increase is realized under constant volume conditions.also explained in Figure 4. the composition of the latter is also determined. The development or disappearance of phases is observed. If the effect of pressure maintenance by gas injection is being studied after each step of the differential measurement. The differential process is repeatedly carried out. the values of characteristic parameters will differ from those determined by the flash process. the differential process is carried out at constant volume and by variation of the system composition. In depleted (gas condensate or oil) reservoirs. usually down to the lowest possible reservoir (separator) pressure. The highest pressure to be . except the starting point representing the intact system under initial reservoir conditions.146 pVTMeasurements 6.3. 6. p is reestablished by injecting a measured volume of the preselected gas. • partial or total evaporation of the immobile liquid condensate is needed. As a result of the gradually changing composition. where any decrease in the reservoir pressure results from the stepwise production of one of the equilibrium phases (usually liberated or free gas). the volumes of the system and those of the existing phases are measured. In contrast to the flash process during which the composition is maintained and the volume is varied.2 Differential Process The differential process describes the change in the volumetric and phase behavior of the system at reservoir temperature and under equilibrium conditions. Both are constant volume processes. There are two subgroups of the reverse differential process. the effect of increasing the pressure by preselected gas can also be studied.3 Reverse Differential Process The reverse differential process is used to study the effect of any gas injection process on the volumetric and phase behavior of systems. the phase volumes are measured and phase transition(s) are observed at reservoir temperature. when • in the initial state of a gas condensate system. This method is used. further information on the volume (and composition) of phases separated can be given. the liquid phase exists beyond the restriction of retrograde condensation. This can be done under initial conditions (gas cycling) or by starting from a given depleted state (gas injection with pressure increase). In the course of the differential study .3. By use of a laboratory separator in which pressure and temperature correspond to those during the field separation process.3 . In the course of gas injection (using a preselected gas).
If some 1phase system develops below this pressure. it is obviously no need to continue the process in a higher pressure region.pVTMeasurements 147 reached corresponds to the fracturing pressure of the formation. or .if this information is not available .it is about 10% higher than the initial reservoir pressure. The effect of gas injection is often the combination of the methods discussed above. .
This action might not be possible to undo. Are you sure you want to continue?
Use one of your book credits to continue reading from where you left off, or restart the preview.