, 102
Liquid Heat
Capacity
THRSWT/6 DIPPR, PPDS, BARIN, IK
CAPE, NIST
100
, 114
Ideal Gas Heat
Capacity
THRSWT/7 Aspen, DIPPR, BARIN,
PPDS, IKCAPE, NIST
107, 127
Second Virial
Coefficient
THRSWT/8 DIPPR 104
Liquid Viscosity TRNSWT/1 Aspen (Andrade), DIPPR,
PPDS, IKCAPE, NIST
101
, 115
Vapor Viscosity TRNSWT/2 Aspen (ChapmanEnskog
Brokaw), DIPPR, PPDS, IK
CAPE, NIST
102
Liquid Thermal
Conductivity
TRNSWT/3 Aspen (SatoRiedel),
DIPPR, PPDS, IKCAPE,
NIST
100
Vapor Thermal
Conductivity
TRNSWT/4 Aspen (StielThodos),
DIPPR, PPDS, IKCAPE,
NIST
102
Liquid Surface
Tension
TRNSWT/5 Aspen (HakimSteinberg
Stiel), DIPPR, PPDS, IK
CAPE, NIST
106
Which equation is actually used to calculate the property for a given
component depends on which parameters are available. If parameters are
available for more than one equation, the Aspen Physical Property System
uses the parameters that were entered or retrieved first from the databanks.
The selection of submodels is driven by the data hierarchy, and controlled by
the submodelselection parameters.
The thermodynamic properties use the THRSWT (thermo switch) submodel
selection parameter, and the transport properties use the TRNSWT (transport
switch) submodelselection parameter.
8 1 Introduction
As the previous table shows, a property is associated with an element of the
submodelselection parameter. For example, THRSWT element 1 controls the
submodel for solid volume.
The following table shows the values for THRSWT or TRNSWT, and the
corresponding submodels.
Parameter Values
(Equation Number) Submodel
0 Aspen
1 to 127 DIPPR
200 to 211 BARIN
301 to 302 PPDS or propertyspecific methods
400 PML
401 to 404 IKCAPE
501 to 515 NIST
All builtin databank components have modelselection parameters (THRSWT,
TRNSWT) that are set to use the correct equations that are consistent with
the available parameters. For example, suppose that parameters for the
DIPPR equation 106 are available for liquid surface tension. For that
component, TRNSWT element 5 is set to 106 in the databank. If you are
retrieving data from an inhouse or user databank, you should store the
appropriate values for THRSWT and TRNSWT in the databank, using the
appropriate equation number. Otherwise, the Aspen Physical Property System
will search for the parameters needed for the Aspen form of the equations.
If a component is available in more than one databank, the Aspen Physical
Property System uses the data and equations based on the databank list
order on the Components Specifications Selection sheet. For example,
suppose the databank search order is ASPENPCD, then PURE25, and that the
Aspen Physical Property System cannot find the parameters for a particular
submodel (equation) in the ASPENPCD databank. If the PURE25 databank
contains parameters for another equation, the Aspen Physical Property
System will use that equation (most likely the DIPPR equation) to calculate
the property for that component.
If your calculation contains any temperaturedependent property parameters,
(such as CPIGDP for DIPPR ideal gas heat capacity, entered on the Methods
 Parameters  Pure Component form), the Aspen Physical Property
System sets the corresponding THRSWT and TRNSWT elements for that
component to the default values shown in the table above. This default
setting might not always be correct. If you know the equation number, you
should enter it directly on the Methods  Parameters  Pure Component
form. For example, suppose you want to use the:
 DIPPR equation form of heat of vaporization (DHVLDP) for a component
 Aspen equations for the remaining temperature dependent properties
Set the fourth element of the THRSWT parameter to 106, and the 13 and 58
elements to 0. If you want to set the other temperaturedependent properties
to use what is defined for that component in the databank, leave the element
blank.
The following table lists the available DIPPR equations and the corresponding
equation (submodel) number.
1 Introduction 9
Available DIPPR Equations
Equation
Number Equation Form
100
101
102
103
104
105
106
107
114
115
116
127
For equations 114 and 116, t = 1T
r
.
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
The following sections describe the Aspen, DIPPR, BARIN, IKCAPE, PPDS, and
NIST equations for each property. For descriptions of the the BARIN equations
for heat capacity and enthalpy, see BARIN Equations for Gibbs Energy,
Enthalpy, Entropy, and Heat Capacity.
Extrapolation Methods
Many temperature dependent property models have upper and lower
temperature limits. The Aspen Physical Property System usually extrapolates
linearly beyond such limits. It calculates the slope of the propertyversus
temperature curve, or the ln(property)versustemperature curve for models
expressed in logarithmic form, at the upper or lower temperature limit. For
temperatures beyond the limit, it uses a linear model with this slope which
meets the curve from the equation at the temperature limit. Thus the model
is:
10 1 Introduction
For T beyond the upper or lower limit, where T
lim
is that limit and Z is the
property or ln(property) as appropriate. Some liquid molar volume models are
actually molar density models which then return the reciprocal of the density
as the liquid molar volume. In these models, the extrapolation occurs for the
density calculation.
There are certain exceptions, detailed below.
Exception 1: Logarithmic Properties Based on Reciprocal
Temperature
This applies to property models expressed in the form (where a(T) includes
any additional dependency on temperature):
For these models, the extrapolation maintains the slope of ln(property) versus
1/T. This applies to the Extended Antoine vapor pressure equation and the
Andrade and DIPPR liquid viscosity equations. Note that the equation for
Henry's Constant is extrapolated by ln(Henry) versus T.
Exception 2: Aspen Ideal Gas Heat Capacity
The Aspen Ideal Gas Heat Capacity model has an explicit equation for
extrapolation at temperatures below the lower limit, which is described in the
model. At high temperatures it follows the usual rule of extrapolating
propertyversustemperature linearly.
Exception 3: No Extrapolation
The equations for certain models are used directly at all temperatures, so that
no extrapolation is performed. These models are the Wagner vapor pressure
equation, the Aly and Lee equation for the DIPPR Ideal Gas Heat Capacity
(using the CPIGDP parameter), and the Water Solubility and Hydrocarbon
Solubility models. The equations for temperaturedependent binary
interaction parameters are also used directly at all temperatures with no
extrapolation.
2 Thermodynamic Property Models 11
2 Thermodynamic Property
Models
This section describes the available thermodynamic property models in the
Aspen Physical Property System. The following table provides a list of
available models, with corresponding Aspen Physical Property System model
names. The table provides phase types for which the model can be used and
information on use of the model for pure components and mixtures.
Aspen Physical Property System thermodynamic property models include
classical thermodynamic property models, such as activity coefficient models
and equations of state, as well as solids and electrolyte models. The models
are grouped according to the type of property they describe.
Thermodynamic Property Models
Phases: V = Vapor; L = Liquid; S = Solid. An X indicates applicable to Pure or
Mixture.
EquationofState Models
A pure component equation of state model calculates PHIL, PHIV, DHL, DHV,
DGL, DGV, DSL, DSV, VL, and VV. Most mixture equation of state models
calculate PHILMX, PHIVMX, DHLMX, DHVMX, DGLMX, DGVMX, DSLMX,
DSVMX, VLMX, and VVMX. Those marked with * only calculate DHLMX,
DHVMX, DGLMX, DGVMX, DSLMX, DSVMX, VLMX, and VVMX. The alpha
functions and mixing rules are options available in some of the models.
Property Model Model Name(s) Phase(s)Pure Mixture
ASME Steam Tables ESH2O0,ESH2O V L X
BWRLeeStarling ESBWR0, ESCSTBWR V L X X
BenedictWebbRubinStarling ESBWRS, ESBWRS0 V L X X
GERG2008 V L X X
HaydenO'Connell ESHOC0,ESHOC V X X
HF equationofstate ESHF0, ESHF V X X
IAPWS95 Steam Tables V L X
Ideal Gas ESIG0, ESIG V X X
LeeKesler * ESLK V L X
LeeKeslerPlcker ESLKP0,ESLKP V L X X
12 2 Thermodynamic Property Models
Property Model Model Name(s) Phase(s)Pure Mixture
NBS/NRC Steam Tables ESSTEAM0,ESSTEAM V L X
Nothnagel ESNTH0,ESNTH V X X
PCSAFT ESPSAFT, ESPSAFT0 V L X X
PengRobinson ESPR0, ESPR V L X X
Standard PengRobinson ESPRSTD0,ESPRSTD V L X X
PengRobinsonWongSandler * ESPRWS0,ESPRWS V L X X
PengRobinsonMHV2 * ESPRV20,ESPRV2 V L X X
Predictive SRK * ESRKSV10, ESRKSV1 V L X X
RedlichKwong ESRK0, ESRK V X X
RedlichKwongAspen ESRKA0,ESRKA V L X X
RedlichKwongSoave ESRKSTD0,ESRKSTD V L X X
RedlichKwongSoaveBoston
Mathias
ESRKS0, ESRKS V L X X
RedlichKwongSoaveMHV2 * ESRKSV20, ESRKSV2 V L X X
RedlichKwongSoaveWong
Sandler *
ESRKSWS0, ESRKSWSV L X X
SchwartzentruberRenon ESRKU0,ESRKU V L X X
SoaveRedlichKwong ESSRK0, ESSRK V L X X
SRKKabadiDanner ESSRK0, ESSRK V L X X
SRKML ESRKSML0, ESRKSML V L X X
VPA/IKCAPE equationofstate ESVPA0, ESVPA V X X
PengRobinson Alpha functions V L X
RKSoave Alpha functions V L X
HuronVidal mixing rules V L X
MHV2 mixing rules V L X
PSRK mixing rules V L X
WongSandler mixing rules V L X
Activity Coefficient Models (Including Electrolyte Models)
These models calculate GAMMA.
Property Model Model Name Phase(s)Pure Mixture
BromleyPitzer GMPT2 L X
ChienNull GMCHNULL L X
Constant Activity Coefficient GMCONS S X
COSMOSAC COSMOSAC L X
Electrolyte NRTL GMENRTL, GMELC,
GMENRHG
L L1 L2 X
ENRTLSAC (patent pending) ENRTLSAC L X
Hansen HANSEN L X
Ideal Liquid GMIDL L X
NRTL (NonRandomTwoLiquid) GMRENON L L1 L2 X
NRTLSAC (patent pending) NRTLSAC L X
Pitzer GMPT1 L X
2 Thermodynamic Property Models 13
Property Model Model Name Phase(s)Pure Mixture
Polynomial Activity Coefficient GMPOLY L S X
RedlichKister GMREDKIS L S X
ScatchardHildebrand GMXSH L X
Symmetric Electrolyte NRTL GMENRTLS L X
ThreeSuffix Margules GMMARGUL L S X
UNIFAC GMUFAC L L1 L2 X
UNIFAC (Lyngby modified) GMUFLBY L L1 L2 X
UNIFAC (Dortmund modified) GMUFDMD L L1 L2 X
UNIQUAC GMUQUAC L L1 L2 X
Unsymmetric Electrolyte NRTL GMENRTLQ L X
van Laar GMVLAAR L X
Wagner interaction parameter GMWIP S X
Wilson GMWILSON L X
Wilson model with liquid molar
volume
GMWSNVOL L X
Vapor Pressure and Liquid Fugacity Models
Property Model Model
Name
Property Phase(s)Pure Mixture
General Pure Component Liquid
Vapor Pressure
PL0XANT PL L L1 L2 X
API Sour SWEQ PHILMX L X
Braun K10 BK10 PHILMX L X
ChaoSeader PHL0CS PHIL L X
GraysonStreed PHL0GS PHIL L X
KentEisenberg ESAMINE PHILMX,
GLMX,
HLMX,
SLMX
L X
MaxwellBonnell PL0MXBN PL L L1 L2 X
Solid Antoine PS0ANT PS S X
Heat of Vaporization Models
These models calculate DHVL.
Property Model Model Name Phase(s)Pure Mixture
General Pure Component Heat
of Vaporization
DHVLWTSN L X
ClausiusClapeyron Equation DHVLCC L X
Molar Volume and Density Models
These models calculate VL (pure liquid), VLMX (liquid mixture), or VS (pure
solid), except for BrelviO'Connell which calculates VLPM.
Property Model Model Name Phase(s)Pure Mixture
API Liquid Volume VL2API L X
BrelviO'Connell VL1BROC L X
14 2 Thermodynamic Property Models
Property Model Model Name Phase(s)Pure Mixture
ChuehPrausnitz VL0CPRKT, VL2CPRKT L X X
Clarke Aqueous Electrolyte
Volume
VAQCLK L X
COSTALD Liquid Volume VL0CTD,VL2CTD L X X
DebyeHckel Volume VAQDH L X
Liquid Constant Molar Volume VL0CONS L X
General Pure Component Liquid
Molar Volume
VL0RKT,VL2RKT L X
Rackett/CampbellThodos
Mixture Liquid Volume
VL2RKT L X X
Modified Rackett VL2MRK L X X
General Pure Component Solid
Molar Volume
VS0POLY S X
Liquid Volume Quadratic Mixing
Rule
VL2QUAD L X
Heat Capacity Models
Property Model Model
Name
Property Phase(s)Pure Mixture
Aqueous Infinite Dilution Heat
Capacity Polynomial
L X
CrissCobble Aqueous Infinite
Dilution Ionic Heat Capacity
L X
General Pure Component Liquid
Heat Capacity
HL0DIP HL, DHL L X
General Pure Component Ideal
Gas Heat Capacity
CPIG V X X
General Pure Component Solid
Heat Capacity
HS0POLY HS S X
Solubility Correlation Models
Property Model Model
Name
Property Phase(s)Pure Mixture
Henry's constant HENRY1 HNRY,
WHNRY
L X
Water solubility L X
Hydrocarbon solubility L X
Other Models
Property Model Model
Name
PropertyPhase(s)Pure Mixture
Cavett Liquid Enthalpy DepartureDHL0CVT,
DHL2CVT
DHL,
DHLMX
L X X
BARIN Equations for Gibbs
Energy, Enthalpy, Entropy and
Heat Capacity
S L V X
BromleyPitzer Enthalpy HAQPT2 HLMX L X
BromleyPitzer Gibbs Energy GAQPT2 GLMX L X
2 Thermodynamic Property Models 15
Property Model Model
Name
PropertyPhase(s)Pure Mixture
Electrolyte NRTL Enthalpy HMXENRTL,
HAQELC,
HMXELC,
HMXENRHG
HLMX L X
Electrolyte NRTL Gibbs Energy GMXENRTL,
GAQELC,
GMXELC,
GMXENRHG
HLMX L X
Liquid Enthalpy from Liquid Heat
Capacity Correlation
DHL0DIP L X X
Enthalpies Based on Different
Reference States
DHL0HREF DHL L V X X
Helgeson Equations of State L X
Pitzer Enthalpy HAQPT1 HLMX L X
Pitzer Gibbs Energy GAQPT1 GLMX L X
Quadratic Mixing Rule L V X
EquationofState Models
The Aspen Physical Property System has the following builtin equationof
state property models. This section describes the equationofstate property
models available.
Model Type
ASME Steam Tables Fundamental
BWRLeeStarling Virial
BenedictWebbRubinStarling Virial
GERG2008 Fundamental/mixing rules
HaydenO'Connell Virial and association
HF EquationofState Ideal and association
HuronVidal mixing rules Mixing rules
IAPWS95 Steam Tables Fundamental
Ideal Gas Ideal
LeeKesler Virial
LeeKeslerPlcker Virial
MHV2 mixing rules Mixing rules
NBS/NRC Steam Tables Fundamental
Nothnagel Ideal
PCSAFT Association
PengRobinson Cubic
Standard PengRobinson Cubic
PengRobinson Alpha functions Alpha functions
PengRobinsonMHV2 Cubic
PengRobinsonWongSandler Cubic
Predictive SRK Cubic
16 2 Thermodynamic Property Models
Model Type
PSRK mixing rules Mixing rules
RedlichKwong Cubic
RedlichKwongAspen Cubic
Standard RedlichKwongSoave Cubic
RedlichKwongSoaveBostonMathias Cubic
RedlichKwongSoaveMHV2 Cubic
RedlichKwongSoaveWongSandler Cubic
RKSoave Alpha functions Alpha functions
SchwartzentruberRenon Cubic
SoaveRedlichKwong Cubic
SRKKabadiDanner Cubic
SRKML Cubic
VPA/IKCAPE equationofstate Ideal and association
WongSandler mixing rules Mixing rules
ASME Steam Tables
The ASME steam tables are implemented like any other equationofstate in
the Aspen Physical Property System. The steam tables can calculate any
thermodynamic property of water or steam and form the basis of the STEAM
TA property method. There are no parameter requirements. The ASME steam
tables are less accurate than the NBS/NRC steam tables.
References
ASME Steam Tables, Thermodynamic and Transport Properties of Steam,
(1967).
K. V. Moore, Aerojet Nuclear Company, prepared for the U.S. Atomic Energy
Commision, ASTEM  A Collection of FORTRAN Subroutines to Evaluate the
1967 ASME equations of state for water/steam and derivatives of these
equations.
BWRLeeStarling
The BenedictWebbRubinLeeStarling equationofstate is the basis of the
BWRLS property method. It is a generalization by Lee and Starling of the
virial equationofstate for pure fluids by Benedict, Webb and Rubin. The
equation is used for nonpolar components, and can manage hydrogen
containing systems.
General Form:
Where:
Mixing Rules:
2 Thermodynamic Property Models 17
Where:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCBWR T
ci
TC X 5.0 2000.0 TEMPERATURE
VCBWR V
ci
*
VC X 0.001 3.5 MOLE
VOLUME
BWRGMA
i
OMEGA X 0.5 3.0
BWRKV
c
ij
0 X 5.0 1.0
BWRKT
q
ij
0 X 5.0 1.0
Binary interaction parameters BWRKV and BWRKT are available in the Aspen
Physical Property System for a large number of components from Brul et al.
(1982) and from Watanasiri et al. (1982). (See Physical Property Data,
Chapter 1).
References
M.R. Brul, C.T. Lin, L.L. Lee, and K.E. Starling, AIChE J., Vol. 28, (1982) p.
616.
Brul et al., Chem. Eng., (Nov., 1979) p. 155.
Watanasiri et al., AIChE J., Vol. 28, (1982) p. 626.
BenedictWebbRubinStarling
The BenedictWebbRubinStarling equationofstate is the basis of the BWRS
property method. It is a modification by Han and Starling of the virial
equationofstate for pure fluids by Benedict, Webb and Rubin. This equation
ofstate can be used for hydrocarbon systems that include the common light
gases, such as H
2
S, CO
2
and N
2
.
The form of the equationofstate is:
18 2 Thermodynamic Property Models
Where:
k
ij
= k
ji
In the mixing rules given above, A
0i
, B
0i
, C
0i
, D
0i
, E
0i
, a
i
, b
i
, c
i
, d
i
, o
i
,
i
are pure
component constants which can be input by the user. For methane, ethane,
propane, isobutane, nbutane, isopentane, npentane, nhexane, n
heptane, noctane, ethylene, propylene, nitrogen, carbon dioxide, and
hydrogen sulfide, values of the parameters in the table below are available in
the EOSLIT databank in the Aspen Properties Enterprise Database.
2 Thermodynamic Property Models 19
If the values of these parameters are not given, and not available from the
databank, the Aspen Physical Property System will calculate them using the
critical temperature, the critical volume (or critical density), the acentric
factor and generalized correlations given by Han and Starling.
When water is present, by default BenedictWebbRubinStarling uses the
steam table to calculate the enthalpy, entropy, Gibbs energy, and molar
volume of water. The total properties are molefraction averages of these
values with the properties calculated by the equation of state for other
components. Fugacity coefficient is not affected. An option code can disable
this use of the steam table.
For best results, the binary parameter k
ij
must be regressed using phase
equilibrium data such as VLE data.
Parameter
Name/
Element
SymbolDefault MDS Lower
Limit
Upper
Limit
Units
BWRSTC T
ci
TC x 5.0 2000.0TEMPERATURE
BWRSVC V
ci
VC x 0.001 3.5 MOLEVOLUME
BWRSOM e
i
OMEGA x 0.5 2.0
BWRSA/1 B
0i
fcn(e
i
,V
ci
, T
ci
) x MOLEVOLUME
BWRSA/2 A
0i
fcn(e
i
,V
ci
, T
ci
) x PRESSURE * MOLE
VOL^2
BWRSA/3 C
0i
fcn(e
i
,V
ci
, T
ci
) x PRESSURE *
TEMPERATURE^2 *
MOLEVOLUME^2
BWRSA/4
i
fcn(e
i
,V
ci
, T
ci
) x MOLEVOLUME^2
BWRSA/5 b
i
fcn(e
i
,V
ci
, T
ci
) x MOLEVOLUME^2
BWRSA/6 a
i
fcn(e
i
,V
ci
, T
ci
) x PRESSURE * MOLE
VOL^3
BWRSA/7 o
i
fcn(e
i
,V
ci
, T
ci
) x MOLEVOLUME^3
BWRSA/8 c
i
fcn(e
i
,V
ci
, T
ci
) x PRESSURE *
TEMPERATURE^2 *
MOLEVOLUME^3
BWRSA/9 D
0i
fcn(e
i
,V
ci
, T
ci
) x PRESSURE *
TEMPERATURE^3 *
MOLEVOLUME^2
BWRSA/10 d
i
fcn(e
i
,V
ci
, T
ci
) x PRESSURE *
TEMPERATURE * MOLE
VOLUME^3
BWRSA/11 E
0i
fcn(e
i
,V
ci
, T
ci
) x PRESSURE *
TEMPERATURE^4 *
MOLEVOLUME^2
BWRAIJ k
ij
x
Constants Used with the correlations of Han and Starling
Parameter Methane Ethane Propane nButane
B
0i
0.723251 0.826059 0.964762 1.56588
A
0i
7520.29 13439.30 18634.70 32544.70
C
0i
2.71092x10
8
2.95195x10
9
7.96178x10
9
1.37436x10
10
20 2 Thermodynamic Property Models
Parameter Methane Ethane Propane nButane
D
0i
1.07737x10
10
2.57477x10
11
4.53708x10
11
3.33159x10
11
E
0i
3.01122x10
10
1.46819x10
13
2.56053x10
13
2.30902x10
12
b
i
0.925404 3.112060 5.462480 9.140660
a
i
2574.89 22404.50 40066.40 71181.80
d
i
47489.1 702189.0 1.50520x10
7
3.64238x10
7
o
i
0.468828 0.909681 2.014020 4.009850
c
i
4.37222x10
8
6.81826x10
9
2.74461x10
10
7.00044x10
10
i
1.48640 2.99656 4.56182 7.54122
Parameter nPentane nHexane nHeptane nOctane
B
0i
2.44417 2.66233 3.60493 4.86965
A
0i
51108.20 45333.10 77826.90 81690.60
C
0i
2.23931x10
10
5.26067x10
10
6.15662x10
10
9.96546x10
10
D
0i
1.01769x10
12
5.52158x10
12
7.77123x10
12
7.90575x10
12
E
0i
3.90860x10
13
6.26433x10
14
6.36251x10
12
3.46419x10
13
b
i
16.607000 29.498300 27.441500 10.590700
a
i
162185.00 434517.00 359087.00 131646.00
d
i
3.88521x10
7
3.27460x10
7
8351150.0 1.85906x10
8
o
i
7.067020 9.702300 21.878200 34.512400
c
i
1.35286x10
11
3.18412x10
11
3.74876x10
11
6.42053x10
11
i
11.85930 14.87200 24.76040 21.98880
References
M. Benedict, G. B. Webb, and L. C. Rubin, J. Chem. Phys., Vol. 8, (1940), p.
334.
M. S. Han, and K. E. Starling , "Thermo Data Refined for LPG. Part 14:
Mixtures", Hydrocarbon Processing, Vol. 51, No. 5, (1972), p. 129.
K. E. Starling, "Fluid Themodynamic Properties for Light Petroleum Systems",
Gulf Publishing Co., Houston, Texas (1973).
References for Parameter Data
K.E. Starling and M.S. Han, "Thermo data refined for LPG, part 14 Mixtures,"
Hydrocarbon Processing, (May 1972), pp. 129132.
K.E. Starling and M.S. Han, "Thermo data refined for LPG, part 15 Industrial
applications," Hydrocarbon Processing, (June 1972), pp. 107115.
K.E. Starling, "Fluid Thermodynamic Properties for Light Petroleum Systems,"
Gulf Publishing Co., Houston, Texas (1973).
GERG2008 Equation of State
The GERG2008 equationofstate model is the basis for the GERG2008
property method.
2 Thermodynamic Property Models 21
The equation of state is based on a multifluid approximation explicit in the
reduced Helmholtz free energy:
(1)
Where the idealgas contribution o
o
and residual contribution o
r
at a given
mixture density, temperature, and molar composition are:
(2)
(3)
Where the reduced mixture density o and inverse reduced mixture
temperature t are:
(4)
In eq. (2), the idealgas contribution of the reduced Helmholtz free energy for
component i is given by:
(5)
In eq. (3), the pure substance contribution to the residual part of the reduced
Helmholtz free energy for component i is given by:
(6)
In eq. (3), the mixture contribution to the residual part of the reduced
Helmholtz free energy is given by:
(7
)
The reduced mixture density is given by:
(8)
And the reduced mixture temperature is given by:
22 2 Thermodynamic Property Models
(9)
Where:
R = molar gas constant = 8.314472 J/molK.
c,i
and T
c,i
= critical density and critical temperature
n
o
oi,k
and 0
o
oi,k
= Coefficients and parameters of eq. (5) for pure components
n
oi,k
, d
oi,k
, t
oi,k
, and c
oi,k
= coefficients and exponents of eq. (6) for pure
components
F
ij
= Composition dependent factor
n
ij,k
= Coefficients and d
ij,k
, t
ij,k
, q
ij,k
, c
ij,k
, 
ij,k
, and
ij,k
= the exponents in eq.
(7) for all binary specific and generalized departure functions

v,ij
and
v,ij
in eq. (8) and 
T,ij
and
T,ij
in eq. (9) = Binary interaction
parameters
Reference
"The GERG2004 WideRange Equation of State for Natural Gases and Other
Mixtures" O. Kunz, R. Klimeck, W. Wagner, M. Jaeschke; GERG TM15 2007;
ISBN 9783183557066; Published for GERG and printed in Germany by
VDI Verlag GmbH (2007).
Kunz, O., Wagner, W., "The new GERG2004 XT08 widerange equation of
state for natural gases and other mixtures." To be submitted to Fluid Phase
Equilibria (beginning of 2010).
HaydenO'Connell
The HaydenO'Connell equationofstate calculates thermodynamic properties
for the vapor phase. It is used in property methods NRTLHOC, UNIFHOC,
UNIQHOC, VANLHOC, and WILSHOC, and is recommended for nonpolar,
polar, and associating compounds. HaydenO'Connell incorporates the
chemical theory of dimerization. This model accounts for strong association
and solvation effects, including those found in systems containing organic
acids, such as acetic acid. The equationofstate is:
Where:
 For nonpolar, nonassociating species:
, with
2 Thermodynamic Property Models 23
, where
 For polar, associating species:
, with
, where
 For chemically bonding species:
, and
CrossInteractions
The previous equations are valid for dimerization and crossdimerization if
these mixing rules are applied:
q = 0 unless a special solvation contribution can be justified (for example, i
and j are in the same class of compounds). Many q values are present in the
Aspen Physical Property System.
Chemical Theory
When a compound with strong association is present in a mixture,
the entire mixture is treated according to the chemical theory of dimerization.
24 2 Thermodynamic Property Models
The chemical reaction for the general case of a mixture of dimerizing
components i and j is:
Where i and j refer to the same component.
The equationofstate becomes:
with
In this case, molar volume is equal to V/n
t
.
This represents true total volume over the true number of species n
t
.
However, the reported molar volume is V/n
a
.
This represents the true total volume over the apparent number of species n
a
.
If dimerization does not occur, n
a
is defined as the number of species. V/n
a
reflects the apparently lower molar volume of an associating gas mixture.
The chemical equilibrium constant for the dimerization reaction on pressure
basis K
p
, is related to the true mole fractions and fugacity coefficients:
Where:
y
i
and y
j
= True mole fractions of monomers
y
ij
= True mole fraction of dimer
i
= True fugacity coefficient of component i
K
ij
= Equilibrium constant for the dimerization of i and j, on a
pressure basis
=
o
ij
= 1 for i=j
=
0 for
Apparent mole fractions y
i
a
are reported, but in the calculation real mole
fractions y
i
, y
j
, and y
ij
are used.
The heat of reaction due to each dimerization is calculated according to:
The sum of the contributions of all dimerization reactions, corrected for the
ratio of apparent and true number of moles is added to the molar enthalpy
departure .
2 Thermodynamic Property Models 25
Parameter Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
RGYR r
i
gyr
10
11
5x10
9
LENGTH
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
HOCETA q 0.0 X
The binary parameters HOCETA for many component pairs are available in the
Aspen Physical Property System. These parameters are retrieved
automatically when you specify any of the following property methods: NRTL
HOC, UNIFHOC, UNIQHOC, VANLHOC, and WILSHOC.
References
J.G. Hayden and J.P. O'Connell, "A Generalized Method for Predicting Second
Virial Coefficients," Ind. Eng. Chem., Process Des. Dev., Vol. 14,No. 3,
(1975), pp. 209 216.
HF EquationofState
HF forms oligomers in the vapor phase. The nonideality in the vapor phase is
found in important deviations from ideality in all thermodynamic properties.
The HF equation accounts for the vapor phase nonidealities. The model is
based on chemical theory and assumes the formation of hexamers.
Species like HF that associate linearly behave as single species. For example,
they have a vapor pressure curve, like pure components. The component on
which a hypothetical unreacted system is based is often called the apparent
(or parent) component. Apparent components react to the true species.
Electrolyte Calculation in Physical Property Methods discusses apparent and
true species. Abbott and van Ness (1992) provide details and basic
thermodynamics of reactive systems.
The temperaturedependent hexamerization equilibrium constant, can fit the
experimentally determined association factors. The builtin functionality is:
(1)
The constants C
0
and C
1
are from Long et al. (1943), and C
2
and C
3
are set to
0. The correlation is valid between 270 and 330 K, and can be extrapolated to
about 370 K (cf. sec. 4). Different sets of constants can be determined by
experimental data regression.
Molar Volume Calculation
The nonideality of HF is often expressed using the association factor, f,
indicating the ratio of apparent number of species to the real number or
species. Assuming the ideal gas law for all true species in terms of (p, V, T)
behavior implies:
26 2 Thermodynamic Property Models
(2)
Where the true number of species is given by 1/f. The association factor is
easily determined from (p, V, T) experiments. For a critical evaluation of data
refer to Vanderzee and Rodenburg (1970).
If only one reaction is assumed for a mixture of HF and its associated species,
(refer to Long et al., 1943), then:
(3)
If p
1
represents the true partial pressure of the HF monomer, and p
6
represents the true partial pressure of the hexamer, then the equilibrium
constant is defined as:
(4)
The true total pressure is:
p = p
1
+ p
6
(5)
If all hexamer were dissociated, the apparent total pressure would be the
hypothetical pressure where:
p
a
= p
1
+ 6p
6
= p + 5p
6
(6)
When physical ideality is assumed, partial pressures and mole fractions are
proportional. The total pressure in equation 5 represents the true number of
species. The apparent total pressure from equation 6 represents the apparent
number of species:
(7)
Note that the outcome of equation 7 is independent of the assumption of
ideality. Equation 7 can be used to compute the number of true species 1/f
for a mixture containing HF, but the association factor is defined differently.
If p
1
and p
6
are known, the molar volume or density of a vapor containing HF
can be calculated using equations 2 and 7. The molar volume calculated is the
true molar volume for 1 apparent mole of HF. This is because the volume of 1
mole of ideal gas (the true molar volume per true number of moles) is always
equal to about 0.0224 m3/mol at 298.15 K.
True Mole Fraction (Partial Pressure) Calculation
If you assume the ideal gas law for a mixture containing HF, the apparent HF
mole fraction is:
(8)
2 Thermodynamic Property Models 27
The denominator of equation 8 is given by equation 6. The numerator (the
apparent partial pressure of HF) is the hypothetical partial pressure only if all
of the hexamer was dissociated. If you substitute equation 4, then equation 8
becomes:
(9)
K is known from Long et al., or can be regressed from (p,V,T) data. The
apparent mole fraction of HF, y
a
, is known to the user and the simulator, but
p
1
, or y = p
1
/p must also be known in order to calculate the thermodynamic
properties of the mixture. Equation 9 must be solved for p
1
.
Equation 9 can be written as a polynomial in p
1
of degree 6:
K(6  5y
a
)(p
1
)
6
+ p
1
 py
a
= 0 (9a)
A second order NewtonRaphson technique is used to determine p
1
. Then p
6
can be calculated by equation 5, and f is known (equation 7).
Gibbs Energy and Fugacity
The apparent fugacity coefficient is related to the true fugacity coefficient and
mole fractions:
(10)
Equation 10 represents a correction to the ideal mixing term of the fugacity.
The ratio of the true number of species to the apparent number of species is
similar to the correction applied in equation 2. Since the ideal gas law is
assumed, the apparent fugacity coefficient is given by the equation. All
variables on the right side are known.
(11)
For pure HF, y
a
= 1:
From the fugacity coefficient, the Gibbs energy departure of the mixture or
pure apparent components can be calculated:
(12)
(12a)
Enthalpy and Entropy
For the enthalpy departure, the heat of reaction is considered. For an
arbitrary gas phase reaction:
28 2 Thermodynamic Property Models
(13)
(14)
Where
i
*
is the pure component thermodynamic potential or molar Gibbs
energy of a component. Equation 4 represents the first two terms of the
general equation 14. The second or third equality relates the equilibrium
constant to the Gibbs energy of reaction, which is thus related to the enthalpy
of reaction:
(15)
All components are assumed to be ideal. The enthalpy departure is equal to
the heat of reaction, per apparent number of moles:
(16)
(17)
From the Gibbs energy departure and enthalpy departure, the entropy
departure can be calculated:
(18)
Temperature derivatives for the thermodynamic properties can be obtained
by straightforward differentiation.
Usage
The HF equationofstate should only be used for vapor phase calculations. It
is not suited for liquid phase calculations.
The HF equationofstate can be used with any activity coefficient model for
nonelectrolyte VLE. Using the Electrolyte NRTL model and the data package
MHF2 is strongly recommended for aqueous mixtures (de Leeuw and
Watanasiri, 1993).
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
ESHFK/1 C
0
43.65
ESHFK/2 C
1
8910
ESHFK/3 C
2
0
ESHFK/4 C
3
0
References
M. M. Abbott and H. C. van Ness, "Thermodynamics of Solutions Containing
Reactive Species, a Guide to Fundamentals and Applications," Fluid Phase Eq.,
Vol. 77, (1992) pp. 53 119.
2 Thermodynamic Property Models 29
V. V. De Leeuw and S. Watanasiri, "Modelling Phase Equilibria and Enthalpies
of the System Water and Hydroflouric Acid Using an HF Equationofstate in
Conjunction with the Electrolyte NRTL Activity Coefficient Model," Paper
presented at the 13th European Seminar on Applied Thermodynamics, June 9
12, CarryleRouet, France, 1993.
R. W. Long, J. H. Hildebrand, and W. E. Morrell, "The Polymerization of
Gaseous Hydrogen and Deuterium Flourides," J. Am. Chem. Soc., Vol. 65,
(1943), pp. 182 187.
C. E. Vanderzee and W. WM. Rodenburg, "Gas Imperfections and
Thermodynamic Excess Properties of Gaseous Hydrogen Fluoride," J. Chem.
Thermodynamics, Vol. 2, (1970), pp. 461 478.
IAPWS95 Steam Tables
The IAPWS95 Steam Tables are implemented like any other equationof
state in the Aspen Physical Property System. These steam tables can calculate
any thermodynamic property of water. The tables form the basis of the
IAPWS95 property method. There are no parameter requirements. They are
the most accurate steam tables in the Aspen Physical Property System.
References
Wanger W. and A. Pru, The IAPWS Formation 1995 for the Thermodynamic
Properties of Ordinary Water Substance for General and Scientific Use, J.
Phys. Chem. Ref. Data, 31(2), 387 535, 2002.
Ideal Gas
The ideal gas law (ideal gas equationofstate) combines the laws of Boyle
and GayLussac. It models a vapor as if it consisted of point masses without
any interactions. The ideal gas law is used as a reference state for equation
ofstate calculations, and can be used to model gas mixtures at low pressures
(without specific gas phase interactions).
The equation is:
p = RT / V
m
LeeKesler
This equationofstate model is based on the work of Lee and Kesler (1975).
In this equation, the volumetric and thermodynamic properties of fluids based
on the Curl and Pitzer approach (1958) have been analytically represented by
a modified BenedictWebbRubin equationofstate (1940). The model
calculates the molar volume, enthalpy departure, Gibbs free energy
departure, and entropy departure of a mixture at a given temperature,
pressure, and composition for either a vapor or a liquid phase. Partial
derivatives of these quantities with respect to temperature can also be
calculated.
Unlike the other equationofstate models, this model does not calculate
fugacity coefficients.
30 2 Thermodynamic Property Models
The compressibility factor and other derived thermodynamic functions of
nonpolar and slightly polar fluids can be adequately represented, at constant
reduced temperature and pressure, by a linear function of the acentric factor.
In particular, the compressibility factor of a fluid whose acentric factor is e, is
given by the following equation:
Z = Z
(0)
+ eZ
(1)
Where:
Z
(0)
= Compressibility factor of a simple fluid (e = 0)
Z
(1)
= Deviation of the compressibility factor of the real fluid from Z
(0)
Z
(0)
and Z
(1)
are assumed universal functions of the reduced temperature and
pressure.
Curl and Pitzer (1958) were quite successful in correlating thermodynamic
and volumetric properties using the above approach. Their application
employed tables of properties in terms of reduced temperature and pressure.
A significant weakness of this method is that the various properties (for
example, entropy departure and enthalpy departure) will not be exactly
thermodynamically consistent with each other. Lee and Kesler (1975)
overcame this drawback by an analytic representation of the tables with an
equationofstate. In addition, the range was extended by including new data.
In the LeeKesler implementation, the compressibility factor of any fluid has
been written in terms of a simple fluid and a reference as follows:
In the above equation both Z
(0)
and Z
(1)
are represented as generalized
equations of the BWR form in terms of reduced temperature and pressure.
Thus,
Equations for the enthalpy departure, Gibbs free energy departure, and
entropy departure are obtained from the compressibility factor using standard
thermodynamic relationships, thus ensuring thermodynamic consistency.
In the case of mixtures, mixing rules (without any binary parameters) are
used to obtain the mixture values of the critical temperature and pressure,
and the acentric factor.
This equation has been found to provide a good description of the volumetric
and thermodynamic properties of mixtures containing nonpolar and slightly
polar components.
Symbol Parameter Name Default Definition
T
c
TCLK TC Critical temperature
P
c
PCLK PC Critical pressure
e OMGLK OMEGA Acentric factor
2 Thermodynamic Property Models 31
References
M. Benedict, G. B. Webb, and L. C. Rubin, J. Chem. Phys., Vol. 8, (1940), p.
334.
R. F. Curl and K.S. Pitzer, Ind. Eng. Chem., Vol. 50, (1958), p. 265.
B. I. Lee and M.G. Kesler, AIChE J., Vol. 21, (1975), p. 510.
LeeKeslerPlcker
The LeeKeslerPlcker equationofstate is the basis for the LKPLOCK
property method. This equationofstate applies to hydrocarbon systems that
include the common light gases, such as H
2
S and CO
2
. It can be used in gas
processing, refinery, and petrochemical applications.
The general form of the equation is:
Where:
The f
o
and f
R
parameters are functions of the BWR form. The f
o
parameter is
for a simple fluid, and f
R
is for reference fluid noctane.
The mixing rules are:
V
cm
=
=
e =
Z
m
=
Where:
V
cij
=
T
cij
=
Z
ci
=
k
ij
= k
ji
32 2 Thermodynamic Property Models
The binary parameter k
ij
is determined from phaseequilibrium data
regression, such as VLE data. The Aspen Physical Property System stores the
binary parameters for a large number of component pairs. These binary
parameters are used automatically with the LKPLOCK property method. If
binary parameters for certain component pairs are not available, they can be
estimated using builtin correlations. The correlations are designed for binary
interactions among the components CO, CO
2
, N
2
, H
2
, CH
4
alcohols and
hydrocarbons. If a component is not CO, CO
2
, N
2
, H
2
, CH
4
or an alcohol, it is
assumed to be a hydrocarbon.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCLKP T
ci
TC x 5.0 2000.0 TEMPERATURE
PCLKP p
ci
PC x PRESSURE
VCLKP V
ci
VC x 0.001 3.5 MOLE
VOLUME
OMGLKP e
I
OMEGA x 0.5 2.0
LKPZC
Z
ci
Method 1: fcn(e)
Method 2:
fcn(p
ci,
V
ci,
T
ci
)
x 0.1 0.5
LKPKIJ
k
ij
fcn(T
ci
V
ci
/ T
cj
V
cj
) x 5.0 5.0
Method 1 is the default for LKPZC; Method 2 can be invoked by setting the
value of LKPZC equal to zero.
Binary interaction parameters LKPKIJ are available for a large number of
components in the Aspen Physical Property System, from Knapp et al.
References
B.I. Lee and M.G. Kesler, AIChE J., Vol. 21, (1975) p. 510; errata: AIChE J.,
Vol. 21, (1975) p. 1040.
V. Plcker, H. Knapp, and J.M. Prausnitz, Ind. Eng. Chem., Process Des. Dev.,
Vol. 17, (1978), p. 324.
H. Knapp, R. Dring, L. Oellrich, U. Plcker, and J. M. Prausnitz. "VaporLiquid
Equilibria for Mixtures of Low Boiling Substances." Dechema Chemistry Data
Series, Vol. VI.
NBS/NRC Steam Tables
The NBS/NRC Steam Tables are implemented like any other equationofstate
in the Aspen Physical Property System. These steam tables can calculate any
thermodynamic property of water. The tables form the basis of the
STEAMNBS and STMNBS2 property methods. There are no parameter
requirements. The STMNBS2 model uses the same equations as STEAMNBS
but with different root search method. The STEAMNBS method is
recommended for use with the SRK, BWRS, MXBONNEL and GRAYSON2
property methods.
2 Thermodynamic Property Models 33
References
L. Haar, J.S. Gallagher, and J.H. Kell, "NBS/NRC Steam Tables," (Washington:
Hemisphere Publishing Corporation, 1984).
Nothnagel
The Nothnagel equationofstate calculates thermodynamic properties for the
vapor phase. It is used in property methods NRTLNTH, UNIQNTH, VANL
NTH, and WILSNTH. It is recommended for systems that exhibit strong vapor
phase association. The model incorporates the chemical theory of
dimerization to account for strong association and solvation effects, such as
those found in organic acids, like acetic acid. The equationofstate is:
Where:
b =
b
ij
=
nc = Number of components in the mixture
The chemical reaction for the general case of a mixture of dimerizing
components i and j is:
The chemical equilibrium constant for the dimerization reaction on pressure
basis K
p
is related to the true mole fractions and fugacity coefficients:
Where:
y
i
and y
j
= True mole fractions of monomers
y
ij
= True mole fraction of dimer
i
= True fugacity coefficient of component i
K
ij
= Equilibrium constant for the dimerization of i and j, on a
pressure basis
When accounting for chemical reactions, the number of true species n
t
in the
mixture changes. The true molar volume V/n
t
is calculated from the
equationofstate. Since both V and n
t
change in about the same proportion,
this number does not change much. However, the reported molar volume is
the total volume over the apparent number of species: V/n
a
. Since the
apparent number of species is constant and the total volume decreases with
34 2 Thermodynamic Property Models
association, the quantity V/n
a
reflects the apparent contraction in an
associating mixture.
The heat of reaction due to each dimerization can be calculated:
The heat of reaction for the mixed dimerization of components i and j is by
default the arithmetic mean of the heats of reaction for the dimerizations of
the individual components. Parameter is a small empirical correction
factor to this value:
The sum of the contributions of all dimerization reactions, corrected for the
ratio of apparent and true number of moles, is added to the molar enthalpy
departure:
The equilibrium constants can be computed using either builtin calculations
or parameters you entered.
 Builtin correlations:
The pure component parameters b, d, and p are stored in the Aspen
Physical Property System for many components.
Parameters you entered:
In this method, you enter A
i
, B
i
, C
i
, and D
i
on the Methods  Parameters 
Pure Component  TDependent form. The units for K
ii
is pressure
1
; use
absolute units for temperature. If you enter K
ii
and K
jj
, then K
ij
is computed
from
If you enter A
i
, B
i
, C
i
, and D
i
, the equilibrium constants are computed using
the parameters you entered. Otherwise the equilibrium constants are
computed using builtin correlations.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper Limit Units
TC T
ci
5.0 2000.0 TEMPERATURE
TB T
bi
4.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
NTHA/1 b
i
0.199 RT
ci
/ p
ci
0.01 1.0 MOLEVOLUME
NTHA/2 d
i
0.33 0.01 3.0
NTHA/3 p
i
0 0.0 1.0
NTHK/1 A
i
PRESSURE
NTHK/2 B
i
0 TEMPERATURE
NTHK/3 C
i
0 TEMPERATURE
2 Thermodynamic Property Models 35
Parameter
Name/Element
Symbol Default Lower
Limit
Upper Limit Units
NTHK/4 D
i
0 TEMPERATURE
NTHDDH 0
MOLE
ENTHALPY
For the following systems, the values given in Nothnagel et al., 1973 are
used by default:
 Methyl chloride/acetone
 Acetonitrile/acetaldehyde
 Acetone/chloroform
 Chloroform/diethyl amine
 Acetone/benzene
 Benzene/chloroform
 Chloroform/diethyl ether
 Chloroform/propyl formate
 Chloroform/ethyl acetate
 Chloroform/methyl acetate
 Chloroform/methyl formate
 Acetone/dichloro methane
 nButane/nperfluorobutane
 nPentane/nperfluoropentane
 nPentane/nperfluorohexane
References
K.H. Nothnagel, D. S. Abrams, and J.M. Prausnitz, "Generalized Correlation
for Fugacity Coefficients in Mixtures at Moderate Pressures," Ind. Eng. Chem.,
Process Des. Dev., Vol. 12, No. 1 (1973), pp. 25 35.
Copolymer PCSAFT EOS Model
This section describes the Copolymer PerturbedChain Statistical Associating
Fluid Theory (PCSAFT). This equationofstate model is used through the PC
SAFT property method.
The copolymer PCSAFT represents the completed PCSAFT EOS model
developed by Sadowski and coworkers (Gross and Sadowski, 2001, 2002a,
2002b; Gross et al., 2003; Becker et al., 2004; Kleiner et al., 2006). Unlike
the PCSAFT EOS model (POLYPCSF) in Aspen Plus, the copolymer PCSAFT
includes the association and polar terms and does not apply mixing rules to
calculate the copolymer parameters from its segments. Its applicability covers
fluid systems from small to large molecules, including normal fluids, water,
alcohols, and ketones, polymers and copolymers and their mixtures.
36 2 Thermodynamic Property Models
References
Gross, J., & Sadowski, G. (2001). PerturbedChain SAFT: An equation of state
based on a perturbation theory for chain molecules. Ind. Eng. Chem. Res.,
40, 12441260.
Gross, J., & Sadowski, G. (2002a). Modeling polymer systems using the
perturbedchain statistical associating fluid theory equation of state. Ind. Eng.
Chem. Res., 41, 10841093.
Gross, J., & Sadowski, G. (2002b). Application of the PerturbedChain SAFT
Equation of State to Associating Systems. Ind. Eng. Chem. Res., 41, 5510
5515.
Gross, J., Spuhl, O., Tumakaka, F., & Sadowski, G. (2003). Modeling
Copolymer Systems Using the PerturbedChain SAFT Equation of State. Ind.
Eng. Chem. Res., 42, 12661274.
Becker, F., Buback, M., Latz, H., Sadowski, G., & Tumakaka, F. (2004).
CloudPoint Curves of Ethylene(Meth)acrylate Copolymers in Fluid Ethene up
to High Pressures and Temperatures Experimental Study and PCSAFT
Modeling. Fluid Phase Equilibria, 215, 263282.
Kleiner, M., Tumakaka, F., Sadowski, G., Latz, H., & Buback, M. (2006).Phase
Equilibria in Polydisperse and Associating Copolymer Solutions: Poly(ethane
co(meth)acrylic acid) Monomer Mixtures. Fluid Phase Equilibria, 241, 113
123..
Copolymer PCSAFT Fundamental Equations
The copolymer PCSAFT model is based on the perturbation theory. The
underlying idea is to divide the total intermolecular forces into repulsive and
attractive contributions. The model uses a hardchain reference system to
account for the repulsive interactions. The attractive forces are further divided
into different contributions, including dispersion, polar, and association.
Using a generated function, , the copolymer PCSAFT model in general can
be written as follows:
where
hc
,
disp
,
assoc
, and
polar
are contributions due to hardchain fluids,
dispersion, association, and polarity, respectively.
The generated function is defined as follows:
where a
res
is the molar residual Helmholtz energy of mixtures, R is the gas
constant, T is the temperature, is the molar density, and Z
m
is the
compressibility factor; a
res
is defined as:
2 Thermodynamic Property Models 37
where a is the Helmholtz energy of a mixture and a
ig
is the Helmholtz energy
of a mixture of ideal gases at the same temperature, density and composition
x
i
. Once is known, any other thermodynamic function of interest can be
easily derived. For instance, the fugacity coefficient
i
is calculated as follows:
with
where is a partial derivative that is always done to the mole fraction
stated in the denominator, while all other mole fractions are considered
constant.
Applying to the departure equations, departure functions of enthalpy,
entropy, and Gibbs free energy can be obtained as follows:
Enthalpy departure:
Entropy departure:
Gibbs free energy departure:
The following thermodynamic conditions must be satisfied:
Hardchain Fluids and Chain Connectivity
In PCSAFT model, a molecule is modeled as a chain molecule by a series of
freelyjointed tangent spheres. The contribution from hardchain fluids as a
reference system consists of two parts, a nonbonding contribution (i.e., hard
38 2 Thermodynamic Property Models
sphere mixtures prior to bonding to form chains) and a bonding contribution
due to chain formation:
where is the mean segment in the mixture,
hs
is the contribution from
hardsphere mixtures on a persegment basis, and
chain
is the contribution
due to chain formation. Both and
hs
are welldefined for mixtures
containing polymers, including copolymers; they are given by the following
equations:
where m
io
, o
io
, and c
io
are the segment number, the segment diameter, and
the segment energy parameter of the segment type o in the copolymer
component i, respectively. The segment number m
io
is calculated from the
segment ratio parameter r
io
:
where M
io
is the total molecular weight of the segment type o in the
copolymer component i and can be calculated from the segment weight
fraction within the copolymer:
where w
io
is the weight fraction of the segment type o in the copolymer
component i, and M
i
is the molecular weight of the copolymer component i.
Following Sadowski and coworkers work ( Gross et al., 2003; Becker et al.,
2004), the contribution from the chain connectivity can be written as follows:
2 Thermodynamic Property Models 39
with
where B
io,i
is defined as the bonding fraction between the segment type o
and the segment type  within the copolymer component i, is the number of
the segment types within the copolymer component i, and g
hs
io,j
(d
io,j
) is the
radial distribution function of hardsphere mixtures at contact.
However, the calculation for B
io,i
depends on the type of copolymers. We
start with a pure copolymer system which consists of only two different types
of segments o and ; this gives:
with
We now apply these equations to three common types of copolymers; a)
alternating, b) block, and c) random.
For an alternating copolymer, m
o
= m

; there are no oo or  adjacent
sequences. Therefore:
For a block copolymer, there is only one o pair and the number of oo and
 pairs depend on the length of each block; therefore:
For a random copolymer, the sequence is only known in a statistical sense. If
the sequence is completely random, then the number of o adjacent pairs is
proportional to the product of the probabilities of finding a segment of type o
and a segment of type  in the copolymer. The probability of finding a
segment of type o is the fraction of segments z
o
in the copolymer:
The bonding fraction of each pair of types can be written as follows:
40 2 Thermodynamic Property Models
where C is a constant and can be determined by the normalization condition
set by Equation 2.70; the value for C is unity. Therefore:
A special case is the Sadowskis model for random copolymer with two types
of segments only ( Gross et al., 2003; Becker et al., 2004). In this model, the
bonding fractions are calculated as follows:
When z

< z
o
When z
o
< z

The generalization of three common types of copolymers from two types of
different segments to multi types of different segments within a copolymer
is straightforward.
For a generalized alternative copolymer, m
o
= m

= ... = m
r
= m/ ; there
are no adjacent sequences for the same type of segments. Therefore,
For a generalized block copolymer, there is only one pair for each adjacent
type of segment pairs (o = ) and the number of pairs for a same type
depends on the length of the block; therefore:
2 Thermodynamic Property Models 41
For a generalized random copolymer, the sequence is only known in a
statistical sense. If the sequence is completely random, then the number of
o adjacent pairs is proportional to the product of the probabilities of finding
a segment of type o and a segment of type  in the copolymer. The
probability of finding a segment of type o is the fraction of o segments z
o
in
the copolymer:
The bonding fraction of each pair of types can be written as follows:
where C is a constant and can be determined by the normalization condition.
Therefore,
That is,
Put C into the equation above, we obtain:
Copolymer PCSAFT Dispersion Term
The equations for the dispersion term are given as follows:
42 2 Thermodynamic Property Models
where o
io,j
and c
io,j
are the cross segment diameter and energy parameters,
respectively; only one adjustable binary interaction parameter, k
io,j
is
introduced to calculate them:
In above equations, the model constants a
1i
, a
2i
, a
3i
, b
1i
, b
2i
, and b
3i
are fitted
to purecomponent vapor pressure and liquid density data of nalkanes (
Gross and Sadowski, 2001).
Association Term for Copolymer Mixtures  2B
Model
The association term in PCSAFT model in general needs an iterative
procedure to calculate the fraction of a species (solvent or segment) that are
bounded to each associationsite type. Only in pure or binary systems, the
fraction can be derived explicitly for some specific models. We start with
general expressions for the association contribution for copolymer systems as
follows:
2 Thermodynamic Property Models 43
where A is the associationsite type index, is the associationsite
number of the associationsite type A on the segment type o in the
copolymer component i, and is the mole fraction of the segment type o
in the copolymer component i that are not bonded with the associationsite
type A; it can be estimated as follows:
with
where is the cross effective association volume and is the
cross association energy; they are estimated via simple combination rules:
where and are the effective association volume and the
association energy between the associationsite types A and B, of the
segment type o in the copolymer component i, respectively.
The associationsite number of the site type A on the segment type o in the
copolymer component i is equal to the number of the segment type o in the
copolymer component i,
where N
io
is the number of the segment type o in the copolymer component i
and M
o
is the molecular weight of the segment type o. In other words, the
associationsite number for each site type within a segment is the same;
therefore, we can rewrite the equations above as follows:
44 2 Thermodynamic Property Models
To calculate , this equation has to be solved iteratively for each
associationsite type associated with a species in a component. In practice,
further assumption is needed for efficiency. The commonly used model is the
socalled 2B model ( Huang and Radosz, 1990; Gross and Sadowski, 2002b).
It assumes that an associating species (solvent or segment) has two
association sites, one is designed as the site type A and another as the site
type B. Similarly to the hydrogen bonding, type A treats as a donor site with
positive charge and type B as an acceptor site with negative charge; only the
donoracceptor association bonding is permitted and this concept applies to
both pure systems (selfassociation such as water) and mixtures (both self
association and crossassociation such as watermethanol). Therefore, we can
rewrite these equations as follows:
It is easy to show that
Therefore
References
Huang, S. H., & Radosz, M. (1990). Equation of State for Small, Large,
Polydisperse, and Associating Molecules. Ind. Eng. Chem. Res., 29, 2284.
Gross, J., & Sadowski, G. (2002b). Application of the PerturbedChain SAFT
Equation of State to Associating Systems. Ind. Eng. Chem. Res., 41, 5510
5515.
Polar Term for Copolymer PCSAFT
The equations for the polar term are given by Jog et al. (2001) as follows:
2 Thermodynamic Property Models 45
In the above equations, I
2
(q) and I
3
(q) are the pure fluid integrals and
io
and (x
p
)
io
are the dipole moment and dipolar fraction of the segment type o
within the copolymer component i, respectively. Both and
are dimensionless. In terms of them, we can have:
Rushbrooke et al. (1973) have shown that
Then I
2
(q) and I
3
(q) are computed in terms of q by the expressions:
46 2 Thermodynamic Property Models
Reference
Jog, P. K., Sauer, S. G., Blaesing, J., & Chapman, W. G. (2001), Application
of Dipolar Chain Theory to the Phase Behavior of Polar Fluids and Mixtures.
Ind. Eng. Chem. Res., 40, 4641.
Rushbrooke, G. S., & Stell, G., Hoye, J. S. (1973), Molec. Phys., 26, 1199.
Ominik, A., Chapman, W.G., Kleiner, M., & Sadowski, G. (2005), Modeling of
Polar Systems with the PerturbedChain SAFT Equation of State. Investigation
of the Performance of Two Polar Terms. Ind. Eng. Chem. Res., 44, 6928
Sauer, S.G., & Chapman (2003), A Parametric Study if Dipolar Chain Theory
with Applications to Ketone Mixtures. Ind. Eng. Chem. Res., 42, 5687
Copolymer PCSAFT EOS Model Parameters
Pure parameters
Each nonassociation species (solvent or segment) must have a set of three
purecomponent parameter; two of them are the segment diameter o and the
segment energy parameter c. The third parameter for a solvent is the
segment number m and for a segment is the segment ratio parameter r. For
an association species, two additional parameters are the effective association
volume k
(AB)
and the association energy c
(AB)
. For a polar species, two
additional parameters are the dipole moment and the segment dipolar
fraction x
p
.
Binary parameters
There are three types of binary interactions in copolymer systems: solvent
solvent, solventsegment, and segmentsegment. The binary interaction
parameter k
io,j
allows complex temperature dependence:
with
where T
ref
is a reference temperature and the default value is 298.15 K.
The following table lists the copolymer PCSAFT EOS model parameters
implemented in Aspen Plus. Some values for these parameters can be found
in the PCSAFT and POLYPCSF databanks.
Parameter
Name/
Element
Symbol Default Lower
Limit
Upper
Limit
MDS Units
Keyword
Comments
PCSFTM m X Unary
PCSFTV o X Unary
PCSFTU c/k X TEMP Unary
2 Thermodynamic Property Models 47
Parameter
Name/
Element
Symbol Default Lower
Limit
Upper
Limit
MDS Units
Keyword
Comments
PCSFTR r X Unary
PCSFAU
c
AB
/k
X TEMP Unary
PCSFAV
k
AB
X Unary
PCSFMU
X Unary
PCSFXP x
p
X Unary
PCSKIJ/1
a
io,j
0.0 X Binary,
Symmetric
PCSKIJ/2
b
io,j
0.0 X Binary,
Symmetric
PCSKIJ/3
c
io,j
0.0 X Binary,
Symmetric
PCSKIJ/4
d
io,j
0.0 X Binary,
Symmetric
PCSKIJ/5
e
io,j
0.0 X Binary,
Symmetric
PCSKIJ/6 T
ref
298.15 X TEMP Binary,
Symmetric
Parameter Input and Regression for Copolymer
PCSAFT
Since the copolymer PCSAFT is built based on the segment concept, the
unary (pure) parameters must be specified for a solvent or a segment.
Specifying a unary parameter for a polymer component (homopolymer or
copolymer) will be ignored by the simulation. For a nonassociation and non
polar solvent, three unary parameters PCSFTM, PCSFTU, and PCSFTV must be
specified. For a nonassociation and nonpolar segment, these three unary
parameters PCSFTR, PCSFTU, and PCSFTV must be specified. For an
association species (solvent or segment), two additional unary parameters
PCSFAU and PCSFAV must be specified. For a polar species (solvent or
segment), two additional unary parameters PCSFMU and PCSFXP must be
specified.
Note that the SI units for the segment diameter o (PCSFTV) and dipole
moment (PCSFMU) are much too large to be practical. The implementation
of PCSAFT in Aspen Plus has the unit in Angstroms () for the segment
diameter and in Debye (D) for the dipole moment; these units are not allowed
to be changed in PCSAFT.
The binary parameter PCSKIJ can be specified for each solventsolvent pair,
or each solventsegment pair, or each segmentsegment pair. By default, the
binary parameter is set to be zero.
A databank called PCSAFT contains both unary and binary PCSAFT
parameters available from literature; it must be used with the PCSAFT
property method. The unary parameters available for segments are stored in
48 2 Thermodynamic Property Models
the SEGMENT databank. If unary parameters are not available for a species
(solvent or segment) in a calculation, the user can perform an Aspen Plus
Data Regression Run (DRS) to obtain unary parameters. For nonpolymer
components (mainly solvents), the unary parameters are usually obtained by
fitting experimental vapor pressure and liquid molar volume data. To obtain
unary parameters for a segment, experimental data on liquid density of the
homopolymer that is built by the segment should be regressed. Once the
unary parameters are available for a segment, the idealgas heat capacity
parameter CPIG may be regressed for the same segment using experimental
liquid heat capacity data for the same homopolymer. In addition to unary
parameters, the binary parameter PCSKIJ for each solventsolvent pair, or
each solventsegment pair, or each segmentsegment pair, can be regressed
using vaporliquid equilibrium (VLE) data in the form of TPXY data in Aspen
Plus.
Note: In Data Regression Run, a homopolymer must be defined as an
OLIGOMER type, and the number of the segment that builds the oligomer
must be specified.
PengRobinson
The PengRobinson equationofstate is the basis for the PRBM property
method. The model has been implemented with choices of different alpha
functions (see PengRobinson Alpha Functions) and has been extended to
include advanced asymmetric mixing rules.
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
By default, the PENGROB property method uses the literature version of the
alpha function and mixing rules (see Standard PengRobinson). The PRBM
property method uses the BostonMathias alpha function and standard mixing
rules. These default property methods are recommended for hydrocarbon
processing applications such as gas processing, refinery, and petrochemical
processes. Their results are comparable to those of the property methods that
use the standard RedlichKwongSoave equationofstate.
When advanced alpha function and asymmetric mixing rules are used with
appropriately obtained parameters, the PengRobinson model can be used to
accurately model polar, nonideal chemical systems. Similar capability is also
available for the SoaveRedlichKwong model.
The equation for the PengRobinson model as used in the PRBM property
method is:
Where:
b =
a = a
0
+a
1
2 Thermodynamic Property Models 49
a
0
=
(the standard quadratic mixing term, where k
ij
has
been made temperaturedependent)
k
ij
=
k
ij
= k
ji
a
1
(an additional, asymmetric term used to model
highly nonlinear systems)
l
ij
=
In general, .
a
i
=
b
i
=
If the Peneloux volume correction is used (option code 4), then the molar
volume is calculated from:
V = V
m
 c
Where:
V
m
= Molar volume calculated by the equation of state
without the correction
c =
(the Peneloux volume correction term)
c
i
=
(the Peneloux volume correction term for pure
components, calculated from the critical
temperature and pressure and the Rackett
parameter)
For best results, the binary parameters k
ij
and l
ij
must be determined from
regression of phase equilibrium data such as VLE data. The Aspen Physical
Property System also has builtin k
ij
and l
ij
for a large number of component
pairs in the EOSLIT databank from Knapp et al. These parameters are used
automatically with the PENGROB property method. Values in the databank
can be different than those used with other models such as SoaveRedlich
Kwong or RedlichKwongSoave, and this can produce different results.
The model has option codes which can be used to customize the model, by
selecting a different alpha function and other model options. See Peng
Robinson Alpha Functions for a description of the alpha functions. See Option
50 2 Thermodynamic Property Models
Codes for Equation of State Models (under ESPR and ESPR0) for a list of the
option codes.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
PRTC T
ci
TC x 5.0 2000.0 TEMPERATURE
PRPC p
ci
PC x 10
5
10
8
PRESSURE
OMGPR
e
i
OMEGA x 0.5 2.0
PRZRA z
RA
RKTZRA x
PRKBV/1 k
ij
(1)
0 x
PRKBV/2 k
ij
(2)
0 x TEMPERATURE
PRKBV/3 k
ij
(3)
0 x TEMPERATURE
PRKBV/4 T
lower
0 x TEMPERATURE
PRKBV/5 T
upper
1000 x TEMPERATURE
PRLIJ/1 l
ij
(1)
0 x
PRLIJ/2 l
ij
(2)
0 x TEMPERATURE
PRLIJ/3 l
ij
(3)
0 x TEMPERATURE
PRLIJ/4 T
lower
0 x TEMPERATURE
PRLIJ/5 T
upper
1000 x TEMPERATURE
References
D.Y. Peng and D. B. Robinson, "A New TwoConstant Equationofstate," Ind.
Eng. Chem. Fundam., Vol. 15, (1976), pp. 5964.
P.M. Mathias, H.C. Klotz, and J.M. Prausnitz, "Equation of state mixing rules
for multicomponent mixtures: the problem of invariance," Fluid Phase
Equilibria, Vol 67, (1991), pp. 3144.
H. Knapp, R. Dring, L. Oellrich, U. Plcker, and J. M. Prausnitz. "VaporLiquid
Equilibria for Mixtures of Low Boiling Substances." Dechema Chemistry Data
Series, Vol. VI.
Standard PengRobinson
The Standard PengRobinson equationofstate is the original formulation of
the PengRobinson equation of state with the standard alpha function (see
PengRobinson Alpha Functions). It is the basis for the PENGROB property
method and it is recommended for hydrocarbon processing applications such
as gas processing, refinery, and petrochemical processes. Its results are
comparable to those of the standard RedlichKwongSoave equation of state.
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
The equation for this model is:
Where:
2 Thermodynamic Property Models 51
b =
a =
a
i
=
b
i
=
k
ij
=
If the Peneloux volume correction is used (option code 4), then the molar
volume is calculated from:
V = V
m
 c
Where:
V
m
= Molar volume calculated by the equation of state
without the correction
c =
(the Peneloux volume correction term)
c
i
=
(the Peneloux volume correction term for pure
components, calculated from the critical
temperature and pressure and the Rackett
parameter)
The model has option codes which can be used to customize the model, by
selecting a different alpha function and other model options. See Peng
Robinson Alpha Functions for a description of the alpha functions. See Option
Codes for Equation of State Models (under ESPRSTD and ESPRSTD0) for a list
of the option codes.
For best results, the binary parameter k
ij
must be determined from regression
of phase equilibrium data such as VLE data. The Aspen Physical Property
System also has builtin k
ij
for a large number of component pairs in the EOS
LIT databank. These parameters are used automatically with the PENGROB
property method. Values in the databank can be different than those used
with other models such as SoaveRedlichKwong or RedlichKwongSoave,
and this can produce different results.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCPRS T
ci
TC x 5.0 2000.0 TEMPERATURE
PCPRS p
ci
PC x 10
5
10
8
PRESSURE
OMGPRS
e
i
OMEGA x 0.5 2.0
52 2 Thermodynamic Property Models
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
PRSZRA z
RA
RKTZRA x
PRKBV/1 k
ij
(1)
0 x   
PRKBV/2 k
ij
(2)
0 x   TEMPERATURE
PRKBV/3 k
ij
(3)
0 x   TEMPERATURE
PRKBV/4 T
lower
0 x   TEMPERATURE
PRKBV/5 T
upper
1000 x   TEMPERATURE
PRLIJ/1 l
ij
(1)
0 x
PRLIJ/2 l
ij
(2)
0 x TEMPERATURE
PRLIJ/3 l
ij
(3)
0 x TEMPERATURE
PRLIJ/4 T
lower
0 x TEMPERATURE
PRLIJ/5 T
upper
1000 x TEMPERATURE
References
D.Y. Peng and D. B. Robinson, "A New TwoConstant Equationofstate," Ind.
Eng. Chem. Fundam., Vol. 15, (1976), pp. 5964.
PengRobinsonMHV2
This model uses the PengRobinson equationofstate for pure compounds.
The mixing rules are the predictive MHV2 rules. Several alpha functions can
be used in the PengRobinsonMHV2 equationofstate model for a more
accurate description of the pure component behavior. The pure component
behavior and parameter requirements are described in Standard Peng
Robinson, or in PengRobinson Alpha Functions.
The MHV2 mixing rules are an example of modified HuronVidal mixing rules.
A brief introduction is provided in HuronVidal Mixing Rules. For more details,
see MHV2 Mixing Rules.
Predictive SRK (PSRK)
This model uses the RedlichKwongSoave equationofstate for pure
compounds. The mixing rules are the predictive Holderbaum rules, or PSRK
method. Several alpha functions can be used in the PSRK equationofstate
model for a more accurate description of the pure component behavior. The
pure component behavior and parameter requirements are described in
Standard RedlichKwongSoave and in Soave Alpha Functions.
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
The PSRK method is an example of modified HuronVidal mixing rules. A brief
introduction is provided in HuronVidal Mixing Rules. For more details, see
Predictive SoaveRedlichKwongGmehling Mixing Rules.
2 Thermodynamic Property Models 53
PengRobinsonWongSandler
This model uses the PengRobinson equationofstate for pure compounds.
The mixing rules are the predictive WongSandler rules. Several alpha
functions can be used in the PengRobinsonWongSandler equationofstate
model for a more accurate description of the pure component behavior. The
pure component behavior and parameter requirements are described in Peng
Robinson, and in PengRobinson Alpha Functions.
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
The WongSandler mixing rules are an example of modified HuronVidal
mixing rules. A brief introduction is provided in HuronVidal Mixing Rules. For
more details see WongSandler Mixing Rules., this chapter.
RedlichKwong
The RedlichKwong equationofstate can calculate vapor phase
thermodynamic properties for the following property methods: NRTLRK,
UNIFAC, UNIFLL, UNIQRK, VANLRK, and WILSRK. It is applicable for
systems at low to moderate pressures (maximum pressure 10 atm) for which
the vaporphase nonideality is small. The HaydenO'Connell model is
recommended for a more nonideal vapor phase, such as in systems
containing organic acids. It is not recommended for calculating liquid phase
properties.
The equation for the model is:
p =
Where:
=
b =
a
i
=
b
i
=
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
54 2 Thermodynamic Property Models
References
O. Redlich and J.N.S. Kwong, "On the Thermodynamics of Solutions V. An
Equationofstate. Fugacities of Gaseous Solutions," Chem. Rev., Vol. 44,
(1949), pp. 223 244.
RedlichKwongAspen
The RedlichKwongAspen equationofstate is the basis for the RKASPEN
property method. It can be used for hydrocarbon processing applications. It is
also used for more polar components and mixtures of hydrocarbons, and for
light gases at medium to high pressures.
The equation is the same as RedlichKwongSoave:
p =
A quadratic mixing rule is maintained for:
a =
An interaction parameter is introduced in the mixing rule for:
b =
For a
i
an extra polar parameter is used:
a
i
=
b
i
=
The interaction parameters are temperaturedependent:
k
a,ij
=
k
b,ij
=
For best results, binary parameters k
ij
must be determined from phase
equilibrium data regression, such as VLE data.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKA T
ci
TC x 5.0 2000.0 TEMPERATURE
PCRKA p
ci
PC x 10
5
10
8
PRESSURE
OMGRKA
e
i
OMEGA x 0.5 2.0
RKAPOL
q
i
0 x 2.0 2.0
RKAKA0 k
a,ij
0
0 x 5.0 5.0
RKAKA1 k
a,ij
1
0 x 15.0 15.0 TEMPERATURE
RKAKB0 k
b,ij
0
0 x 5.0 5.0
2 Thermodynamic Property Models 55
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
RKAKB1 k
b,ij
1
0 x 15.0 15.0 TEMPERATURE
References
Mathias, P.M., "A Versatile Phase Equilibrium Equationofstate", Ind. Eng.
Chem. Process Des. Dev., Vol. 22, (1983), pp. 385 391.
RedlichKwongSoave
This is the standard RedlichKwongSoave equationofstate, and is the basis
for the RKSOAVE property method. It is recommended for hydrocarbon
processing applications, such as gasprocessing, refinery, and petrochemical
processes. Its results are comparable to those of the PengRobinson
equationofstate.
The equation is:
Where:
a
0
is the standard quadratic mixing term:
a
1
is an additional, asymmetric (polar) term:
b =
a
i
=
b
i
=
k
ij
= k
ji
; ;
The parameter a
i
is calculated according to the standard Soave formulation
(see Soave Alpha Functions, equations 1, 2, 3, 5, and 6).
56 2 Thermodynamic Property Models
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
If the Peneloux volume correction is used (option code 4), then the molar
volume is calculated from:
V = V
m
 c
Where:
V
m
is the molar volume calculated from the equation of state without the
correction
The model uses option codes which are described in SoaveRedlichKwong
Option Codes.
For best results, binary parameters k
ij
must be determined from phase
equilibrium data regression (for example, VLE data). The Aspen Physical
Property System also has builtin k
ij
for a large number of component pairs in
the EOSLIT databank from Knapp et al. These binary parameters are used
automatically with the RKSOAVE property method. Values of k
ij
in the
databank can be different than those used with other models such as Soave
RedlichKwong or Standard PengRobinson, and this can produce different
results.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKSS T
ci
TC x 5.0 2000.0 TEMPERATURE
PCRKSS p
ci
PC x 10
5
10
8
PRESSURE
OMRKSS
e
i
OMEGA x 0.5 2.0
RKSKBV/1 k
ij
(1)
0 x 5.0 5.0
RKSKBV/2 k
ij
(2)
0 x TEMPERATURE
RKSKBV/3 k
ij
(3)
0 x TEMPERATURE
RKSKBV/4 T
k,lower
0 x TEMPERATURE
RKSKBV/5 T
k,upper
1000 x TEMPERATURE
RKSLBV/1 l
ij
(1)
0 x
RKSLBV/2 l
ij
(2)
0 x TEMPERATURE
RKSLBV/3 l
ij
(3)
0 x TEMPERATURE
RKSLBV/4 T
l,lower
0 x TEMPERATURE
RKSLBV/5 T
l,upper
1000 x TEMPERATURE
References
G. Soave, "Equilibrium Constants for Modified RedlichKwong Equationof
state," Chem. Eng. Sci., Vol. 27, (1972), pp. 1196 1203.
2 Thermodynamic Property Models 57
J. Schwartzentruber and H. Renon, "Extension of UNIFAC to High Pressures
and Temperatures by the Use of a Cubic Equationofstate," Ind. Eng. Chem.
Res., Vol. 28, (1989), pp. 1049 1955.
A. Peneloux, E. Rauzy, and R. Freze, "A Consistent Correction For Redlich
KwongSoave Volumes", Fluid Phase Eq., Vol. 8, (1982), pp. 723.
H. Knapp, R. Dring, L. Oellrich, U. Plcker, and J. M. Prausnitz. "VaporLiquid
Equilibria for Mixtures of Low Boiling Substances." Dechema Chemistry Data
Series, Vol. VI.
RedlichKwongSoaveBostonMathias
The RedlichKwongSoaveBostonMathias equationofstate is the basis for
the RKSBM property method. It is the RedlichKwongSoave equationof
state with the BostonMathias alpha function (see Soave Alpha Functions). It
is recommended for hydrocarbon processing applications, such as gas
processing, refinery, and petrochemical processes. Its results are comparable
to those of the PengRobinsonBostonMathias equationofstate.
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
The equation is:
p =
Where:
a
0
is the standard quadratic mixing term:
a
1
is an additional, asymmetric (polar) term:
b =
a
i
=
b
i
=
k
ij
= k
ji
58 2 Thermodynamic Property Models
; ;
The parameter a
i
is calculated by the standard Soave formulation at
supercritical temperatures. If the component is supercritical, the Boston
Mathias extrapolation is used (see Soave Alpha Functions).
If the Peneloux volume correction is used (option code 4), then the molar
volume is calculated from:
V = V
m
 c
Where:
V
m
is the molar volume calculated from the equation of state without the
correction
The model uses option codes which are described in SoaveRedlichKwong
Option Codes.
For best results, binary parameters k
ij
must be determined from phase
equilibrium data regression (for example, VLE data).
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKS T
ci
TC x 5.0 2000.0 TEMPERATURE
PCRKS p
ci
PC x 10
5
10
8
PRESSURE
OMGRKS
e
i
OMEGA x 0.5 2.0
RKSKBV/1 k
ij
(1)
0 x 5.0 5.0
RKSKBV/2 k
ij
(2)
0 x TEMPERATURE
RKSKBV/3 k
ij
(3)
0 x TEMPERATURE
RKSKBV/4 T
k,lower
0 x TEMPERATURE
RKSKBV/5 T
k,upper
1000 x TEMPERATURE
RKSLBV/1 l
ij
(1)
0 x
RKSLBV/2 l
ij
(2)
0 x TEMPERATURE
RKSLBV/3 l
ij
(3)
0 x TEMPERATURE
RKSLBV/4 T
l,lower
0 x TEMPERATURE
RKSLBV/5 T
l,upper
1000 x TEMPERATURE
References
G. Soave, "Equilibrium Constants for Modified RedlichKwong Equationof
state," Chem. Eng. Sci., Vol. 27, (1972), pp. 1196 1203.
RedlichKwongSoaveWongSandler
This equationofstate model uses the RedlichKwongSoave equationofstate
for pure compounds. The predictive WongSandler mixing rules are used.
Several alpha functions can be used in the RedlichKwongSoaveWong
2 Thermodynamic Property Models 59
Sandler equationofstate model for a more accurate description of the pure
component behavior. The pure component behavior and parameter
requirements are described in Standard RedlichKwongSoave, and in Soave
Alpha Functions.
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
The WongSandler mixing rules are an example of modified HuronVidal
mixing rules. A brief introduction is provided in HuronVidal Mixing Rules. For
more details, see WongSandler Mixing Rules.
RedlichKwongSoaveMHV2
This equationofstate model uses the RedlichKwongSoave equationofstate
for pure compounds. The predictive MHV2 mixing rules are used. Several
alpha functions can be used in the RKSoaveMHV2 equationofstate model
for a more accurate description of the pure component behavior. The pure
component behavior and its parameter requirements are described in
Standard RedlichKwongSoave, and in Soave Alpha Functions.
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
The MHV2 mixing rules are an example of modified HuronVidal mixing rules.
A brief introduction is provided in HuronVidal Mixing Rules. For more details,
see MHV2 Mixing Rules.
SchwartzentruberRenon
The SchwartzentruberRenon equationofstate is the basis for the SRPOLAR
property method. It can be used to model chemically nonideal systems with
the same accuracy as activity coefficient property methods, such as the
WILSON property method. This equationofstate is recommended for highly
nonideal systems at high temperatures and pressures, such as in methanol
synthesis and supercritical extraction applications.
The equation for the model is:
p =
Where:
a =
b =
60 2 Thermodynamic Property Models
c =
a
i
=
b
i
=
c
i
=
(for T < T
ci
)
(for T > T
ci
)
The latter equation is used to ensure b
i
c
i
remains
positive in supercritical conditions.
g
0i
= b
i
(covolume)
g
2i
=
g
1i
=
k
a,ij
=
l
ij
=
k
b,ij
=
k
a,ij
= k
a,ji
l
ij
= l
ji
k
b,ij
= k
b,ji
The binary parameters k
a,ij
, k
b,ij
, and l
ij
are temperaturedependent. In most
cases, k
a,ij
0
and l
ij
0
are sufficient to represent the system of interest.
VLE calculations are independent of c. However, c does influence the fugacity
values and can be adjusted to (liquid) molar volumes. For a wide temperature
range, adjust c
i0
to the molar volume at 298.15K or at boiling temperature.
The a
i
are calculated using the Extended Mathias Alpha Function, as described
in Soave Alpha Functions.
Warning: Using c
1i
and c
2i
can cause anomalous behavior in enthalpy and
heat capacity. Their use is strongly discouraged.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKU T
ci
TC x 5.0 2000.0 TEMPERATURE
PCRKU p
ci
PC x 105 108 PRESSURE
OMGRKU
e
i
OMEGA x 0.5 2.0
2 Thermodynamic Property Models 61
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
RKUPP0 q
0i
x
RKUPP1 q
1i
0 x
RKUPP2 q
2i
0 x
RKUC0 c
0i
0 x
RKUC1 c
1i
0 x
RKUC2 c
2i
0 x
RKUKA0 k
a,ij
0
0 x
RKUKA1 k
a,ij
1
0 x TEMPERATURE
RKUKA2 k
a,ij
2
0 x TEMPERATURE
RKULA0 l
ij
0
0 x
RKULA1 l
ij
1
0 x TEMPERATURE
RKULA2 l
ij
2
0 x TEMPERATURE
RKUKB0 k
b,ij
0
0 x
RKUKB1 k
b,ij
1
0 x TEMPERATURE
RKUKB2 k
b,ij
2
0 x TEMPERATURE
Absolute temperature units are assumed for k
a,ij
2
, l
ij
2
, and k
b,ij
2
.
For polar components (dipole moment >> 0), if you do not enter q
0i
, the
system estimates q
0i
, q
1i
, q
2i
from vapor pressures using the Antoine vapor
pressure model.
If you do not enter at least one of the binary parameters k
a,ij
0
, k
a,ij
2
, l
ij
0
,
l
ij
2
, k
b,ij
0
, or k
b,ij
2
the system estimates k
a,ij
0
, k
a,ij
2
, l
ij
0
, and l
ij
2
from the UNIFAC
or Hayden O'Connell models.
RKUC0, RKUC1, and RKUC2 are treated as having units m
3
/kmol. No unit
conversion to other molar volume units is done.
References
G. Soave, "Equilibrium Constants for Modified RedlichKwong Equationof
state," Chem. Eng. Sci., Vol. 27, (1972), pp. 1196  1203.
J. Schwartzentruber and H. Renon, "Extension of UNIFAC to High Pressures
and Temperatures by the Use of a Cubic EquationofState," Ind. Eng. Chem.
Res., Vol. 28, (1989), pp. 1049 1955.
A. Peneloux, E. Rauzy, and R. Freze, "A Consistent Correction For Redlich
KwongSoave Volumes", Fluid Phase Eq., Vol. 8, (1982), pp. 723.
SoaveRedlichKwong
The SoaveRedlichKwong equationofstate is the basis of the SRK property
method. This model is based on the same equation of state as the Redlich
KwongSoave model. However, this model has several important differences.
 A volume translation concept introduced by Peneloux and Rauzy is used to
improve molar liquid volume calculated from the cubic equation of state.
 Improvement in water properties is achieved by using the NBS Steam
Table.
62 2 Thermodynamic Property Models
 Improvement in speed of computation for equation based calculation is
achieved by using composition independent fugacity.
 Optional KabadiDanner mixing rules for improved phase equilibrium
calculations in waterhydrocarbon systems (see SRKKabadiDanner)
 Optional Mathias alpha function
Note: You can choose any of the available alpha functions, but you cannot
define multiple property methods based on this model using different alpha
functions within the same run.
This equationofstate can be used for hydrocarbon systems that include the
common light gases, such as H
2
S, CO
2
and N
2
.
The form of the equationofstate is:
Where:
a
0
is the standard quadratic mixing term:
Where:
;
a
1
is an additional, asymmetric (polar) term:
Where:
; ;
If the Peneloux volume correction is used (option code 4), then the molar
volume is calculated from:
V = V
m
 c
Where:
V
m
is the molar volume calculated from the equation of state without the
correction
2 Thermodynamic Property Models 63
The enthalpy, entropy, Gibbs energy, and molar volume of water are
calculated from the steam tables. The total properties are molefraction
averages of these values with the properties calculated by the equation of
state for other components. Fugacity coefficient is not affected.
For best results, the binary parameter k
ij
must be determined from phase
equilibrium data regression (for example, VLE data). The Aspen Physical
Property System also has builtin k
ij
for a large number of component pairs in
the SRKASPEN databank, regressed by AspenTech. These parameters are
used automatically with the SRK property method. Values of k
ij
in the
databank can be different than those used with other models such as
Standard RedlichKwongSoave or Standard RedlichKwongSoave, and this
can produce different results.
The model uses option codes which are described in SoaveRedlichKwong
Option Codes.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SRKTC T
ci
TC x 5.0 2000.0 TEMPERATURE
SRKPC p
ci
PC x 10
5
10
8
PRESSURE
SRKOMG
e
i
OMEGA x 0.5 2.0
SRKZRA z
RA
RKTZRA x
SRKKIJ/1 k
ij
(1)
0 x
SRKKIJ/2 k
ij
(2)
0 x TEMPERATURE
SRKKIJ/3 k
ij
(3)
0 x TEMPERATURE
SRKKIJ/4 T
lower
0 x TEMPERATURE
SRKKIJ/5 T
upper
1000 x TEMPERATURE
SRKLIJ/1 l
ij
(1)
0 x
SRKLIJ/2 l
ij
(2)
0 x TEMPERATURE
SRKLIJ/3 l
ij
(3)
0 x TEMPERATURE
SRKLIJ/4 T
lower
0 x TEMPERATURE
SRKLIJ/5 T
upper
1000 x TEMPERATURE
References
G. Soave, "Equilibrium Constants for Modified RedlichKwong Equationof
state," Chem. Eng. Sci., Vol. 27, (1972), pp. 1196  1203.
A. Peneloux, E. Rauzy, and R. Freze, "A Consistent Correction For Redlich
KwongSoave Volumes", Fluid Phase Eq., Vol. 8, (1982), pp. 723.
P.M. Mathias, H.C. Klotz, and J.M. Prausnitz, "Equation of state mixing rules
for multicomponent mixtures: the problem of invariance," Fluid Phase
Equilibria, Vol 67, (1991), pp. 3144.
64 2 Thermodynamic Property Models
SRKKabadiDanner
The SRKKabadiDanner property model uses the SRK equationofstate with
improved phase equilibrium calculations for mixtures containing water and
hydrocarbons. These improvements are achieved by using the KabadiDanner
mixing rules.
The form of the equationofstate is:
Where:
a
0
is the standard quadratic mixing term:
Where:
;
The best values of k
wj
(w = water) were obtained from experimental data.
Results are given for seven homologous series.
Best Fit Values of k
wj
for Different Homologous Series
with Water
Homologous series k
wj
Alkanes 0.500
Alkenes 0.393
Dialkenes 0.311
Acetylenes 0.348
Naphthenes 0.445
Cycloalkenes 0.355
Aromatics 0.315
a
KD
is the KabadiDanner term for water:
Where:
G
i
is the sum of the group contributions of different groups which make up a
molecule of hydrocarbon i.
2 Thermodynamic Property Models 65
g
l
is the group contribution parameter for groups constituting hydrocarbons.
Groups Constituting Hydrocarbons and Their Group
Contribution Parameters
Group l g
l
, atm m
6
x 10
5
CH4 1.3580
CH3 0.9822
CH2 1.0780
> CH 0.9728
> C < 0.8687
CH2 (cyclic) 0.7488
> CH (cyclic) 0.7352
CH = CH (cyclic) 0.6180
CH2 = CH2 1.7940
CH2 = CH 1.3450
CH2 = C< 0.9066
CH CH 1.6870
CH C 1.1811
CH = 0.5117
> C = (aromatic) 0.3902
This value is obtained from very little data. Might not be reliable.
If the Peneloux volume correction is used (option code 4), then the molar
volume is calculated from:
V = V
m
 c
Where:
V
m
is the molar volume calculated from the equation of state without the
correction
The model uses option codes which are described in SoaveRedlichKwong
Option Codes.
SRKKabadiDanner uses the same parameters as SRK, with added
interaction parameters. Do not specify values for the SRKLIJ parameters
when using SRKKD.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SRKTC T
ci
TC x 5.0 2000.0 TEMPERATURE
SRKPC p
ci
PC x 10
5
10
8
PRESSURE
66 2 Thermodynamic Property Models
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SRKOMG e
i
OMEGA x 0.5 2.0
SRKWF G
i
0 x
SRKZRA z
RA
RKTZRA x
SRKKIJ/1 k
ij
(1)
0 x
SRKKIJ/2 k
ij
(2)
0 x TEMPERATURE
SRKKIJ/3 k
ij
(3)
0 x TEMPERATURE
SRKKIJ/4 T
lower
0 x TEMPERATURE
SRKKIJ/5 T
upper
0 x TEMPERATURE
References
V. Kabadi and R. P. Danner, "A Modified SoaveRedlichKwong Equation of
State for WaterHydrocarbon Phase Equilibria", Ind. Eng. Chem. Process Des.
Dev., Vol. 24, No. 3, (1985), pp. 537541.
SRKML
The SRKML property model is the same as the SoaveRedlichKwong model
with these exceptions:
 k
ij
does not equal k
ji
for nonideal systems; they are unsymmetric, and a
different set of parameters is used, as shown below.
 The l
ij
are calculated from the equation l
ij
= k
ji
 k
ij
 The Peneloux volume correction is disabled by default. (It can be enabled
using option code 4.)
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SMLTC T
ci
TC x 5.0 2000.0 TEMPERATURE
SMLPC p
ci
PC x 10
5
10
8
PRESSURE
SMLOMG
e
i
OMEGA x 0.5 2.0
SMLZRA z
RA
RKTZRA x
SMLCMP 0 x
SRKMLP x 2.0 2.0
SRKGLP x
SMLKIJ/1 k
ij
(1)
0 x
SMLKIJ/2 k
ij
(2)
0 x TEMPERATURE
SMLKIJ/3 k
ij
(3)
0 x TEMPERATURE
SMLKIJ/4 T
lower
0 x TEMPERATURE
SMLKIJ/5 T
upper
1000 x TEMPERATURE
The model uses option codes which are described in SoaveRedlichKwong
Option Codes.
2 Thermodynamic Property Models 67
Additional Parameters
SMLZRA is used in PenelouxRauzy volume translation as described in Soave
RedlichKwong. This volume translation feature can be enabled by setting
option code 4 to 1, but is disabled by default (in which case this parameter is
not used).
SMLCMP is the Composition Independent Fugacity Calculation Flag. You can
use this flag to indicate the component for which its fugacity coefficient in the
mixture will be calculated such that it is independent of composition. The
calculated fugacity coefficient of that component is simply the pure
component fugacity coefficient. This is a simplification. Enter a value of 1 for
the element i to indicate that the fugacity coefficient of component i is
independent of composition. This parameter defaults to zero.
SRKMLP is a polar parameter for the Mathias alpha function.
SRKGLP is a vector of parameters for the GibbonsLaughton alpha function.
VPA/IKCAPE EquationofState
The VPA/IKCAPE equation of state is similar to the HF equation of state but
allows dimerization, tetramerization and hexamerization to occur
simultaneously. The main assumption of the model is that only molecular
association causes the gas phase nonideality. Attractive forces between the
molecules and the complexes are neglected.
There are three kinds of associations, which can be modeled:
 Dimerization (examples: formic acid, acetic acid)
 Tetramerization (example: acetic acid)
 Hexamerization (example: hydrogen fluoride)
To get the largest possible flexibility of the model all these kinds of
association can occur simultaneously, for example, in a mixture containing
acetic acid and HF. Up to five components can associate, and any number of
inert components are allowed. This is the only difference between this model
and the HF equation of state, which account only the hexamerization of HF.
Symbols
In the following description, these symbols are used:
y
i
= Apparent concentration
z
in
= True concentration, for component i and degree of
association n=1, 2, 4, 6
z
Mij
= True concentration of crossdimers of components i
and j, for i,j 1 to 5.
p
0
= Reference pressure
k = Number of components
68 2 Thermodynamic Property Models
Association Equilibria
Every association equilibrium reaction
(1)
(2)
is described by the equilibrium constants
(3)
(4)
By setting
(5)
(6)
their temperature dependence can be reproduced.
To evaluate the true concentration of every complex z
in
, the following
nonlinear systems of equations are to be solved:
Total mass balance:
The sum of true concentrations is unity.
(7)
Mass balance for every component i>1:
The ratio of the monomers of each component i>1 and component i=1
occurring in the various complexes must be equal to the ratio of their
apparent concentrations.
(8)
Thus, a system of k nonlinear equations for k unknowns z
i1
has been
developed. After having solved it, all the z
in
and z
Mij
can be determined using
2 Thermodynamic Property Models 69
equations (3, 4). This is the main step to evaluate all the properties needed
for a calculation.
Specific Volume of the Gas Phase
The compressibility factor is defined by the ratio between the number of
complexes and the number of monomers in the complexes.
(9)
The compressibility factor itself is
(10)
Fugacity Coefficient
As is wellknown from thermodynamics, the fugacity coefficient can be
calculated by
(11)
Isothermal Enthalpy Departure
According to the ASPEN enthalpy model, an equation of state must supply an
expression to compute the isothermal molar enthalpy departure between zero
pressure and actual pressure. In the following section this enthalpy
contribution per mole monomers is abbreviated by Ah
a
.
Taking this sort of gas phase nonideality into account, the specific enthalpy
per mole can be written as
(12)
with
(13)
to evaluate Ah
a
, a mixture consisting of N monomers integrated in the
complexes is considered. The quota of monomers i being integrated in a
complex of degree n is given by
70 2 Thermodynamic Property Models
(14)
and
(16)
respectively. For the reactions mentioned above:
(1)
(2)
the enthalpies of reaction are
(17)
(18)
as the van't Hoff equation
(19)
holds for this case.
For each monomer being integrated in a complex of degree n, its contribution
to the enthalpy departure is Ah
in
/ n or Ah
Mij
/ 2, respectively. Hence, Ah
a
can
easily be derived by
(20)
Isothermal entropy and Gibbs energy departure:
A similar expression for Ag
a
should hold as it does for the enthalpy departure
(eq. 20):
(21)
2 Thermodynamic Property Models 71
using
(22)
and
(23)
(24)
Using the association model, more different species occur than can be
distinguished. Thus, the equivalent expression for the entropy of mixing
should be written with the true concentrations. As eq. 24 refers to 1 mole
monomers, the expression should be weighted by the compressibility factor
representing the true number of moles. The new expression is
(25)
For Ag
a
we obtain
(26)
and, analogously,
(27)
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DMER/1 A
i2
0 X
DMER/2 B
i2
0 X TEMPERATURE
TMER/1 A
i4
0 X
TMER/2 B
i4
0 X TEMPERATURE
HMER/1 A
i6
0 X
HMER/2 B
i6
0 X TEMPERATURE
References
M. M. Abbott and H. C. van Ness, "Thermodynamics of Solutions Containing
Reactive Species, a Guide to Fundamentals and Applications," Fluid Phase Eq.,
Vol. 77, (1992) pp. 53119.
V. V. De Leeuw and S. Watanasiri, "Modeling Phase Equilibria and Enthalpies
of the System Water and Hydrofluoric Acid Using an HF Equationofstate in
Conjunction with the Electrolyte NRTL Activity Coefficient Model," Paper
Presented at the 13th European Seminar on Applied Thermodynamics, June
912, CarryleRouet, France, 1993.
72 2 Thermodynamic Property Models
R. W. Long, J. H. Hildebrand, and W. E. Morrell, "The Polymerization of
Gaseous Hydrogen and Deuterium Fluorides," J. Am. Chem. Soc., Vol. 65,
(1943), pp. 182187.
C. E. Vanderzee and W. Wm. Rodenburg, "Gas Imperfections and
Thermodynamic Excess Properties of Gaseous Hydrogen Fluoride," J. Chem.
Thermodynamics, Vol. 2, (1970), pp. 461478.
PengRobinson Alpha Functions
The pure component parameters for the PengRobinson equationofstate are
calculated as follows:
(1)
(2)
These expressions are derived by applying the critical constraints to the
equationofstate under these conditions:
(3)
The parameter o is a temperature function. It was originally introduced by
Soave in the RedlichKwong equationofstate. This parameter improves the
correlation of the pure component vapor pressure.
Note: You can choose any of the alpha functions described here, but you
cannot define multiple property methods based on this model using different
alpha functions within the same run.
This approach was also adopted by Peng and Robinson:
(4)
Equation 3 is still represented. The parameter m
i
can be correlated with the
acentric factor:
(5)
Equations 1 through 5 are the standard PengRobinson formulation. The
PengRobinson alpha function is adequate for hydrocarbons and other
nonpolar compounds, but is not sufficiently accurate for polar compounds.
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCPR T
ci
TC X 5.0 2000.0 TEMPERATURE
PCPR p
ci
PC X 10
5
10
8
PRESSURE
OMGPR
e
i
OMEGA X 0.5 2.0
2 Thermodynamic Property Models 73
BostonMathias Extrapolation
For light gases at high reduced temperatures (> 5), equation 4 gives
unrealistic results. The boundary conditions are that attraction between
molecules should vanish for extremely high temperatures, and o reduces
asymptotically to zero. Boston and Mathias derived an alternative function for
temperatures higher than critical:
(6)
With
=
=
Where m
i
is computed by equation 5, and equation 4 is used for subcritical
temperatures. Additional parameters are not needed.
Extended GibbonsLaughton Alpha Function
The extended GibbonsLaughton alpha function is suitable for use with both
polar and nonpolar components. It has the flexibility to fit the vapor pressure
of most substances from the triple point to the critical point.
Where T
r
is the reduced temperature; X
i
, Y
i
and Z
i
are substance dependent
parameters.
This function is equivalent to the standard PengRobinson alpha function if
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
PRGLP/1 X X
PRGLP/2 Y 0 X
PRGLP/3 Z 0 X
PRGLP/4 n 2 X
PRGLP/5 T
lower
0 X TEMPERATURE
PRGLP/6 T
upper
1000 X TEMPERATURE
Twu Generalized Alpha Function
The Twu generalized alpha function is a theoreticallybased function that is
currently recognized as the best available alpha function. It behaves better
than other functions at supercritical conditions (T > T
c
) and when the acentric
factor is large. The improved behavior at high values of acentric factor is
important for high molecular weight pseudocomponents. There is no limit on
the minimum value of acentric factor that can be used with this function.
74 2 Thermodynamic Property Models
Where the L, M, and N are parameters that vary depending on the equation of
state and whether the temperature is above or below the critical temperature
of the component.
For PengRobinson equation of state:
Subcritical T Supercritical T
L
(0)
0.272838 0.373949
M
(0)
0.924779 4.73020
N
(0)
1.19764 0.200000
L
(1)
0.625701 0.0239035
M
(1)
0.792014 1.24615
N
(1)
2.46022 8.000000
Twu Alpha Function
The Twu alpha function is a theoreticallybased function that is currently
recognized as the best available alpha function. It behaves better than other
functions at supercritical conditions (T > T
c
).
Where the L, M, and N are substancedependent parameters that must be
determined from regression of purecomponent vapor pressure data or other
data such as liquid heat capacity.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
PRTWUP/1 L X
PRTWUP/2 M 0 X
PRTWUP/3 N 0 X
MathiasCopeman Alpha Function
This is an extension of the PengRobinson alpha function which provides a
more accurate fit of vapor pressure for polar compounds.
(7)
For c
2,i
= 0 and c
3,i
= 0, this expression reduces to the standard Peng
Robinson formulation if c
2,i
= m
i
. You can use vapor pressure data if the
temperature is subcritical to regress the constants. If the temperature is
supercritical, c
2,i
and c
3,i
are set to 0.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCPR T
ci
TC X 5.0 2000.0 TEMPERATURE
PCPR p
ci
PC X 10
5
10
8
PRESSURE
2 Thermodynamic Property Models 75
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
PRMCP/1 c
1,i
X
PRMCP/2 c
2,i
0 X
PRMCP/3 c
3,i
0 X
SchwartzentruberRenonWatanasiri Alpha Function
The SchwartzentruberRenonWatanasiri alpha function is:
(8)
Where m
i
is computed by equation 5. The polar parameters p
1,i
, p
2,i
and p
3,i
are comparable with the c parameters of the MathiasCopeman expression.
Equation 8 reduces to the standard PengRobinson formulation if the polar
parameters are zero. Equation 8 is used only for belowcritical temperatures.
For abovecritical temperatures, the BostonMathias extrapolation is used.
Use equation 6 with:
(9)
(10)
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCPR T
ci
TC X 5.0 2000.0 TEMPERATURE
PCPR p
ci
PC X 10
5
10
8
PRESSURE
OMGPR
e
i
OMEGA X 0.5 2.0
PRSRP/1 X
PRSRP/2 0 X
PRSRP/3 0 X
HYSYS Alpha Function
The HYSYS alpha function uses the same equation as the standard Peng
Robinson alpha function when e < 0.49, and otherwise it uses equation 4
with the following definition of m to correct the behavior for large values of
acentric factor:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCPR T
ci
TC X 5.0 2000.0 TEMPERATURE
PCPR p
ci
PC X 10
5
10
8
PRESSURE
OMGPR
e
i
OMEGA X 0.5 2.0
76 2 Thermodynamic Property Models
Use of Alpha Functions
The alpha functions in PengRobinsonbased equationofstate models is
provided in the following table. You can verify and change the value of
possible option codes on the Methods  Selected Methods  Models sheet.
Alpha function Model name First Option code
Standard PR/
BostonMathias
ESPR0, ESPR
ESPRWS0, ESPRWS
ESPRV20, ESPRV2
0
0
0
Standard Peng Robinson ESPRSTD0, ESPRSTD 1
Extended Gibbons
Laughton
ESPR0, ESPR 2
Twu Generalized alpha
function
ESPR0, ESPR 3
Twu alpha function ESPR0, ESPR 4
HYSYS alpha function ESPRSTD0, ESPRSTD 5
MathiasCopeman ESPRWS0, ESPRWS
ESPRV20, ESPRV2
2
2
Schwartzentruber
Renon
Watanasiri
ESPRWS0, ESPRWS
ESPRV20, ESPRV2
3 (default)
3 (default)
References
J. F. Boston and P.M. Mathias, "Phase Equilibria in a ThirdGeneration Process
Simulator" in Proceedings of the 2nd International Conference on Phase
Equilibria and Fluid Properties in the Chemical Process Industries, West Berlin,
(1721 March 1980) pp. 823849.
D.Y. Peng and D.B. Robinson, "A New TwoConstant Equationofstate," Ind.
Eng. Chem. Fundam., Vol. 15, (1976), pp. 5964.
P.M. Mathias and T.W. Copeman, "Extension of the PengRobinson Equation
ofstate To Complex Mixtures: Evaluation of the Various Forms of the Local
Composition Concept",Fluid Phase Eq., Vol. 13, (1983), p. 91.
J. Schwartzentruber, H. Renon, and S. Watanasiri, "Kvalues for NonIdeal
Systems:An Easier Way," Chem. Eng., March 1990, pp. 118124.
G. Soave, "Equilibrium Constants for a Modified RedlichKwong Equationof
state," Chem Eng. Sci., Vol. 27, (1972), pp. 11961203.
C.H. Twu, J. E. Coon, and J.R. Cunningham, "A New Cubic Equation of State,"
Fluid Phase Equilib., Vol. 75, (1992), pp. 6579.
C.H. Twu, D. Bluck, J.R. Cunningham, and J.E. Coon, "A Cubic Equation of
State with a New Alpha Function and a New Mixing Rule," Fluid Phase Equilib.,
Vol. 69, (1991), pp. 3350.
Soave Alpha Functions
The pure component parameters for the RedlichKwong equationofstate are
calculated as:
2 Thermodynamic Property Models 77
(1)
(2)
These expressions are derived by applying the critical constraint to the
equationofstate under these conditions:
(3)
Note: You can choose any of the alpha functions described here, but you
cannot define multiple property methods based on this model using different
alpha functions within the same run.
In the RedlichKwong equationofstate, alpha is:
(4)
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
Soave Modification
The parameter o
i
is a temperature function introduced by Soave in the
RedlichKwong equationofstate to improve the correlation of the pure
component vapor pressure:
(5)
Equation 3 still holds. The parameter m
i
can be correlated with the acentric
factor:
(6)
Equations 1, 2, 3, 5 and 6 are the standard RedlichKwongSoave
formulation. The Soave alpha function is adequate for hydrocarbons and other
nonpolar compounds, but is not sufficiently accurate for polar compounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKS T
ci
TC X 5.0 2000.0 TEMPERATURE
PCRKS p
ci
PC X 10
5
10
8
PRESSURE
OMGRKS e
i
OMEGA X 0.5 2.0
78 2 Thermodynamic Property Models
BostonMathias Extrapolation
For light gases at high reduced temperatures (> 5), equation 5 gives
unrealistic results. The boundary conditions are that attraction between
molecules should vanish for extremely high temperatures, and reduces
asymptotically to zero. Boston and Mathias derived an alternative function for
temperatures higher than critical:
(7)
With
d
i
=
c
i
=
Where:
m
i
= Computed by equation 6
Equation 5 = Used for subcritical temperatures
Additional parameters are not needed.
Mathias Alpha Function
This is an extension of the Soave alpha function which provides a more
accurate fit of vapor pressure for polar compounds.
(8)
For q
i
=0, equation 8 reduces to the standard RedlichKwongSoave
formulation, equations 5 and 6. For temperatures above critical, the Boston
Mathias extrapolation is used, that is, equation 7 with:
(9)
(10)
The Mathias alpha function is used in the RedlichKwongAspen model, which
is the basis for the RKASPEN property method. This alpha function is also
available as an option for SRK, SRKKD, SRKML, RKSOAVE, and RKSBM.
See SoaveRedlichKwong Option Codes for more information.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKA T
ci
TC X 5.0 2000.0 TEMPERATURE
PCRKA p
ci
PC X 10
5
10
8
PRESSURE
OMGRKA e
i
OMEGA X 0.5 2.0
q
i
X 2.0 2.0
2 Thermodynamic Property Models 79
RKAPOL for RedlichKwongAspen, SRKPOL for SRK and SRKKD, SRKMLP for
SRKML, RKSPOL for RKSBM, or RKSSPO for RKSOAVE.
Extended Mathias Alpha Function
An extension of the Mathias approach is:
(11)
Where m
i
is computed by equation 6. If the polar parameters p
1,i
, p
2,i
and p
3,i
are zero, equation 11 reduces to the standard RedlichKwongSoave
formulation. You can use vapor pressure data to regress the constants if the
temperature is subcritical. Equation 11 is used only for temperatures below
critical.
The BostonMathias extrapolation is used for temperatures above critical, that
is, with:
(12)
(13)
This alpha function is used in the RedlichKwongUNIFAC model which is the
basis for the SRPOLAR property method.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKU T
ci
TC X 5.0 2000.0 TEMPERATURE
PCRKU p
ci
PC X 10
5
10
8
PRESSURE
OMGRKU e
i
OMEGA X 0.5 2.0
RKUPP0 p
1,i
X
RKUPP1 p
2,i
0 X
RKUPP2 p
3,i
0 X
MathiasCopeman Alpha Function
The MathiasCopeman alpha function is suitable for use with both polar and
nonpolar components. It has the flexibility to fit the vapor pressure of most
substances from the triple point to the critical point.
(14)
For c
2,i
=0 and c
3,i
=0 this expression reduces to the standard RedlichKwong
Soave formulation if c
1,i
=m
i
. If the temperature is subcritical, use vapor
pressure data to regress the constants. If the temperature is supercritical, set
c
2,i
and c
3,i
to 0.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKS T
ci
TC X 5.0 2000.0 TEMPERATURE
PCRKS p
ci
PC X 10
5
10
8
PRESSURE
RKSMCP/1 c
1,i
X
RKSMCP/2 c
2,i
0 X
80 2 Thermodynamic Property Models
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
RKSMCP/3 c
3,i
0 X
SchwartzentruberRenonWatanasiri Alpha Function
The SchwartzentruberRenonWatanasiri alpha function is:
(15)
Where m
i
is computed by equation 6 and the polar parameters p
1,i
, p
2,i
and
p
3,i
are comparable with the c parameters of the MathiasCopeman
expression. Equation 15 reduces to the standard RedlichKwongSoave
formulation if the polar parameters are zero. Equation 15 is very similar to
the extended Mathias equation, but it is easier to use in data regression. It is
used only for temperatures below critical. The BostonMathias extrapolation is
used for temperatures above critical, that is, use equation 7 with:
(16)
(17)
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCRKS T
ci
TC X 5.0 2000.0 TEMPERATURE
PCRKS p
ci
PC X 10
5
10
8
PRESSURE
OMGRKS e
i
OMEGA X 0.5 2.0
RKSSRP/1 p
1,i
X
RKSSRP/2 p
2,i
0 X
RKSSRP/3 p
3,i
0 X
Extended GibbonsLaughton Alpha Function
The extended GibbonsLaughton alpha function is suitable for use with both
polar and nonpolar components. It has the flexibility to fit the vapor pressure
of most substances from the triple point to the critical point.
Where T
r
is the reduced temperature; X
i
, Y
i
and Z
i
are substance dependent
parameters.
This function is equivalent to the standard Soave alpha function if
This function is not intended for use in supercritical conditions. To avoid
predicting negative alpha, when T
ri
>1 the BostonMathias alpha function is
used instead.
2 Thermodynamic Property Models 81
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SRKGLP/1 X X
SRKGLP/2 Y 0 X
SRKGLP/3 Z 0 X
SRKGLP/4 n 2 X
SRKGLP/5 T
lower
0 X TEMPERATURE
SRKGLP/6 T
upper
1000 X TEMPERATURE
Twu Generalized Alpha Function
The Twu generalized alpha function is a theoreticallybased function that is
currently recognized as the best available alpha function. It behaves better
than other functions at supercritical conditions (T > T
c
) and when the acentric
factor is large. The improved behavior at high values of acentric factor is
important for high molecular weight pseudocomponents. There is no limit on
the minimum value of acentric factor that can be used with this function.
Where the L, M, and N are parameters that vary depending on the equation of
state and whether the temperature is above or below the critical temperature
of the component.
For SoaveRedlichKwong:
Subcritical T Supercritical T
L
(0)
0.544000 0.379919
M
(0)
1.01309 5.67342
N
(0)
0.935995 0.200000
L
(1)
0.544306 0.0319134
M
(1)
0.802404 1.28756
N
(1)
3.10835 8.000000
Use of Alpha Functions
The use of alpha functions in SoaveRedlichKwong based equationofstate
models is given in the following table. You can verify and change the value of
possible option codes on the Methods  Selected Methods  Models sheet.
Alpha Function Model Name First Option Code
original RK ESRK0, ESRK
standard RKS ESRKSTD0, ESRKSTD
*
1, 2 (default)
standard RKS/BostonMathias *
ESRKSWS0, ESRKSWS
ESRKSV10, ESRKV1
ESRKSV20, ESRKSV2
0
1
1
1
82 2 Thermodynamic Property Models
Alpha Function Model Name First Option Code
Mathias/BostonMathias ESRKA0, ESRKA
Extended Mathias/Boston
Mathias
ESRKU0, ESRKU
MathiasCopeman ESRKSV10, ESRKSV1
ESRKSV20, ESRKSV2
2
2
SchwartzentruberRenon
Watanasiri
ESPRWS0, ESPRWS
ESRKSV10, ESRKSV1
ESRKSV20, ESRKSV2
3 (default)
3 (default)
3 (default)
Twu generalized * 5
GibbonsLaughton with Patel
extension
* 3
Mathias for T < Tc; Boston
Mathias for T > Tc
* 4
* ESRKSTD0, ESRKSTD, ESRKS0, ESRKS, ESSRK, ESSRK0, ESRKSML,
ESRKSML0. The default alpha function (option code 2) for these models is the
standard RKS alpha function, except that the GrabovskyDaubert alpha
function is used for H2: o = 1.202 exp(0.30228xT
ri
)
References
J. F. Boston and P.M. Mathias, "Phase Equilibria in a ThirdGeneration Process
Simulator" in Proceedings of the 2nd International Conference on Phase
Equilibria and Fluid Properties in the Chemical Process Industries, West Berlin,
(1721 March 1980), pp. 823849.
P. M. Mathias, "A Versatile Phase Equilibrium Equationofstate", Ind. Eng.
Chem. Process Des. Dev., Vol. 22, (1983), pp. 385391.
P.M. Mathias and T.W. Copeman, "Extension of the PengRobinson Equation
ofstate To Complex Mixtures: Evaluation of the Various Forms of the Local
Composition Concept", Fluid Phase Eq., Vol. 13, (1983), p. 91.
O. Redlich and J. N. S. Kwong, "On the Thermodynamics of Solutions V. An
Equationofstate. Fugacities of Gaseous Solutions," Chem. Rev., Vol. 44,
(1949), pp. 223244.
J. Schwartzentruber and H. Renon, "Extension of UNIFAC to High Pressures
and Temperatures by the Use of a Cubic Equationofstate," Ind. Eng. Chem.
Res., Vol. 28, (1989), pp. 10491055.
J. Schwartzentruber, H. Renon, and S. Watanasiri, "Kvalues for NonIdeal
Systems:An Easier Way," Chem. Eng., March 1990, pp. 118124.
G. Soave, "Equilibrium Constants for a Modified RedlichKwong Equationof
state," Chem Eng. Sci., Vol. 27, (1972), pp. 11961203.
C.H. Twu, W.D. Sim, and V. Tassone, "Getting a Handle on Advanced Cubic
Equations of State", Chemical Engineering Progress, Vol. 98 #11 (November
2002) pp. 5865.
2 Thermodynamic Property Models 83
HuronVidal Mixing Rules
Huron and Vidal (1979) used a simple thermodynamic relationship to equate
the excess Gibbs energy to expressions for the fugacity coefficient as
computed by equations of state:
(1)
Equation 1 is valid at any pressure, but cannot be evaluated unless some
assumptions are made. If Equation 1 is evaluated at infinite pressure, the
mixture must be liquidlike and extremely dense. It can be assumed that:
(2)
(3)
Using equations 2 and 3 in equation 1 results in an expression for a/b that
contains the excess Gibbs energy at an infinite pressure:
(4)
Where:
(5)
The parameters
1
and
2
depend on the equationofstate used. In general a
cubic equationofstate can be written as:
(6)
Values for
1
and
2
for the PengRobinson and the SoaveRedlichKwong
equations of state are:
Equationofstate
1
2
PengRobinson
RedlichKwongSoave 1 0
This expression can be used at any pressure as a mixing rule for the
parameter. The mixing rule for b is fixed by equation 3. Even when used at
other pressures, this expression contains the excess Gibbs energy at infinite
pressure. You can use any activity coeffecient model to evaluate the excess
Gibbs energy at infinite pressure. Binary interaction coefficients must be
regressed. The mixing rule used contains as many binary parameters as the
activity coefficient model chosen.
This mixing rule has been used successfully for polar mixtures at high
pressures, such as systems containing light gases. In theory, any activity
coefficient model can be used. But the NRTL equation (as modified by Huron
and Vidal) has demonstrated better performance.
84 2 Thermodynamic Property Models
The HuronVidal mixing rules combine extreme flexibility with thermodynamic
consistency, unlike many other molefractiondependent equationofstate
mixing rules. The HuronVidal mixing rules do not allow flexibility in the
description of the excess molar volume, but always predict reasonable excess
volumes.
The HuronVidal mixing rules are theoretically incorrect for low pressure,
because quadratic mole fraction dependence of the second virial coefficient (if
derived from the equationofstate) is not preserved. Since equations of state
are primarily used at high pressure, the practical consequences of this
drawback are minimal.
The Gibbs energy at infinite pressure and the Gibbs energy at an arbitrary
high pressure are similar. But the correspondence is not close enough to
make the mixing rule predictive.
There are several methods for modifying the HuronVidal mixing rule to make
it more predictive. The following three methods are used in Aspen Physical
Property System equationofstate models:
 The modified HuronVidal mixing rule, second order approximation
(MHV2)
 The Predictive SRK Method (PSRK)
 The WongSandler modified HuronVidal mixing rule (WS)
These mixing rules are discussed separately in the following sections. They
have major advantages over other compositiondependent equationofstate
mixing rules.
References
M. J. Huron and J. Vidal, "New Mixing Rules in Simple Equations of State for
representing Vapourliquid equilibria of strongly nonideal mixtures," Fluid
Phase Eq., Vol. 3, (1979), pp. 255271.
MHV2 Mixing Rules
Dahl and Michelsen (1990) use a thermodynamic relationship between excess
Gibbs energy and the fugacity computed by equations of state. This
relationship is equivalent to the one used by Huron and Vidal:
(1)
The advantage is that the expressions for mixture and pure component
fugacities do not contain the pressure. They are functions of compacity V/b
and o:
(2)
Where:
2 Thermodynamic Property Models 85
(3)
and
(4)
With:
(5)
The constants
1
and
2
, which depend only on the equationofstate (see
HuronVidal Mixing Rules) occur in equations 2 and 4.
Instead of using infinite pressure for simplification of equation 1, the condition
of zero pressure is used. At p = 0 an exact relationship between the
compacity and o can be derived. By substitution the simplified equation q(o)
is obtained, and equation 1 becomes:
(6)
However, q(o) can only be written explicitly for o = 5.8. Only an
approximation is possible below that threshold. Dahl and Michelsen use a
second order polynomial fitted to the analytical solution for 10 < o < 13 that
can be extrapolated to low alpha:
(7)
Since q(o)is a universal function (for each equationofstate), the
combination of equations 6 and 7 form the MHV2 mixing rule. Excess Gibbs
energies, from any activity coefficient model with parameters optimized at
low pressures, can be used to determine o, if o
i
, b
i
, and b are known. To
compute b, a linear mixing rule is assumed as in the original HuronVidal
mixing rules:
(8)
This equation is equivalent to the assumption of zero excess molar volume.
The MHV2 mixing rule was the first successful predictive mixing rule for
equations of state. This mixing rule uses previously determined activity
coefficient parameters for predictions at high pressures. UNIFAC was chosen
as a default for its predictive character. The Lyngby modified UNIFAC
formulation was chosen for optimum performance (see UNIFAC (Lyngby
Modified)). However, any activity coefficient model can be used when its
binary interaction parameters are known.
Like the HuronVidal mixing rules, the MHV2 mixing rules are not flexible in
the description of the excess molar volume. The MHV2 mixing rules are
theoretically incorrect at the low pressure limit. But the practical
86 2 Thermodynamic Property Models
consequences of this drawback are minimal (see HuronVidal Mixing Rules,
this chapter).
Reference: S. Dahl and M.L. Michelsen, "HighPressure VaporLiquid
Equilibrium with a UNIFACbased Equationofstate," AIChE J., Vol. 36,
(1990), pp. 18291836.
Predictive SoaveRedlichKwongGmehling
Mixing Rules
These mixing rules by Holderbaum and Gmehling (1991) use a relationship
between the excess Helmholtz energy and equationofstate. They do not use
a relationship between equationofstate properties and excess Gibbs energy,
as in the HuronVidal mixing rules. The pressureexplicit expression for the
equationofstate is substituted in the thermodynamic equation:
(1)
The Helmholtz energy is calculated by integration. A
E
is obtained by:
(2)
Where both A
i
* and A
m
are calculated by using equation 1. A
i
* and A
m
are
written in terms of equationofstate parameters.
The simplification of constant packing fraction (V
m
/ b) is used:
(3)
With:
(4)
Therefore:
(5)
The mixing rule is:
(6)
Where A' is slightly different from A for the HuronVidal mixing rule:
(7)
2 Thermodynamic Property Models 87
Where
1
and
2
, depend on the equationofstate (see HuronVidal Mixing
Rules). If equation 6 is applied at infinite pressure, the packing fraction goes
to 1. The excess Helmholtz energy is equal to the excess Gibbs energy. The
HuronVidal mixing rules are recovered.
The goal of these mixing rules is to be able to use binary interaction
parameters for activity coefficient models at any pressure. These parameters
have been optimized at low pressures. UNIFAC is chosen for its predictive
character. Two issues exist: the packing fraction is not equal to one, and the
excess Gibbs and Helmholtz energy are not equal at the low pressure where
the UNIFAC parameters have been derived.
Fischer (1993) determined that boiling point, the average packing fraction for
about 80 different liquids with different chemical natures was 1.1. Adopting
this value, the difference between liquid excess Gibbs energy and liquid
excess Helmholtz energy can be computed as:
(8)
The result is a predictive mixing rule for cubic equations of state. But the
original UNIFAC formulation gives the best performance for any binary pair
with interactions available from UNIFAC. Gassolvent interactions are
unavailable. However, it has poor accuracy for highly asymmetric such as CH4
nC10H22. To address the issue, the Li correction (Li et al., 1998) is
applied.
With the Li correction, the R and Q parameters for groups CH3, CH2, CH, and
C in UNIFAC are modified. R* and Q*, the effective values of R and Q, are
calculated based on the original values and n
c
, the number of alkyl carbons
(singlebonded carbons in CH3, CH2, CH, and C groups), as follows, for
n
c
<45:
At the University of Oldenburg in Germany, the UNIFAC groups were
extended with oftenoccurring gases. New group interactions were
determined from gassolvent data, specific to the RedlichKwongSoave
equationofstate. The new builtin parameters to the Aspen Physical Property
System are activated when using the PSRK equationofstate model.
The PSRK method has a lot in common with the HuronVidal mixing rules. The
mole fraction is dependent on the second virial coefficient and excess volume
is predicted. These are minor disadvantages.
References
K. Fischer, "Die PSRKMethode: Eine Zustandsgleichung unter Verwendung
des UNIFACGruppenbeitragsmodells," (Dsseldorf: VDI Fortschrittberichte,
Reihe 3: Verfahrenstechnik, Nr. 324, VDI Verlag GmbH, 1993).
88 2 Thermodynamic Property Models
T. Holderbaum and J. Gmehling, "PSRK: A Group Contribution Equationof
state based on UNIFAC," Fluid Phase Eq., Vol. 70, (1991), pp. 251265.
J. Li, K. Fischer, and J. Gmehling, "Prediction of vaporliquid equilibria for
asymmetric systems at low and high pressures with the PSRK model," Fluid
Phase Equilib., 1998, 143, 7182
WongSandler Mixing Rules
These mixing rules use a relationship between the excess Helmholtz energy
and equationofstate. They do not use a relationship between equationof
state properties and excess Gibbs energy, as in the HuronVidal mixing rules.
The pressureexplicit expression for the equationofstate is substituted in the
thermodynamic equation:
(1)
The Helmholtz energy is obtained by integration, A
E
is obtained by:
(2)
Where both A
i
* and A
m
are calculated by using equation 1. A
i
* and A
m
are
written in terms of equationofstate parameters.
Like Huron and Vidal, the limiting case of infinite pressure is used. This
simplifies the expressions for A
i
* and A
m
. Equation 2 becomes:
(3)
Where A depends on the equationofstate (see HuronVidal Mixing Rules).
Equation 3 is completely analogous to the HuronVidal mixing rule for the
excess Gibbs energy at infinite pressure. (See equation 4, HuronVidal Mixing
Rules.) The excess Helmholtz energy can be approximated by the excess
Gibbs energy at low pressure from any liquid activity coefficient model. Using
the Helmholtz energy permits another mixing rule for b than the linear mixing
rule. The mixing rule for b is derived as follows. The second virial coefficient
must depend quadratically on the mole fraction:
(4)
With:
(5)
The relationship between the equationofstate at low pressure and the virial
coefficient is:
(6)
2 Thermodynamic Property Models 89
(7)
Wong and Sandler discovered the following mixing rule to satisfy equation 4
(using equations 6 and 7):
The excess Helmholtz energy is almost independent of pressure. It can be
approximated by the Gibbs energy at low pressure. The difference between
the two functions is corrected by fitting k
ij
until the excess Gibbs energy from
the equationofstate (using the mixing rules 3 and 8) is equal to the excess
Gibbs energy computed by an activity coeffecient model. This is done at a
specific mole fraction and temperature.
This mixing rule accurately predicts the VLE of polar mixtures at high
pressures. UNIFAC or other activity coeffecient models and parameters from
the literature are used. Gas solubilities are not predicted. They must be
regressed from experimental data.
Unlike other (modified) HuronVidal mixing rules, the Wong and Sandler
mixing rule meets the theoretical limit at low pressure. The use of k
ij
does
influence excess molar volume behavior. For calculations where densities are
important, check whether they are realistic.
References
D. S. Wong and S. I. Sandler, "A Theoretically Correct New Mixing Rule for
Cubic Equations of State for Both Highly and Slightly Nonideal Mixtures,"
AIChE J., Vol. 38, (1992), pp. 671 680.
D. S. Wong, H. Orbey, and S. I. Sandler, "Equationofstate Mixing Rule for
Nonideal Mixtures Using Available Activity Coefficient Model Parameters and
That Allows Extrapolation over Large Ranges of Temperature and Pressure",
Ind Eng Chem. Res., Vol. 31, (1992), pp. 2033 2039.
H. Orbey, S. I. Sandler and D. S. Wong, "Accurate Equationofstate
Predictions at High Temperatures and Pressures Using the Existing UNIFAC
Model," Fluid Phase Eq., Vol. 85, (1993), pp. 41 54.
Activity Coefficient Models
The Aspen Physical Property System has the following builtin activity
coefficient models. This section describes the activity coefficient models
available.
Model Type
BromleyPitzer Electrolyte
ChienNull Regular solution, local composition
Constant Activity Coefficient Arithmetic
90 2 Thermodynamic Property Models
Model Type
COSMOSAC Regular solution
Electrolyte NRTL Electrolyte
ENRTLSAC Segment contribution, electrolyte
Hansen Regular solution
Ideal Liquid Ideal
NRTL (NonRandomTwoLiquid) Local composition
NRTLSAC Segment contribution
Pitzer Electrolyte
Polynomial Activity Coefficient Arithmetic
RedlichKister Arithmetic
ScatchardHildebrand Regular solution
Symmetric Electrolyte NRTL Electrolyte, pure fused salt
reference state for ions
ThreeSuffix Margules Arithmetic
UNIFAC Group contribution
UNIFAC (Lyngby modified) Group contribution
UNIFAC (Dortmund modified) Group contribution
UNIQUAC Local composition
Unsymmetric Electrolyte NRTL Electrolyte
Van Laar Regular solution
Wagner interaction parameter Arithmetic
Wilson Local composition
Wilson with Liquid Molar Volume Local composition
BromleyPitzer Activity Coefficient Model
The BromleyPitzer activity coefficient model is a simplified Pitzer activity
coefficient model with Bromley correlations for the interaction parameters.
See Working Equations for a detailed description. This model has predictive
capabilities. It can be used to compute activity coefficients for aqueous
electrolytes up to 6 molal ionic strength, but is less accurate than the Pitzer
model if the parameter correlations are used. The model should not be used
for mixedsolvent electrolyte systems.
The BromleyPitzer model in the Aspen Physical Property System involves
usersupplied parameters, used in the calculation of binary parameters for the
electrolyte system.
Parameters 
(0)
, 
(1)
, 
(2)
, 
(3)
, and u have five elements to account for
temperature dependencies. Elements P1 through P5 follow the temperature
dependency relation:
Where:
T
ref
= 298.15K
2 Thermodynamic Property Models 91
The user must:
 Supply these elements using a Methods  Parameters  Binary  T
Dependent form.
 Specify Comp ID i and Comp ID j on this form, using the same order that
appears on the Components  Specifications  Selection sheet form.
Parameter Name Symbol No. of ElementsDefault Units
Ionic Unary Parameters
GMBPB

ion
1 0
GMBPD
o
ion
1 0
CationAnion Parameters
GMBPB0

(0)
5 0
GMBPB1

(1)
5 0
GMBPB2

(2)
5 0
GMBPB3

(3)
5 0
CationCation Parameters
GMBPTH
u
cc'
5 0
AnionAnion Parameters
GMBPTH
u
aa'
5 0
MoleculeIon and MoleculeMolecule Parameters
GMBPB0

(0)
5 0
GMBPB1

(1)
5 0
Working Equations
The complete Pitzer equation (Frst and Renon, 1982) for the excess Gibbs
energy is (see also equation 4):
(1)
Where:
G
E
= Excess Gibbs energy
R = Gas constant
T = Temperature
n
w
= Kilograms of water
z
i
= Charge number of ion i
= molality of ion i
92 2 Thermodynamic Property Models
Where:
x
i
= Mole fraction of ion i
x
w
= Mole fraction of water
M
w
= Molecular weight of water (g/mol)
n
i
= Moles of ion i
B, C, u and + are interaction parameters, and f(I) is an electrostatic term as
a function of ionic strength; these terms are discussed in Pitzer Activity
Coefficient Model, which has a detailed discussion of the Pitzer model.
The C term and the + term are dropped from equation 1 to give the
simplified Pitzer equation.
(2)
Where:
B
ij
=
f(
ij
(0)
,
ij
(1)
,
ij
(2)
,
ij
(3)
)
Therefore, the simplified Pitzer equation has two types of binary interaction
parameters,  's and u''s. There are no ternary interaction parameters with
the simplified Pitzer equation.
Note that the Pitzer model parameter databank described in Physical Property
Data, Chapter 1, is not applicable to the simplified Pitzer equation.
A builtin empirical correlation estimates the 
(0)
and 
(1)
parameters for
cationanion pairs from the Bromley ionic parameters, 
ion
and o
ion
(Bromley,
1973). The estimated values of 
(0)
's and 
(1)
's are overridden by the user's
input. For parameter naming and requirements, see BromleyPitzer Activity
Coefficient Model.
References
L.A. Bromley, "Thermodynamic Properties of Strong Electrolytes in Aqueous
Solution, " AIChE J., Vol. 19, No. 2, (1973), pp. 313 320.
W. Frst and H. Renon, "Effects of the Various Parameters in the Application
of Pitzer's Model to SolidLiquid Equilibrium. Preliminary Study for Strong 11
Electrolytes," Ind. Eng. Chem. Process Des. Dev., Vol. 21, No. 3, (1982),
pp. 396400.
Parameter Conversion
For nm electrolytes, n and m>1 (22, 23, 34, and so on), the parameter

(3)
corresponds to Pitzer's 
(1)
; 
(2)
is the same in both Aspen Physical
Property System and original Pitzer models. Pitzer refers to the nm
electrolyte parameters as 
(1)
, 
(2)
, 
(0)
. 
(0)
and 
(2)
retain their meanings in
both models, but Pitzer's 
(1)
is Aspen Physical Property System 
(3).
Be
careful to make this distinction when entering nm electrolyte parameters.
2 Thermodynamic Property Models 93
ChienNull
The ChienNull model calculates liquid activity coefficients and it can be used
for highly nonideal systems. The generalized expression used in its derivation
can be adapted to represent other well known formalisms for the activity
coefficient by properly defining its binary terms. This characteristic allows the
model the use of already available binary parameters regressed for those
other liquid activity models with thermodynamic consistency.
The equation for the ChienNull liquid activity coeficient is:
Where:
R
ji
= A
ji
/ A
ij
A
ii
= 0
A
ij
= a
ij
+ b
ij
/ T
Subscripts i and j are component indices.
The choice of model and parameters can be set for each binary pair
constituting the process mixture by assigning the appropriate value to the
ICHNUL parameter.
The Regular Solution and ScatchardHamer models are regained by
substituting in the general expression (ICHNUL = 1 or 2).
V
ji
= S
ji
= V
j
*,l
/ V
i
*,l
Where:
V
j
*,l
= Liquid molar volume of component i
The ChienNull activity coefficient model collapses to the Margules liquid
activity coefficient expression by setting (ICHNUL = 3):
V
ji
= S
ji
= 1
The Van Laar Liquid activity coefficient model is obtained when the V and S
parameters in the ChienNull models are set to the ratio of the cross terms of
A (ICHNUL = 4:)
V
ji
= S
ji
= A
ji
/ A
ij
Finally, the Renon or NRTL model is obtained when we make the following
susbstitutions in the ChienNull expression for the liquid activity (ICHNUL =
5).
S
ji
= R
ji
A
ji
/ A
ij
A
ji
= 2t
ji
G
ji
V
ji
= G
ji
94 2 Thermodynamic Property Models
The following are defined for the NonRandom TwoLiquid activity coefficient
model, where:
t
ij
= a
ij
+ b
ij
/ T
C
ij
= c
ij
+ d
ij
(T  273.15 K)
c
ji
= c
ij
d
ji
= d
ij
The binary parameters CHNULL/1, CHNULL/2, and CHNULL/3 can be
determined from regression of VLE and/or LLE data. Also, if you have
parameters for many of the mixture pairs for the Margules, Van Laar,
ScatchardHildebrand, and NRTL (NonRandomTwoLiquid) activity models,
you can use them directly with the ChienNull activity model after selecting
the proper code (ICHNUL) to identify the source model and entering the
appropriate activity model parameters.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
ICHNUL 3 1 6
CHNULL/1 a
ij
0
CHNULL/2 b
ij
0
CHNULL/3 V
ij
0
The parameter ICHNUL is used to identify the activity model parameters
available for each binary pair of interest. The following values are allowed for
ICHNUL:
ICHNUL = 1 or 2, sets the model to the ScatchardHamer or regular solution
model for the associated binary;
ICHNUL = 3, sets the model to the ThreeSuffix Margules activity model for
the associated binary;
ICHNUL = 4, sets the model to the Van Laar formalism for the activity model
for the associated binary;
ICHNUL = 5, sets the model to the NRTL (Renon) formalism for the activity
model for the associated binary.
ICHNUL = 6, sets the model to the full ChienNull formalism for the activity
model for the associated binary.
When you specify a value for the ICHNUL parameter that is different than the
default, you must enter the appropriate binary model parameters for the
chosen activity model directly. The routine will automatically convert the
expressions and parameters to conform to the ChienNull formulation.
Constant Activity Coefficient
This approach is used exclusively in metallurgical applications where multiple
liquid and solid phases can coexist. You can assign any value to the activity
coefficient of component i. Use the Properties  Parameters  Pure
2 Thermodynamic Property Models 95
Component  Scalar form.
The equation is:
i
= a
i
Parameter
Name/Element
Symbol Default MDS Upper
Limit
Lower
Limit
Units
GMCONS a
i
1.0 x
COSMOSAC
CosmoSAC is a solvation model that describes the electric fields on the
molecular surface of species that are polarizable. It requires a fairly
complicated quantum mechanical calculation, but this calculation must be
done only once for a particular molecule; then the results can be stored. In its
final form, it uses individual atoms as the building blocks for predicting phase
equilibria instead of functional groups. This model formulation provides a
considerably larger range of applicability than groupcontribution methods.
The calculation for liquid nonideality is only slightly more computationally
intensive than activitycoefficient models such as NRTL or UNIQUAC. Cosmo
SAC complements the UNIFAC groupcontribution method, because it is
applicable to virtually any mixture.
The CosmoSAC model calculates liquid activity coefficients. The equation for
the CosmoSAC model is:
With
96 2 Thermodynamic Property Models
Where:
i
=Activity coefficient of component i
i
SG
=StavermanGuggenheim model for combinatorial
contribution to
i
I
i
(o
m
)
=
Segment activity coefficient of segment o
m
in
component i
I
S
(o
m
)
=
Segment activity coefficient of segment o
m
in
solvent mixture
p
i
(o
m
)
=Sigma profile of component i
p
i
(o
m
)
=Sigma profile of solvent mixture
o
=Surface charge density
AW(o
m
,o
n
)
=
Exchange energy between segments o
m
and o
n
AW
HB
(o
m
,o
n
)
=Hydrogenbonding contribution to exchange energy
between segments o
m
and o
n
z =Coordination number, 10
V
i
=Molecular volume of component i
A
i
=Molecular surface area of component i
a
eff
=Standard segment surface area, 7.50
2
V
eff
=Standard component volume, 66.69
3
A
eff
=Standard component surface area, 79.53
2
o'
=Misfit energy constant
2 Thermodynamic Property Models 97
The CosmoSAC model does not require binary parameters. For each
component, it has six input parameters. CSACVL is the component volume
parameter which is always defined in cubic angstroms, regardless of chosen
units sets. SGPRF1 to SGPRF5 are five molecular component sigma profile
parameters; each can store up to 12 points of sigma profile values. All six
input parameters are obtained from COSMO calculation. The Aspen Physical
Property System includes a database of sigma profiles for over 1400
compounds from Mullins et al. (2006). The parameters were obtained by
permission from the Virginia Tech Sigma Profile Database website
(http://www.design.che.vt.edu/VT2004.htm). Aspen Technology, Inc. does
not claim proprietary rights to these parameters.
Note: Starting in version V7.2, additional parameters SGPRF6 and SGPRF7
are available in Aspen Plus and Aspen Properties, as part of a planned
expansion to allow larger sigma profiles to be used. In V7.3.2, three
additional sets of parameters are added as part of further expansion to model
electrolyte systems. They are the component ionic sigma profile from
SGPRE1 through SGPRE7, the molecular sigma profile for ionic pair
from SGPBF1 through SGPBF7, and the ionic sigma profile for ionic pair from
SGPBE1 through SGPBE7. However, the model is not yet updated to use
these parameters and you should not yet try to use sigma profiles larger than
51 elements.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CSACVL V
i
x VOLUME (
3
)
SGPRF1 A
i
p
i
(112) x
SGPRF2 A
i
p
i
(1324) x
SGPRF3 A
i
p
i
(2536) x
SGPRF4 A
i
p
i
(3748) x
SGPRF5 A
i
p
i
(4951) x
Option Codes
The primary version of COSMOSAC is the model by Lin and Sandler (2002).
Two other versions are available using an option code, as detailed in the table
below:
Option
Code
Description
1 COSMOSAC model by Lin and Sandler (2002)
2 COSMORS model by Klamt (1995)
3 Lin and Sandler model with modified exchange energy (Lin
et al., 2002)
References
A. Klamt, "Conductorlike Screening Model for Real Solvents: A New Approach
to the Quantitative Calculation of Solvation Phenomena," J. Phys. Chem. 99,
2224 (1995).
S.T. Lin, P. M. Mathias, Y. Song, C.C. Chen, and S. I. Sandler,
"Improvements of PhaseEquilibrium Predictions for HydrogenBonding
98 2 Thermodynamic Property Models
Systems from a New Expression for COSMO Solvation Models," presented at
the AIChE Annual Meeting, Indianapolis, IN, 38 November (2002).
S.T. Lin and S. I. Sandler, "A Priori Phase Equilibrium Prediction from a
Segment Contribution Solvation Model," Ind. Eng. Chem. Res. 41, 899
(2002).
E. Mullins, et al. "SigmaProfile Database for Using COSMOBased
Thermodynamic Methods," Ind. Eng. Chem. Res. 45, 4389 (2006).
Electrolyte NRTL Activity Coefficient Model
(GMENRTL)
The Electrolyte NRTL activity coefficient model (GMENRTL) is a versatile
model for the calculation of activity coefficients. Using binary and pair
parameters, the model can represent aqueous electrolyte systems as well as
mixed solvent electrolyte systems over the entire range of electrolyte
concentrations. This model can calculate activity coefficients for ionic species
and molecular species in aqueous electrolyte systems as well as in mixed
solvent electrolyte systems. The model reduces to the wellknown NRTL
model when electrolyte concentrations become zero (Renon and Prausnitz,
1969).
The electrolyte NRTL model uses the infinite dilution aqueous solution as the
reference state for ions. It adopts the Born equation to account for the
transformation of the reference state of ions from the infinite dilution mixed
solvent solution to the infinite dilution aqueous solution.
Water must be present in the electrolyte system in order to compute the
transformation of the reference state of ions. Thus, it is necessary to
introduce a trace amount of water to use the model for nonaqueous
electrolyte systems.
The Aspen Physical Property System uses the electrolyte NRTL model to
calculate activity coefficients, enthalpies, and Gibbs energies for electrolyte
systems. Model development and working equations are provided in
Theoretical Basis and Working Equations.
The adjustable parameters for the electrolyte NRTL model include the:
 Pure component dielectric constant coefficient of nonaqueous solvents
 Born radius of ionic species
 NRTL parameters for moleculemolecule, moleculeelectrolyte, and
electrolyteelectrolyte pairs
The pure component dielectric constant coefficients of nonaqueous solvents
and Born radius of ionic species are required only for mixedsolvent
electrolyte systems. The temperature dependency of the dielectric constant of
solvent B is:
2 Thermodynamic Property Models 99
Each type of electrolyte NRTL parameter consists of both the nonrandomness
factor, o, and energy parameters, t. The temperature dependency relations
of the electrolyte NRTL parameters are:
 MoleculeMolecule Binary Parameters:
 ElectrolyteMolecule Pair Parameters:
 ElectrolyteElectrolyte Pair Parameters:
For the electrolyteelectrolyte pair parameters, the two electrolytes must
share either one common cation or one common anion:
Where:
T
ref
= 298.15K
Many parameter pairs are included in the electrolyte NRTL model parameter
databank (see Physical Property Data, Chapter 1).
Parameter
Name
Symbol No. of
Elements
Default MDS Units
Dielectric Constant Unary Parameters
CPDIEC A
B
1
B
B
1 0
C
B
1 298.15 TEMPERATURE
B
B'B
0 x TEMPERATURE
NRTL/3
o
BB'
= o
B'B
.3 x
NRTL/4 0 x TEMPERATURE
NRTL/5 F
BB'
0 x TEMPERATURE
F
B'B
0 x TEMPERATURE
100 2 Thermodynamic Property Models
Parameter
Name
Symbol No. of
Elements
Default MDS Units
Dielectric Constant Unary Parameters
NRTL/6 G
BB'
0 x TEMPERATURE
G
B'B
0 x TEMPERATURE
ElectrolyteMolecule Pair Parameters
GMELCC C
ca,B
1 0 x
C
B,ca
1 0 x
GMELCD D
ca,B
1 0 x TEMPERATURE
D
B,ca
1 0 x TEMPERATURE
GMELCE E
ca,B
1 0 x
E
B,ca
1 0 x
GMELCN
o
ca,B
= o
B,ca
1 .2 x
ElectrolyteElectrolyte Pair Parameters
GMELCC C
ca',ca''
1 0 x
C
ca'',ca'
1 0 x
C
c'a,c''a
1 0 x
C
c''a,c'a
1 0 x
GMELCD D
ca',ca''
1 0 x TEMPERATURE
D
ca'',ca'
1 0 x TEMPERATURE
D
c'a,c''a
1 0 x TEMPERATURE
D
c''a,c'a
1 0 x TEMPERATURE
GMELCE E
ca',ca''
1 0 x
E
ca'',ca'
1 0 x
E
c'a,c''a
1 0 x
E
c''a,c'a
1 0 x
GMELCN
o
ca',ca''
= o
ca'',ca'
1 .2 x
o
c'a,c''a
= o
c''a,c'a
1 .2 x
Certain Electrolyte NRTL activity coefficient model parameters are used with
reciprocal temperature terms:
 CPDIEC
 NRTL/2
 GMELCD for electrolyteelectrolyte or electrolytemolecule pairs
When any of these parameters is specified, absolute temperature units are
used for the calculations in this model.
Reference: H. Renon, and J.M. Prausnitz, "Local Compositions in
Thermodynamic Excess Functions for Liquid Mixtures", AIChE J., Vol. 14, No.
1, (1968), pp. 135144.
Theoretical Basis and Working Equations
In this section, the theoretical basis of the model is explained and the working
equations are given. The different ways parameters can be obtained are
discussed with references to the databank directories and the Data
2 Thermodynamic Property Models 101
Regression System (DRS). The parameter requirements of the model are
given in Electrolyte NRTL Activity Coefficient Model.
Development of the Model
The Electrolyte NRTL model was originally proposed by Chen et al., for
aqueous electrolyte systems. It was later extended to mixed solvent
electrolyte systems (Mock et al., 1984, 1986). The model is based on two
fundamental assumptions:
 The likeion repulsion assumption: states that the local composition of
cations around cations is zero (and likewise for anions around anions).
This is based on the assumption that the repulsive forces between ions of
like charge are extremely large. This assumption may be justified on the
basis that repulsive forces between ions of the same sign are very strong
for neighboring species. For example, in salt crystal lattices the immediate
neighbors of any central ion are always ions of opposite charge.
 The local electroneutrality assumption: states that the distribution of
cations and anions around a central molecular species is such that the net
local ionic charge is zero. Local electroneutrality has been observed for
interstitial molecules in salt crystals.
Chen proposed an excess Gibbs energy expression which contains two
contributions: one contribution for the longrange ionion interactions that
exist beyond the immediate neighborhood of a central ionic species, and the
other related to the local interactions that exist at the immediate
neighborhood of any central species.
The unsymmetric PitzerDebyeHckel model and the Born equation are used
to represent the contribution of the longrange ionion interactions, and the
NonRandom Two Liquid (NRTL) theory is used to represent the local
interactions. The local interaction contribution model is developed as a
symmetric model, based on reference states of pure solvent and pure
completely dissociated liquid electrolyte. The model is then normalized by
infinite dilution activity coefficients in order to obtain an unsymmetric model.
This NRTL expression for the local interactions, the PitzerDebyeHckel
expression, and the Born equation are added to give equation 1 for the
excess Gibbs energy (see the following note).
(1)
This leads to
(2)
Note: The notation using * to denote an unsymmetric reference state is well
accepted in electrolyte thermodynamics and will be maintained here. The
reader should be warned not to confuse it with the meaning of * in classical
thermodynamics according to IUPAC/ISO, referring to a pure component
property. In fact in the context of G or , the asterisk as superscript is never
used to denote pure component property, so the risk of confusion is minimal.
For details on notation, see Chapter 1 of Physical Property Methods.
102 2 Thermodynamic Property Models
References
C.C. Chen, H.I. Britt, J.F. Boston, and L.B. Evans, "Local Compositions Model
for Excess Gibbs Energy of Electrolyte Systems: Part I: Single Solvent, Single
Completely Dissociated Electrolyte Systems:, AIChE J., Vol. 28, No. 4, (1982),
p. 588596.
C.C. Chen, and L.B. Evans, "A Local Composition Model for the Excess Gibbs
Energy of Aqueous Electrolyte Systems," AIChE J., Vol. 32, No. 3, (1986), p.
444459.
B. Mock, L.B. Evans, and C.C. Chen, "Phase Equilibria in MultipleSolvent
Electrolyte Systems: A New Thermodynamic Model," Proceedings of the 1984
Summer Computer Simulation Conference, p. 558.
B. Mock, L.B. Evans, and C.C. Chen, "Thermodynamic Representation of
Phase Equilibria of MixedSolvent Electrolyte Systems," AIChE J., Vol. 32, No.
10, (1986), p. 16551664.
LongRange Interaction Contribution
The PitzerDebyeHckel formula, normalized to mole fractions of unity for
solvent and zero for electrolytes, is used to represent the longrange
interaction contribution.
with
(3)
(4)
(5)
Where:
x
i
= Mole fraction of component i
M
s
= Molecular weight of the solvent
A
= DebyeHckel parameter
N
A
= Avogadro's number
d
s
= Mass density of solvent
Q
e
= Electron charge
c
s
= Dielectric constant of the solvent
T = Temperature
k = Boltzmann constant
I
x
= Ionic strength (mole fraction scale)
x
i
= Mole fraction of component i
2 Thermodynamic Property Models 103
z
i
= Charge number of ion i
I
*
= Unsymmetric activity coefficient of an ionic species I
I
*lc
= NRTL term
I
*PDH
= PitzerDebyeHckel term
I
*FH
= FloryHuggins term
= Aqueousphase infinitedilution reference state
I
i
= Activity coefficient of conceptual segment i
r
I
= Total segment number of component I
x
I
= Mole fraction of component I
r
I,i
= Number of conceptual segment i containing in component I
x
i
= Segment mole fraction of conceptual segment i in mixtures
o
ij
= NRTL binary nonrandomness factor parameter for
conceptual segments
t
ij
= NRTL binary interaction energy parameter for conceptual
segments
A
= DebyeHckel parameter
i
= Activity coefficient of component i
V
i
= Molar volume of component i
o
i
d
= Hansen solubility parameter of component i for nonpolar
effect
o
i
p
= Hansen solubility parameter of component i for polar effect
o
i
h
= Hansen solubility parameter of component i for hydrogen
bonding effect

i
= Volume fraction of component i
2 Thermodynamic Property Models 117
x
i
= Mole fraction of component i
R = Gas constant
T = Temperature
The Hansen model does not require binary parameters. For each component,
it has four input parameters.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DELTAD
o
i
d
x PRESSURE^0.5
DELTAP
o
i
p
x PRESSURE^0.5
DELTAH
o
i
h
x PRESSURE^0.5
HANVOL V
i
x VOLUME
Option codes
The Hansen volume is implemented as an input parameter. If the Hansen
volume is not input by the user it will be calculated by an Aspen Plus internal
method. You can also request the Aspen Plus method using Option Codes in
Aspen Plus Interface. The table below lists the option codes.
First Option Code in Hansen model
0 Hansen volume input by user (default)
Other values Hansen volume calculated by Aspen Plus
Reference
Frank, T. C.; Downey, J. R.; Gupta, S. K. "Quickly Screen Solvents for
Organic Solids," Chemical Engineering Progress 1999, December, 41.
Hansen, C. M. Hansen Solubility Parameters: A Users Handbook; CRC Press,
2000.
Ideal Liquid
This model is used in Raoult's law. It represents ideality of the liquid phase.
This model can be used for mixtures of hydrocarbons of similar carbon
number. It can be used as a reference to compare the results of other activity
coefficient models.
The equation is:
ln
i
= 0
NRTL (NonRandom TwoLiquid)
The NRTL model calculates liquid activity coefficients for the following
property methods: NRTL, NRTL2, NRTLHOC, NRTLNTH, and NRTLRK. It is
recommended for highly nonideal chemical systems, and can be used for VLE
118 2 Thermodynamic Property Models
and LLE applications. The model can also be used in the advanced equation
ofstate mixing rules, such as WongSandler and MHV2.
The equation for the NRTL model is:
for T
lower
s T s T
upper
Where:
G
ij
=
t
ij
=
o
ij
=
t
ii
= 0
G
ii
= 1
a
ij
, b
ij
, e
ij
, and f
ij
are unsymmetrical. That is, a
ij
may not be equal to a
ji
, etc.
Recommended c
ij
Values for Different Types of Mixtures
c
ij
Mixtures
0.30 Nonpolar substances; nonpolar with polar nonassociated liquids; small
deviations from ideality
0.20 Saturated hydrocarbons with polar nonassociated liquids and systems that
exhibit liquidliquid immiscibility
0.47 Strongly selfassociated substances with nonpolar substances
The binary parameters a
ij
, b
ij
, c
ij
, d
ij
, e
ij
, and f
ij
can be determined from VLE
and/or LLE data regression. The Aspen Physical Property System has a large
number of builtin binary parameters for the NRTL model. The binary
parameters have been regressed using VLE and LLE data from the Dortmund
Databank. The binary parameters for the VLE applications were regressed
using the ideal gas, RedlichKwong, and Hayden O'Connell equations of state.
See Physical Property Data, Chapter 1, for details.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
NRTL/1 a
ij
0 x 100.0 100.0
NRTL/2 b
ij
0 x 30000 30000.0 TEMPERATURE
NRTL/3 c
ij
0.30 x 0.0 1.0
NRTL/4 d
ij
0 x 0.02 0.02 TEMPERATURE
NRTL/5 e
ij
0 x TEMPERATURE
NRTL/6 f
ij
0 x TEMPERATURE
NRTL/7 T
lower
0 x TEMPERATURE
NRTL/8 T
upper
1000 x TEMPERATURE
2 Thermodynamic Property Models 119
Note: If any of b
ij
, d
ij
, or e
ij
is nonzero, absolute temperature units are
assumed for b
ij
, d
ij
, e
ij
, and f
ij
. Otherwise, user input units for temperature are
used. The temperature limits are always interpreted in user input units.
The NRTL2 property method uses data set 2 for NRTL. All other NRTL
methods use data set 1.
References
H. Renon and J.M. Prausnitz, "Local Compositions in Thermodynamic Excess
Functions for Liquid Mixtures," AIChE J., Vol. 14, No. 1, (1968), pp. 135
144.
NRTLSAC Model
NRTLSAC (patent pending) is a segment contribution activity coefficient
model, derived from the Polymer NRTL model and extended to handle
electrolytes, but usable in Aspen Plus or Aspen Properties without Aspen
Polymers. NRTLSAC can be used for fast, qualitative estimation of the
solubility of complex organic compounds in common solvents. It can also be
used as a general activity coefficient model in Aspen Plus, Aspen Properties,
and HYSYS.
Conceptually, the model treats the liquid nonideality of mixtures containing
complex organic molecules (solute) and small molecules (solvent) in terms of
interactions between three pairwise interacting conceptual segments:
hydrophobic segment (x), hydrophilic segment (z), and polar segments (y
and y+). In practice, these conceptual segments become the molecular
descriptors used to represent the molecular surface characteristics of each
solute or solvent molecule. Hexane, water, and acetonitrile are selected as
the reference molecules for the hydrophobic, hydrophilic, and polar segments,
respectively. The molecular parameters for all other solvents can be
determined by regression of available VLE or LLE data for binary systems of
solvent and the reference molecules or their substitutes.
The treatment results in four componentspecific molecular parameters:
hydrophobicity X, hydrophilicity Z, and polarity Y and Y+. The two types of
polar segments, Y and Y+, are used to reflect the wide variations of
interactions between polar molecules and water.
NRTLSAC can also be used to model electrolyte systems. In this case, an
electrolyte segment e, corresponding to the electrolyte parameter E, is
introduced. This conceptual segment e completely dissociates to a cationic
segment (c) and an anionic segment (a), both of unit charge. All electrolytes,
organic or inorganic, symmetric or unsymmetric, univalent or multivalent, are
to be represented with this conceptual 11 electrolyte segment e together
with the previously defined hydrophobic segment x, polar segments y and
y+, and hydrophilic segment z in NRTLSAC. The reference state for ions is by
default an unsymmetric state based on infinite dilution in aqueous solution,
but an option code is available to select the symmetric state of pure fused
salts. When there are no electrolytes present, the segment e is unused and
the current model reduces to the nonelectrolyte version of NRTLSAC present
in earlier releases.
120 2 Thermodynamic Property Models
The conceptual segment contribution approach in NRTLSAC represents a
practical alternative to the UNIFAC functional group contribution approach.
This approach is suitable for use in the industrial practice of carrying out
measurements for a few selected solvents and then using NRTLSAC to
quickly predict other solvents or solvent mixtures and to generate a list of
suitable solvent systems.
The NRTLSAC model calculates liquid activity coefficients.
Note: This is the updated version of NRTLSAC, represented with property
model GMNRTLS and property method NRTLSAC. This version does not
require the specification of components as oligomers. For the old version, see
NRTLSAC for Segments/Oligomers and ENRTLSAC.
For the model equations, see NRTLSAC Model Derivation.
Parameters used in NRTLSAC
Each component can have up to five parameters, r
x,I
, r
y,I
, r
y+,I
, r
z,I
, and r
e,I
,
representing the equivalent number of segments of each type for the NRTL
activity coefficient model. Only one or two of these molecular parameters are
needed for most solvents in practice. These parameters are implemented
together as pure parameter XYZE with five elements representing these five
parameters. Values for this parameter are available for many common
solvents in the NRTLSAC databank.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper Limit Units
XYZE/1 r
x,I
XYZE/2 r
y,I
XYZE/3 r
y+,I
XYZE/4 r
z,I
XYZE/5 r
e,I
Electrolytes must be modeled as ion pairs in this system, while the individual
ions are components in the Aspen Physical Property System, so for these ion
pairs, the five parameters are stored in binary parameter BXYZE which has
elements corresponding to those of XYZE.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper Limit Units
BXYZE/1 r
x,CA
BXYZE/2 r
y,CA
BXYZE/3 r
y+,CA
BXYZE/4 r
z,CA
BXYZE/5 r
e,CA
The conceptual segment numbers of a cationic component and
an anionic component come from the dissociation of their
corresponding electrolyte component CA, as defined by the chemical equation
describing the dissociation of the electrolyte:
2 Thermodynamic Property Models 121
with
Since an electrolyte component CA can be measured by up to five conceptual
segments , we can calculate r
i,C
and r
i,A
as follows for systems of
single electrolyte.
For an electrolyte system where multielectrolytes may be generated, a
simple mixing rule is used:
where Y
C
is a cationic charge composition fraction and Y
A
is an anionic charge
composition fraction; they are defined as follows:
Notice that electrolyte here is meant to represent an ionpair composed of a
cationic component and an anionic component in the solutions. The result is
that electrolytes are generated from all possible combinations of ions in the
solution; each generated electrolyte is not necessarily associated with an ion
pair through the dissociation.
Option Codes in NRTLSAC
Three option codes are available for NRTLSAC to select the reference state
and to optionally exclude the FloryHuggins and longrange interaction terms:
122 2 Thermodynamic Property Models
Option CodeValue Meaning
1 0 Reference state for ions is unsymmetric: infinite dilution in
aqueous solution (default)
2 Reference state for ions is symmetric: pure fused salts
2 0 FloryHuggins term included (default)
1 FloryHuggins term removed
3 0 Longrange interaction term included (default)
1 Longrange interaction term removed
References
C.C. Chen and Y. Song, "Extension of Nonrandom TwoLiquid Segment
Activity Coefficient Model for Electrolytes," Ind. Eng. Chem. Res., 2005, 44,
8909.
Y. Song and C.C. Chen, "Symmetric Nonrandom TwoLiquid Segment Activity
Coefficient Model for Electrolytes," Ind. Eng. Chem. Res., 2009, 48, 5522.
NRTLSAC Reference States
The NRTLSAC activity coefficient model for component I is composed of the
local composition term, ln
I
lc
, the PitzerDebyeHckel longrange interaction
term, ln
I
PDH
, and the FloryHuggins term,
I
FH
:
(1)
This equation needs to be normalized based on the reference states of
molecular and ionic components.
Reference state for molecular components
The reference state for a molecular component is defined as follows:
(2)
This definition is the socalled standard state of pure liquids for molecular
components and it is also called the symmetric reference state for molecular
components.
Reference state for ionic components
The standard state of pure liquids is hypothetical for ionic components in
electrolyte systems. The symmetric reference state is defined as the pure
fused salt state of each electrolyte component in the system.
However, the conventional reference state for ionic components is the
infinitedilution activity coefficient in pure water; it is also called the
unsymmetric reference state for ionic components. In NRTLSAC model, we
will consider both of these reference states; the unsymmetric state is the
default.
2 Thermodynamic Property Models 123
Pure fused salt state of an electrolyte component
For an electrolyte component CA, the pure fused salt state can be defined as
follows:
(3)
(4)
where
I
= Activity coefficient of component I
I
C
=
I
FH
=
FloryHuggins term for combinatorial contribution to
I
I
R
=
I
lc
= NRTL term for local composition interaction contribution to
I

I
= Segment mole fraction of component I
p
I
= Effective component size parameter
s
I
and c
I
= Empirical parameters for p
I
r
I
= Total segment number of component I
x
I
= Mole fraction of component I
r
I,m
= Number of conceptual segment m containing in component
I
x
i
= Segment mole fraction of conceptual segment i in mixtures
o
im
= NRTL binary nonrandomness factor parameter for
conceptual segments
t
im
= NRTL binary interaction energy parameter for conceptual
segments
136 2 Thermodynamic Property Models
NRTL binary parameters for conceptual segments
The NRTL binary parameters between conceptual segments in NRTLSAC are
determined by available VLE and LLE data between reference molecules
defined above.
Segment 1 x x y y+ x
Segment 2 y z z z y+
t
12
1.643 6.547 2.000 2.000 1.643
t
21
1.834 10.949 1.787 1.787 1.834
o
12
= o
21
0.2 0.2 0.3 0.3 0.2
Parameters used in NRTLSAC
Each component can have up to four parameters, r
I,x
, r
I,y
, r
I,y+
, and r
I,z
although only one or two of these molecular parameters are needed for most
solvents in practice. Since conceptual segments apply to all molecules, these
four molecular parameters are implemented together as a binary parameter,
NRTLXY(I, m) where I represents a component (molecule) index and m
represents a conceptual segment index.
In addition, the FloryHuggins size parameter, FHSIZE , is used in NRTLSAC
to calculate the effective component size parameter, p
I
. The FloryHuggins
combinatorial term can be turned off by setting c
I
= 0 for each component in
mixtures.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units Comment
NRTLXY r
I,m
Binary,
symmetric
FHSIZE/1 s
I
1.0 1E15 1E15 Unary
FHSIZE/2
c
I
1.0 1E10 1E10 Unary
Option codes
The FloryHuggins term is included by default in the NRTLSAC model. You can
remove this term using the first option code. The table below lists the values
for this option code.
0 FloryHuggins term included (default)
Others FloryHuggins term removed
NRTLSAC molecular parameters for common solvents
The molecular parameters are identified for 62 solvents and published.
Solvent name r
I,x
r
I,y
r
I,y+
r
I,z
ACETICACID 0.045 0.164 0.157 0.217
ACETONE 0.131 0.109 0.513
ACETONITRILE 0.018 0.131 0.883
ANISOLE 0.722
BENZENE 0.607 0.190
2 Thermodynamic Property Models 137
Solvent name r
I,x
r
I,y
r
I,y+
r
I,z
1BUTANOL 0.414 0.007 0.485
2BUTANOL 0.335 0.082 0.355
NBUTYLACETATE 0.317 0.030 0.330
METHYLTERTBUTYLETHER 1.040 0.219 0.172
CARBONTETRACHLORIDE 0.718 0.141
CHLOROBENZENE 0.710 0.424
CHLOROFORM 0.278 0.039
CUMENE 1.208 0.541
CYCLOHEXANE 0.892
1,2DICHLOROETHANE 0.394 0.691
1,1DICHLOROETHYLENE 0.529 0.208
1,2DICHLOROETHYLENE 0.188 0.832
DICHLOROMETHANE 0.321 1.262
1,2DIMETHOXYETHANE 0.081 0.194 0.858
N,NDIMETHYLACETAMIDE 0.067 0.030 0.157
N,NDIMETHYLFORMAMIDE 0.073 0.564 0.372
DIMETHYLSULFOXIDE 0.532 2.890
1,4DIOXANE 0.154 0.086 0.401
ETHANOL 0.256 0.081 0.507
2ETHOXYETHANOL 0.071 0.318 0.237
ETHYLACETATE 0.322 0.049 0.421
ETHYLENEGLYCOL 0.141 0.338
DIETHYLETHER 0.448 0.041 0.165
ETHYLFORMATE 0.257 0.280
FORMAMIDE 0.089 0.341 0.252
FORMICACID 0.707 2.470
NHEPTANE 1.340
NHEXANE 1.000
ISOBUTYLACETATE 1.660 0.108
ISOPROPYLACETATE 0.552 0.154 0.498
METHANOL 0.088 0.149 0.027 0.562
2METHOXYETHANOL 0.052 0.043 0.251 0.560
METHYLACETATE 0.236 0.337
3METHYL1BUTANOL 0.419 0.538 0.314
METHYLBUTYLKETONE 0.673 0.224 0.469
METHYLCYCLOHEXANE 1.162 0.251
METHYLETHYLKETONE 0.247 0.036 0.480
METHYLISOBUTYLKETONE 0.673 0.224 0.469
ISOBUTANOL 0.566 0.067 0.485
138 2 Thermodynamic Property Models
Solvent name r
I,x
r
I,y
r
I,y+
r
I,z
NMETHYL2PYRROLIDONE 0.197 0.322 0.305
NITROMETHANE 0.025 1.216
NPENTANE 0.898
1PENTANOL 0.474 0.223 0.426 0.248
1PROPANOL 0.375 0.030 0.511
ISOPROPYLALCOHOL 0.351 0.070 0.003 0.353
NPROPYLACETATE 0.514 0.134 0.587
PYRIDINE 0.205 0.135 0.174
SULFOLANE 0.210 0.457
TETRAHYDROFURAN 0.235 0.040 0.320
1,2,3,4TETRAHYDRONAPHTHALENE0.443 0.555
TOLUENE 0.604 0.304
1,1,1TRICHLOROETHANE 0.548 0.287
TRICHLOROETHYLENE 0.426 0.285
MXYLENE 0.758 0.021 0.316
WATER 1.000
TRIETHYLAMINE 0.557 0.105
1OCTANOL 0.766 0.032 0.624 0.335
Reference
C.C. Chen and Y. Song, "Solubility Modeling with a Nonrandom TwoLiquid
Segment Activity Coefficient Model," Ind. Eng. Chem. Res. 43, 8354 (2004).
Using NRTLSAC
NRTLSAC (patent pending) is a segment contribution activity coefficient
model, derived from the Polymer NRTL model, usable in Aspen Properties and
Aspen Polymers. NRTLSAC can be used for fast, qualitative estimation of the
solubility of complex organic compounds in common solvents. For more
information about the model, see NRTLSAC for Segments/Oligomers.
Note: A newer version of NRTLSAC comes with its own property method
named NRTLSAC and does not require the specification of a method and
oligomer components as described below.
The NRTLSAC model for Segments/Oligomers in the Aspen Physical Property
System is a liquid activity coefficient model called NRTLSAC. To specify it:
1 On the Methods  Specifications sheet, specify method filter ALL.
2 Specify base method NRTLSAC.
In order to use this version of NRTLSAC, all components must be defined as
oligomers. Four conceptual segments also must be defined. On the
Components  Polymers  Oligomers sheet, enter a number for at least
one conceptual segment for each oligomer component, as required by the
definition of an oligomer. These numbers are not used by NRTLSAC.
2 Thermodynamic Property Models 139
On the Methods  Parameters  Binary Interaction  NRTL1 form, enter
the binary parameters between conceptual segments. In the following
example, the conceptual segments are named X, Y, Y+, and Z.
Segment 1 X X Y Y+ X
Segment 2 Y Z Z Z Y+
AIJ 1.643 6.547 2.000 2.000 1.643
AJI 1.834 10.949 1.787 1.787 1.834
CIJ 0.2 0.2 0.3 0.3 0.2
On the Methods  Parameters  Binary Interaction  NRTLXY1 form,
enter a nonzero value for at least one of the four parameters for each
component.
Pitzer Activity Coefficient Model
The Pitzer model was developed as an improvement upon an earlier model
proposed by Guggenheim ( 1935, 1955). The earlier model worked well at low
electrolyte concentrations, but contained discrepancies at higher
concentrations (>0.1M). The Pitzer model resolved these discrepancies,
without resorting to excessive arrays of higherorder terms.
Important: The model can be used for aqueous electrolyte systems, up to 6
molal ionic strength. It cannot be used for systems with any other solvent or
mixed solvents. Any nonwater molecular components are considered solutes
and treated as Henry components.
This section provides theoretical background for the model. All model
equations and parameter requirements are included.
The Pitzer model is commonly used in the calculation of activity coefficients
for aqueous electrolytes up to 6 molal ionic strength. Do not use this model if
a nonaqueous solvent exists. Henry's law parameters are required for all
other components in the aqueous solution. The model development and
working equations are provided in the following sections. Parameter
conversion between the Pitzer notation and our notation is also provided.
The Pitzer model in the Aspen Physical Property System involves user
supplied parameters that are used in the calculation of binary and ternary
parameters for the electrolyte system.
Five elements (P
1
through P
5
) account for the temperature dependencies of
parameters 
(0)
, 
(1)
, 
(2)
, 
(3)
, C
5 0 x
CationCation Parameters
GMPTTH
u
cc'
5 0 x
AnionAnion Parameters
GMPTTH
u
aa'
5 0 x
Ternary Parameters
GMPTPS,
GMPTP1,
GMPTP2,
GMPTP3,
GMPTP4
+
ijk
1 (in each
parameter)
0 x
MoleculeIon and MoleculeMolecule Parameters
GMPTB0

(0)
5 0 x
GMPTB1

(1)
5 0 x
GMPTC
C
5 0 x
Model Development
The Pitzer model analyzes "hardcore" effects in the DebyeHckel theory. It
uses the following expansion as a radial distribution function:
(1)
Where:
g
ij
= Distribution function
r = Radius
q
ij
=
(pair potential of mean force)
2 Thermodynamic Property Models 141
With:
z
i
= Charge of ion i
Q
e
= Electron charge
+
j
(r)
= Average electric potential for ion j
k = Boltzmann's constant
T = Temperature
This radial distribution function is used in the socalled pressure equation that
relates this function and the intermolecular potential to thermodynamic
properties. From this relation you can obtain an expression for the osmotic
coefficient.
Pitzer proposes a general equation for the excess Gibbs energy. The basic
equation is:
(2)
Where:
G
E
= Excess Gibbs energy
R = Gas constant
T = Temperature
n
w
= Kilograms of water
m
i
=
(molality of ion i)
With:
x
i
= Mole fraction of ion i
x
w
= Mole fraction of water
M
w
= Molecular weight of water (g/mol)
n
i
= Moles of ion i
The function f(I) is an electrostatic term that expresses the effect of long
range electrostatic forces between ions. This takes into account the hardcore
effects of the DebyeHckel theory. This term is discussed in detail in the
following section. The parameters
ij
are second virial coefficients that
account for the shortrange forces between solutes i and j. The parameters
ijk
account for the interactions between solutes, i, j, k. For ionion
interactions,
ij
is a function of ionic strength. For moleculeion or molecule
molecule interactions this ionic strength dependency is neglected. The
dependence of
ijk
on ionic strength is always neglected. The matrices
ij
and
ijk
are also taken to be symmetric (that is,
ij
=
ji
).
Pitzer modified this expression for the Gibbs energy by identifying
combinations of functions. He developed interaction parameters that can be
142 2 Thermodynamic Property Models
evaluated using experimental data. He selected mathematical expressions for
these parameters that best fit experimental data.
Pitzer's model can be applied to aqueous systems of strong electrolytes and
to aqueous systems of weak electrolytes with molecular solutes. These
applications are discussed in the following section.
In the Aspen Physical Property System, this model is applied using the
reference state of infinite dilution solution in water for nonwater molecular
solutes and ionic species. The properties such as DHAQFM are obtained at 25
C and 1 atm.
Application of the Pitzer Model to Aqueous Strong
Electrolyte Systems
Pitzer modified his basic equation to make it more useful for data correlation
of aqueous strong electrolytes. He defined a set of more directly observable
parameters to represent combinations of the second and third virial
coefficients. The modified Pitzer equation is:
(3)
z
i
= Charge of ion i
Subscripts c, c', and a, a' denote cations and anions of the solution. B, C, u,
and + are interaction parameters. f(I) is an electrostatic term as a function of
ionic strength. The cationanion parameters B and C are characteristic for an
aqueous singleelectrolyte system. These parameters can be determined by
the properties of pure (apparent) electrolytes. B is expressed as a function of

(0)
and 
(1)
, or of 
(0)
, 
(2)
, and 
(3)
(see equations 11 through 15).
The parameters u and + are for the difference of interaction of unlike ions of
the same sign from the mean of like ions. These parameters can be measured
from commonion mixtures. Examples are NaCl + KCl + H
2
O or NaCl + NaNO
3
+ H
2
O (sic, Pitzer, 1989). These terms are discussed in detail later in this
section.
Frst and Renon (1982) propose the following expression as the Pitzer
equation for the excess Gibbs energy:
(4)
2 Thermodynamic Property Models 143
The difference between equations 3 and 4 is that Pitzer orders cation before
anions. Frst and Renon do not. All summations are taken over all ions i and j
(both cations and anions). This involves making the parameter matrices B
ij
,
C
ij
, u
ij
, and +
ijk
symmetric, as follows:
Secondorder parameters are written B
ij
if i and j are ions of different sign. B
ij
= 0 if the sign of z
i
= sign of z
j
, and B
ii
= 0. Since cations are not ordered
before anions, B
ij
= B
ji
. This eliminates the 2 in the second term in brackets in
Pitzer's original expression (equation 3). Secondorder parameters are written
u
ij
if i and j are ions of the same sign. Thus u
ij
= 0 if the sign of z
i
is different
from the sign of z
j
, and u
ii
= 0 with u
ij
= u
ji
.
Thirdorder parameters are written C
ij
if i and j are ions with different signs.
C
ij
= 0 if the sign of z
i
= sign of z
j
, and C
ii
= 0 with C
ij
= C
ji
. The factor of 2 in
the fifth bracketed term in Pitzer's original expression (equation 3) becomes
1/2 in equation 4. The matrix C is symmetric and is extended to all
ions to make the equation symmetric.
+
ijk
is written for three different ions +
ijk
= +
kij
= +
jki
, and +
ikk
= 0. +
ijk
= 0
if the sign of z
i
=sign of z
j
=sign of z
k
. The factor of 1/6 is different from 1/2
in the last term in brackets in Pitzer's original expression. Pitzer distinguishes
between cations and anions. In Pitzer's original model this parameter appears
twice, as +
cc'a
and +
c'ca
. In this modified model, it appears six times, as +
cc'a
;
+
c'ca
; +
acc'
; +
ac'c
; +
cac'
; and +
c'ac
. Frst and Renon's expression, equation 4,
calculates the expressions for activity coefficients and osmotic coefficients.
Pitzer (1975) modified his model by adding the electrostatic unsymmetrical
mixing effects, producing this modified Pitzer equation for the excess Gibbs
energy:
(4a)
Calculation of Activity Coefficients
The natural logarithm of the activity coefficient for ions is calculated from
equation 4a to give:
144 2 Thermodynamic Property Models
(5)
Where u is neglected and u
ij
and u
'
ij
are the electrostatic unsymmetric mixing
effects:
The X parameters are calculated differently on the option code.
For option code = 1, there is no unsymmetric mixing correction term:
For option code = 0 (default), the unsymmetric mixing correction term is in
polynomial form:
For option code = 1, the unsymmetric mixing correction term is in integral
form:
2 Thermodynamic Property Models 145
For water the logarithm of the activity coefficient is calculated similarly, as
follows:
Applying:
to equation 3 and using:
Where N
w
= moles water, gives:
(6)
f(I), the electrostatic term, is expressed as a function of ionic strength I :
(7)
I, the ionic strength, is defined as:
(8)
Taking the derivative of equation 7 with respect to I, gives:
(9)
So that:
(10)
This equation is used in equation 6. In equations 7 and 9, is the usual Debye
Hckel constant for the osmotic coefficient, determined from:
146 2 Thermodynamic Property Models
(11)
Where:
N
A
= Avogadro's constant
d
w
= Water density
c
B
= Dielectric constant of solvent B
b is an adjustable parameter, which has been optimized in this model to equal
1.2.
B and B' need expressions so that equations 5 and 6 can completely be solved
for the activity coefficients. The parameter B is determined differently for
different electrolyte pairings. For 1n electrolytes (11, 12, 21, and so on)
the following expression gives the parameter B:
(12)
with o
1
=2.0.
For nm electrolytes, n and m>1 (22, 23, 34, and so on), B is determined
by the following expression:
(13)
with o
2
= 12.0 and o
3
= 1.4.
By taking appropriate derivatives, expressions for B' can be derived for 1n
electrolytes:
(14)
and for nm electrolytes:
(15)
The parameters 
(0)
, 
(1)
, 
(2)
, 
(3)
and also C, u, and + can be found in
Pitzer's articles .
After the activity coefficients are calculated, they can be converted to the
mole fraction scale from the molality scale by the following relations:
For solutes:
2 Thermodynamic Property Models 147
(16)
For water as a solvent:
(17)
Where:
m
= Activity coefficient (molality scale)
x
= Activity coefficient (mole fraction scale)
Application of the Pitzer Model to Aqueous
Electrolyte Systems with Molecular Solutes
In aqueous weak electrolyte systems with molecular solutes, the second and
third virial coefficients in the basic Pitzer equation for moleculeion and
moleculemolecule interactions must be considered. The following extensions
of Pitzer's interaction parameters are made.
The secondorder parameters B
ij
are extended to include moleculemolecule
and moleculeion interaction parameters.
The thirdorder parameters +
ijk
are extended to moleculemoleculemolecule
interactions. The following expressions relate +
ijk
to Pitzer's original
ijk
:
+
iii
= 6
iii
However, moleculemolecule interactions were not taken into account by
Pitzer and coworkers. So
iii
is an artificially introduced quantity.
The equations for activity coefficients and the Gibbs free energy are the same
as equations 3 through 6.
Parameters
The Pitzer model in the Aspen Physical Property System involves user
supplied parameters. These parameters are used in the calculation of binary
and ternary parameters for the electrolyte system. These parameters include
the cationanion parameters 
(0)
, 
(1)
, 
(2)
, 
(3)
and C
, cationcation
parameter u
cc'
, anionanion parameter u
aa'
, cation1cation2common anion
parameter +
cc'a
, anion1anion2common cation parameter +
caa'
, and the
moleculeion and moleculemolecule parameters 
(0)
, 
(1)
, and, C
. The
parameter names in the Aspen Physical Property System and their
requirements are discussed in Pitzer Activity Coefficient Model.
148 2 Thermodynamic Property Models
Parameter Conversion
For nm electrolytes, n and m>1 (22, 23, 34, and so on), the parameter

(3)
corresponds to Pitzer's 
(1)
. 
(2)
is the same in both the Aspen Physical
Property System and original Pitzer models. Pitzer refers to the nm
electrolyte parameters as 
(1)
, 
(2)
, 
(0)
. 
(0)
and 
(2)
retain their meanings in
both models, but Pitzer's 
(1)
is 
(3)
in the Aspen Physical Property System. Be
careful to make this distinction when entering nm electrolyte parameters.
Pitzer often gives values of 
(0)
, 
(1)
, 
(2)
, 
(3)
, and C
) for 13
dilute aqueous electrolytes
(Pitzer and Mayorga, 1973) Binary parameters for 11 inorganic
electrolytes, salts of carboxylic acids (11),
tetraalkylammonium halids, sulfonic acids
and salts, additional 11 organic salts, 21
inorganic compounds, 21 organic
electrolytes, 31 electrolytes, 41 and 51
electrolytes
(Pitzer and Mayorga, 1974) Binary parameters for 22 electrolytes in
water at 25C
(Pitzer and Kim, 1974) Binary and ternary parameters for mixed
electrolytes, binary mixtures without a
common ion, mixed electrolytes with three
or more solutes
(Pitzer, 1975) Ternary parameters for systems mixing
doubly and singly charged ions
(Pitzer and Silvester, 1976) Parameters for phosphoric acid and its buffer
solutions
(Pitzer, Roy and Silvester, 1977) Parameters and thermodynamic properties
for sulfuric acid
(Silvester and Pitzer, 1977) Data for NaCl and aqueous NaCl solutions
(Pitzer, Silvester, and Peterson, 1978) Rare earth chlorides, nitrates, and
perchlorates
(Peiper and Pitzer, 1982) Aqueous carbonate solutions, including
mixtures of sodium carbonate, bicarbonate,
and chloride
(Phutela and Pitzer, 1983) Aqueous calcium chloride
(Conceicao, de Lima, and Pitzer, 1983) Saturated aqueous solutions, including
mixtures of sodium chloride, potassium
chloride, and cesium chloride
(Pabalan and Pitzer, 1987) Parameters for polynomial unsymmetric
mixing term
(Kim and Frederick, 1988) Parameters for integral unsymmetric mixing
term
Pitzer References
Conceicao, M., P. de Lima, and K.S. Pitzer, "Thermodynamics of Saturated
Aqueous Solutions Including Mixtures of NaCl, KCl, and CsCl, "J. Solution
Chem, Vol. 12, No. 3, (1983), pp. 171185.
Frst, W. and H. Renon, "Effects of the Various Parameters in the Application
of Pitzer's Model to SolidLiquid Equilibrium. Preliminary Study for Strong 11
Electrolytes," Ind. Eng. Chem. Process Des. Dev., Vol. 21, No. 3, (1982),
pp. 396400.
Guggenheim, E.A., Phil. Mag., Vol. 7, No. 19, (1935), p. 588.
Guggenheim, E.A. and J.C. Turgeon, Trans. Faraday Soc., Vol. 51, (1955), p.
747.
150 2 Thermodynamic Property Models
Kim, H. and W.J. Frederick, "Evaluation of Pitzer Ion Interaction Parameters
of Aqueous Mixed Electrolyte Solution at 25C, Part 2: Ternary Mixing
Parameters," J. Chem. Eng. Data, 33, (1988), pp. 278283.
Pabalan, R.T. and K.S. Pitzer, "Thermodynamics of Concentrated Electrolyte
Mixtures and the Prediction of Mineral Solubilities to High Temperatures for
Mixtures in the system NaKMgClSO4OHH2O," Geochimica Acta, 51,
(1987), pp. 24292443.
Peiper, J.C. and K.S. Pitzer, "Thermodynamics of Aqueous Carbonate
Solutions Including Mixtures of Sodium Carbonate, Bicarbonate, and
Chloride," J. Chem. Thermodynamics, Vol. 14, (1982), pp. 613638.
Phutela, R.C. and K.S. Pitzer, "Thermodynamics of Aqueous Calcium
Chloride," J. Solution Chem., Vol. 12, No. 3, (1983), pp. 201207.
Pitzer, K.S., "Thermodynamics of Electrolytes. I. Theoretical Basis and
General Equations, " J. Phys. Chem., Vol. 77, No. 2, (1973), pp. 268277.
Pitzer, K.S., J. Solution Chem., Vol. 4, (1975), p. 249.
Pitzer, K.S., "Fluids, Both Ionic and NonIonic, over Wide Ranges of
Temperature and Composition," J. Chen. Thermodynamics, Vol. 21, (1989),
pp. 117. (Seventh Rossini lecture of the commission on Thermodynamics of
the IUPAC, Aug. 29, 1988, Prague, exCzechoslovakia).
Pitzer, K.S. and J.J. Kim, "Thermodynamics of Electrolytes IV; Activity and
Osmotic Coefficients for Mixed Electrolytes," J.Am. Chem. Soc., Vol. 96
(1974), p. 5701.
Pitzer, K.S. and G. Mayorga, "Thermodynamics of Electrolytes II; Activity and
Osmotic Coefficients for Strong Electrolytes with One or Both Ions Univalent,"
J. Phys. Chem., Vol. 77, No. 19, (1973), pp. 23002308.
Pitzer, K.S. and G. Mayorga, J. Solution Chem., Vol. 3, (1974), p. 539.
Pitzer, K.S., J.R. Peterson, and L.F. Silvester, "Thermodynamics of
Electrolytes. IX. Rare Earth Chlorides, Nitrates, and Perchlorates, "J. Solution
Chem., Vol. 7, No. 1, (1978), pp. 4556.
Pitzer, K.S., R.N. Roy, and L.F. Silvester, "Thermodynamics of Electrolytes 7
Sulfuric Acid," J. Am. Chem. Soc., Vol. 99, No. 15, (1977), pp. 49304936.
Pitzer, K.S. and L.F. Silvester, J. Solution Chem., Vol. 5, (1976), p. 269.
Silvester, L.F. and K.S. Pitzer, "Thermodynamics of Electrolytes 8 High
Temperature Properties, Including Enthalpy and Heat Capacity, With
Application to Sodium Chloride," J. Phys. Chem., Vol. 81, No. 19, (1977), pp.
18221828.
Polynomial Activity Coefficient
This model represents activity coefficient as an empirical function of
composition and temperature. It is used frequently in metallurgical
applications where multiple liquid and solid solution phases can exist.
The equation is:
2 Thermodynamic Property Models 151
Where:
A
i
=
B
i
=
C
i
=
D
i
=
E
i
=
For any component i, the value of the activity coefficient can be fixed:
i
= f
i
This model is not part of any property method. To use it:
1 On the Methods  Specifications sheet, specify an activity coefficient
model, such as NRTL.
2 Click the Methods  Selected Methods folder.
3 In the Object Manager, click New.
4 In the Create New ID dialog box, enter a name for the new method.
5 In the Base Property Method field, select NRTL.
6 Click the Models tab.
7 Change the Model Name for GAMMA from GMRENON to GMPOLY.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
GMPLYP/1 a
i1
0 x
GMPLYP/2 a
i2
0 x
GMPLYP/3 a
i3
0 x
GMPLYP/4 b
i1
0 x
GMPLYP/5 b
i2
0 x
GMPLYP/6 b
i3
0 x
GMPLYP/7 c
i1
0 x
GMPLYP/8 c
i2
0 x
GMPLYP/9 c
i3
0 x
GMPLYP/10 d
i1
0 x
GMPLYP/11 d
i2
0 x
GMPLYP/12 d
i3
0 x
GMPLYP/13 e
i1
0 x
GMPLYP/14 e
i2
0 x
GMPLYP/15 e
i3
0 x
GMPLYO f
i
x
Note: If you specify GMPLYP on the Methods  Parameters  Pure
Component  TDependent sheet, you can only enter the first 12 elements.
If you want to specify values for elements 13 to 15, you should go to the
Customize  AddInput  Add After sheet and enter the values of all 15
elements as in the following example:
152 2 Thermodynamic Property Models
PROPDATA GMPLYP1
INUNITS SI
PROPLIST GMPLYP
PVAL WATER 0.0 1.5 0.0 &
0.0 0.0 0.0 &
0.0 0.0 0.0 &
0.0 0.0 0.0 &
0.0 16. 0.0
RedlichKister
This model calculates activity coefficients. It is a polynomial in the difference
between mole fractions in the mixture. It can be used for liquid and solid
mixtures (mixed crystals).
The equation is:
Where:
nc = Number of components
A
1,ij
= a
ij
/ T + b
ij
A
2,ij
= c
ij
/ T + d
ij
A
3,ij
= e
ij
/ T + f
ij
A
4,ij
= g
ij
/ T + h
ij
A
5,ij
= m
ij
/ T + n
ij
A
n,ii
= A
n,jj
= 0.0
A
n,ji
= A
n,ij
(1)
(n1)
A
n,kj
= A
n,jk
(1)
(n1)
For any component i, the value of the activity coefficient can be fixed:
i
= v
i
Parameter Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
GMRKTB/1 a
ij
0 x
GMRKTB/2 b
ij
0 x
GMRKTB/3 c
ij
0 x
GMRKTB/4 d
ij
0 x
GMRKTB/5 e
ij
0 x
GMRKTB/6 f
ij
0 x
GMRKTB/7 g
ij
0 x
GMRKTB/8 h
ij
0 x
GMRKTB/9 m
ij
0 x
GMRKTB/10 n
ij
0 x
GMRKTO v
i
x
2 Thermodynamic Property Models 153
Reference
J. P. Novak, J. Matous, and J. Pick, Studies in Modern Thermodynamics,
Volume 7: LiquidLiquid Equilibria, Appendix 4, Elsevier, 1987.
ScatchardHildebrand
The ScatchardHildebrand model calculates liquid activity coefficients. It is
used in the CHAOSEA property method and the GRAYSON property method.
The equation for the ScatchardHildebrand model is:
Where:
A
ij
=
i
=
V
m
*,l
=
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
x 5.0 2000.0 TEMPERATURE
DELTA
o
i
x 10
3
10
5
SOLUPARAM
VLCVT1 V
i
*,CVT
x 0.0005 1.0 MOLE
VOLUME
GMSHVL V
i
*,l
x 0.01 1.0 MOLE
VOLUME
GMSHXL k
ij
0.0 x 5 5
ThreeSuffix Margules
This model can be used to describe the excess properties of liquid and solid
solutions. It does not find much use in chemical engineering applications, but
is still widely used in metallurgical applications. Note that the binary
parameters for this model do not have physical significance.
The equation is:
Where k
ij
is a binary parameter:
154 2 Thermodynamic Property Models
For any component i, the value of the activity coefficient can be fixed:
i
= d
i
Parameter Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
GMMRGB/1 a
ij
0 x TEMPERATURE
GMMRGB/2 b
ij
0 x
GMMRGB/3 c
ij
0 x
GMMRGO d
i
x
References
M. Margules, "ber die Zusammensetzung der gesttigten Dmpfe von
Mischungen," Sitzungsber. Akad. Wiss. Vienna, Vol. 104, (1895), p. 1293.
D.A. Gaskell, Introduction to Metallurgical Thermodyanics, 2nd ed., (New
York: Hemisphere Publishing Corp., 1981), p. 360.
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987).
Symmetric and Unsymmetric Electrolyte
NRTL Activity Coefficient Model
The Symmetric Electrolyte NRTL activity coefficient model (GMENRTLS) uses a
symmetric reference state for ions as pure fused salts, rather than infinite
dilution in aqueous solution. This basis is easier to use for nonaqueous and
mixedsolvent systems, and eliminates the need to introduce water into
otherwise waterfree systems. It also allows the model to reduce to the
standard NRTL model when there are no electrolytes in the system. The
ENRTLSR property method is based on this model.
Chemical constants, enthalpy, and Gibbs free energy are calculated with
respect to the symmetric ionic reference state.
The Unsymmetric Electrolyte NRTL activity coefficient model (GMENRTLQ)
uses the same equations as GMENRTLS, but the unsymmetric reference state
for ions (infinite dilution in aqueous solution). The ENRTLRK property method
is based on this model. Unlike the original Electrolyte NRTL activity coefficient
model, GMENRTLQ is also used to calculate enthalpy and Gibbs free energy
from thermodynamics based on the unsymmetric ionic reference state.
These models also handle zwitterions.
Parameters
Both symmetric and unsymmetric models share the same binary and pair
parameters. The adjustable model parameters are the symmetric nonrandom
factor parameters, o, and the asymmetric binary interaction energy
parameters, t. These parameters exist for moleculemolecule pairs (o
mm'
=
o
m'm
while t
mm'
= t
m'm
), moleculeelectrolyte pairs (o
m,ca
= o
ca,m
while t
m,ca
=
2 Thermodynamic Property Models 155
t
ca,m
where ca represents an ion pair), and electrolyteelectrolyte pairs (o
ca,ca'
= o
ca',ca
and o
ca,c'a
= o
c'a,ca
while t
ca,ca'
= t
ca',ca
and t
ca,c'a
= t
c'a,ca
)
The parameters for ion pairs are temperaturedependent:
where .
The NRTL parameters are used for moleculemolecule parameters, with
temperature dependence:
The temperature dependency of the dielectric constant is given by:
Option codes can affect the performance of this model. See Option Codes for
Activity Coefficient Models for details.
The following table lists the parameters used by GMENRTLS and GMENRTLQ:
156 2 Thermodynamic Property Models
Parameter
Name
Symbol No. of
Elements
Default MDS Units
Dielectric Constant Unary Parameters
CPDIEC a 1
b 1 0
T
ref
1 298.15 TEMPERATURE
b
ji
0 x TEMPERATURE
NRTL/3 c
ij
=c
ji
.3 x
NRTL/4 d
ij
=d
ji
0 x TEMPERATURE
NRTL/5 e
ij
0 x TEMPERATURE
e
ji
0 x TEMPERATURE
NRTL/6 f
ji
0 x TEMPERATURE
f
ji
0 x TEMPERATURE
ElectrolyteMolecule Pair Parameters
GMENCC C
ca,m
1 0 x
C
m,ca
1 0 x
GMENCD D
ca,m
1 0 x TEMPERATURE
D
m,ca
1 0 x TEMPERATURE
GMENCE E
ca,m
1 0 x
E
m,ca
1 0 x
GMENCN
o
ca,m
= o
m,ca
1 .2 x
ElectrolyteElectrolyte Pair Parameters
GMENCC C
ca,c'a
1 0 x
C
c'a,ca
1 0 x
C
ca,ca'
1 0 x
C
ca',ca
1 0 x
GMENCD D
ca,c'a
1 0 x TEMPERATURE
D
c'a,ca
1 0 x TEMPERATURE
D
ca,ca'
1 0 x TEMPERATURE
D
ca',ca
1 0 x TEMPERATURE
GMENCE E
ca,c'a
1 0 x
E
c'a,ca
1 0 x
E
ca,ca'
1 0 x
E
ca',ca
1 0 x
GMENCN
o
ca,c'a
= o
c'a,ca
1 .2 x
o
ca,ca'
= o
ca',ca
1 .2 x
2 Thermodynamic Property Models 157
Zwitterions in Symmetric and Unsymmetric
Electrolyte NRTL
The Symmetric and Unsymmetric Electrolyte NRTL models support
zwitterions, compounds with both positive and negative charges but net
charge of zero. Zwitterions are defined as:
 Components of type Conventional, similar to solvents
 Parameter ZWITTER set to 1; all other components in the Aspen Physical
Property System default to zero for ZWITTER
 Parameter PLXANT/1 less than 1.0E10 so that they are nonvolatile.
Zwitterions are handled as follows:
 A zwitterion interacts with other molecular species through NRTL
parameters only, excluding any interactions through Henry constants and
pair parameters.
 The activity coefficient is calculated as a solute.
 The contribution from zwitterions to the solution enthalpy and Gibbs free
energy are calculated as solutes using the infinite dilution heat capacity
model CPAQ0, DGAQFM, and DHAQFM.
Working Equations for Symmetric and
Unsymmetric Electrolyte NRTL
The symmetric and unsymmetric electrolyte NRTL models have two
contributions, one from local interactions that exist at the immediate
neighbourhood of any species, and the other from the longrange ionion
interactions that exist beyond the immediate neighbourhood of an ionic
species. To account for the local interactions, the model uses the electrolyte
NRTL expression. To account for the longrange interactions, the model uses
the PitzerDebyeHckel (PDH) formula (Pitzer, 1980, 1986). The following
equation is the basis of the electrolyte NRTL model for the excess Gibbs free
energy of electrolyte systems:
(1)
The excess Gibbs free energy G
m
ex
is defined as:
(2)
where G
m
is the Gibbs free energy of electrolyte systems and G
m
id
is the Gibbs
free energy of an ideal solution at the same conditions of temperature,
pressure, and composition.
The activity coefficient
i
of component i can be derived from Eq. 1 as follows:
(3)
or
158 2 Thermodynamic Property Models
(4)
Each of these terms will be discussed in subsequent sections.
Reference States in Electrolyte Systems
The activity coefficient needs to be normalized by choosing a reference state
for any molecular and ionic component, respectively.
Reference state for molecular components
The reference state for a molecular component m is defined as follows:
(5)
This definition is the socalled standard state of pure liquids for molecular
components and it is also called the symmetric reference state for molecular
components.
Reference state for ionic components
The standard state of pure liquids is hypothetical for ionic components in
electrolyte systems. Instead, the symmetric reference state is defined as the
pure fused salt state of each electrolyte component in the system.
However, the conventional reference state for ionic components is the
infinitedilution activity coefficients; it is also called the unsymmetric
reference state for ionic components.
GMENRTLS uses the symmetric reference state, while GMENRTLQ uses the
unsymmetric reference state.
Pure fused salt state of an electrolyte component
For an electrolyte component ca, the pure fused salt state can be defined as
follows:
(6)
(7)
Where
is the DebyeHckel
parameter, I
x
is the ionic strength, is the closest approach parameter, N
A
is
Avogadro's number, v
s
is the molar volume of the solvent, Q
e
is the electron
charge, c
s
is the dielectric constant of the solvent, k
B
is the Boltzmann
constant, z
i
is the charge number of component i, and I
x
0
represents I
x
at the
reference state.
For the unsymmetric reference state, I
x
0
= 0.
For the symmetric reference state, the DebyeHckel theory is originally
based on a single electrolyte with water as the solvent. Therefore, we can
obtain I
x
0
from Eqs. 13 and 14:
(56)
For multielectrolyte systems with mixedsolvents, I
x
0
can take this form:
(57)
(58)
2 Thermodynamic Property Models 165
(59)
so that
(60)
where x
m
0 applies to all molecular components in the solution. This
definition ensures that the excess Gibbs free energy from the long range
interactions will be zero at the symmetric reference state regardless of how
many electrolytes are present in the solution.
The contribution to the activity coefficient from component i can be derived
as:
(61)
For a solvent component, this is:
(62)
For a cation or anion component:
(63)
For the unsymmetric reference state:
(64)
And for the symmetric reference state:
(65)
(66)
166 2 Thermodynamic Property Models
(67)
or
(68)
(69)
The DebyeHckel theory is original based on a single electrolyte with water
as the solvent. The molar volume v
s
and the dielectric constant c
s
for the
single solvent need to be extended for mixedsolvents; a simple composition
average mixing rule is adequate to calculate them as follows:
(70)
(71)
where s is a solvent component in the mixture and M
s
is the solvent molecular
weight. Each sum is over all solvent components in the solution.
Born term correction
If the infinite dilution aqueous solution is chosen as the reference state, we
need to correct the the DebyeHckel term for the change of the reference
state from the mixedsolvent composition to aqueous solution. The Born term
(Robinson and Stokes, 1970; Rashin and Honig, 1985) is used for this
purpose:
(72)
where N
A
is the Avogadro constant and R is the gas constant. is the
Born term correction to the unsymmetric PitzerDebyeHckel formula,
, and c
w
is the dielectric constant of water, and r
i
is the Born radius
of species i.
The Born contribution to the activity coefficient of component i can be derived
as follows:
2 Thermodynamic Property Models 167
(73)
For a cation or anion component, this is:
(74)
The correction to the activity coefficient for a solvent component is zero:
(75)
Henry Components in the Symmetric and
Unsymmetric Electrolyte NRTL Models
Light gases (Henry components) are usually supercritical at the temperature
and pressure of the system. In that case, pure component vapor pressure is
meaningless and therefore the pure liquid state at the temperature and
pressure of the system cannot serve as the reference state. The reference
state for a Henry component is redefined to be at infinite dilution (that is, at
x
i
0) and at the temperature and pressure of the system.
The liquid phase reference fugacity, f
i
*,l
, becomes the Henrys constant for
Henry components in the solution, H
i
, and the activity coefficient,
i
, is
converted to the infinite dilution reference state through the relationship:
(76)
where
i
of Henry
components: aqueous and mixedsolvent. This can be specified on the Setup
 Calculation Options  Reactions sheet. Mixedsolvent is the default.
For mixedsolvent basis,
h
is calculated as follows:
(83)
where x
h
0 applies to all Henry components and x
i
0 to all ionic
components in the solution.
For aqueous basis, the unsymmetric activity coefficients of Henry components
are calculated as follows:
(84)
Electrolyte Chemical Equilibria
In determining the composition of an electrolyte system, it is important to
know the equilibrium constants of the reactions taking place. An equilibrium
constant is expressed as the product of the activity of each species raised to
its stoichiometric coefficients. Two different scales are used in Aspen Plus: the
molality scale and the mole fraction scale but both are based on aqueous
electrolyte chemical equilibrium. For instance, the equilibrium constant for the
mole fraction scale in Aspen Plus is written in one of these:
(85)
(86)
2 Thermodynamic Property Models 169
Where:
K = Equilibrium constant
x
w
= Water mole fraction
w
= Water activity coefficient
x
s
= Nonwater solvent mole fraction
s
= Nonwater solvent activity coefficient
x
i
= Mole fraction of solute component (Henry or
ion)
i
*
= Unsymmetric activity coefficient of solute
component
v
i
= Stoichiometric coefficient
The above equations are limited to aqueous electrolyte chemical equilibria
only. Therefore, the chemical constants in Aspen Plus databanks for
electrolyte systems with infinite dilution aqueous reference state cannot be
used for electrolyte systems with the symmetric reference state for ionic
components.
The chemical constants for electrolyte systems with the symmetric reference
state can be written in these forms:
(87)
(88)
where i indicates a molecular or ionic component and h represents a Henry
component. Eq. 83 or 84 is used for calculation of the unsymmetric activity
coefficients of Henry components. However, the calculation for the activity
coefficients of ionic components is carried out with the symmetric reference
state.
Other Thermophysical Properties
The activity coefficient model can be related to other properties through
fundamental thermodynamic equations. These properties (called excess liquid
functions) are relative to the ideal liquid mixture at the same condition:
Excess molar liquid Gibbs free energy
(89)
Excess molar liquid enthalpy
(90)
Excess molar liquid entropy
170 2 Thermodynamic Property Models
(91)
The excess liquid functions given by Equations 8991 are calculated from the
same activity coefficient model. In practice, however, the activity coefficient
i
is often derived first from the excess liquid Gibbs free energy of a mixture
from an activity coefficient model:
(92)
with
(93)
(94)
Where is the liquid Gibbs free energy of mixing; it is defined as the
difference between the Gibbs free energy of the mixture and that of the pure
component and is the ideal Gibbs free energy of mixing. Once the
excess liquid functions are known, the thermodynamic properties of liquid
mixtures can be computed as follows:
(95)
(96)
(97)
where H
i
l
and H
i
ig
are the enthalpy and ideal gas enthalpy of component i at
the system conditions. Similarly, G
i
l
and G
i
ig
are the Gibbs free energy and
ideal gas Gibbs free energy of component i at the system conditions. In Aspen
Plus, both H
i
ig
and G
i
ig
are computed by the expressions:
(98
)
(99
)
2 Thermodynamic Property Models 171
where is the standard enthalpy of formation of ideal gas at
, is the ideal gas heat capacity, and is the
standard Gibbs free energy of formation of ideal gas at .
However, the above equations are directly not applicable to mixtures
containing ionic components because the ideal gas model becomes invalid for
ionic components.
The formulation to calculate the enthalpy and Gibbs free energy for
electrolyte systems can be carried out as follows:
(100)
(101)
(102)
where indexes s, h, and ca are meant to represent the contributions from
solvents, Henry components and ionic components, respectively.
Solvents
The contribution from solvents can be written as follows:
(103)
(104)
(105)
(106)
(107)
(108)
where is the liquid fugacity coefficient of pure solvent
component i, p
i
sat
is the vapor pressure of pure component at the system
temperature T, 
i
v
is the vapor fugacity coefficient of pure component at T
and p
i
sat
(normally calculated from an equationofstate model), and
172 2 Thermodynamic Property Models
is the Poynting pressure correction from p
i
sat
to p,
and V
i
l
is the liquid molar volume at T and p.
Henry components
The contribution from Henry components can be written as follows:
(109)
(110)
(111)
(112)
(113)
(114)
(115)
Ionic components
The contribution from ionic components can be written as follows:
(116)
(117)
(118)
(119)
(120)
2 Thermodynamic Property Models 173
(121)
where H
i
ref
and G
i
ref
are the enthalpy and Gibbs free energy of ionic
component i at the system conditions with a specified reference state
(symmetric or unsymmetric).
The property calculations for solvents and Henry components are the same
when the reference state is changed. Only for ionic components do the
property methods need to be specified with a reference state. The methods
with the unsymmetric reference state are available already in Aspen Plus and
Aspen Properties. New methods are needed only for the symmetric reference
state. Overall, the calculated total enthalpy or Gibbs free energy for the same
electrolyte solution should be the same, regardless of the reference state
specified.
Enthalpy and Gibbs free energy of ionic components with the symmetric
reference state can be written as follows:
(122)
(123)
where and are enthalpy and Gibbs free energy of ionic component
i with the unsymmetric reference state, respectively, and available already in
Aspen Plus and Aspen Properties. AH
i
and AG
i
are the new contributions from
the symmetric reference state.
The unsymmetric enthalpy for an ionic component is calculated from
the infinite dilution aqueous phase heat capacity as follows:
(124)
where T
ref
= 298.15K. By default, is calculated from the aqueous
infinite dilution heat capacity polynomial. If the polynomial model parameters
are not available, is calculated from the CrissCobble correlation.
The unsymmetric Gibbs free energy for an ionic component is
calculated from the infinite dilution aqueous phase heat capacity as
follows:
(125)
174 2 Thermodynamic Property Models
(126)
where the term RT ln (1000/M
w
) is added because and
are based on a molality scale, and is based on a mole fraction scale.
AH
i
and AG
i
for the symmetric reference state
The reference state for ionic components in the symmetric electrolyte NRTL
model is the pure fused salts containing these ions, so the enthalpy or Gibbs
free energy of the ionic components at the symmetric reference state is the
enthalpy contributions or the Gibbs free energy contributions of these ions to
the system of the pure fused salts. The condition of the pure fused salts can
be defined as follows:
(127)
which applies to all molecular components in the solution.
Given that the calculated total enthalpy or total Gibbs free energy of an
electrolyte solution by any reference state should be the same, the
formulation of enthalpy or Gibbs free energy of the ionic components at the
symmetric reference state can be derived from the unsymmetric electrolyte
NRTL enthalpy or Gibbs free energy calculations at the condition that all
molecular components (solvents and Henry components) approach zero, i.e.
the pure fused salts. Applying Eq. 127, the system enthalpy and Gibbs free
energy with the unsymmetric reference state can be expressed as:
(128
)
(129
)
where and are the total enthalpy and total Gibbs
free energy of the solution calculated by the unsymmetric electrolyte NRTL
model at the limit of all molecular components approach zero, that is, the
pure fused salts state, and and are the
excess enthalpy and excess Gibbs free energy of ion i at the same limit
condition.
Applying Eq. 127, the system enthalpy and Gibbs free energy with the
symmetric reference state can be expressed as:
(130)
2 Thermodynamic Property Models 175
(131)
Comparing Eqs. 130 and 131 to Eqs. 128 and 129, we can get:
(132)
(133)
Therefore, the enthalpy and Gibbs free energy of ionic components with the
symmetric reference state can be written as:
(134)
(135)
One of advantages with the symmetric reference state is its capability to
model nonaqueous electrolyte systems without introducing a trace amount of
water. However, for calculating system enthalpy, entropy, and Gibbs free
energy correctly, water must be defined in the component list so that the
corrections showed by Eqs. 132 and 133 can be calculated from the
unsymmetric reference state.
References for Symmetric and Unsymmetric
Electrolyte NRTL
C.C. Chen and L.B. Evans, "A Local Composition Model for the Excess Gibbs
Energy of Aqueous Electrolyte Systems," AIChE Journal, 1986, 32, 444.
Y. Song and C.C. Chen, "Symmetric Nonrandom TwoLiquid Activity
Coefficient Model for Electrolytes" (to be published).
K.S. Pitzer, "Electrolytes. From Dilute Solutions to Fused Salts," J. Am. Chem.
Soc., 1980, 102, 29022906.
K.S. Pitzer and J.M. Simonson, "Thermodynamics of Multicomponent, Miscible,
Ionic Systems: Theory and Equations," J. Phys. Chem., 1986, 90, 30053009.
R.A. Robinson and R.H. Stokes, Electrolyte Solutions, Second Revised Edition.
Butterworths: London, 1970.
A.A. Rashin and B. Honig, "Reevaluation of the Born Model of Ion Hydration,"
J. Phys. Chem., 1985, 89 (26), pp 55885593.
UNIFAC Activity Coefficient Model
The UNIFAC model calculates liquid activity coefficients for the following
property methods: UNIFAC, UNIFHOC, and UNIFLL. Because the UNIFAC
model is a groupcontribution model, it is predictive. All published group
176 2 Thermodynamic Property Models
parameters and group binary parameters are stored in the Aspen Physical
Property System.
The equation for the original UNIFAC liquid activity coefficient model is made
up of a combinatorial and residual term:
ln
=
ln
i
c
+ ln
i
r
ln
i
c
=
Where the molecular volume and surface fractions are:
and
Where nc is the number of components in the mixture. The coordination
number z is set to 10. The parameters r
i
and q
i
are calculated from the group
volume and area parameters:
and
Where v
ki
is the number of groups of type k in molecule i, and ng is the
number of groups in the mixture.
The residual term is:
I
k
is the activity coefficient of a group at mixture composition, and I
k
i
is the
activity coefficient of group k in a mixture of groups corresponding to pure i.
The parameters I
k
and I
k
i
are defined by:
With:
And:
2 Thermodynamic Property Models 177
The parameter X
k
is the group mole fraction of group k in the liquid:
Parameter
Name/Element
Symbol DefaultMDS Lower
Limit
Upper
Limit
Units
UFGRP
(k,v
k
, m, v
m
, ...)
GMUFQ Q
k
GMUFR R
k
GMUFB b
kn
TEMPERATURE
The parameter UFGRP stores the UNIFAC functional group number and
number of occurrences of each group. UFGRP is stored in the Aspen Physical
Property System pure component databank for most components. For
nondatabank components, enter UFGRP on the Components  Molecular
Structure  Functional Group sheet. See Physical Property Data, Chapter 3,
for a list of the UNIFAC functional groups.
UNIFACPSRK
The PSRK property method uses GMUFPSRK, the UNIFACPSRK model, which
is a variation on the standard UNIFAC model. UNIFACPSRK has special
groups defined for the light gases CO
2
, H
2
, NH
3
, N
2
, O
2
, CO, H
2
S, and argon,
and the group binary interaction parameters are temperaturedependent,
using the values in parameter UNIFPS, instead of the constant value from
GMUFB used above, so that:
Where a, b, and c are the three elements of UNIFPS.
References
Aa. Fredenslund, J. Gmehling and P. Rasmussen, "VaporLiquid Equilibria
using UNIFAC," (Amsterdam: Elsevier, 1977).
Aa. Fredenslund, R.L. Jones and J.M. Prausnitz, AIChE J., Vol. 21, (1975), p.
1086.
H.K. Hansen, P. Rasmussen, Aa. Fredenslund, M. Schiller, and J. Gmehling,
"VaporLiquid Equilibria by UNIFAC Group Contribution. 5 Revision and
Extension", Ind. Eng. Chem. Res., Vol. 30, (1991), pp. 23522355.
UNIFAC (Dortmund Modified)
The UNIFAC modification by Gmehling and coworkers (Weidlich and
Gmehling, 1987; Gmehling et al., 1993), is slightly different in the
combinatorial part. It is otherwise unchanged compared to the original
UNIFAC:
178 2 Thermodynamic Property Models
With:
The temperature dependency of the interaction parameters is:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
UFGRPD
(k,v
k
, m, v
m
, ...)
GMUFDQ Q
k
GMUFDR R
k
UNIFDM/1 a
mn,1
0 TEMPERATURE
UNIFDM/2 a
mn,2
0 TEMPERATURE
UNIFDM/3 a
mn,3
0 TEMPERATURE
The parameter UFGRPD stores the group number and the number of
occurrences of each group. UFGRPD is stored in the Aspen Physical Property
System pure component databank. For nondatabank components, enter
UFGRPD on the Components  Molecular Structure  Functional Group
sheet. See Physical Property Data, Chapter 3, for a list of the Dortmund
modified UNIFAC functional groups.
References
U. Weidlich and J. Gmehling, "A Modified UNIFAC Model 1. Prediction of VLE,
h
E
and ," Ind. Eng. Chem. Res., Vol. 26, (1987), pp. 13721381.
J. Gmehling, J. Li, and M. Schiller, "A Modified UNIFAC Model. 2. Present
Parameter Matrix and Results for Different Thermodynamic Properties," Ind.
Eng. Chem. Res., Vol. 32, (1993), pp. 178193.
UNIFAC (Lyngby Modified)
The equations for the "temperaturedependent UNIFAC" (Larsen et al., 1987)
are similar to the original UNIFAC:
,
Volume fractions are modified:
2 Thermodynamic Property Models 179
With:
Where I
k
and I
k
i
have the same meaning as in the original UNIFAC, but
defined as:
With:
The temperature dependency of a is described by a function instead of a
constant:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
UFGRPL
(k,v
k
, m, v
m
, ...)
GMUFLQ Q
k
GMUFLR R
k
UNIFLB/1 a
mn,1
0 TEMPERATURE
UNIFLB/2 a
mn,2
0 TEMPERATURE
UNIFLB/3 a
mn,3
0 TEMPERATURE
The parameter UFGRPL stores the modified UNIFAC functional group number
and the number of occurrences of each group. UFGRPL is stored in the Aspen
Physical Property System pure component databank. For nondatabank
components, enter UFGRP on the Components  Molecular Structure 
Functional Group sheet. See Physical Property Data, Chapter 3, for a list of
the Larsen modified UNIFAC functional groups.
180 2 Thermodynamic Property Models
Reference: B. Larsen, P. Rasmussen, and Aa. Fredenslund, "A Modified
UNIFAC GroupContribution Model for Prediction of Phase Equilibria and Heats
of Mixing," Ind. Eng. Chem. Res., Vol. 26, (1987), pp. 2274 2286.
UNIQUAC Activity Coefficient Model
The UNIQUAC model calculates liquid activity coefficients for these property
methods: UNIQUAC, UNIQ2, UNIQHOC, UNIQNTH, and UNIQRK. It is
recommended for highly nonideal chemical systems, and can be used for VLE
and LLE applications. This model can also be used in the advanced equations
of state mixing rules, such as WongSandler and MHV2.
The equation for the UNIQUAC model is:
Where:
u
i
=
u
i
' =
u
i
=
l
i
=
t
i
' =
t
ij
=
z = 10
a
ij
, b
ij
, c
ij
, and d
ij
are unsymmetrical. That is, a
ij
may not be equal to a
ji
, etc.
Absolute temperature units are assumed for the binary parameters a
ij
, b
ij
, c
ij
,
d
ij
, and e
ij
.
can be determined from VLE and/or LLE data regression. The Aspen Physical
Property System has a large number of builtin parameters for the UNIQUAC
model. The binary parameters have been regressed using VLE and LLE data
from the Dortmund Databank. The binary parameters for VLE applications
were regressed using the ideal gas, RedlichKwong, and HaydenO'Connell
equations of state. See Physical Property Data, Chapter 1, for details.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
GMUQR r
i
x
2 Thermodynamic Property Models 181
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
GMUQQ q
i
x
GMUQQ1 q
i
' q x
UNIQ/1 a
ij
0 x 50.0 50.0
UNIQ/2 b
ij
0 x 15000.0 15000.0 TEMPERATURE
UNIQ/3 c
ij
0 x TEMPERATURE
UNIQ/4 d
ij
0 x TEMPERATURE
UNIQ/5 T
lower
0 K x TEMPERATURE
UNIQ/6 T
upper
1000 K x TEMPERATURE
UNIQ/7 e
ij
0 x TEMPERATURE
Absolute temperature units are assumed for elements 2 through 4 and 7 of
UNIQ.
The UNIQ2 property method uses data set 2 for UNIQ. All other UNIQUAC
methods use data set 1.
References
D.S. Abrams and J.M. Prausnitz, "Statistical Thermodynamics of liquid
mixtures: A new expression for the Excess Gibbs Energy of Partly or
Completely Miscible Systems," AIChE J., Vol. 21, (1975), p. 116.
A. Bondi, "Physical Properties of Molecular Crystals, Liquids and Gases," (New
York: Wiley, 1960).
Simonetty, Yee and Tassios, "Prediction and Correlation of LLE," Ind. Eng.
Chem. Process Des. Dev., Vol. 21, (1982), p. 174.
Van Laar Activity Coefficient Model
The Van Laar model (Van Laar 1910) calculates liquid activity coefficients for
the property methods: VANLAAR, VANL2, VANLHOC, VANLNTH, and VANL
RK. It can be used for highly nonideal systems.
Where:
z
i
=
A
i
=
B
i
=
C
i
=
182 2 Thermodynamic Property Models
A
ij
=
C
ij
=
C
ij
= C
ji
A
ii
= B
ii
= C
ii
= 0
a
ij
and b
ij
are unsymmetrical. That is, a
ij
may not be equal to a
ji
, and b
ij
may
not be equal to b
ji
.
Parameters
Name/Element
Symbol DefaultMDS Lower Limit Upper
Limit
Units
VANL/1 a
ij
0 x 50.0 50.0
VANL/2 b
ij
0 x 15000.0 15000.0 TEMPERATURE
VANL/3 c
ij
0 x 50.0 50.0
VANL/4 d
ij
0 x 15000.0 15000.0 TEMPERATURE
The VANL2 property method uses data set 2 for VANL. All other Van Laar
methods use data set 1.
References
J.J. Van Laar, "The Vapor Pressure of Binary Mixtures," Z. Phys. Chem., Vol.
72, (1910), p. 723.
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed. (New York: McGrawHill, 1987).
Wagner Interaction Parameter
The Wagner Interaction Parameter model calculates activity coefficients. This
model is used for dilute solutions in metallurgical applications.
The relative activity coefficient with respect to the reference activity
coefficient of a solute i (in a mixture of solutes i, j, and l and solvent A) is:
Where:
The parameter
i
ref
is the reference activity coefficient of solute i:
k
ij
is a binary parameter:
For any component i, the value of the activity coefficient can be fixed:
i
= g
i
2 Thermodynamic Property Models 183
This model is recommended for dilute solutions.
Parameter Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
GMWIPR/1 a
i
0 x TEMPERATURE
GMWIPR/2 b
i
0 x
GMWIPR/3 c
i
0 x
GMWIPB/1 d
ij
0 x TEMPERATURE
GMWIPB/2 e
ij
0 x
GMWIPB/3 f
ij
0 x
GMWIPO g
i
x
GMWIPS 0 x
GMWIPS is used to identify the solvent component. You must set GMWIPS to
1.0 for the solvent component. This model allows only one solvent.
References
A.D. Pelton and C. W. Bale, "A Modified Interaction Parameter Formalism for
NonDilute Solutions," Metallurgical Transactions A, Vol. 17A, (July 1986),
p. 1211.
Wilson Activity Coefficient Model
The Wilson model calculates liquid activity coefficients for the following
property methods: WILSON, WILS2, WILSHOC, WILSNTH, WILSRK, WILS
HF, WILSLR, and WILSGLR. It is recommended for highly nonideal systems,
especially alcoholwater systems. It can also be used in the advanced
equationofstate mixing rules, such as WongSandler and MHV2. This model
cannot be used for liquidliquid equilibrium calculations.
The equation for the Wilson model is:
Where:
ln A
ij
=
The extended form of ln A
ij
provides more flexibility in fitting phase
equilibrium and enthalpy data. a
ij
, b
ij
, c
ij
, d
ij
, and e
ij
are unsymmetrical. That
is, a
ij
may not be equal to a
ji
, etc.
The binary parameters a
ij
, b
ij
, c
ij
, d
ij
, and e
ij
must be determined from data
regression or VLE and/or heatofmixing data. The Aspen Physical Property
System has a large number of builtin binary parameters for the Wilson
model. The binary parameters have been regressed using VLE data from the
Dortmund Databank. The binary parameters were regressed using the ideal
gas, RedlichKwong, and HaydenO'Connell equations of state. See Physical
Property Data, Chapter 1, for details.
184 2 Thermodynamic Property Models
Parameter
Name/Element
Symbol Default MDS Lower Limit Upper
Limit
Units
WILSON/1 a
ij
0 x 50.0 50.0
WILSON/2 b
ij
0 x 15000.0 15000.0 TEMPERATURE
WILSON/3 c
ij
0 x  TEMPERATURE
WILSON/4 d
ij
0 x TEMPERATURE
WILSON/5 T
lower
0 K x TEMPERATURE
WILSON/6 T
upper
1000 K x TEMPERATURE
WILSON/7 e
ij
0 x TEMPERATURE
The WILS2 property method uses data set 2 for WILSON. All other Wilson
methods use data set 1.
In the original formulation of the Wilson model, a
ij
= ln V
j
/V
i
, c
ij
= d
ij
= e
ij
=
0, and
b
ij
= (
ij

ii
)/R, where V
j
and V
i
are pure component liquid molar volume at
25C.
If any of b
iA
, c
iA
, and e
iA
are nonzero, absolute temperature units are
assumed for all coefficients. If b
iA
, c
iA
, and e
iA
are all zero, the others are
interpreted in input units. The temperature limits are always interpreted in
input units.
References
G.M. Wilson, J. Am. Chem. Soc., Vol. 86, (1964), p. 127.
Wilson Model with Liquid Molar Volume
This Wilson model (used in the method WILSVOL) calculates liquid activity
coefficients using the original formulation of Wilson (Wilson 1964) except that
liquid molar volume is calculated at system temperature, instead of at 25C.
It is recommended for highly nonideal systems, especially alcohol water
systems. It can be used in any activity coefficient property method or in the
advanced equation of state mixing rules, such as Wong Sandler and MHV2.
This model cannot be used for liquid liquid equilibrium calculations.
The equation for the Wilson model is:
Where:
ln A
ij
=
V
j
and V
i
are pure component liquid molar volume at the system temperature
calculated using the General Pure Component Liquid Molar Volume model. The
extended form of ln A
ij
provides more flexibility in fitting phase equilibrium
2 Thermodynamic Property Models 185
and enthalpy data. a
ij
, b
ij
, c
ij
, d
ij
, and e
ij
are unsymmetrical. That is, a
ij
may
not be equal to a
ji
, etc.
The binary parameters a
ij
, b
ij
, c
ij
, d
ij
, and e
ij
must be determined from data
regression of VLE and/or heatofmixing data. There are no built in binary
parameters for this model.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
WSNVOL/1 a
ij
0 x 50.0 50.0
WSNVOL/2 b
ij
0 x 15000.0 15000.0 TEMPERATURE
WSNVOL/3 c
ij
0 x TEMPERATURE
WSNVOL/4 d
ij
0 x TEMPERATURE
WSNVOL/5 e
ij
0 x TEMPERATURE
WSNVOL/6 T
lower
0 K x TEMPERATURE
WSNVOL/7 T
upper
1000 K x TEMPERATURE
Pure component parameters for the General Pure Component Liquid Molar
Volume model are also required.
In the original formulation of the Wilson model, a
ij
= c
ij
= d
ij
= e
ij
= 0 and
. V
j
and V
i
are calculated at 25C.
If any of b
iA
, c
iA
, and e
iA
are nonzero, absolute temperature units are
assumed for all coefficients. If b
iA
, c
iA
, and e
iA
are all zero, the others are
interpreted in input units. The temperature limits are always interpreted in
input units.
Reference: G.M. Wilson, J. Am. Chem. Soc., Vol. 86, (1964), p. 127.
Vapor Pressure and Liquid
Fugacity Models
The Aspen Physical Property System has the following builtin vapor pressure
and liquid fugacity models. This section describes the vapor pressure and
liquid fugacity models available.
Model Type
General Pure Component Liquid Vapor
Pressure
Vapor pressure
API Sour Vapor pressure
Braun K10 Vapor pressure
ChaoSeader Fugacity
GraysonStreed Fugacity
KentEisenberg Fugacity
MaxwellBonnell Vapor pressure
186 2 Thermodynamic Property Models
Model Type
Solid Antoine Vapor pressure
General Pure Component Liquid Vapor
Pressure
The Aspen Physical Property System has several submodels for calculating
vapor pressure of a liquid. It uses parameter THRSWT/3 to determine which
submodel is used. See Pure Component TemperatureDependent Properties
for details.
If THRSWT/3
is
Then this equation is
used
And this parameter is used
0 Extended Antoine PLXANT
200211 Barin CPLXP1, CPLXP2, CPIXP1, CPIXP2,
and CPIXP3
301 Wagner WAGNER
302 PPDS Modified Wagner WAGNER
400 PML LNVPEQ and one of LNVP1, LOGVP1,
LNPR1, LOGPR1, LNPR2, LOGPR2
401 IKCAPE PLPO
501 NIST TDE Polynomial PLTDEPOL
502 NIST Wagner 25 WAGNER25
Extended Antoine Equation
Parameters for many components are available for the extended Antoine
equation from the Aspen Physical Property System pure component databank.
This equation can be used whenever the parameter PLXANT is available.
The equation for the extended Antoine vapor pressure model is:
Extrapolation of ln p
i
*,l
versus 1/T occurs outside of temperature bounds.
Parameter
Name/ Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
PLXANT/1 C
1i
x PRESSURE,
TEMPERATURE
PLXANT/2 C
2i
x TEMPERATURE
PLXANT/3, . . . , 7 C
3i
, ..., C
7i
0 x TEMPERATURE
PLXANT/8 C
8i
0 x TEMPERATURE
PLXANT/9 C
9i
1000 x TEMPERATURE
If C
5i
, C
6i
, or C
7i
is nonzero, absolute temperature units are assumed for all
coefficients C
1i
through C
7i
. The temperature limits are always in user input
units.
2 Thermodynamic Property Models 187
Barin
See Barin Equations for Gibbs Energy, Enthalpy, Entropy, and Heat Capacity
for details about this submodel.
Wagner Vapor Pressure Equation
The Wagner vapor pressure equation is the best equation for correlation. The
equation can be used if the parameter WAGNER is available:
Where:
T
ri
= T / T
ci
p
ri
*,l
= p
i
*,l
/ p
ci
Linear extrapolation of ln p
i
*,l
versus T occurs outside of temperature bounds.
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
PPDS Modified Wagner Vapor Pressure Equation
The PPDS equation also uses the same parameter WAGNER as the standard
Wagner equation:
Where:
T
ri
= T / T
ci
p
ri
*,l
= p
i
*,l
/ p
ci
Linear extrapolation of ln p
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
WAGNER/1 C
1i
x
WAGNER/2 C
2i
0 x
WAGNER/3 C
3i
0 x
WAGNER/4 C
4i
0 x
WAGNER/5 C
5i
0 x TEMPERATURE
WAGNER/6 C
6i
1000 x TEMPERATURE
TC T
ci
TEMPERATURE
PC p
ci
PRESSURE
188 2 Thermodynamic Property Models
NIST Wagner 25 Liquid Vapor Pressure Equation
This equation is the same as the PPDS Modified Wagner equation above, but
it uses parameter WAGNER25 instead of WAGNER, and it uses critical
properties from this parameter set also.
Where:
T
ri
= T / T
ci
Linear extrapolation of ln p
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
WAGNER25/1 C
1i
x
WAGNER25/2 C
2i
0 x
WAGNER25/3 C
3i
0 x
WAGNER25/4 C
4i
0 x
WAGNER25/5 ln p
ci
0 x
WAGNER25/6 T
ci
x TEMPERATURE
WAGNER25/7 T
lower
0 x TEMPERATURE
WAGNER25/8 T
upper
1000 x TEMPERATURE
IKCAPE Vapor Pressure Equation
The IKCAPE model is a polynomial equation. If the parameter PLPO is
available, the Aspen Physical Property System can use the IKCAPE vapor
pressure equation:
Linear extrapolation of ln p
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/Element
Symbol DefaultMDS Lower
Limit
Upper
Limit
Units
PLPO/1 C
1i
X PRESSURE
TEMPERATURE
PLPO/2, ..., 10 C
2i
, ..., C
10i
0 X TEMPERATURE
PLPO/11 C
11i
0 X TEMPERATURE
PLPO/12 C
12i
1000 X TEMPERATURE
2 Thermodynamic Property Models 189
PML Vapor Pressure Equations
The PML vapor pressure equations are modified versions of the Antoine and
Wagner equations. Each equation comes in two alternate forms, identical
except for the use of natural or base10 logarithms. Parameter LNVPEQ/1
specifies the number of the equation used. Each equation uses a separate
parameter: LNVP1 for equation 1, LOGVP1 for 2, LNPR1 for 3, LOGPR1 for 4,
LNPR2 for 5, and LOGPR2 for 6.
Equation 1 (natural logarithm) and 2 (base 10 logarithm):
Equation 3 (natural logarithm) and 4 (base 10 logarithm):
Equation 5 (natural logarithm) and 6 (base 10 logarithm):
Where:
T
ri
= T / T
ci
p
ri
*,l
= p
i
*,l
/ p
ci
LNVPEQ/2 and LNVPEQ/3 are the lower and upper temperature limits,
respectively.
In equations 14, if elements 4, 7, or 8 are nonzero, absolute temperature
units are assumed for all elements.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
LNVP1/1, ...,8 C
1i
, ..., C
8i
x PRESSURE
TEMPERATURE
LOGVP1/1, ..., 8 C
1i
, ..., C
8i
x PRESSURE
TEMPERATURE
LNPR1/1, ..., 8 C
1i
, ..., C
8i
x PRESSURE
TEMPERATURE
LOGPR1/1, ..., 8 C
1i
, ..., C
8i
x PRESSURE
TEMPERATURE
LNPR2/1,2 C
1i
, C
2i
x
LOGPR2/1,2 C
1i
, C
2i
x
LNVPEQ/1 (equation
number)
LNVPEQ/2 T
lower
0 X TEMPERATURE
LNVPEQ/3 T
upper
1000 X TEMPERATURE
TC T
ci
TEMPERATURE
PC p
ci
PRESSURE
190 2 Thermodynamic Property Models
NIST TDE Polynomial for Liquid Vapor Pressure
This equation can be used for calculating vapor pressure when parameter
PLTDEPOL is available.
Linear extrapolation of ln p
i
*,l
versus T occurs outside of temperature bounds.
If elements 2, 3, 6, or 8 are nonzero, absolute temperature units are
assumed for all elements.
Parameter
Name/Element
Symbol DefaultMDS Lower
Limit
Upper
Limit
Units
PLTDEPOL/1 C
1i
X
PLTDEPOL/2 C
2i
0 X TEMPERATURE
PLTDEPOL/3 C
3i
0 X
PLTDEPOL/4, ..., 8 C
4i
, ..., C
8i
0 X TEMPERATURE
PLTDEPOL/9 T
lower
0 X TEMPERATURE
PLTDEPOL/10 T
upper
1000 X TEMPERATURE
References
Reid, Prausnitz, and Poling, The Properties of Gases and Liquids, 4th ed.,
(New York: McGrawHill, 1987).
Harlacher and Braun, "A FourParameter Extension of the Theorem of
Corresponding States," Ind. Eng. Chem. Process Des. Dev., Vol. 9, (1970), p.
479.
W. Wagner, Cryogenics, Vol. 13, (1973), pp. 470482.
D. Ambrose, M. B. Ewing, N. B. Ghiassee, and J. C. Sanchez Ochoa "The
ebulliometric method of vaporpressure measurement: vapor pressures of
benzene, hexafluorobenzene, and naphthalene," J. Chem. Thermodyn. 22
(1990), p. 589.
API Sour Model
The API Sour model is based on the API sour water model for correlating the
ammonia, carbon dioxide, and hydrogen sulfide volatilities from aqueous sour
water system. The model assumes aqueous phase chemical equilibrium
reactions involving CO
2
, H
2
S, and NH
3
. The model is not usable with chemistry
in the true component approach. Use the apparent component approach with
this model.
The model is applicable from 20 C to 140 C. The authors developed the model
using available phase equilibrium data and reported average errors between
the model and measured partial pressure data as follows
Compound Average Error, %
Up to 60 C Above 60 C
2 Thermodynamic Property Models 191
Compound Average Error, %
Up to 60 C Above 60 C
Ammonia 10 36
Carbon dioxide 11 24
Hydrogen sulfide 12 29
Detail of the model is described in the reference below and is too involved to
be reproduced here.
Reference
New Correlation of NH3, CO2, and H2S Volatility Data from Aqueous Sour
Water Systems, API Publication 955, March 1978 (American Petroleum
Institute).
Braun K10 Model
The BK10 model uses the Braun K10 Kvalue correlations, which were
developed from the K10 charts (Kvalues at 10 psia) for both real and pseudo
components. The form of the equation used is an extended Antoine vapor
pressure equation with coefficients specific to real components and pseudo
component boiling ranges.
This method is strictly applicable to heavy hydrocarbons at low pressures.
However, our model includes coefficients for a large number of hydrocarbons
and light gases. For pseudocomponents the model covers boiling ranges 450
700 K (350 800F). Heavier fractions can also be handled using the
methods developed by AspenTech.
References
B.C. Cajander, H.G. Hipkin, and J.M. Lenoir, "Prediction of Equilibrium Ratios
from Nomograms of Improved Accuracy," Journal of Chemical Engineering
Data, vol. 5, No. 3, July 1960, p. 251259.
J.M. Lenoir, "Predict K Values at Low Temperatures, part 1," Hydrocarbon
Processing, p. 167, September 1969.
J.M. Lenoir, "Predict K Values at Low Temperatures, part 2," Hydrocarbon
Processing, p. 121, October 1969.
ChaoSeader Pure Component Liquid
Fugacity Model
The ChaoSeader model calculates pure component fugacity coefficient, for
liquids. It is used in the CHAOSEA property method. This is an empirical
model with the CurlPitzer form. The general form of the model is:
Where:
192 2 Thermodynamic Property Models
=
Parameter
Name/Element
Symbol Default MDS Lower Limit Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
OMEGA
e
i
0.5 2.0
References
K.C. Chao and J.D. Seader, "A General Correlation of VaporLiquid Equilibria
in Hydrocarbon Mixtures," AIChE J., Vol. 7, (1961), p. 598.
GraysonStreed Pure Component Liquid
Fugacity Model
The GraysonStreed model calculates pure component fugacity coefficients for
liquids, and is used in the GRAYSON/GRAYSON2 property methods. It is an
empirical model with the CurlPitzer form. The general form of the model is:
Where:
Parameter
Name/Element
Symbol Default MDS Lower Limit Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
OMEGA
e
i
0.5 2.0
References
H.G. Grayson and C.W. Streed, Paper 20PO7, Sixth World Petroleum
Conference, Frankfurt, June 1963.
KentEisenberg Liquid Fugacity Model
The KentEisenberg model calculates liquid mixture component fugacity
coefficients and liquid enthalpy for the AMINES property method.
The chemical equilibria in H
2
S + CO
2
+ amine systems are described using
these chemical reactions:
2 Thermodynamic Property Models 193
Where:
R' = Alcohol substituted alkyl groups
The equilibrium constants are given by:
The chemical equilibrium equations are solved simultaneously with the
balance equations. This obtains the mole fractions of free H
2
S and CO
2
in
solution. The equilibrium partial pressures of H
2
S and CO
2
are related to the
respective free concentrations by Henry's constants:
The apparent fugacities and partial molar enthalpies, Gibbs energies and
entropies of H
2
S and CO
2
are calculated by standard thermodynamic
relationships. The chemical reactions are always considered.
The values of the coefficients for the seven equilibrium constants (A
1i
, ... A
5i
)
and for the two Henry's constants B
1i
and B
2i
are built into the Aspen Physical
Property System. The coefficients for the equilibrium constants were
determined by regression. All available data for the four amines were used:
monoethanolamine, diethanolamine, disopropanolamine and diglycolamine.
You are not required to enter any parameters for this model.
References
R.L. Kent and B. Eisenberg, Hydrocarbon Processing, (February 1976),
pp. 8792.
194 2 Thermodynamic Property Models
MaxwellBonnell Vapor Pressure Model
The MaxwellBonnell model calculates vapor pressure using the Maxwell
Bonnell vapor pressure correlation for all hydrocarbon pseudocomponents as
a function of temperature. This is an empirical correlation based on API
procedure 5A1.15, 5A1.13. This model is used in property method MXBONNEL
for calculating vapor pressure and liquid fugacity coefficients (Kvalues).
References
API procedure 5A1.15 and 5A1.13.
Solid Antoine Vapor Pressure Model
The vapor pressure of a solid can be calculated using the Antoine equation.
Parameters for some components are available for the extended Antoine
equation from the Aspen Physical Property System pure component databank.
This equation can be used whenever the parameter PSANT is available.
The equation for the solid Antoine vapor pressure model is:
Extrapolation of ln p
i
*,s
versus 1/T occurs outside of temperature bounds.
Parameter
Name/
Element
Symbol Default MDS Lower Limit Upper
Limit
Units
PSANT/1 C
1i
x PRESSURE,
TEMPERATURE
PSANT/2 C
2i
x TEMPERATURE
PSANT/3 C
3i
0 x TEMPERATURE
PSANT/4 C
4i
0 x TEMPERATURE
PSANT/5 C
5i
1000 x TEMPERATURE
General Pure Component Heat
of Vaporization
The Aspen Physical Property System has several submodels for calculating
pure component heat of vaporization. It uses parameter THRSWT/4 to
determine which submodel is used. See Pure Component Temperature
Dependent Properties for details.
If THRSWT/4
is
Then this equation is usedAnd this parameter is
used
0 Watson DHVLWT
106 DIPPR DHVLDP
301 PPDS DHVLDS
401 IKCAPE DHVLPO
2 Thermodynamic Property Models 195
If THRSWT/4
is
Then this equation is usedAnd this parameter is
used
505 NIST TDE Watson equation DHVLTDEW
DIPPR Heat of Vaporization Equation
The DIPPR equation is used to calculate heat of vaporization when THRSWT/4
is set to 106. (Other DIPPR equations may sometimes be used. See Pure
Component TemperatureDependent Properties for details.)
The equation for the DIPPR heat of vaporization model is:
Where:
T
ri
= T / T
ci
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Linear extrapolation of A
vap
H
i
*
versus T occurs outside of temperature bounds,
using the slope at the temperature bound, except that A
vap
H
i
*
is zero for
.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DHVLDP/1 C
1i
x MOLEENTHALPY
DHVLDP/2, ..., 5 C
2i
, ..., C
5i
0 x
DHVLDP/6 C
6i
0 x TEMPERATURE
DHVLDP/7 C
7i
1000 x TEMPERATURE
TC T
ci
5.0 2000.0 TEMPERATURE
Watson Heat of Vaporization Equation
The Watson equation is used to calculate heat of vaporization when
THRSWT/4 is set to 0. See Pure Component TemperatureDependent
Properties for details.
The equation for the Watson model is:
Where:
A
vap
H
i
*
(T
1
)
= Heat of vaporization at temperature T
1
Linear extrapolation of A
vap
H
i
*
versus T occurs below the minimum
temperature bound, using the slope at the temperature bound.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
196 2 Thermodynamic Property Models
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
DHVLWT/1
A
vap
H
i
*
(T
1
)
5x10
4
5x10
8
MOLEENTHALPY
DHVLWT/2 T
1
4.0 3500.0 TEMPERATURE
DHVLWT/3 a
i
0.38 2.0 2.0
DHVLWT/4 b
i
0 2.0 2.0
DHVLWT/5 T
min
0 0.0 1500.0 TEMPERATURE
PPDS Heat of Vaporization Equation
The PPDS equation is used to calculate heat of vaporization when THRSWT/4
is set to 301. See Pure Component TemperatureDependent Properties for
details.
The equation for the PPDS model is:
where R is the gas constant.
Linear extrapolation of A
vap
H
i
*
versus T occurs outside of temperature bounds,
using the slope at the temperature bound.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
DHVLDS/1 C
1i
DHVLDS/2 C
2i
0
DHVLDS/3 C
3i
0
DHVLDS/4 C
4i
0
DHVLDS/5 C
5i
0
DHVLDS/6 C
6i
0 TEMPERATURE
DHVLDS/7 C
7i
1000 TEMPERATURE
IKCAPE Heat of Vaporization Equation
The IKCAPE equation is used to calculate heat of vaporization when
THRSWT/4 is set to 401. See Pure Component TemperatureDependent
Properties for details.
The equation for the IKCAPE model is:
Linear extrapolation of A
vap
H
i
*
versus T occurs outside of temperature bounds,
using the slope at the temperature bound
2 Thermodynamic Property Models 197
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DHVLPO/1 C
1i
X MOLE
ENTHALPY
DHVLPO/2, ...,
10
C
2i
, ..., C
10i
0 X MOLE
ENTHALPY
TEMPERATURE
DHVLPO/11 C
11i
0 X TEMPERATURE
DHVLPO/12 C
12i
1000 X TEMPERATURE
NIST TDE Watson Heat of Vaporization
Equation
The NIST TDE Watson equation is used to calculate heat of vaporization when
THRSWT/4 is set to 505. See Pure Component TemperatureDependent
Properties for details.
The equation is:
Linear extrapolation of A
vap
H
i
*
versus T occurs outside of temperature bounds,
using the slope at the temperature bound
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DHVLTDEW/1 C
1i
X
DHVLTDEW/2, 3,
4
C
2i
, C
3i
, C
4i
0 X
DHVLTDEW/5 T
ci
X TEMPERATURE
DHVLTDEW/6 nTerms 4 X
DHVLTDEW/7 T
lower
0 X TEMPERATURE
DHVLTDEW/8 T
upper
1000 X TEMPERATURE
ClausiusClapeyron Equation
The Aspen Physical Property System can calculate heat of vaporization using
the Clausius Clapeyron equation:
Where:
= Slope of the vapor pressure curve calculated from the Extended
Antoine equation
V
i
*,v
= Vapor molar volume calculated from the Redlich Kwong
equation of state
198 2 Thermodynamic Property Models
V
i
*,l
= Liquid molar volume calculated from the Rackett equation
For parameter requirements, see General Pure Component Liquid Vapor
Pressure, the General Pure Component Liquid Molar Volume model, and
Redlich Kwong.
Molar Volume and Density
Models
The Aspen Physical Property System has the following builtin molar volume
and density models available. This section describes the molar volume and
density models.
Model Type
API Liquid Volume Liquid volume
BrelviO'Connell Partial molar liquid
volume of gases
ChuehPrausnitz Liquid volume
Clarke Aqueous Electrolyte Volume Liquid volume
COSTALD Liquid Volume Liquid volume
DebyeHckel Volume Electrolyte liquid volume
Liquid Constant Molar Volume Model Liquid volume
General Pure Component Liquid Molar
Volume
Liquid volume/liquid
density
Rackett/CampbellThodos Mixture Liquid
Volume
Liquid volume
Modified Rackett Liquid volume
General Pure Component Solid Molar Volume Solid volume
Liquid Volume Quadratic Mixing Rule Liquid volume
API Liquid Molar Volume
This model calculates liquid molar volume for a mixture, using the API
procedure and the Rackett model. Ideal mixing is assumed:
Where:
x
p
= Mole fraction of pseudocomponents
x
r
= Mole fraction of real components
For pseudocomponents, the API procedure is used:
Where:
2 Thermodynamic Property Models 199
fcn = A correlation based on API procedure 6A3.5 (API Technical Data
Book, Petroleum Refining, 4th edition)
At high density, the Ritter equation is used (adapted from Ritter, Lenoir, and
Schweppe, Petrol. Refiner 37 [11] 225 (1958)):
Where SG is the specific gravity, T
b
is the mean average boiling point in
Rankine, T is the temperature of the system in Rankine, and the mass volume
is produced in units of cubic feet per poundmass.
This procedure is valid over the following conditions:
 UOPK: 10.5  12.5
 API: 0  95
 Mean Average Boiling Point: 0  1100 F
 Temperature: 0  1000 F
 Calculated density: 0.4  1.05 g/cc
The effect of pressure is automatically accounted for using procedure 6A3.10.
This procedure has the following validity range:
 Density at low pressure: 0.7  1.0 g/cc
 Pressure: 0  100,000 psig
For real components, the mixture Rackett model is used:
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
See the Rackett model for descriptions.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TB T
b
4.0 2000.0 TEMPERATURE
API API 60.0 500.0
TC T
c
5.0 2000.0 TEMPERATURE
PC p
c
10
5
10
8
PRESSURE
RKTZRA Z
RA
ZC 0.1 0.5
There are two versions of this model: VL2API and VL2API5. Model VL2API is
used in route VLMX20, while model VL2API5 is used in route VLMX24.
The main difference between the VL2API and VL2API5 models is as follows:
 The VL2API model calculates the liquid density for each pseudocomponent
using the API procedure 6A3.5 (or Ritter equation), then computes the
density of the pseudocomponent mixture as a molefractionweighted
average.
200 2 Thermodynamic Property Models
 The VL2API5 model calculates the liquid density of the mixture of the
pseudocomponents as a whole. It first computes the specific gravity and
mean average boiling point of the pseudocomponent mixture, then uses the
API procedure 6A3.5 (or Ritter equation) to compute the mixture liquid
density.
Both models use the same procedure 6A3.10 for pressure correction.
Experience shows that VL2API5 is less sensitive to how the
pseudocomponents are generated from the same assay (number of cut
points, etc.).
BrelviO'Connell
The BrelviO'Connell model calculates partial molar volume of a supercritical
component i at infinite dilution in pure solvent A. Partial molar volume at
infinite dilution is required to compute the effect of pressure on Henry's
constant. (See Henry's Constant.)
The general form of the BrelviO'Connell model is:
Where:
i = Solute or dissolvedgas component
A = Solvent component
V
BO
= V
1
+ TV
2
The above correlation applies to both solute and solvent.
The liquid molar volume of solvent is obtained from the Rackett model:
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
cA
5.0 2000.0 TEMPERATURE
PC p
cA
10
5
10
8
PRESSURE
RKTZRA Z
A
RA
ZC x 0.1 1.0
VLBROC/1 V
1
VC x 1.0 1.0 MOLEVOLUME
VLBROC/2 V
2
0 x 0.1 0.1 TEMPERATURE
References
S.W. Brelvi and J.P. O'Connell, AIChE J., Vol. 18, (1972), p. 1239.
S.W. Brelvi and J.P. O'Connell, AIChE J., Vol. 21, (1975), p. 157.
2 Thermodynamic Property Models 201
ChuehPrausnitz Liquid Molar Volume
Model
The ChuehPrausnitz model is used to calculate liquid molar volume for pure
components and mixtures. It uses the Rackett model to calculate the
saturated molar volume then applies a pressure correction term. The mixture
form of the model is described below. The equations reduce to the pure
component version.
Where
V
m
=Liquid molar volume of the mixture
V
m
s
=Liquid molar volume at saturation, calculated using the Rackett
model for mixtures or the General liquid volume model for pure
components
Z
cm
=
Critical compressibility factor of the mixture = Ex
i
Z
ci
Z
ci
=Critical compressibility factor of component i
N =
P =System pressure
P
vp
=
Saturation pressure = for
mixtures, or calculated using the General vapor pressure model
for pure components
P
cm
=Critical pressure of the mixture
T =Temperature
T
rm
=Reduced temperature = T/T
cm
T
cm
=Critical temperature of the mixture
e
m
=
Acentric factor of the mixture = Ex
i
e
i
e
i
=Acentric factor of component i
T
cm
and P
cm
are calculated by:
V
cm
=
Ex
i
Z
ci
T
ci
/P
ci
T
cm
=
[Ex
i
Z
ci
T
ci
3/2
/P
ci
]
2
/V
cm
2
P
cm
= Z
cm
T
cm
/V
cm
T
ci
= Critical temperature of component i
P
ci
= Critical pressure of component i
202 2 Thermodynamic Property Models
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Note: Above T
r
=0.99 an extrapolation method is used to smooth the
transition to constant molar volume equal to the critical volume.
The parameters required for the model are the same as those for the Rackett
model with the addition of the acentric factor, e
i
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
VCRKT V
ci
VC x 0.001 3.5 MOLE
VOLUME
RKTZRA Z
i
*,RA
ZC x 0.1 1.0
RKTKIJ k
ij
x 5.0 5.0
OMEGA
e
i
0.5 2.0
Reference
Chueh, P.L. and J.M. Prausnitz, AIChE J. 15, 471 (1969).
Clarke Aqueous Electrolyte Volume
The Clarke model calculates liquid molar volume for electrolytes solutions.
The model is applicable to mixed solvents and is based on:
 Molar volume of molecular solvents (equation 2)
 The relationship between the partial molar volume of an electrolyte and its
mole fraction in the solvent (equation 4)
All quantities are expressed in terms of apparent components.
If option code 1 is set to 1, the liquid volume quadratic mixing rule is used
instead. The default option uses this equation to calculate the liquid molar
volume for electrolyte solutions:
(1)
Where:
V
m
l
= Liquid molar volume for electrolyte solutions.
V
s
l
= Liquid molar volume for solvent mixtures.
V
e
l
= Liquid molar volume for electrolytes.
Apparent Component Approach
For molecular solvents, the liquid molar volume is calculated by:
2 Thermodynamic Property Models 203
(2)
Where:
x
w
= Mole fraction of water
V
w
*
= Molar volume of water from the steam table.
x
nws
= Sum of the mole fractions of all nonwater
solvents.
V
nws
l
= Liquid molar volume for the mixture of all non
water solvents. It is calculated using the
Rackett equation.
For electrolytes:
(3)
(4)
(5)
(6)
Where:
x
ca
= Apparent mole fraction of electrolyte ca
V
ca
= Liquid molar volume for electrolyte ca
The mole fractions x
ca
are reconstituted arbitrarily from the true ionic
concentrations, even if you use the apparent component approach. This
technique is explained in Electrolyte Calculation in Physical Property Methods.
The result is that electrolytes are generated from all possible combinations of
ions in solution. The following equation is consistently applied to determine
the amounts of each possible apparent electrolyte n
ca
:
(7)
Where:
n
ca
= Number of moles of apparent electrolyte ca
z
c
= Charge of c
z
factor
= z
c
if c and a have the same number of
charges; otherwise 1.
n
c
= Number of moles of cation c
n
a
= Number of moles of anion a
204 2 Thermodynamic Property Models
For example: given an aqueous solution of Ca
2+
, Na
+
, SO
4
2
, Cl

four
electrolytes are found: CaCl
2
, Na
2
SO
4
, CaSO
4
, and NaCl. The Clarke
parameters of all four electrolytes are used. You can rely on the default,
which calculates the Clarke parameters from ionic parameters. Otherwise, you
must enter parameters for any electrolytes that may not exist in the
components list. If you do not want to use the default, the first step in using
the Clarke model is to add any needed components for electrolytes not in the
components list.
True Component Approach
The true molar volume is obtained from the apparent molar volume:
(8)
Where:
V
m
l,t
= Liquid volume per number of true species
V
m
l,a
= Liquid volume per number of apparent species, V
m
l
of equation 1
n
a
= Number of apparent species
n
t
= Number of true species
The apparent molar volume is calculated as explained in the preceding
subsection.
Temperature Dependence
The temperature dependence of the molar volume of the solution is
approximately equal to the temperature dependence of the molar volume of
the solvent mixture:
(9)
Where V
m
l
(298.15K) is calculated from equation 1.
Parameter
Name/Element
Applicable
Components
Symbol Default Units
VLCLK/1 CationAnion MOLEVOLUME
VLCLK/2 CationAnion A
ca
0.020 MOLEVOLUME
If VLCLK/1 is missing, it is calculated based on VLBROC and CHARGE as
follows:
(10)
If VLBROC/1 is missing, the default value of 0.0012 is used. See the Brelvi
O'Connell model for VLBROC and also Rackett/CampbellThodos Mixture
Liquid Volume for additional parameters used in the Rackett equation.
2 Thermodynamic Property Models 205
Reference
C.C. Chen, private communication.
COSTALD Liquid Volume
The equation for the COSTALD liquid volume model is:
Where:
V
m
R,0
and V
m
R,o
are functions or T
r
for
For , there is a linear interpolation between the liquid density at
T
r
= 0.95 and the vapor density at T
r
= 1.05. This model can be used to
calculate saturated and compressed liquid molar volume. The compressed
liquid molar volume is calculated using the Tait equation:
Where B and C are functions of T, e, T
c
, P
c
and P
sat
is the saturated pressure
at T.
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Mixing Rules:
Where:
To improve results, the Aspen Physical Property System uses a special
correlation for water when this model is used. Changing the VSTCTD and
OMGCTD parameters for water will not affect the results of the special
correlation.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
VSTCTD V
r
*,CTD
VC X 0.001 3.5 MOLEVOLUME
OMGCTD
e
r
OMEGA X 0.5 2.0
206 2 Thermodynamic Property Models
References
R.W. Hankinson and G.H. Thomson, AIChE J., Vol. 25, (1979), p. 653.
G.H. Thomson, K.R. Brobst, and R.W. Hankinson, AIChE J., Vol. 28, (1982),
p. 4, p. 671.
DebyeHckel Volume
The DebyeHckel model calculates liquid molar volume for aqueous
electrolyte solutions.
The equation for the DebyeHckel volume model is:
Where:
V
w
* is the molar volume for water and is calculated from the ASME steam
table.
V
k
is calculated from the DebyeHckel limiting law for ionic species:
Where:
V
k
w
= Density of water (kg / m
3
)
2 Thermodynamic Property Models 207
c
w
= Dielectric constant of water (Fm
1
), a function
of pressure and temperature (Bradley and
Pitzer, 1979)
The above equation for V
k
can also apply to nonwater molecular species
since the second term on the right is zero; it is assumed to be the infinite
dilution partial volume for molecular species.
V
k
= V
1
+ TV
2
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
VLBROC/1 V
1
0 x 1.0 1.0 MOLEVOLUME
VLBROC/2 V
2
0 x 0.1 0.1 TEMPERATURE
References
D. J. Bradley, K. S. Pitzer, "Thermodynamics of Electrolytes," J. Phys. Chem.,
83 (12), 1599 (1979).
H.C. Helgeson and D.H. Kirkham, "Theoretical prediction of the
thermodynamic behavior of aqueous electrolytes at high pressure and
temperature. I. Thermodynamic/electrostatic properties of the solvent", Am.
J. Sci., 274, 1089 (1974).
K.S. Pitzer, "Theory: Ion Interaction Approach," Activity Coefficients in
Electrolyte Solutions, Pytkowicz, R. ed., Vol. I, (CRC Press Inc., Boca Raton,
Florida, 1979).
Liquid Constant Molar Volume Model
This model, VL0CONS, uses a constant value entered by the user as the pure
component liquid molar volume. It is not a function of temperature or
pressure. This is used with the solids handling property method for modeling
nonconventional solids.
Parameter Name Default MDS Units
VLCONS * x MOLEVOLUME
* When no value is provided for VLCONS, the General Pure Component Liquid
Molar Volume model is used to calculate the liquid molar volume.
General Pure Component Liquid Molar
Volume
The Aspen Physical Property System has several submodels for calculating
liquid molar volume. It uses parameter THRSWT/2 to determine which
submodel is used. See Pure Component TemperatureDependent Properties
for details.
If THRSWT/2 is This equation is used And this parameter is used
0 Rackett RKTZRA
100116 DIPPR DNLDIP
208 2 Thermodynamic Property Models
If THRSWT/2 is This equation is used And this parameter is used
301 PPDS DNLPDS
401 IKCAPE VLPO
503 NIST ThermoML
Polynomial
DNLTMLPO
504 NIST TDE expansion DNLEXSAT
514 NIST TDE Rackett DNLRACK
515 NIST COSTALD DNLCOSTD
For liquid molar volume of mixtures, the Rackett mixture equation is always
used by default. This is not necessarily consistent with the pure component
molar volume or density. If you need consistency, select route VLMX26 on the
Methods  Selected Methods form. This route calculates mixture molar
volume from the molefraction average of pure component molar volumes.
Many of these equations calculate density first, and return calculate liquid
molar volume based on that density:
DIPPR
DIPPR equation 105 is the default DIPPR equation for most substances:
This equation is similar to the Rackett equation.
DIPPR equation 116 is the default equation for water.
t = 1  T / T
c
Other DIPPR equations, such as equation 100, may be used for some
substances. Check the value of THRSWT/2 to determine the equation used.
See Pure Component TemperatureDependent Properties for details about
DIPPR equations.
In either case, linear extrapolation of
i
*,l
versus T occurs outside of
temperature bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNLDIP/1 C
1i
x MOLEDENSITY
DNLDIP/2 C
2i
0 x
DNLDIP/3 C
3i
T
ci
x TEMPERATURE
DNLDIP/4 C
4i
0 x
DNLDIP/5 C
5i
0 x
DNLDIP/6 C
6i
0 x TEMPERATURE
DNLDIP/7 C
7i
1000 x TEMPERATURE
2 Thermodynamic Property Models 209
For equation 116, the units are MOLEDENSITY for parameters DNLDIP/1
through DNLDIP/5 and the default for DNLDIP/3 is 0. For equation 105,
DNLDIP/5 is not used, and absolute temperature units are assumed for
DNLDIP/3.
PPDS
The PPDS equation is:
Linear extrapolation of
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/Element
Symbol DefaultMDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
VC V
ci
0.001 3.5 MOLEVOLUME
TC T
ci
5.0 2000.0 TEMPERATURE
DNLPDS/1 C
1i
MASSDENSITY
DNLPDS/2 C
2i
0 MASSDENSITY
DNLPDS/3 C
3i
0 MASSDENSITY
DNLPDS/4 C
4i
0 MASSDENSITY
DNLPDS/5 C
5i
0 x TEMPERATURE
DNLPDS/6 C
6i
1000 x TEMPERATURE
IKCAPE
The IKCAPE equation is:
Linear extrapolation of V
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
VLPO/1 C
1i
X MOLE
VOLUME
VLPO/2, ..., 10 C
2i
, ..., C
10i
0 X MOLE
VOLUME
TEMPERATURE
VLPO/11 C
11i
0 X TEMPERATURE
VLPO/12 C
12i
1000 X TEMPERATURE
NIST ThermoML Polynomial
This equation can be used when parameter DNLTMLPO is available.
210 2 Thermodynamic Property Models
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNLTMLPO/1 C
1i
X MOLE
DENSITY
DNLTMLPO/2, 3,
4
C
2i
, C
3i
, C
4i
0 X MOLE
DENSITY
TEMPERATURE
DNLTMLPO/5 nTerms 4 X
DNLTMLPO/6 0 X TEMPERATURE
DNLTMLPO/7 1000 X TEMPERATURE
Rackett
The Rackett equation is:
Where:
T
r
= T / T
ci
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Note: Above T
r
=0.99 an extrapolation method is used to smooth the
transition to constant molar volume equal to the critical volume.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
RKTZRA Z
i
*,RA
ZC x 0.1 1.0
NIST TDE Rackett Parameters
This equation can be used when the parameter DNLRACK is available.
Linear extrapolation of V
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNLRACK/1 Z
c
x
DNLRACK/2 n 2/7 x
DNLRACK/3 T
ci
x TEMPERATURE
2 Thermodynamic Property Models 211
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNLRACK/4 p
ci
0 x PRESSURE
DNLRACK/5 T
lower
0 x TEMPERATURE
DNLRACK/6 T
upper
1000 x TEMPERATURE
NIST COSTALD Parameters
This equation can be used when the parameter DNLCOSTD is available.
Linear extrapolation of V
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNLCOSTD/1 V
oi
x VOLUME
DNLCOSTD/2
O
0 x
DNLCOSTD/3 T
ci
x TEMPERATURE
DNLCOSTD/4 T
lower
0 x TEMPERATURE
DNLCOSTD/5 T
upper
1000 x TEMPERATURE
NIST TDE Expansion
This equation can be used when the parameter DNLEXSAT is available.
Linear extrapolation of V
i
*,l
versus T occurs outside of temperature bounds.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNLEXSAT/1
ci
x MOLE
DENSITY
DNLEXSAT/2 C
1i
x MOLE
DENSITY
DNLEXSAT/3,
..., DNLEXSAT/7
C
2i
, ..., C
6i
0 x MOLE
DENSITY
DNLEXSAT/8 T
ci
x TEMPERATURE
212 2 Thermodynamic Property Models
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNLEXSAT/9 nTerms 6 x
DNLEXSAT/10 T
lower
0 x TEMPERATURE
DNLEXSAT/11 T
upper
1000 x TEMPERATURE
References
H.G. Rackett, J.Chem. Eng. Data., Vol. 15, (1970), p. 514.
C.F. Spencer and R.P. Danner, J. Chem. Eng. Data, Vol. 17, (1972), p. 236.
Rackett/CampbellThodos Mixture Liquid
Volume
The Rackett equation calculates liquid molar volume for all activity coefficient
based and petroleum tuned equation of state based property methods. In the
last category of property methods, the equation is used in conjunction with
the API model. The API model is used for pseudocomponents, while the
Rackett model is used for real components. (See API Liquid Volume .)
CampbellThodos is a variation on the Rackett model which allows the
compressibility term Z
i
*,RA
to vary with temperature.
Rackett
The equation for the Rackett model is:
Where:
T
c
=
=
Z
m
RA
=
V
cm
=
T
r
= T / T
c
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Note: Above T
r
=0.99 an extrapolation method is used to smooth the
transition to constant molar volume equal to the critical volume.
2 Thermodynamic Property Models 213
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
VCRKT V
ci
VC x 0.001 3.5 MOLE
VOLUME
RKTZRA Z
i
*,RA
ZC x 0.1 1.0
RKTKIJ k
ij
x 5.0 5.0
CampbellThodos
The CampbellThodos model uses the same equation as the Rackett model,
above, except that Z
m
RA
is allowed to vary with temperature:
Z
m
RA
=
CampbellThodos uses a separate set of parameters, RACKET. T
min
and T
max
define the temperature range where the equation is applicable.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
RACKET/1 R*T
ci
/P
ci
R*Tci/Pci MOLE
VOLUME
RACKET/2 Z
i
*,RA
RKTZRA x 0.1 1.0
RACKET/3 d
i
0 x 0 0.11
RACKET/4 T
min
0 x TEMPERATURE
RACKET/5 T
max
1000 x TEMPERATURE
The CampbellThodos model is used when RACKET/3 is set to a value less
than 0.11. Do not change this parameter unless you intend to use this model.
References
H.G. Rackett, J.Chem, Eng. Data., Vol. 15, (1970), p. 514.
C.F. Spencer and R.P. Danner, J. Chem. Eng. Data, Vol. 17, (1972), p. 236.
Modified Rackett Liquid Molar Volume
The Modified Rackett equation improves the accuracy of liquid mixture molar
volume calculation by introducing additional parameters to compute the pure
component parameter RKTZRA and the binary parameter k
ij
.
The equation for the Modified Rackett model is:
214 2 Thermodynamic Property Models
Where:
T
c
=
k
ij
=
=
Z
m
RA
=
Z
i
*,RA
=
V
cm
=
T
r
= T / T
c
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Note: Above T
r
=0.99 an extrapolation method is used to smooth the
transition to constant molar volume equal to the critical volume.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MRKZRA/1 a
i
RKTZRA x 0.1 0.5
MRKZRA/2 b
i
0 x
MRKZRA/3 c
i
0 x
MRKKIJ/1 A
ij
x
MRKKIJ/2 B
ij
0 x
MRKKIJ/3 C
ij
0 x
References
H.G. Rackett, J.Chem, Eng. Data., Vol. 15, (1970), p. 514.
C.F. Spencer and R.P. Danner, J. Chem. Eng. Data, Vol. 17, (1972), p. 236.
Rackett Extrapolation Method
The Rackett equation has a formula involving an exponent of 1+(1T
r
)
2/7
which is invalid above the critical temperature. As a result, a special
extrapolation method is required for this equation. This method involves the
2 Thermodynamic Property Models 215
calculation of an intermediate temperature T
00
near the critical temperature.
When the temperature exceeds T
00
, the volume is constant at the critical
volume. When the temperature is between 0.99T
c
and T
00
, a circle equation is
used to smoothly interpolate the volume between the value and slope at
0.99T
c
and the constant value at T
00
.
Details
First the volume V
0
at 0.99T
c
and the critical volume V
c
are calculated:
Then the volume difference is calculated, as well as the temperature
difference required for the circle equation:
From this, the required intermediate temperature T
00
can be calculated:
Then the volume V
00
at the circle's center can be calculated:
216 2 Thermodynamic Property Models
Finally, the equation of the circle is used to determine any point (T,V) for
0.99T
c
< T < T
00
:
General Pure Component Solid Molar
Volume
The Aspen Physical Property System has several submodels for calculating
solid molar volume. It uses parameter THRSWT/1 to determine which
submodel is used. See Pure Component TemperatureDependent Properties
for details.
If THRSWT/1 is This equation is used And this parameter is
used
0 Aspen VSPOLY
100 DIPPR DNSDIP
401 IKCAPE VSPO
503 NIST ThermoML
polynomial
DNSTMLPO
Aspen Polynomial
The equation for the Aspen solids volume polynomial is:
Linear extrapolation of V
i
*,s
versus T occurs outside of temperature bounds.
Parameter
Name
Applicable
Components
Symbol MDS Default Units
VSPOLY/1 Salts, CI solids C
1i
x MOLEVOLUME
TEMPERATURE
VSPOLY/2, ..., 5 Salts, CI solids C
2i
, ..., C
5i
x 0 MOLEVOLUME
TEMPERATURE
VSPOLY/6 Salts, CI solids C
6i
x 0 MOLEVOLUME
TEMPERATURE
VSPOLY/7 Salts, CI solids C
7i
x 1000 MOLEVOLUME
TEMPERATURE
IKCAPE Equation
The IKCAPE equation is:
2 Thermodynamic Property Models 217
Linear extrapolation of V
i
*,s
versus T occurs outside of temperature bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
VSPO/1 C
1i
X MOLE
VOLUME
VSPO/2, ..., 10 C
2i
, ..., C
10i
0 X MOLE
VOLUME
TEMPERATURE
VSPO/11 C
11i
0 X TEMPERATURE
VSPO/12 C
12i
1000 X TEMPERATURE
DIPPR
The DIPPR equation is:
Linear extrapolation of
i
*,s
versus T occurs outside of temperature bounds.
V
i
*,s
= 1 /
i
*,s
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
The model returns solid molar volume for pure components.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNSDIP/1 C
1i
x MOLE
DENSITY
DNSDIP/2, ..., 5 C
2i
, ..., C
5i
0 x MOLE
DENSITY,
TEMPERATURE
DNSDIP/6 C
6i
0 x TEMPERATURE
DNSDIP/7 C
7i
1000 x TEMPERATURE
NIST ThermoML Polynomial
Linear extrapolation of
i
*,s
versus T occurs outside of temperature bounds.
V
i
*,s
= 1 /
i
*,s
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DNSTMLPO/1 C
1i
x MOLE
DENSITY
DNSTMLPO/2,...,8 C
2i
, ..., C
8i
0 x MOLE
DENSITY,
TEMPERATURE
DNSTMLPO/9 nTerms 8 x
DNSTMLPO/10 T
lower
0 x TEMPERATURE
DNSTMLPO/11 T
upper
1000 x TEMPERATURE
218 2 Thermodynamic Property Models
Liquid Volume Quadratic Mixing Rule
With i and j being components, the liquid volume quadratic mixing rule is:
Option Codes
Option Code Value Descriptions
1 0 Use normal pure component liquid volume model for all
components (default)
1 Use steam tables for water
2 0 Use mole basis composition (default)
1 Use mass basis composition
Parameter
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
VLQKIJ K
ij
0 x  
This model is not part of any property method. To use it, you will need to
define a property method on the Methods  Selected Methods form.
Specify the route VLMXQUAD for VLMX on the Routes sheet of this form, or
the model VL2QUAD for VLMX on the Models sheet.
Heat Capacity Models
The Aspen Physical Property System has five builtin heat capacity models.
This section describes the heat capacity models available.
Model Type
Aqueous Infinite Dilution Heat Capacity
Polynomial
Electrolyte liquid
CrissCobble Aqueous Infinite Dilution
Ionic Heat Capacity
Electrolyte liquid
General Pure Component Liquid Heat
Capacity
Liquid
General Pure Component Ideal Gas Heat
Capacity
Ideal gas
General Pure Component Solid Heat
Capacity
Solid
Aqueous Infinite Dilution Heat Capacity
The aqueous phase infinite dilution enthalpies, entropies, and Gibbs energies
are calculated from the heat capacity polynomial. The values are used in the
calculation of aqueous and mixed solvent properties of electrolyte solutions:
2 Thermodynamic Property Models 219
versus T is linearly extrapolated using the slope at C
7i
for T < C
7i
versus T is linearly extrapolated using the slope at C
8i
for T < C
8i
Parameter
Name/Element
Applicable
Components
Symbol Default Units
CPAQ0/1 Ions, molecular solutes C
1i
TEMPERATURE and
HEAT CAPACITY
CPAQ0/2,,6 Ions, molecular solutes C
2i
, ..., C
6i
0 TEMPERATURE and
HEAT CAPACITY
CPAQ0/7 Ions, molecular solutes C
7i
0 TEMPERATURE
CPAQ0/8 Ions, molecular solutes C
8i
1000 TEMPERATURE
If any of C
4i
through C
6i
is nonzero, absolute temperature units are
assumed for C
1i
through C
6i
. Otherwise, user input units for temperature are
used. The temperature limits are always interpreted in user input units.
CrissCobble Aqueous Infinite Dilution Ionic
Heat Capacity
The CrissCobble correlation for aqueous infinite dilution ionic heat capacity is
used if no parameters are available for the aqueous infinite dilution heat
capacity polynomial. From the calculated heat capacity, the thermodynamic
properties entropy, enthalpy and Gibbs energy at infinte dilution in water are
derived:
Parameter
Name
Applicable
Components
Symbol Default Units
IONTYP Ions Ion Type 0
SO25C Anions MOLEENTROPY
Cations MOLEENTROPY
General Pure Component Liquid Heat
Capacity
The Aspen Physical Property System has several submodels for calculating
liquid heat capacity. It uses parameter THRSWT/6 to determine which
submodel is used. See Pure Component TemperatureDependent Properties
for details.
If THRSWT/6 is This equation is used And this parameter is used
100 DIPPR CPLDIP
220 2 Thermodynamic Property Models
If THRSWT/6 is This equation is used And this parameter is used
200211 Barin CPLXP1, CPLXP2
301 PPDS CPLPDS
401 IKCAPE heat capacity
polynomial
CPLPO
403 IKCAPE liquid heat
capacity
CPLIKC
503 NIST ThermoML
polynomial
CPLTMLPO
506 NIST TDE equation CPLTDECS
This liquid heat capacity model is used to calculate pure component liquid
heat capacity and pure component liquid enthalpy. To use this model, two
conditions must exist:
 One of the parameters for calculating heat capacity (see table) is
available.
 The component is not supercritical (HENRYCOMP).
The model uses a specific method (see Methods in Property Calculation
Methods and Routes):
Where
= Reference enthalpy calculated at T
ref
= Ideal gas enthalpy
= Vapor enthalpy departure
= Enthalpy of vaporization
T
ref
is the reference temperature; it defaults to 298.15 K. You can enter a
different value for the reference temperature. This is useful when you want to
use this model for very light components or for components that are solids at
298.15K.
Activate this model by specifying the route DHL09 for the property DHL on
the Methods  Selected Methods  Routes sheet. For equation of state
property method, you must also modify the route for the property DHLMX to
use a route with method 2 or 3, instead of method 1. For example, you can
use the route DHLMX00 or DHLMX30. You must ascertain that the route for
DHLMX that you select contains the appropriate vapor phase model and heat
of mixing calculations. Click the View button on the form to see details of the
route.
Optionally, you can specify that this model is used for only certain
components. The properties for the remaining components are then
calculated by the standard model. Use the parameter COMPHL to specify the
2 Thermodynamic Property Models 221
components for which this model is used. By default, all components with the
CPLDIP or CPLIKC parameters use this model.
Barin
See Barin Equations for Gibbs Energy, Enthalpy, Entropy, and Heat Capacity
for details about this submodel.
DIPPR Liquid Heat Capacity
The DIPPR equation is used to calculate liquid heat capacity when parameter
THRSWT/6 is 100.
The DIPPR equation is:
Linear extrapolation occurs for C
p
*,l
versus T outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLDIP/1 C
1i
x MOLEHEAT
CAPACITY,
TEMPERATURE
CPLDIP/2, ..., 5 C
2i
, ..., C
5i
0 x MOLEHEAT
CAPACITY,
TEMPERATURE
CPLDIP/6 C
6i
0 x TEMPERATURE
CPLDIP/7 C
7i
1000 x TEMPERATURE
TREFHL T
ref
298.15 TEMPERATURE
COMPHL
To specify that the model is used for a component, enter a value of 1.0 for
COMPHL.
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
PPDS Liquid Heat Capacity
The PPDS equation is used to calculate liquid heat capacity when parameter
THRSWT/6 is 301.
The PPDS equation is:
Where R is the gas constant.
Linear extrapolation occurs for C
p
*,l
versus T outside of bounds.
222 2 Thermodynamic Property Models
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLPDS/1 C
1i
CPLPDS/2 C
2i
0
CPLPDS/3 C
3i
0
CPLPDS/4 C
4i
0
CPLPDS/5 C
5i
0
CPLPDS/6 C
6i
0
CPLPDS/7 C
7i
0 x TEMPERATURE
CPLPDS/8 C
8i
0.99 TC x TEMPERATURE
IKCAPE Liquid Heat Capacity
Two IKCAPE equations can be used to calculate liquid heat capacity. Linear
extrapolation occurs for C
p
*,l
versus T outside of bounds for either equation.
When THRSWT/6 is 403, the IKCAPE liquid heat capacity equation is used.
The equation is:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLIKC/1 C
1i
T
x MOLEHEAT
CAPACITY
CPLIKC/2,...,4 C
2i
, ..., C
4i
0 x MOLEHEAT
CAPACITY,
TEMPERATURE
CPLIKC/5 C
5i
0 x MOLEHEAT
CAPACITY,
TEMPERATURE
CPLIKC/6 C
6i
0 x TEMPERATURE
CPLIKC/7 C
7i
1000 x TEMPERATURE
If C
5i
is nonzero, absolute temperature units are assumed for C
2i
through
C
5i
. Otherwise, user input units for temperature are used. The temperature
limits are always interpreted in user input units.
When THRSWT/6 is 401, the IKCAPE heat capacity polynomial is used.The
equation is:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLPO/1 C
1i
X MOLECAPACITY
CPLPO/2,,10 C
2i
, ..., C
10i
0 X MOLECAPACITY
TEMPERATURE
CPLPO/11 C
11i
0 X TEMPERATURE
2 Thermodynamic Property Models 223
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLPO/12 C
12i
1000 X TEMPERATURE
See Pure Component TemperatureDependent Properties for details on the
THRSWT parameters.
NIST Liquid Heat Capacity
Two NIST equations can be used to calculate liquid heat capacity. Linear
extrapolation occurs for C
p
*,l
versus T outside of bounds for either equation.
When THRSWT/6 is 503, the ThermoML polynomial is used to calculate liquid
heat capacity:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLTMLPO/1 C
1i
X J/K^2/mol
CPLTMLPO/2,,5 C
2i
, ..., C
5i
0 X J/K^2/mol
CPLTMLPO/6 nTerms 5 X
CPLTMLPO/7 T
lower
0 X TEMPERATURE
CPLTMLPO/8 T
upper
1000 X TEMPERATURE
When THRSWT/6 is 506, the TDE equation is used to calculate liquid heat
capacity:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLTDECS/1 C
1i
X MOLEHEAT
CAPACITY
CPLTDECS/2,,4 C
2i
, ..., C
4i
0 X MOLEHEAT
CAPACITY
CPLTDECS/5 B 0 X MOLEHEAT
CAPACITY
CPLTDECS/6 T
ci
X TEMPERATURE
CPLTDECS/7 nTerms 4 X
CPLTDECS/8 T
lower
0 X TEMPERATURE
CPLTDECS/9 T
upper
1000 X TEMPERATURE
See Pure Component TemperatureDependent Properties for details on the
THRSWT parameters.
224 2 Thermodynamic Property Models
General Pure Component Ideal Gas Heat
Capacity
The Aspen Physical Property System has several submodels for calculating
ideal gas heat capacity. It uses parameter THRSWT/7 to determine which
submodel is used. See Pure Component TemperatureDependent Properties
for details.
If THRSWT/7 is This equation is used And this parameter is
used
0 Ideal gas heat capacity
polynomial
CPIG
107, 127 DIPPR 107 or 127 CPIGDP
200211 Barin CPIXP1, CPIXP2, CPIXP3
301 PPDS CPIGDS
401 IKCAPE heat capacity
polynomial
CPIGPO
503 NIST ThermoML
polynomial
CPITMLPO
513 NIST AlyLee CPIALEE
These equations are also used to calculate ideal gas enthalpies, entropies,
and Gibbs energies.
Aspen Ideal Gas Heat Capacity Polynomial
The ideal gas heat capacity polynomial is available for components stored in
ASPENPCD, AQUEOUS, and SOLIDS databanks. This model is also used in
PCES.
C
p
*,ig
is linearly extrapolated using slope at
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPIG/1 C
1i
MOLEHEATCAPACITY,
TEMPERATURE
CPIG/2, ..., 6 C
2i
, ..., C
6i
0 MOLEHEATCAPACITY,
TEMPERATURE
CPIG/7 C
7i
0 TEMPERATURE
CPIG/8 C
8i
1000 TEMPERATURE
CPIG/9, 10,
11
C
9i
, C
10i
, C
11i
MOLEHEATCAPACITY,
TEMPERATURE
If C
10i
or C
11i
is nonzero, then absolute temperature units are assumed for
C
9i
through C
11i
. Otherwise, user input temperature units are used for all
parameters. User input temperature units are always used for C
1i
through C
8i
.
2 Thermodynamic Property Models 225
NIST ThermoML Polynomial
This equation can be used when parameter CPITMLPO is available.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPITMLPO /1 C
1i
x J/K^2/mol
CPITMLPO /2,
..., 6
C
2i
, ..., C
6i
0 x J/K^2/mol
CPITMLPO/7 nTerms 6 x
CPITMLPO /8 T
lower
0 x TEMPERATURE
CPITMLPO /9 T
upper
1000 x TEMPERATURE
DIPPR Equation 107
The DIPPR ideal gas heat capacity equation 107 by Aly and Lee 1981 is:
No extrapolation is used with this equation.
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPIGDP/1 C
1i
x MOLEHEAT
CAPACITY
CPIGDP/2 C
2i
0 x MOLEHEAT
CAPACITY
CPIGDP/3 C
3i
0 x TEMPERATURE
CPIGDP/4 C
4i
0 x MOLEHEAT
CAPACITY
CPIGDP/5 C
5i
0 x TEMPERATURE
CPIGDP/6 C
6i
0 x TEMPERATURE
CPIGDP/7 C
7i
1000 x TEMPERATURE
Absolute temperature units are assumed for C
3i
and C
5i
. The temperature
limits are always interpreted in user input units.
DIPPR Equation 127
The DIPPR ideal gas heat capacity equation 127 is:
226 2 Thermodynamic Property Models
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPIGDP/1 C
1i
x MOLEHEAT
CAPACITY
CPIGDP/2 C
2i
0 x MOLEHEAT
CAPACITY
CPIGDP/3 C
3i
0 x TEMPERATURE
CPIGDP/4 C
4i
0 x MOLEHEAT
CAPACITY
CPIGDP/5 C
5i
0 x TEMPERATURE
CPIGDP/6 C
6i
0 x MOLEHEAT
CAPACITY
CPIGDP/7 C
7i
0 x TEMPERATURE
Absolute temperature units are assumed for C
3i
, C
5i
and C
7i
.
Barin
See Barin Equations for Gibbs Energy, Enthalpy, Entropy, and Heat Capacity
for details about this submodel.
NIST AlyLee
This equation is the same as the DIPPR Aly and Lee equation 107 above, but
it uses a different parameter set. Note that elements 6 and 7 of the CPIALEE
parameter are not used in the equation.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPIALEE/1 C
1i
x MOLEHEAT
CAPACITY
CPIALEE/2 C
2i
0 x MOLEHEAT
CAPACITY
CPIALEE/3 C
3i
0 x TEMPERATURE
CPIALEE/4 C
4i
0 x MOLEHEAT
CAPACITY
CPIALEE/5 C
5i
0 x TEMPERATURE
CPIALEE/8 C
6i
0 x TEMPERATURE
CPIALEE/9 C
7i
1000 x TEMPERATURE
Absolute temperature units are assumed for C
3i
and C
5i
. The temperature
limits are always interpreted in user input units.
PPDS
The PPDS equation is:
2 Thermodynamic Property Models 227
where R is the gas constant.
Linear extrapolation of C
p
*,ig
versus T is performed outside temperature
bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPIGDS/1 C
1i
TEMPERATURE
CPIGDS/2, , 8 C
2i
, ..., C
8i
0
CPIGDS/9 C
9i
0 TEMPERATURE
CPIGDS/10 C
10i
1000 TEMPERATURE
IKCAPE Heat Capacity Polynomial
The equation is:
Linear extrapolation of C
p
*,ig
versus T is performed outside temperature
bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPIGPO/1 C
1i
X MOLECAPACITY
CPIGPO/2,,10 C
2i
, ..., C
10i
0 X MOLECAPACITY
TEMPERATURE
CPIGPO/11 C
11i
0 X TEMPERATURE
CPIGPO/12 C
12i
1000 X TEMPERATURE
References
Data for the Ideal Gas Heat Capacity Polynomial: Reid, Prausnitz and Poling,
The Properties of Gases and Liquids, 4th ed., (New York: McGrawHill, 1987).
The Aspen Physical Property System combustion data bank, JANAF
Thermochemical Data, Compiled and calculated by the Thermal Research
Laboratory of Dow Chemical Company.
F. A. Aly and L. L. Lee, "SelfConsistent Equations for Calculating the Ideal
Gas Heat Capacity, Enthalpy, and Entropy, Fluid Phase Eq., Vol. 6, (1981), p.
169.
228 2 Thermodynamic Property Models
General Pure Component Solid Heat
Capacity
The Aspen Physical Property System has several submodels for calculating
solid heat capacity. It uses parameter THRSWT/5 to determine which
submodel is used. See Pure Component TemperatureDependent Properties
for details.
If THRSWT/5 is This equation is used And this parameter is
used
0 Aspen solid heat capacity
polynomial
CPSPO1
100 or 102 DIPPR CPSDIP
200211 Barin CPSXP1, CPSXP2, ,
CPSXP7
401 IKCAPE heat capacity
polynomial
CPSPO
503 NIST ThermoML
polynomial
CPSTMLPO
The enthalpy, entropy, and Gibbs energy of solids are also calculated from
these equations:
Aspen Solid Heat Capacity Polynomial
The Aspen equation is:
Linear extrapolation occurs for C
p,i
*,s
versus T outside of bounds.
Parameter
Name
Applicable
Components
Symbol MDS Default Units
CPSPO1/1 Solids, Salts C
1i
x
CPSPO1/2, ..., 6 Solids, Salts C
2i
, ..., C
6i
x 0
CPSPO1/7 Solids, Salts C
7i
x 0
CPSPO1/8 Solids, Salts C
8i
x 1000
The units are TEMPERATURE and HEATCAPACITY for all elements. If any of
C
4i
through C
6i
are nonzero, absolute temperature units are assumed for
elements C
1i
through C
6i
. Otherwise, user input temperature units are
assumed for all elements. The temperature limits are always interpreted in
user input units.
Barin
See Barin Equations for Gibbs Energy, Enthalpy, Entropy, and Heat Capacity
for details about this submodel.
DIPPR
There are two DIPPR equations that may generally be used.
The more common one, DIPPR equation 100, is:
2 Thermodynamic Property Models 229
Linear extrapolation occurs for C
p,i
*,s
versus T outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPSDIP/1 C
1i
x MOLEHEAT
CAPACITY,
TEMPERATURE
CPSDIP/2,...,5 C
2i
, ..., C
5i
0 x MOLEHEAT
CAPACITY,
TEMPERATURE
CPSDIP/6 C
6i
0 x TEMPERATURE
CPSDIP/7 C
7i
1000 x TEMPERATURE
DIPPR equation 102 is:
Linear extrapolation occurs for C
p,i
*,s
versus T outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPSDIP/1 C
1i
x MOLEHEAT
CAPACITY,
TEMPERATURE
CPSDIP/2 C
2i
x
CPSDIP/3,4 C
3i
, C
4i
0 x TEMPERATURE
CPSDIP/6 C
6i
0 x TEMPERATURE
CPSDIP/7 C
7i
1000 x TEMPERATURE
If C
3i
or C
4i
are nonzero or C
2i
is negative, absolute temperature units are
assumed for elements C
1i
through C
4i
. Otherwise, user input temperature
units are assumed for all elements. The temperature limits are always
interpreted in user input units.
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
NIST ThermoML Polynomial
The equation is:
Linear extrapolation occurs for C
p,i
*,s
versus T outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPSTMLPO/1 C
1i
x J/K^2/mol
CPSTMLPO/2,...,5C
2i
, ..., C
5i
0 x J/K^2/mol
CPSTMLPO/6 nTerms 5 x
CPSTMLPO/7 T
lower
0 x TEMPERATURE
CPSTMLPO/8 T
upper
1000 x TEMPERATURE
230 2 Thermodynamic Property Models
IKCAPE Heat Capacity Polynomial
The equation is:
Linear extrapolation occurs for C
p,i
*,s
versus T outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPSPO/1 C
1i
X MOLECAPACITY
CPSPO/2,,10 C
10i
0 X MOLECAPACITY
TEMPERATURE
CPSPO/11 C
11i
0 X TEMPERATURE
CPSPO/12 C
12i
1000 X TEMPERATURE
Solubility Correlations
The Aspen Physical Property System has three builtin solubility correlation
models. This section describes the solubility correlation models available.
Model Type
Henry's constant Gas solubility in liquid
Water solubility Water solubility in organic liquid
Hydrocarbon solubility Hydrocarbon solubility in waterrich
liquid
Henry's Constant
The Henry's constant model is used when Henry's Law is applied to calculate
Kvalues for dissolved gas components in a mixture. Henry's Law is available
in all activity coefficient property methods, such as the WILSON property
method. The model calculates Henry's constant for a dissolved gas component
(i) in one or more solvents (A or B):
Where:
w
A
=
ln H
iA
(T, p
A
*,l
) =
Linear extrapolation occurs for ln H
iA
versus T outside of bounds.
2 Thermodynamic Property Models 231
H
iA
(T, P) =
The parameter is obtained from the BrelviO'Connell model. p
A
*,l
is
obtained from the Antoine model. is obtained from the appropriate activity
coefficient model.
The Henry's constants a
iA
, b
iA
, c
iA
, d
iA
, and e
iA
are specific to a solutesolvent
pair. They can be obtained from regression of gas solubility data. The Aspen
Physical Property System has a large number of builtin Henry's constants for
many solutes in solvents. These parameters were obtained using data from
the Dortmund Databank. In addition, a small number of Henry's constants
from the literature are available in the BINARY databank. See Physical
Property Data, Chapter 1, for details.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
VC V
cA
0.001 3.5 MOLEVOLUME
HENRY/1 a
iA
x PRESSURE,
TEMPERATURE
HENRY/2 b
iA
0 x TEMPERATURE
HENRY/3 c
iA
0 x TEMPERATURE
HENRY/4 d
iA
0 x TEMPERATURE
HENRY/5 T
L
0 x TEMPERATURE
HENRY/6 T
H
2000 x TEMPERATURE
HENRY/7 e
iA
0 x TEMPERATURE
If any of b
iA
, c
iA
, and e
iA
are nonzero, absolute temperature units are
assumed for all coefficients. If b
iA
, c
iA
, and e
iA
are all zero, the others are
interpreted in input units. The temperature limits are always interpreted in
input units.
If a
iA
is missing, is set to zero and the weighting factor w
A
is
renormalized.
Reference for BINARY Databank
E. Wilhelm, R. Battino, and R.J. Wilcock, "LowPressure Solubility of Gases in
Liquid Water," Chemical Reviews, 1977, Vol. 77, No. 2, pp 219  262.
Water Solubility
This model calculates solubility of water in a hydrocarbonrich liquid phase.
The model is used automatically when you model a hydrocarbonwater
system with the freewater option. See FreeWater Immiscibility Simplification
in FreeWater and ThreePhase Calculations for details.
The expression for the liquid mole fraction of water in the ith hydrocarbon
species is:
232 2 Thermodynamic Property Models
No extrapolation is used with this equation.
The parameters for about 60 components are stored in the Aspen Physical
Property System pure component databank.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
WATSOL/1 C
1i
fcn(T
bi
, SG
i
, M
i
) 10.0 33.0
WATSOL/2 C
2i
fcn(T
bi
, SG
i
, M
i
) 10000.0 3000.0 TEMPERATURE
WATSOL/3 C
3i
0 0.05 0.05 TEMPERATURE
WATSOL/4 C
4i
0 0.0 500 TEMPERATURE
WATSOL/5 C
5i
1000 4.0 1000 TEMPERATURE
Absolute temperature units are assumed for elements 2 through 5.
Hydrocarbon Solubility
This model calculates solubility of hydrocarbon in a waterrich liquid phase.
The model is used automatically when you model a hydrocarbonwater
system with the dirtywater option. See FreeWater Immiscibility
Simplification in FreeWater and Rigorous ThreePhase Calculations for
details.
The expression for the liquid mole fraction of the ith hydrocarbon species in
water is:
No extrapolation is used with this equation.
The parameters for about 40 components are stored in the Aspen Physical
Property System pure component databank.
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
HCSOL/1 C
1i
fcn(carbon
number)
1000.0 1000.0
HCSOL/2 C
2i
0 100000.0 100000.0 TEMPERATURE
HCSOL/3 C
3i
0 100.0 100.0 TEMPERATURE
HCSOL/4 C
4i
0 0.0 500 TEMPERATURE
HCSOL/5 C
5i
1000 4.0 1000 TEMPERATURE
For Hydrocarbons and pseudocomponents, the default values are estimated
by the method given by API Procedure 9A2.17 at 25 C.
Absolute temperature units are assumed for elements 2 and 3. The
temperature limits are always interpreted in user input units.
2 Thermodynamic Property Models 233
Reference
C. Tsonopoulos, Fluid Phase Equilibria, 186 (2001), 185206.
Other Thermodynamic Property
Models
The Aspen Physical Property System has some builtin additional
thermodynamic property models that do not fit in any other category. This
section describes these models:
 Cavett Liquid Enthalpy Departure
 Barin Equations for Gibbs Energy, Enthalpy, Entropy and Heat Capacity
 Electrolyte NRTL Enthalpy
 Electrolyte NRTL Gibbs Energy
 Liquid Enthalpy from Liquid Heat Capacity Correlation
 Enthalpies Based on Different Reference States
 Helgeson Equations of State
 Quadratic Mixing Rule
Cavett
The general form for the Cavett model is:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
DHLCVT ZC X 0.1 0.5
Barin Equations for Gibbs Energy, Enthalpy,
Entropy, and Heat Capacity
The following equations are used when parameters from the Aspen Physical
Property System inorganic databank are retrieved.
 Gibbs energy:
(1)
 Enthalpy:
(2)
234 2 Thermodynamic Property Models
 Entropy:
(3)
 Heat capacity:
(4)
o refers to an arbitrary phase which can be solid, liquid, or ideal gas. For each
phase, multiple sets of parameters from 1 to n are present to cover multiple
temperature ranges. The value of the parameter n depends on the phase.
(See tables that follow.) When the temperature is outside all these
temperature ranges, linear extrapolation of properties versus T is performed
using the slope at the end of the nearest temperature bound.
The four properties C
p
, H, S, and G are interrelated as a result of the
thermodynamic relationships:
There are analytical relationships between the expressions describing the
properties C
p
, H, S, and G (equations 1 to 4). The parameters a
n,i
to h
n,i
can
occur in more than one equation.
The Aspen Physical Property System has other models which can be used to
calculate temperaturedependent properties which the BARIN equations can
calculate. The Aspen Physical Property System uses the parameters in
THRSWT to determine which model is used. See Pure Component
TemperatureDependent Properties for details.
If this parameter is 200 to 211Then the BARIN equations are
used to calculate
THRSWT/3 Liquid vapor pressure
THRSWT/5 Solid heat capacity
THRSWT/6 Liquid heat capacity
THRSWT/7 Ideal gas heat capacity
The liquid vapor pressure is computed from Gibbs energy as follows:
ln p = (G
L
 G
V
)/RT + ln p
ref
where p is the vapor pressure, and p
ref
is the reference pressure 101325 Pa.
Thus, parameters for both liquid and vapor are necessary to calculate vapor
pressure.
Solid Phase
The parameters in range n are valid for temperature: T
n,l
s
< T < T
n,h
s
2 Thermodynamic Property Models 235
When you specify this parameter, be sure to specify at least elements 1
through 3.
Parameter Name
/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPSXPn/1 T
n,l
s
x TEMPERATURE
CPSXPn/2 T
n,h
s
x TEMPERATURE
CPSXPn/3 a
n,i
s
x
CPSXPn/4 b
n,i
s
0 x
CPSXPn/5 c
n,i
s
0 x
CPSXPn/6 d
n,i
s
0 x
CPSXPn/7 e
n,i
s
0 x
CPSXPn/8 f
n,i
s
0 x
CPSXPn/9 g
n,i
s
0 x
CPSXPn/10 h
n,i
s
0 x
n is 1 through 7. CPSXP1 vector stores solid parameters for the first
temperature range. CPSXP2 vector stores solid parameters for the second
temperature range, and so on.
TEMPERATURE, ENTHALPY, ENTROPY
Liquid Phase
The parameters in range n are valid for temperature: T
n,l
l
< T < T
n,h
l
When you specify this parameter, be sure to specify at least elements 1
through 3.
Parameter Name
/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPLXPn/1 T
n,l
l
x TEMPERATURE
CPLXPn/2 T
n,h
l
x TEMPERATURE
CPLXPn/3 a
n,i
l
x
CPLXPn/4 b
n,i
l
0 x
CPLXPn/5 c
n,i
l
0 x
CPLXPn/6 d
n,i
l
0 x
CPLXPn/7 e
n,i
l
0 x
CPLXPn/8 f
n,i
l
0 x
CPLXPn/9 g
n,i
l
0 x
CPLXPn/10 h
n,i
l
0 x
n is 1 through 2. CPLXP1 stores liquid parameters for the first temperature
range. CPLXP2 stores liquid parameters for the second temperature range.
TEMPERATURE, ENTHALPY, ENTROPY
Ideal Gas Phase
The parameters in range n are valid for temperature: T
n,l
ig
< T < T
n,h
ig
When you specify this parameter, be sure to specify at least elements 1
through 3.
236 2 Thermodynamic Property Models
Parameter Name
/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
CPIXPn/1 T
n,l
ig
x TEMPERATURE
CPIXPn/2 T
n,h
ig
x TEMPERATURE
CPIXPn/3 a
n,i
ig
x
CPIXPn/4 b
n,i
ig
0 x
CPIXPn/5 c
n,i
ig
0 x
CPIXPn/6 d
n,i
ig
0 x
CPIXPn/7 e
n,i
ig
0 x
CPIXPn/8 f
n,i
ig
0 x
CPIXPn/9 g
n,i
ig
0 x
CPIXPn/10 h
n,i
ig
0 x
n is 1 through 3. CPIXP1 vector stores ideal gas parameters for the first
temperature range. CPIXP2 vector stores ideal gas parameters for the second
temperature range, and so on.
TEMPERATURE, ENTHALPY, ENTROPY
Electrolyte NRTL Enthalpy Model
(HMXENRTL)
The equation for the electrolyte NRTL enthalpy model (HMXENRTL) is:
The molar enthalpy H
m
*
and the molar excess enthalpy H
m
*E
are defined with
the asymmetrical reference state: the pure solvent water and infinite dilution
of molecular solutes and ions. (here * refers to the asymmetrical reference
state.)
H
w
* is the pure water molar enthalpy, calculated from the Ideal Gas model
and the ASME Steam Table equationofstate. (here * refers to pure
component.)
H
s
*,l
is the enthalpy contribution from a nonwater solvent. It is calculated as
usual for components in activity coefficient models:
H
s
*,l
(T) = H
s
*,ig
+ DHV
s
(T,p)  AH
s,vap
(T).
The term DHV
s
(T,p) = H
s
*,v
 H
s
*,ig
is the vapor enthalpy departure
contribution to liquid enthalpy; option code 5 determines how this is
calculated.
The aqueous infinite dilution thermodynamic enthalpy H
k
is calculated from
the infinite dilution aqueous phase heat capacity as follows:
2 Thermodynamic Property Models 237
where the subscript k refers to any ion or molecular solute. By default,
is calculated from the aqueous infinite dilution heat capacity polynomial. If
the polynomial model parameters are not available, is calculated from
the CrissCobble correlation for ionic solutes.
For molecular solutes (e.g. Henry components), if the aqueous infinite dilution
heat capacity polynomial model parameters are not available, H
k
is
calculated from Henry's law:
H
m
*E
is calculated from the electrolyte NRTL activity coefficient model.
See CrissCobble correlation model and Henry's law model for more
information.
Parameter
Name
Applicable
Components
Symbol Default Units
IONTYP Ions Ion 0
SO25C Cations
S
c
,aq
(T=298)
MOLEENTROPY
Anions
S
a
,aq
(T=298)
MOLEENTROPY
DHAQFM Ions, Molecular Solutes
A
f
H
k
,aq
MOLEENTHALPY
CPAQ0 Ions, Molecular Solutes
C
p,k
,aq
HEATCAPACITY
DHFORM Molecular Solutes
A
f
H
i
*,ig
MOLEENTHALPY
Water, Solvents
A
f
H
w
*,ig
MOLEENTHALPY
CPIG Molecular Solutes C
p,i
*,ig
Water, Solvents C
p,w
*,ig
IONTYP is not needed if CPAQ0 is given for ions.
DHFORM is not used if DHAQFM and CPAQ0 are given for molecular solutes
(components declared as Henry's components). If CPAQ0 is missing, DHFORM
and Henry's constants are used to calculate infinite dilution enthalpy for
solutes.
The unit keywords for CPIG are TEMPERATURE and HEATCAPACITY. If
CPIG/10 or CPIG/11 is nonzero, then absolute temperature units are
assumed for CPIG/9 through CPIG/11. Otherwise, user input temperature
units are used for all elements of CPIG. User input temperature units are
always used for other elements of CPIG.
238 2 Thermodynamic Property Models
Option Codes for Electrolyte NRTL Enthalpy Model
(HMXENRTL)
The electrolyte NRTL enthalpy model (HMXENRTL) has seven option codes
and the option codes can affect the performance of this model.
Option code 1. Use this option code to specify the default values of pair
parameters for water/solute and solvent/solute; the solute represents a
cation/anion pair. The value (1) sets the default values to zero and the value
(3) sets the default values for water/solute to (8,4) and for solvent/solute to
(10,2). The value (3) is the default choice of the option code.
Option code 2. Use this option code to specify the vapor phase equationof
state (EOS) model used for the liquid enthalpy calculation. The value (0) sets
the ideal gas EOS model and the value (1) sets the HF EOS model. The value
(0) is the default.
Option code 3. Always leave this option code set to the value (1) to use the
solvent/solvent binary parameters obtained from NRTL parameters.
Option code 4. Not used.
Option code 5. Use this option code to specify how the vapor phase enthalpy
departure (DHV) is calculated. The value (0) sets DHV = 0, the value (1)
specifies using RedlichKwong equation of state, and the value (3) specifies
using HaydenOConnell equation of state. The value (0) is the default.
Option code 6. Not used.
Option code 7. Use this option code to specify the method for handling
Henry components and multiple solvents. The value (0) sets the pure liquid
enthalpy to that calculated by aqueous infinite dilution heat capacity (only
water as solvent) and the value (1) sets the pure liquid enthalpy for Henry
components using Henrys law. Use value (1) when there are multiple
solvents. The value (0) is the default.
Electrolyte NRTL Gibbs Free Energy Model
(GMXENRTL)
The equation for the NRTL Gibbs free energy model (GMXENRTL) is:
The molar Gibbs free energy and the molar excess Gibbs free energy G
m
*
and
G
m
*E
are defined with the asymmetrical reference state: as pure water and
infinite dilution of molecular solutes and ions. (* refers to the asymmetrical
reference state.) The ideal mixing term is calculated normally, where j refers
to any component. The molar Gibbs free energy of pure water (or
thermodynamic potential)
w
* is calculated from the ideal gas contribution.
This is a function of the ideal gas heat capacity and the departure function.
(here * refers to the pure component.)
The departure function is obtained from the ASME steam tables.
2 Thermodynamic Property Models 239
s
*
,l
is the Gibbs free energy contribution from a nonwater solvent. It is
calculated as usual for components in activity coefficient models.
The aqueous infinite dilution thermodynamic potential
k
is calculated from
the infinite dilution aqueous phase heat capacity polynomial model, by
default. Subscript k refers to any ion or molecular solute. If polynomial model
parameters are not available, it is calculated from the CrissCobble model for
ionic solutes:
where the subscript k refers to any ion or molecular solute and the term RT
ln(1000/M
w
) is added because A
f
H
k
,aq
and A
f
G
k
,aq
are based on a molality
scale, while
k
is
calculated from Henry's law:
G
m
*E
is calculated from the electrolyte NRTL activity coefficient model.
See the CrissCobble correlation model and Henry's law model for more
information.
Parameter
Name
Applicable
Components
Symbol Default Units
IONTYP Ions Ion 0
SO25C Cations
S
c
,aq
(T=298)
MOLEENTROPY
Anions
S
a
,aq
(T=298)
MOLEENTROPY
DGAQFM Ions, Molecular Solutes
A
f
G
k
,aq
MOLEENTHALPY
CPAQ0 Ions, Molecular Solutes
C
p,k
,aq
HEATCAPACITY
DGFORM Molecular Solutes
A
f
G
i
MOLEENTHALPY
Water, Solvents
A
f
G
w
MOLEENTHALPY
240 2 Thermodynamic Property Models
Parameter
Name
Applicable
Components
Symbol Default Units
CPIG Molecular Solutes C
p,i
*,ig
Water, Solvents C
p,w
*,ig
IONTYP and SO25C are not needed if CPAQ0 is given for ions.
DGFORM is not needed if DHAQFM and CPAQ0 are given for molecular
solutes.
The unit keywords for CPIG are TEMPERATURE and HEATCAPACITY. If
CPIG/10 or CPIG/11 is nonzero, then absolute temperature units are
assumed for CPIG/9 through CPIG/11. Otherwise, user input temperature
units are used for all elements of CPIG. User input temperature units are
always used for other elements of CPIG.
Option Codes for Electrolyte NRTL Gibbs Free Energy
Model (GMXENRTL)
The electrolyte NRTL Gibbs free energy model (GMXENRTL) has six option
codes and the option codes can affect the performance of this model.
Option code 1. Use this option code to specify the default values of pair
parameters for water/solute and solvent/solute; the solute represents a
cation/anion pair. The value (1) sets the default values to zero and the value
(3) sets the default values for water/solute to (8,4) and for solvent/solute to
(10,2). The value (3) is the default choice of the option code.
Option code 2. Use this option code to specify the vapor phase equationof
state (EOS) model used for the liquid Gibbs free energy calculation. The value
(0) sets the ideal gas EOS model and the value (1) sets the HF EOS model.
The value (0) is the default.
Option code 3. Always leave this option code set to the value (1) to use the
solvent/solvent binary parameters obtained from NRTL parameters.
Option code 4. Not used.
Option code 5. Use this option code to specify how the pure vapor phase
fugacity coefficient (PHIV) is calculated. The value (0) sets PHIV = 1 (ideal
gas law), the value (1) specifies using RedlichKwong equation of state, and
the value (3) specifies using HaydenOConnell equation of state. The value
(0) is the default.
Option code 6. Use this option code to specify the method for handling
Henry components and multiple solvents. The value (0) sets the pure liquid
Gibbs free energy to that calculated by aqueous infinite dilution heat capacity
(only water as solvent) and the value (1) sets the pure liquid Gibbs free
energy for Henry components using Henrys law. Use value (1) when there
are multiple solvents. The value (0) is the default.
Liquid Enthalpy from Liquid Heat Capacity
Correlation
Liquid enthalpy is directly calculated by integration of liquid heat capacity:
2 Thermodynamic Property Models 241
The reference enthalpy is calculated at T
ref
as:
Where:
H
i
*,ig
= Ideal gas enthalpy
H
i
*,v
 H
i
*,ig
= Vapor enthalpy departure from equation of state
A
vap
H
i
*,l
= Heat of vaporization from General model
T
ref
= Reference temperature, specified by user. Defaults to
298.15 K
See General Pure Component Heat of Vaporization for parameter requirement
and additional details.
Enthalpies Based on Different Reference
States
Two property methods, WILSLR and WILSGLR, are available to calculate
enthalpies based on different reference states. The WILSLR property method
is based on saturated liquid reference state for all components. The WILSGLR
property method allows both ideal gas and saturated liquid reference states
for different components.
These property methods use an enthalpy method that optimizes the accuracy
tradeoff between liquid heat capacity, heat of vaporization, and vapor heat
capacity at actual process conditions. This highly recommended method
eliminates many of the problems associated with accurate thermal properties
for both phases, especially the liquid phase.
The liquid enthalpy of mixture is calculated by the following equation (see the
table labeled Liquid Enthalpy Methods):
Where:
H
m
ig
= Enthalpy of ideal gas mixture
=
H
i
*,ig
= Ideal gas enthalpy of pure component i
(H
m
l
H
m
ig
) = Enthalpy departure of mixture
For supercritical components, declared as Henry's components, the enthalpy
departure is calculated as follows:
242 2 Thermodynamic Property Models
For subcritical components:
H
m
l
H
m
ig
=
H
m
E,l
=
H
A
*,l
H
A
*,ig
= Enthalpy departure of pure component A
H
*,ig
and H
*,l
can be calculated based on either saturated liquid or ideal gas as
reference state as described in the sections that follow.
For the WILSLR property method, H
*,ig
and H
*,l
are calculated based on the
saturated liquid reference state for all components.
For the WILSGLR property method, H
*,ig
and H
*,l
can be calculated based on
the saturated liquid reference state for some components and the ideal gas
reference state for other components. You can set the value of a pure
component parameter called RSTATE to specify the reference state for each
component. RSTATE = 1 denotes ideal gas reference state. RSTATE = 2
denotes saturated liquid reference state. If it is not set, the following default
rules apply based on the normal boiling point of the component, i, TB(i):
 If TB(i) <= 298.15 K, ideal gas reference state is used
 If TB(i) > 298.15 K, saturated liquid reference state is used.
Note: When H
*,ig
(HIG) is calculated based on saturated liquid reference
state, ideal gas Gibbs free energy and entropy (GIG and SIG) are also
calculated based on saturated liquid reference state to maintain
thermodynamic consistency, and liquid Gibbs free energy and entropy are
also calculated based on saturated liquid reference state since they are
calculated based on departure from the ideal gas values.
Saturated Liquid as Reference State
The saturated liquid enthalpy at temperature T is calculated as follows:
Where:
H
i
ref,l
= Reference enthalpy for liquid state at T
i
ref,l
C
p,i
*,l
= Liquid heat capacity of component i
The saturated liquid Gibbs free energy is calculated as follows:
Where:
G
i
ref,l
= Reference Gibbs free energy for liquid state at T
i
ref,l

i
*,l
= Liquid fugacity coefficient of component i
p = System pressure
2 Thermodynamic Property Models 243
p
ref
= Reference pressure (=101325 N/m
2
)
For the WILSLR property method, H
i
ref,l
and G
i
ref,l
default to zero (0). The
reference temperature T
i
ref,l
defaults to 273.15K.
For the WILSGLR property method, the reference temperature T
i
ref,l
defaults
to 298.15K and H
i
ref,l
defaults to:
And G
i
ref,l
defaults to:
Where:
H
i
ref,ig
= Ideal gas enthalpy of formation for liquid state at
298.15K
G
i
ref,ig
= Ideal gas Gibbs free energy of formation for liquid
state at 298.15K
AH
i
*,v
= Vapor enthalpy departure of component i
AG
i
*,v
= Vapor Gibbs free energy departure of component i
A
vap
H
i
*
= Enthalpy of vaporization of component i
Note that we cannot default the liquid reference enthalpy and Gibbs free
energy to zero, as is the case for WILSLR, because it will cause inconsistency
with the enthalpy of components that use idealgas reference state. The
default values used result in the enthalpies of all components being on the
same basis. In fact, if you enter values for H
i
ref,l
and G
i
ref,l
for a liquid
reference state component you must make sure that they are consistent with
each other and are consistent with the enthalpy basis of the remaining
components in the mixture. If you enter a value for H
i
ref,l
, you should also
enter a value for G
i
ref,l
to ensure consistency.
When the liquidreference state is used, the ideal gas enthalpy at
temperature T is not calculated from the integration of the ideal gas heat
capacity equation (see Ideal Gas as Reference State section below). For
consistency, it is calculated from liquid enthalpy as follows:
Where:
T
i
con,l
= Temperature of conversion from liquid to vapor
enthalpy for component i
A
vap
H
i
*
(T
i
con,l
)
= Heat of vaporization of component i at
temperature of T
con,l
AH
i
*,v
(T
i
con,l
, p
i
*,l
)
= Vapor enthalpy departure of component i at the
conversion temperature and vapor pressure p
i
*,l
p
i
*,l
= Liquid vapor pressure of component i
244 2 Thermodynamic Property Models
= Ideal gas heat capacity of component i
T
i
con,l
is the temperature at which one crosses from liquid state to the vapor
state. This is a user defined temperature that defaults to the system
temperature T. T
i
con,l
may be selected such that heat of vaporization for
component i at the temperature is most accurate.
The liquid heat capacity and the ideal gas heat capacity can be calculated
from the General Pure Component Liquid Heat Capacity and General Pure
Component Ideal Gas Heat Capacity, or other available models. The heat of
vaporization can be calculated from the General Pure Component Heat of
Vaporization, or other available models.
The ideal gas heat capacity at temperature T can be obtained by
differentiating the ideal gas enthalpy equation above with respect to
temperature. If the conversion temperature T
i
con,l
is specified and not equal to
T, then only the last term depends on T and the calculated ideal gas heat
capacity is obtained from the general pure component ideal
gas heat capacity or other models.
However, if the conversion temperature is not specified, it defaults to T, and
the ideal gas enthalpy equation becomes:
And then the ideal gas heat capacity, calculated by differentiating this
equation, is:
The vapor enthalpy is calculated from ideal gas enthalpy as follows:
Where:
AH
i
*,v
(T, P)
= Vapor enthalpy departure of pure component i at the system
temperature and pressure
The enthalpy departure is obtained from an equationofstate that is being
used in the property method. For WILSLR and WILSGLR, the ideal gas
equation of state is used.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
RSTATE 2
TREFHL T
i
ref,l
TEMPERATURE
DHLFRM H
i
ref,l
MOLE ENTHALPY
DGLFRM G
i
ref,l
MOLE ENTHALPY
TCONHL T
i
con,l
T TEMPERATURE
Enthalpy reference state given by RSTATE. 2 denotes saturated liquid as
reference state.
2 Thermodynamic Property Models 245
For WILSLR property method TREFHL defaults to 273.15K. For WILSGLR
property method, TREFHL defaults to 298.15 K.
For WILSLR property method, DHLFRM defaults to zero (0). For WILS
GLR property method, DHLFRM defaults to the equation above.
For WILSLR property method, DGLFRM defaults to zero (0). For WILS
GLR property method, DGLFRM defaults to the equation above.
Liquid heat capacity equation is required for all components.
Ideal Gas as Reference State
The saturated liquid enthalpy is calculated as follows:
Where:
H
i
ref,ig
= Reference state enthalpy for ideal gas at T
i
ref,ig
= Heat of formation of ideal gas at 298.15 K by default
T
i
ref,ig
= Reference temperature corresponding to H
i
ref,ig
.
Defaults to 298.15 K
T
i
con,ig
= The temperature at which one crosses from vapor
state to liquid state. This is a user defined
temperature that defaults to the system temperature
T. T
i
con,ig
may be selected such that heat of
vaporization of component i at the temperature is
most accurate.
The ideal gas enthalpy is calculated as follows:
The vapor enthalpy is calculated as follows:
The liquid heat capacity and the ideal gas heat capacity can be calculated
from the General Pure Component Liquid Heat Capacity and General Pure
Component Ideal Gas Heat Capacity, or other available models. The heat of
vaporization can be calculated from the General Pure Component Heat of
Vaporization, or other available models. The enthalpy departure is obtained
from an equation of state that is being used in the property method. For
WILSLR and WILSGLR, the ideal gas equation of state is used.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
RSTATE 1
TREFHI T
i
ref,ig
TEMPERATURE
DHFORM H
i
ref,ig
MOLE ENTHALPY
TCONHI T
i
con,ig
T TEMPERATURE
246 2 Thermodynamic Property Models
Enthalpy reference state RSTATE for a component. Value of 1 denotes ideal
gas.
For components with TB s 298.15 K, RSTATE defaults to 1 (ideal gas).
TREFHI defaults to 298.15 K. For components with TB > 298.15 K, RSTATE
defaults to 2 (liquid) and TREFHI does not apply to these components. See
the Saturated Liquid as Reference State section for more details.
Helgeson Equations of State
The Helgeson equations of state for standard volume , heat capacity ,
entropy , enthalpy of formation , and Gibbs energy of formation
at infinite dilution in aqueous phase are:
2 Thermodynamic Property Models 247
Where:
248 2 Thermodynamic Property Models
Where:
+
= Pressure constant for a solvent (2600 bar for water)
u
= Temperature constant for a solvent (228 K for
water)
e
= Born coefficient
c
= Dielectric constant of a solvent
T
r
= Reference temperature (298.15 K)
P
r
= Reference pressure (1 bar)
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
AHGPAR/1, ... ,
4
a
1
, ..., a
4
0
CHGPAR/1, ... ,
2
c
1
, c
2
x
DHAQHG 0 0.5x10
10
0.5x10
10
MOLEENTHALPY
DGAQHG 0 0.5x10
10
0.5x10
10
MOLEENTHALPY
S25HG 0 0.5x10
10
0.5x10
10
MOLEENTROPY
OMEGHG 0 0.5x10
10
0.5x10
10
MOLEENTHALPY
Units for AHGPAR are complex, and involve temperature, pressure, and
moleenthalpy. If pressure is under 200 bar, AHGPAR may not be required.
References
Tanger, J.C. IV and H.C. Helgeson, "Calculation of the thermodynamic and
transport properties of aqueous species at high pressures and temperatures:
Revised equation of state for the standard partial properties of ions and
electrolytes," American Journal of Science, Vol. 288, (1988), p. 1998.
Shock, E.L. and H.C. Helgeson, "Calculation of the thermodynamic and
transport properties of aqueous species at high pressures and temperatures:
2 Thermodynamic Property Models 249
Correlation algorithms for ionic species and equation of state predictions to 5
kb and 1000 C," Geochimica et Cosmochimica Acta, Vol. 52, p. 20092036.
Shock, E.L., H.C. Helgeson, and D.A. Sverjensky, "Calculation of the
thermodynamic and transport properties of aqueous species at high pressures
and temperatures: Standard partial molal properties of inorganic neutral
species," Geochimica et Cosmochimica Acta, Vol. 53, p. 21572183.
Quadratic Mixing Rule
The quadratic mixing rule is a generalpurpose mixing rule which can be
applied to various properties. For a given property Q, with i and j being
components, the rule is:
The pure component properties Q
i
and Q
j
are calculated by the default model
for that property, unless modified by option codes. Composition x
i
and x
j
is in
mole fraction unless modified by option codes. K
ij
is a binary parameter
specific to the mixing rule for each property. A variation on the equation using
the logarithm of the property is used for viscosity.
250 2 Thermodynamic Property Models
3 Transport Property Models 251
3 Transport Property Models
This section describes the transport property models available in the Aspen
Physical Property System. The following table provides an overview of the
available models. This table lists the Aspen Physical Property System model
names, and their possible use in different phase types, for pure components
and mixtures.
Viscosity models
These models calculate MUL (pure) and/or MULMX (mixture) for the liquid
phase, and MUV (pure) and/or MUVMX (mixture) for the vapor phase, except
models marked with * calculate MUVLP (pure) and/or MUVMXLP (mixture) for
the vapor phase, and models marked with ** calculate MUVPC (pure) and
MUVMXPC (mixture).
Model Model name Phase(s) Pure Mixture
Andrade Liquid Mixture
Viscosity
MUL2ANDR L X
General Pure Component
Liquid Viscosity
MUL0ANDR L X
API Liquid Viscosity MUL2API L X
API 1997 Liquid Viscosity MULAPI97 L X
Aspen Liquid Mixture Viscosity MUASPEN L X
ASTM Liquid Mixture Viscosity MUL2ASTM L X
General Pure Component
Vapor Viscosity *
MUV0CEB V X
ChapmanEnskogBrokaw
Wilke Mixing Rule *
MUV2BROK,
MUV2WILK
V X
ChungLeeStarling Low
Pressure *
MUV0CLSL,
MUV2CLSL
V X X
ChungLeeStarling MUV0CLS2,
MUV2CLS2,
MUL0CLS2,
MUL2CLS2
V L X X
DeanStiel Pressure Correction
**
MUV0DSPC,
MUV2DSPC
V X X
IAPS Viscosity MUV0H2O,
MUL0H2O
V L X
JonesDole Electrolyte
Correction
MUL2JONS L X
252 3 Transport Property Models
Model Model name Phase(s) Pure Mixture
LetsouStiel MUL0LEST,
MUL2LEST
L X X
Lucas MUV0LUC,
MUV2LUC
V X X
TRAPP viscosity MUL0TRAP,
MUL2TRAP,
MUV0TRAP,
MUV2TRAP
V L X X
Twu liquid viscosity MUL2TWU L X
Viscosity quadratic mixing rule MUL2QUAD L X
Thermal conductivity models
These models calculate KL (pure) and/or KLMX (mixture) for the liquid phase,
and KV (pure) and/or KVMX (mixture) for the vapor phase, except models
marked with * calculate KVLP (pure) and/or KVMXLP (mixture) for the vapor
phase, and models marked with ** calculate KVPC (pure) and KVMXPC
(mixture) for the vapor phase.
Model Model name Phase(s) Pure Mixture
ChungLeeStarling Thermal
Conductivtity
KV0CLS2,
KV2CLS2,
KL0CLS2,
KL2CLS2
V L X X
IAPS Thermal Conductivity for
Water
KV0H2O
KL0H2O
V
L
X
X
m
= Absolute viscosity of the mixture (Nsec/m
2
)
i
= Absolute viscosity of component i (Nsec/m
2
)
log = Common logarithm (base 10)
f = An adjustable parameter, typically in the
range of 0.5 to 1.0
There are two significant differences in this implementation as compared with
the version from the book . The book uses kinematic viscosity in mm
2
/s
while the Aspen Physical Property System uses absolute viscosity in cP (=
0.001 Nsec/m
2
) with an equation in the same form. And f is treated as an
adjustable parameter in this model, while the book makes it a function of the
pure component kinematic viscosity.
The individual component viscosities are calculated by the General Pure
Component Liquid Viscosity model. The parameter f can be specified by
setting the value for MULOGF for the first component in the component list
(as defined on the Components  Specifications  Selection sheet).
3 Transport Property Models 261
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MULOGF f 1.0 0.0 2.0
"Petroleum Refining, 1 Crude Oil, Petroleum Products, Process Flowsheets",
Institut Francais du Petrole Publications, 1995, p. 130.
General Pure Component Vapor Viscosity
The Aspen Physical Property System has several submodels for calculating
pure component low pressure vapor viscosity. It uses parameter TRNSWT/2
to determine which submodel is used. See Pure Component Temperature
Dependent Properties for details.
If TRNSWT/2 is This equation is used And this parameter is
used
0 ChapmanEnskog
Brokaw
STKPAR, LJPAR
102 DIPPR MUVDIP
301 PPDS MUVPDS
302 PPDS kinetic theory MUVCEB
401 IKCAPE polynomial
equation
MUVPO
402 IKCAPE Sutherland
equation
MUVSUT
503 NIST ThermoML
polynomial
MUVTMLPO
ChapmanEnskogBrokaw
The equation for the ChapmanEnskog model is:
Where:
O
q
=
A parameter o is used to determine whether to use the Stockmayer or
LennardJones potential parameters for c/k (energy parameter) and o
(collision diameter). To calculate o, the dipole moment p and either the
Stockmayer parameters or T
b
and V
b
are needed. The polarity correction is
from Brokaw.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
STKPAR/1
(c
i
/k)
ST
fcn(T
bi
, V
bi
, p
i
) X TEMPERATURE
STKPAR/2
o
i
ST
fcn(T
bi
, V
bi
, p
i
) X LENGTH
262 3 Transport Property Models
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
LJPAR/1
(c
i
/k)
LJ
fcn(T
ci
, e
i
)
X TEMPERATURE
LJPAR/2
o
i
LJ
fcn(T
ci
, p
ci
, e
i
)
X LENGTH
DIPPR Vapor Viscosity
The equation for the DIPPR vapor viscosity model is:
When necessary, the vapor viscosity is extrapolated beyond this temperature
range linearly with respect to T, using the slope at the temperature limits.
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
PCES uses the DIPPR equation in estimating vapor viscosity.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MUVDIP/1 C
1i
X VISCOSITY
MUVDIP/2 C
2i
0 X
MUVDIP/3, 4 C
3i
, C
4i
0 X TEMPERATURE
MUVDIP/5 0 X
MUVDIP/6 C
6i
0 X TEMPERATURE
MUVDIP/7 C
7i
1000 X TEMPERATURE
If any of C
2i
through C
4i
are nonzero, absolute temperature units are
assumed for C
1i
through C
4i
. Otherwise, all coefficients are interpreted in user
input temperature units. The temperature limits are always interpreted in
user input units.
MUVDIP/5 is not used in this equation. It is normally set to zero. The
parameter is provided for consistency with other DIPPR equations.
PPDS
The PPDS submodel includes both the basic PPDS vapor viscosity equation
and the PPDS kinetic theory vapor viscosity equation. For either equation,
linear extrapolation of viscosity versus T occurs for temperatures outside
bounds.
PPDS Vapor Viscosity
The equation is:
3 Transport Property Models 263
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MUVPDS/1 C
1i
VISCOSITY
MUVPDS/2 C
2i
0
MUVPDS/3 C
3i
0
MUVPDS/4 C
4i
0 TEMPERATURE
MUVPDS/5 C
5i
1000 TEMPERATURE
PPDS Kinetic Theory Vapor Viscosity
The equation is:
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
MUVCEB/1 C
1i
LENGTH
MUVCEB/2 C
2i
TEMPERATURE
MUVCEB/3 C
3i
0
MUVCEB/4 C
4i
0 TEMPERATURE
MUVCEB/5 C
5i
1000 TEMPERATURE
Absolute temperature units are assumed for C
2i
. The temperature limits are
always interpreted in user input units.
IKCAPE Vapor Viscosity
The IKCAPE vapor viscosity model includes both the Sutherland equation and
the polynomial equation. For either equation, linear extrapolation of viscosity
versus T occurs for temperatures outside bounds.
Sutherland Equation
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MUVSUT/1 C
1i
X VISCOSITY
MUVSUT/2 C
2i
0 X TEMPERATURE
MUVSUT/3 C
3i
0 X TEMPERATURE
MUVSUT/4 C
4i
1000 X TEMPERATURE
If C
2i
is nonzero, absolute temperature units are assumed for C
1i
and C
2i
.
Otherwise, all coefficients are interpreted in user input temperature units. The
temperature limits are always interpreted in user input units.
264 3 Transport Property Models
Polynomial
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MUVPO/1 C
1i
X VISCOSITY
MUVPO/2, ..., 10 C
2i
, ..., C
10i
0 X VISCOSITY,
TEMPERATURE
MUVPO/11 C
11i
0 X TEMPERATURE
MUVPO/12 C
12i
1000 X TEMPERATURE
NIST ThermoML Polynomial
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MUVTMLPO/1 C
1i
X Pa*s/T
MUVTMLPO/2, ..., 4 C
2i
, ..., C
4i
0 X Pa*s/T
MUVTMLPO/5 nTerms 4 X
MUVTMLPO/6 T
lower
0 X TEMPERATURE
MUVTMLPO/7 T
upper
1000 X TEMPERATURE
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling. The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 392.
ChapmanEnskogBrokawWilke Mixing
Rule
The low pressure vapor mixture viscosity is calculated by the Wilke
approximation of the ChapmanEnskog equation:
For u
ij
,the formulation by Brokaw is used:
Where:
3 Transport Property Models 265
The Stockmayer or LennardJones potential parameters c/k (energy
parameter) and o (collision diameter) and the dipole moment p are used to
calculate o. The k represents Boltzmann's constant 1.38065 x 10
23
J/K. If the
Stockmayer parameters are not available, o is estimated from T
b
and V
b
:
Where p is in debye.
The pure component vapor viscosity q
i
*,v
(p = 0) can be calculated using the
General Pure Component Vapor Viscosity (or another low pressure vapor
viscosity model).
Ensure that you supply parameters for q
i
*,v
(p = 0).
Parameter
Name/
Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
STKPAR/1
(c
i
/k)
ST
fcn(T
bi
, V
bi
, p
i
) X TEMPERATURE
STKPAR/2
o
i
ST
fcn(T
bi
, V
bi
, p
i
) X LENGTH
References
R.C. Reid, J.M. Prausnitz, and T.K. Sherwood, The Properties of Gases and
Liquids, 3rd ed., (New York: McGrawHill, 1977), pp. 410416.
266 3 Transport Property Models
ChungLeeStarling LowPressure Vapor
Viscosity
The lowpressure vapor viscosity by Chung, Lee, and Starling is:
Where the viscosity collision integral is:
The shape and polarity correction is:
The parameter p
r
is the reduced dipolemoment:
C
1
is a constant of correlation.
The polar parameter k is tabulated for certain alcohols and carboxylic acids.
The previous equations can be used for mixtures when applying these mixing
rules:
Where:
3 Transport Property Models 267
V
cij
=
ij
= 0 (in almost all cases)
T
cij
=
,
ij
= 0 (in almost all cases)
e
ij
=
M
ij
=
k
ij
=
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCCLS T
ci
TC x 5.0 2000.0 TEMPERATURE
VCCLS V
ci
VC x 0.001 3.5 MOLEVOLUME
MW M
i
1.0 5000.0
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
OMGCLS
e
i
OMEGA x 0.5 2.0
CLSK
k
i
0.0 x 0.0 0.5
CLSKV
ij
0.0 x 0.5 0.5
CLSKT
,
ij
0.0 x 0.5 0.5
The model specific parameters also affect the ChungLeeStarling Viscosity
and the ChungLeeStarling Thermal Conductivity models.
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 396, p. 413.
ChungLeeStarling Viscosity
The ChungLeeStarling viscosity equation for vapor and liquid, high and low
pressure is:
With:
268 3 Transport Property Models
f
1
=
f
2
=
F
C
=
The molar density can be calculated using an equationofstate model (for
example, the BenedictWebbRubin). The parameter p
r
is the reduced
dipolemoment:
C
1
and C
2
are constants of correlation.
The polar parameter k is tabulated for certain alcohols and carboxylic acids.
For low pressures, f
1
is reduced to 1.0 and f
2
becomes negligible. The
equation reduces to the low pressure vapor viscosity model by ChungLee and
Starling.
The previous equations can be used for mixtures when applying these mixing
rules:
Where:
V
cij
=
ij
= 0 (in almost all cases)
3 Transport Property Models 269
T
cij
=
,
ij
= 0 (in almost all cases)
e
ij
=
M
ij
=
k
ij
=
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCCLS T
ci
TC x 5.0 2000.0 TEMPERATURE
VCCLS V
ci
VC x 0.001 3.5 MOLEVOLUME
MW M
i
1.0 5000.0
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
OMGCLS
e
i
OMEGA x 0.5 2.0
CLSK
k
i
0.0 x 0.0 0.5
CLSKV
ij
0.0 x 0.5 0.5
CLSKT
,
ij
0.0 x 0.5 0.5
The model specific parameters affect the results of the ChungLeeStarling
Thermal Conductivity and Low Pressure Viscosity models as well.
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 427.
DeanStiel Pressure Correction
The residual vapor viscosity or the pressure correction to low pressure vapor
viscosity by Dean and Stiel is:
Where q
v
(p = 0) is obtained from a low pressure viscosity model (for
example, General Pure Component Vapor Viscosity). The dimensionless
making factor is:
=
270 3 Transport Property Models
T
c
=
M =
p
c
=
V
cm
=
Z
cm
=
rm
= V
cm
/ V
m
v
The parameter V
m
v
is obtained from RedlichKwong equationofstate.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
VC V
ci
0.001 3.5 MOLEVOLUME
IAPS Viscosity for Water
The IAPS viscosity models, developed by the International Association for
Properties of Steam, calculate vapor and liquid viscosity for water and steam.
These models are used in option sets STEAMNBS and STEAMTA.
The general form of the equation for the IAPS viscosity models is:
q
w
=fcn(T, p)
Where:
fcn = Correlation developed by IAPS
The models are only applicable to water. There are no parameters required
for the models.
JonesDole Electrolyte Correction
The JonesDole model calculates the correction to the liquid mixture viscosity
of a solvent mixture, due to the presence of electrolytes:
Where:
q
solv
= Viscosity of the liquid solvent mixture, by default
calculated by the Andrade model
Aq
ca
l
= Contribution to the viscosity correction due to
apparent electrolyte ca
3 Transport Property Models 271
The parameter q
solv
can be calculated by different models depending on
option code 3 for MUL2JONS:
Option Code
Value
Solvent liquid mixture viscosity
model
0
Andrade liquid mixture viscosity
model (default)
1 Viscosity quadratic mixing rule
2 Aspen liquid mixture viscosity
model
The parameter Aq
ca
l
can be calculated by three different equations.
If these parameters are available Use this equation
IONMOB and IONMUB JonesDole
IONMUB BreslauMiller
Carbonell
When the concentration of apparent electrolyte exceeds 0.1 M, the Breslau
Miller equation is used instead.
JonesDole
The JonesDole equation is:
(1)
Where:
= Concentration of apparent electrolyte ca (2)
x
ca
a
= Mole fraction of apparent electrolyte ca (3)
A
ca
= (4)
L
a
= (5)
L
c
= (6)
B
ca
= (7)
When the electrolyte concentration exceeds 0.1 M, the BreslauMiller equation
is used instead. This behavior can be disabled by setting the second option
code for MUL2JONS to 1.
BreslauMiller
The BreslauMiller equation is:
(8)
272 3 Transport Property Models
Where the effective volume V
c
is given by:
for salts involving univalent ions
(9)
for other salts
(9a)
Carbonell
The Carbonell equation is:
(10)
Where:
M
k
= Molecular weight of an apparent electrolyte
component k
You must provide parameters for the model used for the calculation of the
liquid mixture viscosity of the solvent mixture.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
CHARGE z 0.0
MW M 1.0 5000.0
IONMOB/1 l
1
AREA, MOLES
IONMOB/2 l
2
0.0 AREA, MOLES,
TEMPERATURE
IONMUB/1 b
1
MOLEVOLUME
IONMUB/2 b
2
0,0 MOLEVOLUME,
TEMPERATURE
When IONMOB/1 is missing, the JonesDole model uses a nominal value of
5.0 and issues a warning. This parameter should be specified for ions, and not
allowed to default to this value.
References
A. L. Horvath, Handbook of Aqueous Electrolyte Solutions, (Chichester: Ellis
Horwood, 1985).
LetsouStiel
The LetsouStiel model calculates liquid viscosity at high temperatures for
0.76 s T
r
s 0.98. This model is used in PCES.
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
The general form for the model is:
3 Transport Property Models 273
q
l
c = (q
l
c)
0
+ e(q
l
c)
1
Where:
(q
l
c)
0
=
(q
l
c)
1
=
c
=
e
=
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
10
5
10
8
PRESSURE
OMEGA
e
i
0.5 2.0
References
R.C. Reid, J.M. Pransnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 471.
Lucas Vapor Viscosity
The equation for the Lucas vapor viscosity model is:
Where the dimensionless low pressure viscosity is given by:
The dimensionlessmaking group is:
The pressure correction factor Y is:
The polar and quantum correction factors at high and low pressure are:
F
P
=
F
Q
=
F
Pi
(p = 0) =
274 3 Transport Property Models
F
Qi
(p = 0) = fcn(T
ri
), but is only nonunity for the quantum
gates i = H
2
, D
2
, and He.
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
The Lucas mixing rules are:
T
c
=
p
c
=
M =
F
P
(p = 0) =
F
Q
(p = 0) =
Where A differs from unity only for certain mixtures.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCLUC T
ci
TC x 5.0 2000.0 TEMPERATURE
PCLUC p
ci
PC x 10
5
10
8
PRESSURE
ZCLUC Z
ci
ZC x 0.1 0.5
MW M
i
1.0 5000.0
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 421, 431.
TRAPP Viscosity Model
The general form for the TRAPP viscosity model is:
Where:
The parameter is the mole fraction vector; fcn is a corresponding states
correlation based on the model for vapor and liquid viscosity TRAPP, by the
National Bureau of Standards (NBS, currently NIST) . The model can be used
for both pure components and mixtures. The model should be used for
nonpolar components only.
3 Transport Property Models 275
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
TCTRAP T
ci
TC x 5.0 2000.0 TEMPERATURE
PCTRAP p
ci
PC x 10
5
10
8
PRESSURE
VCTRAP V
ci
VC x 0.001 3.5 MOLEVOLUME
ZCTRAP Z
ci
ZC x 0.1 1.0
OMGRAP
e
i
OMEGA x 0.5 3.0
References
J.F. Ely and H.J.M. Hanley, "Prediction of Transport Properties. 1. Viscosities
of Fluids and Mixtures," Ind. Eng. Chem. Fundam., Vol. 20, (1981), pp. 323
332.
Twu Liquid Viscosity
The Twu liquid viscosity model is based upon the work of C.H. Twu (1985).
The correlation uses nalkanes as a reference fluid and is capable of
predicting liquid viscosity for petroleum fractions with normal boiling points up
to 1340 F and API gravity up to 30.
Given the normal boiling point T
b
and the specific gravity SG of the system to
be modeled, the Twu model estimates the viscosity of the nalkane reference
fluid of the same normal boiling point at 100 F and 210 F, and its specific
gravity. These are used to estimates the viscosity of the system to be
modeled at 100 F and at 210 F, and these viscosities are used to estimate the
viscosity at the temperature of interest.
276 3 Transport Property Models
Where:
SG = Specific gravity of petroleum fraction
T
b
= Normal boiling point of petroleum fraction, Rankine
SG = Specific gravity of reference fluid with normal
boiling point T
b
T = Temperature of petroleum fraction, Rankine
q
= Kinematic viscosity of petroleum fraction at T, cSt
q
i
= Kinematic viscosity of petroleum fraction at 100 F
(i=1) and 210 F (i=2), cSt
q
1
, q
2
ij
= 0 (in almost all cases)
T
cij
=
,
ij
= 0 (in almost all cases)
e
ij
=
M
ij
=
k
ij
=
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TCCLS T
ci
TC x 5.0 2000.0 TEMPERATURE
VCCLS V
ci
VC x 0.001 3.5 MOLEVOLUME
MW M
i
1.0 5000.0
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
OMGCLS
e
i
OMEGA x 0.5 2.0
CLSK
k
i
0.0 x 0.0 0.5
CLSKV
ij
0.0 x 0.5 0.5
CLSKT
,
ij
0.0 x 0.5 0.5
The modelspecific parameters also affect the results of the ChungLee
Starling Viscosity models.
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 505, 523.
IAPS Thermal Conductivity for Water
The IAPS thermal conductivity models were developed by the International
Association for Properties of Steam. These models can calculate vapor and
liquid thermal conductivity for water and steam. They are used in option sets
STEAMNBS and STEAMTA.
The general form of the equation for the IAPS thermal conductivity models is:
w
=fcn(T, p)
Where:
280 3 Transport Property Models
fcn = Correlation developed by IAPS
The models are only applicable to water. No parameters are required.
Li Mixing Rule
Liquid mixture thermal conductivity is calculated using Li equation (Reid
et.al., 1987):
Where:
The pure component liquid molar volume V
i
*,l
is calculated from the Rackett
model.
The pure component liquid thermal conductivity
i
*,l
is calculated by the
General Pure Component Liquid Thermal Conductivity model.
Reference: R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases
and Liquids, 4th ed., (New York: McGrawHill, 1987), p. 550.
Riedel Electrolyte Correction
The Riedel model can calculate the correction to the liquid mixture thermal
conductivity of a solvent mixture, due to the presence of electrolytes:
Where:
l
solv
= Thermal conductivity of the liquid solvent mixture,
calculated by the General Pure Component Liquid
Thermal Conductivity model using the Vredeveld
mixing rule
x
ca
a
= Mole fraction of the apparent electrolyte ca
a
c
, a
a
= Riedel ionic coefficient
V
m
l
= Apparent molar volume computed by the Clarke
density model
Apparent electrolyte mole fractions are computed from the true ion mole
fractions and ionic charge number. They can also be computed if you use the
apparent component approach. A more detailed discussion of this method is
found in Electrolyte Calculation.
3 Transport Property Models 281
You must provide parameters for the SatoRiedel model. This model is used
for the calculation of the thermal conductivity of solvent mixtures.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
CHARGE z 0.0
IONRDL a 0.0
THERMAL CONDUCTIVITY, MOLEVOLUME
The behavior of this model can be changed using option codes (these codes
apply to the Vredeveld mixing rule).
Option
Code
Option
Value
Description
1 0 Do not check ratio of KL max / KL min
1 Check ratio. If KL max / KL min > 2, set exponent to 1,
overriding option code 2.
2 0 Exponent is 2
1 Exponent is 0.4
2 Exponent is 1. This uses a weighted average of liquid thermal
conductivities.
General Pure Component Liquid Thermal
Conductivity
The Aspen Physical Property System has several submodels for calculating
pure component liquid thermal conductivity. It uses parameter TRNSWT/3 to
determine which submodel is used. See Pure Component Temperature
Dependent Properties for details.
If TRNSWT/3 is This equation is used And this parameter is used
0 SatoRiedel
100 DIPPR KLDIP
301 PPDS KLPDS
401 IKCAPE KLPO
503 NIST ThermoML
polynomial
KLTMLPO
510 NIST PPDS8 equation KLPPDS8
SatoRiedel
The SatoRiedel equation is (Reid et al., 1987):
Where:
T
bri
= T
bi
/ T
ci
T
ri
= T / T
ci
282 3 Transport Property Models
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
TC T
ci
5.0 2000.0 TEMPERATURE
TB T
bi
4.0 2000.0 TEMPERATURE
PPDS
The equation is:
Linear extrapolation of
*,l
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KLPDS/1 C
1i
THERMAL
CONDUCTIVITY
KLPDS/2 C
2i
0
KLPDS/3 C
3i
0
KLPDS/4 C
4i
0
KLPDS/5 C
5i
0 TEMPERATURE
KLPDS/6 C
6i
1000 TEMPERATURE
NIST PPDS8 Equation
The equation is
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KLPPDS8/1 C
1i
THERMAL
CONDUCTIVITY
KLPPDS8/2, ..., 7 C
2i
, ..., C
7i
0
KLPPDS8/8 T
Ci
TEMPERATURE
KLPPDS8/9 nTerms 7
KLPPDS8/10 T
lower
0 TEMPERATURE
KLPPDS8/11 T
upper
1000 TEMPERATURE
DIPPR Liquid Thermal Conductivity
The DIPPR equation is:
3 Transport Property Models 283
Linear extrapolation of
*,l
versus T occurs outside of bounds.
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
The DIPPR equation is used by PCES when estimating liquid thermal
conductivity.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KLDIP/1 C
1i
x THERMAL
CONDUCTIVITY,
TEMPERATURE
KLDIP/2, ... , 5 C
2i
, ..., C
5i
0 x THERMAL
CONDUCTIVITY,
TEMPERATURE
KLDIP/6 C
6i
0 x TEMPERATURE
KLDIP/7 C
7i
1000 x TEMPERATURE
NIST ThermoML Polynomial
The equation is:
Linear extrapolation of
*,l
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KLTMLPO/1 C
1i
x THERMAL
CONDUCTIVITY,
TEMPERATURE
KLTMLPO/2, ... ,
4
C
2i
, ..., C
4i
0 x THERMAL
CONDUCTIVITY,
TEMPERATURE
KLTMLPO/5 nTerms 4 x
KLTMLPO/6 T
lower
0 x TEMPERATURE
KLTMLPO/7 T
upper
1000 x TEMPERATURE
IKCAPE
The IKCAPE equation is:
Linear extrapolation of
*,l
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KLPO/1 C
1i
x THERMAL
CONDUCTIVITY
284 3 Transport Property Models
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KLPO/2, ... , 10 C
2i
, ..., C
10i
0 x THERMAL
CONDUCTIVITY,
TEMPERATURE
KLPO/11 C
11i
0 x TEMPERATURE
KLPO/12 C
12i
1000 x TEMPERATURE
References
R.C. Reid, J.M. Prausnitz, and T.K. Sherwood, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1977), p. 533.
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 550.
Solid Thermal Conductivity Polynomial
Thermal conductivity for solid pure components is calculated using the solid
thermal conductivity polynomial. For mixtures, the molefraction weighted
average is used.
For pure solids, thermal conductivity is calculated by:
For mixtures:
Linear extrapolation of
i
*,s
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KSPOLY/1 a THERMAL
CONDUCTIVITY
KSPOLY/2, 3, 4, 5 b, c, d, e 0 THERMAL
CONDUCTIVITY,
TEMPERATURE
KSPOLY/6 0 x TEMPERATURE
KSPOLY/7 1000 x TEMPERATURE
General Pure Component Vapor Thermal
Conductivity
The Aspen Physical Property System has several submodels for calculating
pure component low pressure vapor thermal conductivity. It uses parameter
TRNSWT/4 to determine which submodel is used. See Pure Component
TemperatureDependent Properties for details.
3 Transport Property Models 285
If TRNSWT/4 is This equation is used And this parameter is
used
0 StielThodos
102 DIPPR KVDIP
301 PPDS KVPDS
401 IKCAPE KVPO
503 NIST ThermoML
polynomial
KVTMLPO
StielThodos
The StielThodos equation is:
Where:
q
i
*,v
(p = 0) can be obtained from the General Pure Component Vapor
Viscosity model.
C
pi
*,ig
is obtained from the General Pure Component Ideal Gas Heat Capacity
model.
R is the universal gas constant.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
DIPPR Vapor Thermal Conductivity
The DIPPR equation for vapor thermal conductivity is:
Linear extrapolation of
i
*,v
versus T occurs outside of bounds.
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
The DIPPR equation is used in PCES when estimating vapor thermal
conductivity.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KVDIP/1 C
1i
x THERMAL
CONDUCTIVITY
KVDIP/2 C
2i
0 x
KVDIP/3, 4 C
3i
, C
4i
0 x TEMPERATURE
KVDIP/5 0 x
KVDIP/6 C
6i
0 x TEMPERATURE
KVDIP/7 C
7i
1000 x TEMPERATURE
If any of C
2i
through C
4i
are nonzero, absolute temperature units are
assumed for C
1i
through C
4i
. Otherwise, all coefficients are interpreted in user
286 3 Transport Property Models
input temperature units. The temperature limits are always interpreted in
user input units.
PPDS
The equation is:
Linear extrapolation of
i
*,v
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KVPDS/1 C
1i
THERMAL
CONDUCTIVITY
KVPDS/2 C
2i
0 THERMAL
CONDUCTIVITY
KVPDS/3 C
3i
0 THERMAL
CONDUCTIVITY
KVPDS/4 C
4i
0 THERMAL
CONDUCTIVITY
KVPDS/5 C
5i
0 TEMPERATURE
KVPDS/6 C
6i
1000 TEMPERATURE
IKCAPE Polynomial
Linear extrapolation of
i
*,v
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KVPO/1 C
1i
x THERMAL
CONDUCTIVITY
KVPO/2, ... , 10 C
2i
, ..., C
10i
0 x THERMAL
CONDUCTIVITY,
TEMPERATURE
KVPO/11 C
11i
0 x TEMPERATURE
KVPO/12 C
12i
1000 x TEMPERATURE
NIST ThermoML Polynomial
Linear extrapolation of
i
*,v
versus T occurs outside of bounds.
3 Transport Property Models 287
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
KVTMLPO/1 C
1i
x THERMAL
CONDUCTIVITY,
TEMPERATURE
KVTMLPO/2, ... ,
4
C
2i
, ..., C
4i
0 x THERMAL
CONDUCTIVITY,
TEMPERATURE
KVTMLPO/5 nTerms 4 x
KVTMLPO/6 T
lower
0 x TEMPERATURE
KVTMLPO/7 T
upper
1000 x TEMPERATURE
References
R.C. Reid, J.M. Praunitz, and B.E. Poling, The Properties of Gases and Liquid,
4th ed., (New York: McGrawHill, 1987), p. 494.
StielThodos Pressure Correction Model
The pressure correction to a pure component or mixture thermal conductivity
at low pressure is given by:
Where:
rm
=
The parameter V
m
v
can be obtained from RedlichKwong.
v
(p = 0) can be obtained from the low pressure General Pure Component
Vapor Thermal Conductivity.
This model should not be used for polar substances, hydrogen, or helium.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
TC T
ci
5.0 2000.0 TEMPERATURE
PC 10
5
10
8
PRESSURE
VC V
ci
0.001 3.5 MOLEVOLUME
ZC Z
ci
0.1 0.5
References
R.C. Reid, J.M. Praunitz, and B.E. Poling, The Properties of Gases and Liquids,
4th ed., (New York: McGrawHill, 1987), p. 521.
Vredeveld Mixing Rule
Liquid mixture thermal conductivity is calculated using the Vredeveld equation
(Reid et al., 1977):
288 3 Transport Property Models
Where:
w
i
= Liquid phase weight fraction of component i
i
*,l
= Pure component liquid thermal conductivity of component i
Where n is determined from two option codes on model KL2VR:
If option code 1 is Then n is determined by
0 (default) Option code 2, always
1
Option code 2, unless . In this case n is
set to 1.
If option code 2 is Then n is
0 (default) 2
1 0.4
2 1 (This uses a weighted average of liquid thermal
conductivities.)
For most systems, the ratio of maximum to minimum pure component liquid
thermal conductivity is between 1 and 2, where the exponent 2 is
recommended, and is the default value used.
Pure component liquid thermal conductivity
i
*,l
is calculated by the General
Pure Component Liquid Thermal Conductivity model.
Reference: R.C. Reid, J.M. Prausnitz, and T.K. Sherwood, The Properties of
Gases and Liquids, 4th ed., (New York: McGrawHill, 1977), p. 533.
TRAPP Thermal Conductivity Model
The general form for the TRAPP thermal conductivity model is:
Where:
= Mole fraction vector
C
pi
*,ig
= Ideal gas heat capacity calculated using the
General pure component ideal gas heat capacity
model
fcn = Corresponding states correlation based on the
model for vapor and liquid thermal conductivity
made by the National Bureau of standards (NBS,
currently NIST)
The model can be used for both pure components and mixtures. The model
should be used for nonpolar components only.
3 Transport Property Models 289
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
TCTRAP T
ci
TC x 5.0 2000.0 TEMPERATURE
PCTRAP p
ci
PC x 10
5
10
8
PRESSURE
VCTRAP V
ci
VC x 0.001 3.5 MOLEVOLUME
ZCTRAP Z
ci
ZC x 0.1 1.0
OMGRAP
e
i
OMEGA x 0.5 3.0
References
J.F. Ely and H.J. M. Hanley, "Prediction of Transport Properties. 2. Thermal
Conductivity of Pure Fluids and Mixtures," Ind. Eng. Chem. Fundam., Vol. 22,
(1983), pp. 9097.
WassiljewaMasonSaxena Mixing Rule
The vapor mixture thermal conductivity at low pressures is calculated from
the pure component values, using the WassiljewaMasonSaxena equation:
Where:
i
*,v
= Calculated by the General Pure Component Vapor
Thermal Conductivity model
q
i
*,v
(p = 0)
= Obtained from the General Pure Component Vapor
Viscosity model
You must supply parameters for q
i
*,v
(p = 0) and
i
*,v
.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), pp. 530531.
Diffusivity Models
The Aspen Physical Property System has seven builtin diffusivity models. This
section describes the diffusivity models available.
290 3 Transport Property Models
Model Type
ChapmanEnskogWilkeLee (Binary) Low pressure vapor
ChapmanEnskogWilkeLee (Mixture) Low pressure vapor
DawsonKhouryKobayashi (Binary) Vapor
DawsonKhouryKobayashi (Mixture) Vapor
NernstHartley Electrolyte
WilkeChang (Binary) Liquid
WilkeChang (Mixture) Liquid
ChapmanEnskogWilkeLee (Binary)
The binary diffusion coefficient at low pressures is calculated using
the ChapmanEnskogWilkeLee model:
D
ij
v
=D
ji
v
Where:
The collision integral for diffusion is:
O
D
=
The binary size and energy parameters are defined as:
o
ij
=
c
ij
=
A parameter o is used to determine whether to use the Stockmayer or
LennardJones potential parameters for c/k (energy parameter ) and o
(collision diameter). To calculate o, the dipole moment p, and either the
Stockmayer parameters or T
b
and V
b
are needed.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
MW M
i
1.0 5000.0
MUP p
i
0.0 5x10
24
DIPOLEMOMENT
TB T
bi
4.0 2000.0 TEMPERATURE
VB V
b
0.001 3.5 MOLEVOLUME
OMEGA
e
i
0.5 2.0
STKPAR/1
(c
i
/k)
ST
fcn(T
bi
, V
bi
, p
i
) x TEMPERATURE
3 Transport Property Models 291
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
STKPAR/2
o
i
ST
fcn(T
bi
, V
bi
, p
i
) x LENGTH
LJPAR/1
(c
i
/k)
LJ
fcn(T
ci
, e
i
)
x TEMPERATURE
LJPAR/2
o
i
LJ
fcn(T
ci
, p
ci
, e
i
)
x LENGTH
References
R.C. Reid, J.M. Praunitz, and B.E. Poling, The Properties of Gases and Liquids,
4th ed., (New York: McGrawHill, 1987), p. 587.
ChapmanEnskogWilkeLee (Mixture)
The diffusion coefficient of a gas into a gas mixture at low pressures is
calculated using an equation of Bird, Stewart, and Lightfoot by default (option
code 0):
If the first option code is set to 1, Blanc's law is used instead:
The binary diffusion coefficient D
ij
v
(p = 0) at low pressures is calculated using
the ChapmanEnskogWilkeLee model. (See ChapmanEnskogWilkeLee
(Binary).)
You must provide parameters for this model.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DVBLNC 1 x
DVBLNC is set to 1 for a diffusing component and 0 for a nondiffusing
component.
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 597.
DawsonKhouryKobayashi (Binary)
The binary diffusion coefficient D
ij
v
at high pressures is calculated from the
DawsonKhouryKobayashi model:
292 3 Transport Property Models
D
ij
v
=D
ji
v
D
ij
v
(p = 0) is the lowpressure binary diffusion coefficient obtained from the
ChapmanEnskogWilkeLee model.
The parameters
m
v
and V
m
v
are obtained from the RedlichKwong equation
ofstate model.
You must supply parameters for these two models.
Subscript i denotes a diffusing component. j denotes a solvent.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
VC V
ci
x 0.001 3.5 MOLEVOLUME
References
R.C. Reid, J.M. Prausnitz, and T.K. Sherwood. The Properties of Gases and
Liquids, 3rd ed., (New York: McGrawHill, 1977), pp. 560565.
DawsonKhouryKobayashi (Mixture)
The diffusion coefficient of a gas into a gas mixture at high pressure is
calculated using an equation of Bird, Stewart, and Lightfoot by default (option
code 0):
If the first option code is set to 1, Blanc's law is used instead:
The binary diffusion coefficient D
ij
v
at high pressures is calculated from the
DawsonKhouryKobayashi model. (See DawsonKhouryKobayashi (Binary).)
At low pressures (up to 1 atm) the binary diffusion coefficient is instead
calculated by the ChapmanEnskogWilkeLee (Binary) model.
You must provide parameters for this model.
3 Transport Property Models 293
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
DVBLNC 1
DVBLNC is set to 1 for a diffusing component and 0 for a nondiffusing
component.
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 597.
NernstHartley
The effective diffusivity of an ion i in a liquid mixture with electrolytes can be
calculated using the NernstHartley model:
(1)
Where:
F = 9.65x10
7
C/kmole (Faraday's number)
x
k
= Mole fraction of any molecular species k
z
i
= Charge number of species i
The binary diffusion coefficient of the ion with respect to a molecular species
is set equal to the effective diffusivity of the ion in the liquid mixture:
(2)
The binary diffusion coefficient of an ion i with respect to an ion j is set to the
mean of the effective diffusivities of the two ions:
The diffusivity for molecular species is calculated by the WilkeChang
(Mixture) model.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
CHARGE z 0.0
IONMOB/1 l
1
AREA, MOLES
IONMOB/2 l
2
0.0 AREA, MOLES,
TEMPERATURE
When IONMOB/1 is missing, the NernstHartley model uses a nominal value
of 5.0 and issues a warning. This parameter should be specified for ions, and
not allowed to default to this value.
294 3 Transport Property Models
References
A. L. Horvath, Handbook of Aqueous Electrolyte Solutions, (Chichester: Ellis
Horwood, Ltd, 1985).
WilkeChang (Binary)
The WilkeChang model calculates the liquid diffusion coefficient of
component i in a mixture at finite concentrations:
D
ij
l
= D
ji
l
The equation for the WilkeChang model at infinite dilution is:
Where i is the diffusing solute and j the solvent, and:
j
= Association factor of solvent. 2.26 for water, 1.90
for methanol, 1.50 for ethanol, 1.20 for propyl
alchohols and nbutanol, and 1.00 for all other
solvents.
V
bi
= Liquid molar volume at T
b
of solvent i
q
j
l
= Liquid viscosity of the solvent. This can be obtained
from the General Pure Component Liquid Viscosity
model. You must provide parameters for one of
these models.
q
l
= Liquid viscosity of the complete mixture of n
components
x
i
, x
j
= Apparent binary mole fractions. If the actual mole
fractions are then
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
MW M
j
1.0 5000.0
VB V
bi
*,l
0.001 3.5 MOLEVOLUME
3 Transport Property Models 295
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 598600.
WilkeChang (Mixture)
The WilkeChang model calculates the infinitedilution liquid diffusion
coefficient of component i in a mixture.
The equation for the WilkeChang model is:
With:
Where:
j
= Association factor of solvent. 2.26 for water, 1.90 for
methanol, 1.50 for ethanol, 1.20 for propyl alchohols
and nbutanol, and 1.00 for all other solvents.
n
l
= Mixture liquid viscosity of all nondiffusing
components. This can be obtained from the General
Pure Component Liquid Viscosity model. You must
provide parameters for one its submodels.
The parameter DLWC specifies which components diffuse. It is set to 1 for a
diffusing component and 0 for a nondiffusing component.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
MW M
j
1.0 5000.0
VB V
bi
*,l
0.001 3.5 MOLEVOLUME
DLWC 1
References
R.C. Reid, J.M. Praunsnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th ed., (New York: McGrawHill, 1987), p. 618.
Surface Tension Models
The Aspen Physical Property System has the following builtin surface tension
models.This section describes the surface tension models available.
Model Type
Liquid Mixture Surface Tension Liquidvapor
296 3 Transport Property Models
Model Type
API Liquidvapor
IAPS Waterstream
General Pure Component Liquid Surface
Tension
Liquidvapor
OnsagerSamaras Electrolyte Correction Electrolyte liquidvapor
Modified MacLeodSugden Liquidvapor
Liquid Mixture Surface Tension
The liquid mixture surface tension is calculated using a general weighted
average expression (R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties
of Gases and Liquids, 4th. ed., New York: McGrawHill, 1987, p. 643):
Where:
x = Mole fraction
r = Exponent (specified by the option code)
Hadden (S. T. Hadden, Hydrocarbon Process Petrol Refiner, 45(10), 1966, p.
161) suggested that the exponent value r =1 should be used for most
hydrocarbon mixtures. However, Reid recommended the value of r in the
range of 1 to 3. The exponent value r can be specified using the models
Option Code (option code = 1, 1, 2, ..., 9 corresponding to the value of r).
The default value of r for this model is 1.
The pure component liquid surface tension o
i
*,l
is calculated by the General
Pure Component Liquid Surface Tension model.
API Surface Tension
The liquid mixture surface tension for hydrocarbons is calculated using the
API model. This model is recommended for petroleum and petrochemical
applications. It is used in the CHAOSEA, GRAYSON, LKPLOCK, PENGROB,
and RKSOAVE property models. The general form of the model is:
Where:
fcn = A correlation based on API Procedure 10A32 (API Technical Data Book,
Petroleum Refining, 4th edition)
The original form of this model is only designed for petroleum, and treats all
components as pseudocomponents (estimating surface tension from boiling
point, critical temperature, and specific gravity). If option code 1 is set to 0
(the default), it behaves this way. Set option code 1 to 1 for the model to use
the General Pure Component Liquid Surface Tension model to calculate the
surface tension of real components and the API model for pseudocomponents.
3 Transport Property Models 297
The mixture surface tension is then calculated as a molefractionweighted
average of these surface tensions.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
TB T
bi
4.0 2000.0 TEMPERATURE
SG SG 0.1 2.0
TC T
ci
5.0 2000 TEMPERATURE
IAPS Surface Tension for Water
The IAPS surface tension model was developed by the International
Association for Properties of Steam. It calculates liquid surface tension for
water and steam. This model is used in option sets STEAMNBS and STEAM
TA.
The general form of the equation for the IAPS surface tension model is:
o
w
=fcn(T, p)
Where:
fcn = Correlation developed by IAPS
The model is only applicable to water. No parameters are required.
General Pure Component Liquid Surface
Tension
The Aspen Physical Property System has several submodels for calculating
liquid surface tension. It uses parameter TRNSWT/5 to determine which
submodel is used. See Pure Component TemperatureDependent Properties
for details.
If TRNSWT/5 is This equation is used And this parameter is
used
0 HakimSteinbergStiel
106 DIPPR SIGDIP
301 PPDS SIGPDS
401 IKCAPE polynomial
equation
SIGPO
505 NIST TDE Watson
equation
SIGTDEW
511 NIST TDE expansion SIGISTE
512 NIST PPDS14 Equation SIGPDS14
HakimSteinbergStiel
The HakimSteinbergStiel equation is:
Where:
298 3 Transport Property Models
Q
pi
=
m
i
=
The parameter _
i
is the Stiel polar factor.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
TC T
ci
5.0 2000.0 TEMPERATURE
PC p
ci
PRESSURE
OMEGA
e
i
0.5 2.0
CHI
_
i
0
DIPPR Liquid Surface Tension
The DIPPR equation for liquid surface tension is:
Where:
T
ri
= T / T
ci
Linear extrapolation of o
i
*,l
versus T occurs outside of bounds.
(Other DIPPR equations may sometimes be used. See Pure Component
TemperatureDependent Properties for details.)
The DIPPR model is used by PCES when estimating liquid surface tension.
Note: Reduced temperature T
r
is always calculated using absolute
temperature units.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
SIGDIP/1 C
1i
SURFACE
TENSION
SIGDIP/2, ..., 5 C
2i
, ..., C
5i
0
SIGDIP/6 C
6i
0 TEMPERATURE
SIGDIP/7 C
7i
1000 TEMPERATURE
PPDS
The equation is:
Linear extrapolation of o
i
*,l
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SIGPDS/1 C
1i
SURFACE
TENSION
3 Transport Property Models 299
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SIGPDS/2 C
2i
0
SIGPDS/3 C
3i
0
SIGPDS/4 C
4i
0 TEMPERATURE
SIGPDS/5 C
5i
1000 TEMPERATURE
NIST PPDS14 Equation
This equation is the same as the PPDS equation above, but it uses its own
parameter set which includes critical temperature.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SIGPDS14/1 C
1i
x N/m
SIGPDS14/2 C
2i
0 x
SIGPDS14/3 C
3i
0 x
SIGPDS14/4 T
ci
x TEMPERATURE
SIGPDS14/5 C
4i
0 x TEMPERATURE
SIGPDS14/6 C
5i
1000 x TEMPERATURE
IKCAPE Polynomial
The IKCAPE equation is:
Linear extrapolation of o
i
*,l
versus T occurs outside of bounds.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SIGPO/1 C
1i
x SURFACE
TENSION
SIGPO/2, ..., 10 C
2i
, ..., C
10i
0 x SURFACE
TENSION
TEMPERATURE
SIGPO/11 C
11i
0 x TEMPERATURE
SIGPO/12 C
12i
1000 x TEMPERATURE
NIST TDE Watson Equation
This equation can be used when parameter SIGTDEW is available.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SIGTDEW/1 C
1i
x
300 3 Transport Property Models
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SIGTDEW/2, 3, 4 C
2i
, C
3i
, C
4i
0 x
SIGTDEW/5 T
c
x TEMPERATURE
SIGTDEW/6 nTerms 4 x
SIGTDEW/7 T
lower
0 x TEMPERATURE
SIGTDEW/8 T
upper
1000 x TEMPERATURE
NIST TDE Expansion
This equation can be used when parameter SIGISTE is available.
Parameter
Name/Element
Symbol Default MDS Lower
Limit
Upper
Limit
Units
SIGISTE/1 C
1i
x N/m
SIGISTE/2, 3, 4 C
2i
, C
3i
, C
4i
0 x N/m
SIGISTE/5 T
ci
x TEMPERATURE
SIGISTE/6 nTerms 4 x
SIGISTE/7 T
lower
0 x TEMPERATURE
SIGISTE/8 T
upper
1000 x TEMPERATURE
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 4th. ed., (New York: McGrawHill, 1987), p. 638.
OnsagerSamaras
The OnsagerSamaras model calculates the correction to the liquid mixture
surface tension of a solvent mixture, due to the presence of electrolytes:
for salt concentration < 0.03
M
(1)
Where:
o
solv
= Surface tension of the solvent mixture calculated
by the General Pure Component Liquid Surface
Tension model
x
ca
a
= Mole fraction of the apparent electrolyte ca
Ao
ca
= Contribution to the surface tension correction due
to apparent electrolyte ca
For each apparent electrolyte ca, the contribution to the surface tension
correction is calculated as:
3 Transport Property Models 301
(2)
Where:
c
solv
= Dielectric constant of the solvent mixture
c
ca
a
=
V
m
l
= Liquid molar volume calculated by the Clarke
model
Apparent electrolyte mole fractions are computed from the true ion mole
fractions and ionic charge number. They are also computed if you use the
apparent component approach. See Apparent Component and True
Component Approaches in the Electrolyte Calculation chapter for a more
detailed discussion of this method.
Above salt concentration 0.03 M, the slope of surface tension vs. mole
fraction is taken to be constant at the value from 0.03 M.
You must provide parameters for the General Pure Component Liquid Surface
Tension model, used for the calculation of the surface tension of the solvent
mixture.
Parameter
Name/Element
Symbol Default Lower
Limit
Upper
Limit
Units
CHARGE z 0.0
References
A. L. Horvath, Handbook of Aqueous Electrolyte Solutions, (Chichester: Ellis,
Ltd. 1985).
Modified MacLeodSugden
The modified MacLeodSugden equation for mixture liquid surface tension can
be derived from the standard MacLeodSugden equation by assuming that the
density of the vapor phase is zero. The modified MacLeodSugden equation is:
Where:
o
i
*,l
= Surface tension for pure component i, calculated using
the General Pure Component Liquid Surface Tension
model.
V
i
*,l
= Liquid molar volume for pure component i, calculated
using the General Pure Component Liquid Molar Volume
model.
302 3 Transport Property Models
V
l
= Mixture liquid molar volume, calculated using the
Rackett model.
References
R.C. Reid, J.M. Prausnitz, and B.E. Poling, The Properties of Gases and
Liquids, 3rd. ed., (New York: McGrawHill, 1977).
4 Nonconventional Solid Property Models 303
4 Nonconventional Solid
Property Models
This section describes the nonconventional solid density and enthalpy models
available in the Aspen Physical Property System. The following table lists the
available models and their model names. Nonconventional components are
solid components that cannot be characterized by a molecular formula. These
components are treated as pure components, though they are complex
mixtures.
Nonconventional Solid Property Models
General Enthalpy and
Density Models
Model name Phase(s)
General density polynomial DNSTYGEN S
General heat capacity
polynomial
ENTHGEN S
Enthalpy and Density
Models for Coal and Char
Model name Phase(s)
General coal enthalpy model HCOALGEN S
IGT coal density model DCOALIGT S
IGT char density model DCHARIGT S
General Enthalpy and Density
Models
The Aspen Physical Property System has two builtin general enthalpy and
density models. This section describes the general enthalpy and density
models available:
General Density Polynomial
DNSTYGEN is a general model that gives the density of any nonconventional
solid component. It uses a simple mass fraction weighted average for the
reciprocal temperaturedependent specific densities of its individual
constituents. There may be up to twenty constituents with mass percentages.
304 4 Nonconventional Solid Property Models
You must define these constituents, using the general component attribute
GENANAL. The equations are:
Where:
w
ij
= Mass fraction of the jth constituent in component i
ij
s
= Density of the jth consituent in component i
Parameter
Name/Element
Symbol MDS Default Lower
Limit
Upper
Limit
Units
DENGEN/1, 5, 9, , 77 a
i,j1
x MASSENTHALPY
and TEMPERATURE
DENGEN/2, 6, 10, ,
78
a
i,j2
x 0 MASSENTHALPY
and TEMPERATURE
DENGEN/3, 7, 11, ,
79
a
i,j3
x 0 MASSENTHALPY
and TEMPERATURE
DENGEN/4, 8, 12, ,
80
a
i,j4
x 0 MASSENTHALPY
and TEMPERATURE
Use the elements of GENANAL to input the mass percentages of the
constituents. The structure of DENGEN is: Elements 1 to 4 are the four
coefficients for the first constituent, elements 5 to 8 are the coefficients for
the second constitutent, and so on, for up to 20 constituents.
General Heat Capacity Polynomial
ENTHGEN is a general model that gives the specific enthalpy of any
nonconventional component as a simple massfractionweightedaverage for
the enthalpies of its individual constituents. You may define up to twenty
constituents with mass percentages, using the general component attribute
GENANAL. The specific enthalpy of each constituent at any temperature is
calculated by combining specific enthalpy of formation of the solid with a
sensible heat change. (See Nonconventional Component Enthalpy Calculation
in Physical Property Methods.)
The equations are:
Where:
4 Nonconventional Solid Property Models 305
w
ij
= Mass fraction of the jth constituent in
component i
h
i
s
= Specific enthalpy of solid component i
A
f
h
j
s
= Specific enthalpy of formation of constituent j
C
p
s
= Heat capacity of the jth constituent in
component i
Parameter
Name/Element
Symbol MDS Default Lower
Limit
Upper
Limit
Units
DHFGEN/J
A
f
h
j
s
x 0 MASSENTHALPY
HCGEN/1, 5, 9, , 77 a
i,j1
x MASSENTHALPY
and
TEMPERATURE
HCGEN/2, 6, 10, , 78 a
i,j2
x 0 MASSENTHALPY
and
TEMPERATURE
HCGEN/3, 7, 11, , 79 a
i,j3
x 0 MASSENTHALPY
and
TEMPERATURE
HCGEN/4, 8, 12, , 80 a
i,j4
x 0 MASSENTHALPY
and
TEMPERATURE
The elements of GENANAL are used to input the mass percentages of the
constituents. The structure for HCGEN is: Elements 1 to 4 are the four
coefficients for the first constituent, elements 5 to 8 are the coefficients for
the second constitutent, and so on, for up to 20 constituents.
Enthalpy and Density Models
for Coal and Char
Coal is modeled in the Aspen Physical Property System as a nonconventional
solid. Coal models are empirical correlations, which require solid material
characterization information. Component attributes are derived from
constituent analyses. Definitions of coal component attributes are given in the
Aspen Plus User Guide, Chapter 6.
Enthalpy and density are the only properties calculated for nonconventional
solids. This section describes the special models available in the Aspen
Physical Property System for the enthalpy and density of coal and char. The
component attributes required by each model are included. The coal models
are:
 General coal enthalpy
 IGT Coal Density
 IGT Char Density
306 4 Nonconventional Solid Property Models
 User models for density and enthalpy. See User Models for
Nonconventional Properties in Chapter 6 of Aspen Plus User Models, for
details on writing the subroutines for these user models.
Notation
Most correlations for the calculation of coal properties require proximate,
ultimate, and other analyses. These are converted to a dry, mineralmatter
free basis. Only the organic portion of the coal is considered.
Moisture corrections are made for all analyses except hydrogen, according to
the formula:
Where:
w = The value determined for weight fraction
w
d
= The value on a dry basis
= The moisture weight fraction
For hydrogen, the formula includes a correction for freemoisture hydrogen:
The mineral matter content is calculated using the modified Parr formula:
The ash term corrects for water lost by decomposition of clays in the ash
determination. The average water constitution of clays is assumed to be 11.2
percent. The sulfur term allows for loss in weight of pyritic sulfur when pyrite
is burned to ferric oxide. The original Parr formula assumed that all sulfur is
pyritic sulfur. This formula included sulfatic and organic sulfur in the mineral
matter calculation. When information regarding the forms of sulfur is
available, use the modified Parr formula to give a better approximation of the
percent of inorganic material present. Because chlorine is usually small for
United States coals, you can omit chlorine from the calculation.
Correct analyses from a dry basis to a dry, mineralmatterfree basis, using
the formula:
Where:
Aw
d
= Correction factor for other losses, such as the loss
of carbon in carbonates and the loss of hydrogen
present in the water constitution of clays
4 Nonconventional Solid Property Models 307
The oxygen and organic sulfur contents are usually calculated by difference
as:
Where:
C
p
= Heat capacity / (J/kgK)
c
p
= Heat capacity / (cal/gC)
h = Specific enthalpy
A
c
h
= Specific heat of combustion
A
f
h
= Specific heat of formation
R
o
= Meanmaximum relectance in oil
T = Temperature/K
t = Temperature/C
w = Weight fraction
= Specific density
Subscripts:
A = Ash
C = Carbon
Cl = Chlorine
FC = Fixed carbon
H = Hydrogen
H
2
O = Moisture
MM = Mineral matter
N = Nitrogen
O = Oxygen
So = Organic sulfur
Sp = Pyritic sulfur
St = Total sulfur
S = Other sulfur
VM = Volatile matter
Superscripts:
d = Dry basis
308 4 Nonconventional Solid Property Models
m = Mineralmatterfree basis
General Coal Enthalpy Model
The general coal model for computing enthalpy in the Aspen Physical Property
System is HCOALGEN. This model includes a number of different correlations
for the following:
 Heat of combustion
 Heat of formation
 Heat capacity
You can select one of these correlations using an option code in the Methods
 NCProps form. (See the Aspen Plus User Guide, Chapter 6). Use option
codes to specify a calculation method for properties. Each element in the
option code vector is used in the calculation of a different property.
The table labeled HCOALGEN Option Codes (below) lists model option codes
for HCOALGEN. The table is followed by a detailed description of the
calculations used for each correlation.
The correlations are described in the following section. The component
attributes are defined in Aspen Plus User Guide, Chapter 6.
Heat of Combustion Correlations
The heat of combustion of coal in the HCOALGEN model is a gross calorific
value. It is expressed in Btu/lb of coal on a dry mineralmatterfree basis.
ASTM Standard D586507a defines standard conditions for measuring gross
calorific value. (Earlier ASTM Standard D2015 used the same conditions.)
Initial oxygen pressure is 20 to 40 atmospheres. Products are in the form of
ash; liquid water; and gaseous CO
2
, SO
2
, and NO
2
.
You can calculate net calorific value from gross calorific value by making a
deduction for the latent heat of vaporization of water.
Heat of combustion values are converted back to a dry, mineralmatter
containing basis with a correction for the heat of combustion of pyrite. The
formula is:
The heat of combustion correlations were evaluated by the Institute of Gas
Technology (IGT). They used data for 121 samples of coal from the Penn
State Data Base (IGT, 1976) and 457 samples from a USGS report (Swanson,
et al., 1976). These samples included a wide range of United States coal
fields. The constant terms in the HCOALGEN correlations are bias corrections
obtained from the IGT study.
Boie Correlation:
Parameter Name/Element Symbol Default
BOIEC/1 a
1i
151.2
BOIEC/2 a
2i
499.77
4 Nonconventional Solid Property Models 309
Parameter Name/Element Symbol Default
BOIEC/3 a
3i
45.0
BOIEC/4 a
4i
47.7
BOIEC/5 a
5i
27.0
BOIEC/6 a
6i
189.0
Dulong Correlation:
Parameter Name/Element Symbol Default
DLNGC/1 a
1i
145.44
DLNGC/2 a
2i
620.28
DLNGC/3 a
3i
40.5
DLNGC/4 a
4i
77.54
DLNGC/5 a
5i
16.0
Grummel and Davis Correlation:
Parameter Name/Element Symbol Default
GMLDC/1 a
1i
0.3333
GMLDC/2 a
2i
654.3
GMLDC/3 a
3i
0.125
GMLDC/4 a
4i
0.125
GMLDC/5 a
5i
424.62
GMLDC/6 a
6i
2.0
Mott and Spooner Correlation:
Parameter Name/Element Symbol Default
MTSPC/1 a
1i
144.54
MTSPC/2 a
2i
610.2
MTSPC/3 a
3i
40.3
MTSPC/4 a
4i
62.45
MTSPC/5 a
5i
30.96
MTSPC/6 a
6i
65.88
MTSPC/7 a
7i
47.0
IGT Correlation:
310 4 Nonconventional Solid Property Models
Parameter Name/Element Symbol Default
CIGTC/1 a
1i
178.11
CIGTC/2 a
2i
620.31
CIGTC/3 a
3i
80.93
CIGTC/4 a
4i
44.95
CIGTC/5 a
5i
5153.0
Revised IGT Correlation (Perry's, 7th ed., equation 277):
Parameter Name/Element Symbol Default
CIGT2/1 a
1i
146.58
CIGT2/2 a
2i
568.78
CIGT2/3 a
3i
29.4
CIGT2/4 a
4i
6.58
CIGT2/5 a
5i
51.53
User Input Value of Heat Combustion
Parameter Name/Element Symbol Default
HCOMB
A
c
h
i
d
0
Standard Heat of Formation Correlations
There are two standard heat of formation correlations for the HCOALGEN
model:
 Heat of combustionbased
 Direct
Heat of CombustionBased Correlation: This is based on the assumption that
combustion results in complete oxidation of all elements except sulfatic sulfur
and ash, which are considered inert. The numerical coefficients are
combinations of stoichiometric coefficients and heat of formation for CO
2
,
H
2
O, HCl, and NO
2
at 298.15K:
For example, the complete oxidation of hydrogen is based on the reaction
, since the stable phase of water at 298.15 K is
liquid, the heat of vaporization at 298.15 K is needed in the conversion. The
numerical coefficient of is calculated by:
4 Nonconventional Solid Property Models 311
The complete oxidation of carbon is based on the reaction
, and the numerical coefficient of is calculated by:
The complete oxidation of sulfur (pyritic and organic sulfur) is based on the
reaction , and the numerical coefficient of is
calculated by:
The complete oxidation of nitrogen is based on the reaction
, and the numerical coefficient of is calculated by:
The complete oxidation of chlorine is based on the reaction
, and the numerical coefficient of
is calculated by:
Direct Correlation: Normally small, relative to its heat of combustion. An error
of 1% in the heat of a combustionbased correlation produces about a 50%
error when it is used to calculate the heat of formation. For this reason, the
following direct correlation was developed, using data from the Penn State
Data Base. It has a standard deviation of 112.5 Btu/lb, which is close to the
limit, due to measurement in the heat of combustion:
Where:
Parameter Name/Element Symbol Default
HFC/1 a
1i
1810.123
HFC/2 a
2i
502.222
HFC/3 a
3i
329.1087
HFC/4 a
4i
121.766
HFC/5 a
5i
542.393
HFC/6 a
6i
1601.573
312 4 Nonconventional Solid Property Models
Parameter Name/Element Symbol Default
HFC/7 a
7i
424.25
HFC/8 a
8i
525.199
HFC/9 a
9i
11.4805
HFC/10 a
10i
31.585
HFC/11 a
11i
13.5256
HFC/12 a
12i
11.5
HFC/13 a
13i
685.846
HFC/14 a
14i
22.494
HFC/15 a
15i
64836.19
Heat Capacity Kirov Correlations
The Kirov correlation (1965) considered coal to be a mixture of moisture, ash,
fixed carbon, and primary and secondary volatile matter. Secondary volatile
matter is any volatile matter up to 10% on a dry, ashfree basis; the
remaining volatile matter is primary. The correlation developed by Kirov
treats the heat capacity as a weighted sum of the heat capacities of the
constituents:
Where:
i = Component index
j = Constituent index j = 1, 2 , ... , ncn
Where the values of j represent:
1 Moisture
2 Fixed carbon
3 Primary volatile matter
4 Secondary volatile matter
5 Ash
w
j
= Mass fraction of jth constituent on dry basis
This correlation calculates heat capacity in cal/gramC using temperature in
C. The parameters must be specified in appropriate units for this conversion.
Parameter Name/Element Symbol Default
CP1C/1 a
i,11
1.0
CP1C/2 a
i,12
0
CP1C/3 a
i,13
0
CP1C/4 a
i,14
0
CP1C/5 a
i,21
0.165
CP1C/6 a
i,22
4 Nonconventional Solid Property Models 313
Parameter Name/Element Symbol Default
CP1C/7 a
i,23
CP1C/8 a
i,24
0
CP1C/9 a
i,31
0.395
CP1C/10 a
i,32
CP1C/11 a
i,33
0
CP1C/12 a
i,34
0
CP1C/13 a
i,41
0.71
CP1C/14 a
i,42
CP1C/15 a
i,43
0
CP1C/16 a
i,44
0
CP1C/17 a
i,51
0.18
CP1C/18 a
i,52
CP1C/19 a
i,53
0
CP1C/20 a
i,54
0
Cubic Temperature Equation
The cubic temperature equation is:
Parameter Name/Element Symbol Default
CP2C/1 a
1i
0.438
CP2C/2 a
2i
CP2C/3 a
3i
CP2C/4 a
4i
The default values of the parameters were developed by Gomez, Gayle, and
Taylor (1965). They used selected data from three lignites and a
subbituminous B coal, over a temperature range from 32.7 to 176.8C. This
correlation calculates heat capacity in cal/gramC using temperature in C. The
parameters must be specified in appropriate units for this conversion.
HCOALGEN Option Codes
Option Code
Number
Option Code
Value
Calculation
Method
Parameter
Names
Component
Attributes
1 Heat of Combustion
1 Boie correlation BOIEC ULTANAL
SULFANAL
PROXANAL
2 Dulong
correlation
DLNGC ULTANAL
SULFANAL
PROXANAL
3 Grummel and
Davis correlation
GMLDC ULTANAL
SULFANAL
PROXANAL
314 4 Nonconventional Solid Property Models
Option Code
Number
Option Code
Value
Calculation
Method
Parameter
Names
Component
Attributes
1 Heat of Combustion
4 Mott and Spooner
correlation
MTSPC ULTANAL
SULFANAL
PROXANAL
5 IGT correlation CIGTC ULTANAL
PROXANAL
6 User input value HCOMB ULTANAL
PROXANAL
7 Revised IGT
correlation
CIGT2 ULTANAL
PROXANAL
2 Standard Heat of Formation
1 Heatof
combustion
based correlation
ULTANAL
SULFANAL
2 Direct correlation HFC ULTANAL
SULFANAL
PROXANAL
3 Heat Capacity
1 Kirov correlation CP1C PROXANAL
2 Cubic
temperature
equation
CP2C
4 Enthalpy Basis
1 Elements in their
standard states
at 298.15K and 1
atm
2 Component at
298.15 K