PVTsim 13
CALSEP
Contents
Introduction
Introduction ............................................................................................................................... 5
Composition Handling
13
Composition Handling............................................................................................................. 13
Types of fluid analyses.............................................................................................. 13
Handling of pure components heavier than C6 .......................................................... 14
Fluid handling operations .......................................................................................... 15
Mixing ....................................................................................................................... 15
Weaving .................................................................................................................... 15
Recombination........................................................................................................... 15
Characterization to the same pseudocomponents..................................................... 15
Flash Algorithms
17
25
Equations of State
28
37
45
Transport Properties
49
Transport Properties................................................................................................................. 49
Viscosity.................................................................................................................... 49
Thermal Conductivity................................................................................................ 55
Gas/oil Interfacial Tension ........................................................................................ 57
References ................................................................................................................. 58
PVT Experiments
60
PVT Experiments..................................................................................................................... 60
Constant Mass Expansion.......................................................................................... 60
Differential Depletion................................................................................................ 61
Constant Volume Depletion ...................................................................................... 61
Separator Experiments............................................................................................... 62
Viscosity Experiment ................................................................................................ 62
Swelling Experiment ................................................................................................. 62
References ................................................................................................................. 63
63
68
73
Unit Operations
77
Unit Operations........................................................................................................................ 77
Compressor................................................................................................................ 77
Expander.................................................................................................................... 79
Cooler ........................................................................................................................ 80
Heater ........................................................................................................................ 80
Pump.......................................................................................................................... 80
Valve ......................................................................................................................... 80
Separator.................................................................................................................... 80
References ................................................................................................................. 80
81
Hydrate Formation................................................................................................................... 81
Types of Hydrates ..................................................................................................... 81
Hydrate Model........................................................................................................... 82
Hydrate P/T Flash Calculations................................................................................. 85
Calculation of Fugacities ......................................................................................................... 86
Fluid Phases............................................................................................................... 86
Hydrate Phases .......................................................................................................... 86
Ice .............................................................................................................................. 87
Salts ........................................................................................................................... 87
References ................................................................................................................. 88
90
Asphaltenes
96
Asphaltenes.............................................................................................................................. 96
Asphaltene Component Properties ............................................................................ 96
References ................................................................................................................. 97
H2S Simulations
98
H2S Simulations....................................................................................................................... 98
99
114
128
137
Introduction
Introduction
This document describes the calculation procedures used in PVTsim. When installing PVTsim
the Method Documentation is copied to the installation directory as a PDF document
(pvtdoc.pdf). It may further be accessed from the <Help> menu in PVTsim. The <Help> menu
also gives access to a Users Manual. This is during installation copied to the PVTsim installation
directory as the PDF document pvthelp.pdf.
Component Classes
PVTsim distinguishes between the following component classes
Water
Hydrate inhibitors
Salts
Other inorganic
Organic defined
Pseudocomponents
The program is delivered with a pure component database consisting of the following
components
Short Name
Water
H2O
Hydrate inhibitors
MeOH
EtOH
PG
DPGME
MEG
PGME
DPG
DEG
TEG
Glycerol
Salts
NaCl
Systematic Name
Formula Name
Water
H2O
Methanol
Ethanol
Propyleneglycol
Dipropyleneglycolmethylether
Monoethyleneglycol
Propyleneglycolmethylether
Dipropyleneglycol
Diethyleneglycol
Triethyleneglycol
Glycerol
CH4O
C2H6O
C6H8O2
C7H16O3
C2H6O2
C7H10O2
C6H14O3
C4H10O3
C6H14O4
C3H8O3
Sodium chloride
NaCl
KCl
NaBr
CaCl2
HCOONa
HCOOK
KBr
HCOOCs
CaBr2
ZnBr2
Other inorganic
He
H2
N2
Ar
O2
CO2
H2S
Organic defined
C1
C2
C3
cC3
iC4
nC4
2,2dimC3
cC4
iC5
nC5
cC5
2,2dimC4
2,3dimC4
2mC5
3mC5
nC6
C6
mcC5
Benzene
Napht
cC6
223tmC4
3,3dimC5
2mC6
c13dmcC5
t13dmcC5
3mC6
t12dmcC5
nC7
mcC6
etcC5
113trcC5
Potassium chloride
Sodium bromide
Calcium chloride (anhydrous)
Sodium formate (anhydrous)
Potassium formate (anhydrous)
Potassium bromide
Caesium formate (anhydrous)
Calcium bromide (anhydrous)
Zinc bromide
KCl
NaBr
CaCl2
HCOONa
HCOOK
KBr
HCOOCs
CaBr2
ZnBr2
Helium4
Hydrogen
Nitrogen
Argon
Oxygen
Carbon dioxide
Hydrogen sulfide
He(4)
H2
N2
Ar
O2
CO2
H2S
Methane
Ethane
Propane
Cyclopropane
Isobutane
Normalbutane
2,2Dimethylpropane
Cyclopropane
2methylbutane
Normalpentane
Cyclopentane
2,2Dimethylbutane
2,3Dimethylbutane
2Methylpentane
3Methylpentane
Normalhexane
Hexane
Methylcyclopentane
Benzene
Naphthalene
Cyclohexane
2,2,3Trimethylbutane
3,3Dimethylbutane
2Methylhexane
Cis1,3Dimethylcyclopentane
Trans1,3Dimethylcyclopentane
3Methylhexane
Trans1,2Dimethylcyclopentane
Normalheptane
Methylcyclohexane
Ethylcyclopentane
1,1,3Trimethylcyclopentane
CH4
C2H6
C3H8
C3H6
C4H10
C4H10
C5H12
C4H8
C5H12
C5H12
C5H8
C6H14
C6H14
C6H14
C6H14
C6H14
C6H12
C6H6
C10H8
C6H12
C7H16
C7H16
C7H16
C7H14
C7H14
C7H16
C7H14
C7H16
C7H14
C7H14
C8H16
Toluene
2mC7
cC7
3mC7
11dmcC6
c13dmcC6
t12dmcC6
nC8
c12dmcC6
EtcC6
etBenzene
pXylene
mXylene
2mC8
oXylene
1m3ecC6
1m4ecC6
cC8
4mC8
nC9
Mesitylene
PsCumene
nC10
Hemellitol
nC11
nC12
nC13
1mNapht
nC14
nC15
nC16
nC17
nC18
nC19
nC20
nC21
nCn
nC40
Toluene
2Methylheptane
Cycloheptane
3Methylheptane
1,1Dimethylcyclohexane
Cis1,3Dimethylcyclohexane
Trans1,2Dimethylcyclohexane
Normaloctane
Cis1,2Dimethylcyclohexane
Ethylcyclohexane
EthylBenzene
Paraxylene
Metaxylene
2Methyloctane
Orthoxylene
1Methyl3Ethylcyclohexane
1Methyl4Ethylcyclohexane
Cyclooctane
4Methyloctane
Normalnonane
1,3,5TrimethylBenzene
1,2,4TrimethylBenzene
Normaldecane
1,2,3TrimethylBenzene
Normalundecane
Normaldodecane
Normaltridecane
1methylNaphthalene
Normaltetradecane
Normalpentadecane
Normalhexadecane
Normalheptadecane
Normaloctadecane
Normalnonadecane
Normaleicosane
NormalC21
NormalCn
NormalC40
C7H8
C8H18
C7H14
C8H18
C8H16
C8H16
C8H16
C8H18
C8H16
C8H16
C8H10
C8H10
C8H10
C9H20
C8H10
C9H18
C9H18
C8H16
C9H20
C9H20
C9H12
C9H12
C10H22
C9H12
C11H24
C12H26
C13H28
C11H10
C14H30
C15H32
C16H34
C17H36
C18H38
C19H40
C20H42
C21H44
CnH2n+2
C40H82
The database furthermore contains the carbon number fractions from a C21 fraction to a C100
fraction. Each fraction Cn consists of all components with a boiling point in the interval from that
of nCn1 + 0.5C/0.9F to that of nCn + 0.5C/0.9F.
Finally the database contains the components CHCmp_1 to CHCmp_6, which are dummy
pseudocomponents. The only properties given in the database are the molecular weight,
and
, and the molecular weight will usually also have to be modified by the user. Other
component properties must be entered manually.
Component Properties
For each component the database holds the following component properties
Enthalpy of melting (
)
PNA distribution (only for pseudocomponents)
Wax fraction (only for nparaffins and pseudocomponents)
Asphaltene fraction (only for pseudocomponents)
Parachor
Hydrate formation indicator (None, I, II, H and combinations)
Hydrate Langmuir constants
Number of ions in aqueous solution (only for salts)
Number of crystal water molecules per salt molecule (only for salts)
Pc of wax forming fractions (only for nparaffins and pseudocomponents)
and
The component properties needed to calculate various physical properties and transport
properties will usually be established as a part of the fluid characterization. It is however, also
possible to input new components without entering all component properties and it is possible to
input compositions in characterized form.
Tc, Pc, ,
,
, and molecular weight are required input for all components to perform
simulations. Whether the remaining component properties are needed or not depends on the
simulation to be performed.
The below table shows what component properties are needed to calculate a given property for
gas and oil phases.
Physical or transport property
Volume
Density
Z factor
Enthalpy (H)
Entropy (S)
Heat capacity (CP)
Heat capacity (CV)
Kappa (CP/ CV)
JouleThomson coefficient
Velocity of sound
Viscosity
Thermal conductivity
Surface tension
*1)
Component type
Name
Critical temperature (Tc)
Critical pressure (Pc)
Acentric factor ( )
and
Molecular weight (M)
Missing Properties
PVTsim has a <Complete> option for estimating missing component properties for a fluid
composition entered in characterized form. The number of missing properties estimated depends
on the properties entered manually. It is assumed that Tc, Pc, ,
,
, and molecular weight
have all been entered. Below is shown what other properties are needed to estimate a given
missing property and a reference is given to the section in the Method Documentation where the
property correlation is described.
Property
Component properties
Liquid density
Normal boiling point
Weight average molecular
weight
Critical volume
Vapor pressure model
MathiasCopeman coefficients
Tindependent term of SRK
or PR Peneloux parameter
Tdependent term of SRK or
PR Peneloux parameter
Melting point depression
(
Enthalpy of melting (
PNA distribution
Wax fraction
Asphaltene fraction
Parachor
Hydrate former or not
Hydrate Langmuir constants
Number of ions in aqueous
solution (only for salts)
Number of crystal water
molecules per salt molecule
(only for salts)
Not estimated
Not estimated
for defined components.
Liquid density for pseudocomponents
Not estimated for defined
components. Liquid density
for pseudocomponents
Only for pseudocomponents.
Viscosity data for an
uninhibited/inhibited fluid.
Molecular weight
Not estimated for defined
components. Liquid density
for pseudocomponents
Irrelevant for defined
components. None for pseudocomponents
Irrelevant for defined
components. None for pseudocomponents
Irrelevant for defined
components. Liquid density
for pseudocomponents
Irrelevant for defined
components. None for pseudocomponents.
Irrelevant for defined
components. Liquid density
for pseudocomponents
Not estimated for defined
components. Liquid density
for pseudocomponents
Not estimated
Not estimated
Not estimated
Not estimated
Composition Handling
Composition Handling
PVTsim distinguishes between the following fluid types
Characterized compositions
Compositions with plus fraction are compositions as reported by PVT laboratories where the last
component is a plus fraction residue. For this type of compositions the required input is mol%s
of all components and molecular weights and densities of all C7+ components (carbon number
fractions).
Compositions with No plus fraction require the same input as compositions with a plus fraction.
In this case the heaviest component is not a residue but an actual component or a boiling point
cut and no extrapolation is performed. Gas mixtures with only a marginal content of C7+
components are usually classified as compositions with No plus fraction.
In the simulations characterized compositions are used. These are usually generated from a Plus
fraction or No plus fraction type of composition. They may alternatively be entered manually.
heavy ends may be lost in the analysis, especially heavy aromatics such as asphaltenes. The main
problem with a GC analysis is however that no information is retained on molecular weight (M)
and density of the cuts above C6. These are instead estimated from correlations. This in particular
is a problem for the plus fraction residue properties, which are essential for a proper
representation of the heaviest constituents of the fluid. To remedy this problem a GC
composition may be entered into PVTsim as follows.
Often a set of residue properties is available say for the C7+ fraction, while the measured GC
composition often extends to e.g. C30. In this case one may enter the mol%'s to C30 together with
the M and density of the total C7+ fraction leaving the M and density fields blank for the higher
C8  C30 fractions. With this input, the program will be extrapolating from the C7+ fraction
properties, while honoring the reported composition for the fractions up to C30 under the mass
balance constraints. If no information is available on the residue properties, one may as an
alternative lump back the composition to C7+ and estimate properties from there, which will often
provide equally accurate simulation results as with the detailed GC composition.
TBP Distillation
The TBP distillation requires a larger sample volume, typically 50 200 cc and is more time
consuming. The method separates the components heavier than C6 into fractions bracketed by the
boiling points of the normal alkanes. For instance, the C7 fraction refers to all species, which
distil off between the boiling point of nC6 + 0.5C/0.9F, and the boiling point of nC7 +
0.5C/0.9F, regardless of how many carbon atoms these components contain. Each of the
fractions distilled off is weighed and the molecular weights and densities are determined
experimentally. The density and molecular weight in combination provide valuable information
to the characterization procedure on the PNA distribution. Aromatic components for instance
have a higher density and a lower Mw than paraffinic components. The residue from the
distillation is also analyzed for amount, M and density. These properties are important in the
characterization procedure.
Whenever possible, it is recommended that input for PVTsim is generated based on a TBP
analysis. The accuracy of the characterization procedure relies on good values for densities and
molecular weights of the C7+ fractions. Parameters such as the Peneloux volume shift for the
heavier pseudocomponents are estimated based on the input densities, and consequently the
quality of the input directly affects the density predictions of the equation of state (EOS) model.
While the default values in PVTsim are generally considered to be reasonably accurate, they can
never be expected to match the characteristics on any given crude exactly, and thus experimental
values are much to be preferred.
accordingly. This procedure ensures that discrepancies between different component classes are
avoided in the characterization.
Mixing
PVTsim may be used to mix or weave from 2 to 50 fluid compositions. A mixing will not
necessarily retain the pseudocomponents of the individual compositions. Averaging the
properties of the pseudocomponents in the individual compositions generates new pseudocomponents. Mixing may be performed on all types of compositions. For fluids characterized in
PVTsim mixing is done on the level where the fluid has been characterized but not yet lumped.
Each set of discrete fractions is mixed and the properties of the mixed fraction averaged on a
mass basis. Afterwards the mixed fluid is lumped to the specified number of components. If the
total number of C7+ components in the fluids to be mixed exceeds the defaults number of pseudocomponents (12), pseudocomponents of approximately the same weight are lumped to get down
to the desired number of pseudocomponents in the mixed fluid.
Weaving
Weaving will maintain the pseudocomponents of the individual compositions and can only be
performed for characterized compositions. When weaving two fluids, all pseudocomponents
from all the original fluids are maintained in the resulting weaved fluid. This may lead to several
components having the same name, and it is therefore advisable to tag the component names in
order to avoid confusion later on. The weaving option is useful to track specific components in a
process simulation or for allocation studies.
Recombination
Recombination is a mixing on volumetric basis performed for a given P and T (usually separator
conditions). Recombination can only be performed for two compositions, an oil and a gas
composition. The recombination option is often used to combine a separator gas phase and a
separator oil phase to get the feed to the separator. When the two fluids are recombined, the GOR
and liquid density at separator conditions must be input. Alternatively the saturation point of the
recombined fluid can be entered along with the liquid density. When the GOR is specified, the
program determines the number of mols corresponding to the input volumes and simply mixes
the two fluids based on this. When the saturation pressure is specified, the recombination is
iterative (i.e. how much of the gas should be added to yield this saturation pressure).
fashion as the mixing operation with the only difference that the same pseudos logic keeps track
of the molar amount of each pseudocomponent contained in each individual fluid.
The characterization to the same pseudocomponents option is a very powerful tool, and can be
applied for a number of tasks. In compositional pipeline simulations where different streams are
mixed during the calculations or in compositional reservoir simulations where zones with
different PVT behavior are considered, mixing is straightforward when all fluids have the same
pseudocomponents. It is furthermore possible to do regression in combination with the
characterization to the same pseudos, in which case one may put special emphasis on fluids for
which PVT data sets are available. In this case the data sets will also affect the characterization of
the fluids for which no PVT data exist.
Characterization to same pseudocomponents is described in more detail in the section of
Characterization of Heavy Hydrocarbons.
Flash Algorithms
Flash Algorithms
The flash algorithms of PVTsim are the backbone of all equilibrium calculations performed in
the various simulation options. The terminology behind the different flash options are described
in the following.
PT Flash
The input to a PT flash calculation consists of
Number of phases
Amounts and molar compositions of each phase
Compressibility factor (Z) or density of each phase
Flash Algorithms
PVTsim makes use of the following flash algorithms
Specific PT flash options considering the appropriate solid phases are used in the hydrate, wax,
and asphaltene options.
A flash calculation assumes thermodynamic equilibrium. The thermodynamic models available
in PVTsim are the SoaveRedlichKwong (SRK) equation of state, the PengRobinson (PR)
equation of state, and the PengRobinson 78 (PR78) equation of state. These equations are
presented in Equation of State section. To apply an equation of state, a number of properties are
needed for each component contained in the actual mixture. These are established through a C7+characterization as outlined in the section on Characterization of Heavy Hydrocarbons.
PVTsim uses the PT flash algorithms of Michelsen (1982a, 1982b). They are based on the
principle of Gibbs energy minimization. In a flash process a mixture will settle in the state at
which its Gibbs free energy
N
G = n i i
i =1
N
G
G1 = G 0 i i
i =1
n i n
,,
i =1
i =1
G = G I + G II G 0 = i (( i )II ( i )0 ) = yi (( i ) II ( i ) 0 )
, and yi is the mol fraction of component i in phase II. The subindices 0 and II
where
refer to the single phase and to phase II, respectively. Only one phase is formed if
is greater
than zero for all possible trial compositions of phase II. The chemical potential,
expressed in terms of the fugacity, fi, as follows
, may be
i = i0 + RT1n f i = i0 + RT(1n z i + ln i + 1n P )
is a standard state chemical potential, a fugacity coefficient, z a mol fraction, P the
where
pressure, and the subindex i stands for component i. The standard state is in this case the pure
component i at the temperature and pressure of the system. The equation for
may then be
rewritten to
G N
= y i (1n y i +1n( i ) II ln z i 1n( i )0 )
RT i =1
where zi is the mol fraction of component i in the total mixture. The stability criterion can now be
expressed in terms of mol fractions and fugacity coefficients. Only one phase exists if
N
y (ln y + ln( )
i =1
i II
ln z i ln( i ) 0 ) > 0
for all trial compositions of phase II. A minimum in G will at the same time be a stationary point.
A stationary point must satisfy the equation
ln yi + ln( i ) II lnzi ln( i )0 = k
where k is independent of component index. Introducing new variables, Yi, given by
ln Yi = ln yi k
the following equation may be derived
1n Yi = 1n zi + 1n(
)0 1n(
)II
PVTsim uses the following initial estimate for the ratio Ki between the mol fraction of
component i in the vapor phase and in the liquid phase
Ki =
Pci
T
exp 5.42 (1 ci )
P
T
where
Ki= yi/xi
and Tci is the critical temperature and Pci the critical pressure of component i. As initial estimates
for Yi are used Kizi, if phase 0 is a liquid and zi/Ki, if phase 0 is a vapor. The fugacity
coefficients, ( )II, corresponding to the initial estimates for Yi are determined based on these
fugacity coefficients, new Yivalue are determined, and so on. For a singlephase mixture this
direct substitution calculation will either converge to the trivial solution (i.e. to two identical
phases) or to Yivalues fulfilling the criterion
N
Y 1
i =1
yi + (1 )x i = zi ,
(i = 1,2,3,..., N )
Equilibrium equations
yi iV = x i iL ,
(i = 1,2,3,..., N )
(y x ) = 0, (i = 1,2,3,..., N )
i =1
In these equations xi, yi and zi are mol fractions in the liquid phase, the vapor phase and the total
mixture, respectively. is the molar fraction of the vapor phase.
and
are the fugacity
coefficients of component i in the vapor and liquid phases calculated from the equation of state.
There are (2N + 1) equations to solve with (2N + 3) variables, namely (x1, x2, x3,, xN), (y1, y2,
y3,.,yN), , T and P. With T and P specified, the number of variables equals the number of
equations. The equations can be simplified by introducing the equilibrium ratio or Kfactor, Ki =
yi/xi. The following expressions may then be derived for xi and yi
xi =
zi
,
1 + (K i 1)
(i = 1,2,3,..., N )
(i = 1,2,3,..., N )
yi = K i x i ,
and for Ki
iL
Ki = V ,
i
(i = 1,2,3,..., N )
The above (2N+1) equations may then be reduced to the following (N+1) equations
ln K i =
ln iL
,
ln iV
(i = 1,2,3,..., N )
N
(y x ) = z (K
i
i =1
1)/(1 + (K i 1)) = 0
For a given total composition, a given (T, P) and Ki estimated from the stability analysis, an
estimate of
may be derived. This will allow new estimates of xi and yi to be derived and the K
factors to be recalculated. A new value of is calculated and so on. This direct substitution
calculation may be repeated until convergence. For more details on the procedure it is
recommended to consult the articles of Michelsen (1982a, 1982b).
For a system consisting of J phases the mass balance equation is
zi (K im 1)
=0
Hi
i =1
N
where
j 1
H i = 1 + m (K im 1)
m =1
y im =
z i K im
,
Hi
y iJ =
zi
,
Hi
where
(i = 1,2,3,..., N; m = 1,2,3,..., J )
(i = 1,2,3,..., N )
and
g1 = z i (K i 1) i
i =1
g 2 = H H spec
where
i = 1 + (K i 1)
H is total molar enthalpy for the estimated phase compositions, and Hspec is the specified molar
enthalpy. At convergence both g1 and g2 are zero. The iteration procedure is described in
Michelsen (1986).
Other flash specifications are VT, UV and HS. V is the molar volume and T the absolute
temperature. A VT specification is useful to for example determine the pressure in an offshore
pipeline during shutdown. U is the internal energy. A dynamic flow problem may sometimes
more conveniently be expressed in U and V than in P and T.
Phase Identification
If a PT flash calculation for an oil or gas mixture shows existence of two phases, the phase of the
lower density will in general be assumed to be gas or vapor and the phase of the higher density
liquid or oil. In the case of a singlephase solution it is less obvious whether to consider the
single phase to be a gas or a liquid. There exists no generally accepted definition to distinguish a
gas from a liquid. Since the terms gas and oil are very much used in the oil industry, a criterion is
needed for distinguishing between the two types of phases.
The following phase identification criteria are used in PVTsim
Liquid if
1.
The pressure is lower than the critical pressure and the temperature lower than the bubble
point temperature.
2.
The pressure is above the critical pressure and the temperature lower than the critical
temperature.
Gas if
1.
The pressure is lower than the critical pressure and the temperature higher than the dew
point temperature.
2.
The pressure is above the critical pressure and the temperature higher than the critical
temperature.
In the flash options handling water, a phase containing more than 80 mol% total of the
components water, hydrate inhibitors and salts is identified as an aqueous phase.
Other inorganic
Organic defined
Pseudocomponents
Water
Hydrate inhibitors
Other inorganic
Organic defined
Pseudocomponents
Salts
Water
Hydrate inhibitors
Other inorganic
Organic defined
Pseudocomponents
References
Michelsen, M.L., The Isothermal Flash Problem. Part I: Stability, Fluid Phase Equilibria 9,
1982a, 1.
Michelsen, M.L., The Isothermal Flash Problem. Part II: PhaseSplit Calculation, Fluid Phase
Equilibria 9, 1982b, 21.
Michelsen, M.L., Multiphase Isenthalpic and Isentropic Flash Algorithms, SEP Report 8616,
Institut for Kemiteknik, The Technical University of Denmark, 1986
equals 1
Pci
T
exp 5.42(1 ci )
P
T
(y
i =1
x i ) = z i (K i 1)/(1 + (K i 1)) = 0
i =1
are solved for T and equal to the specified vapor mol fraction. The correct value of T is
subsequently calculated by solving this equation in conjunction with
lnK i =
ln iL
ln iV
where the liquid (L) and vapor (V) phase fugacity coefficients,
of state.
An initial estimate of the second point on the phase envelope is calculated using the derivatives
of T and Ki with respect to P calculated in the first point. The correct solution is again found by
solving the above equations.
From the third point and on the extrapolation is based on the two latest calculated points and the
corresponding derivatives. This stepwise calculation is continued until the temperature is below
the specified lower temperature limit.
In simulations of PVT experiments, knowledge of the complete phase envelope is not needed but
only the saturation pressure at the temperature of the experiment. A saturation point is also
located through a phase envelope calculation. A critical point may be considered a special type of
saturation point, and the critical point is easily identified as a point where the lnKi changes sign.
Some fluids have more than one critical point. The critical point is furthermore verified by a
more direct method as described by Michelsen and Heideman (1981).
The basic phase envelope option only considers two phases (one gas and one liquid). For many
reservoir fluid mixtures a PTregion exists with 3 phases (1 gas and 2 liquids). This is for
example often the case for gas condensate mixtures at low temperatures. The phase envelope
option in PVTsim allows a check to be performed of the possible existence of a 3 phase region.
For fluids with no aqueous components (i.e. water, hydrate inhibitors or salts) it is possible to
obtain other phase envelope diagrams than the traditional PTphase envelope diagram. PVTsim
allows combinations of the following properties on the axes of the phase envelope diagram
Pressure (P)
Temperature (T)
Enthalpy (H)
Entropy (S)
Volume (V)
Internal Energy (U)
Other inorganic
Organic defined
Pseudocomponents.
Water (no inner lines)
Hydrate inhibitors (no inner lines)
The saturation point algorithm used in the saturation point option and the PVT simulations is also
based on the phase envelope algorithm, but does not handle water and hydrate inhibitors.
References
Lindeloff, N. and Michelsen, M.L., Phase Envelope Calculations for HydrocarbonWater
Mixtures, SPE 77385, SPE ATCE in San Antonio, Tx, September 29 October 2, 2002.
Michelsen, M.L., Calculation of Phase Envelopes and Critical Points for Multicomponent
Mixtures, Fluid Phase Equilibria, 1980, 4, pp. 110.
Michelsen, M.L. and Heidemann, R.A., Calculation of Critical Points from Cubic TwoConstant
Equations of State, AIChE J., 27, 1981, pp. 521523.
Equations of State
Equations of State
The phase equilibrium calculations in PVTsim are based on one of the following equations
All equations may be used with or without Peneloux volume correction (Peneloux et al., 1982). A
constant or a temperature dependent Peneloux correction may be used. The temperature
dependent volume correction is determined to comply with the ASTM 125080 correlation for
volume correction factors for stable oils (Pedersen et al., 2002).
SRK Equation
The SRK equation takes the form
P=
RT
a(T)
V b V(V + b)
where P is the pressure, T the temperature, V the molar volume, R the gas constant and a and b
are equation of state parameters, which for a pure component are determined by imposing the
critical conditions
P
2P
(( )T = ( 2 )T = 0)crit. point
V
V
The following relation is then obtained for parameter a of component i at the critical point
a ci = a
R 2 Tci2
Pci
bi = b
R Tci
Pci
where
a = 0.42748
b = 0.08664
Tci is the critical temperature of component i and Pci the critical pressure. Values for Tc, Pc and
may be seen from the PVTsim pure component database. All the values except those for salts are
taken from Reid et al. (1977). The values for the salts are chosen to ensure that these components
remain in the aqueous phase (Srensen et al., 2002).
The temperature dependence of the aparameter is expressed in the form of a term ai(T), which
multiplied with aci gives the final expression for the aparameter of the SRKequation
ai(T) = aci i(T)
The parameter
T 0.5
i (T) = 1 + m1
Tc
where
m i = 0.480 +1.574i 0.176i2
It is seen that i(T) equals 1 at critical temperature at which ai(T) therefore becomes equal to aci.
is the acentric factor that is defined as follows (Pitzer, 1955)
i = log10 PriVap
Tr = 0.7
where
is the reduced vapor pressure of component i (vapor pressure divided by critical
pressure).
An alternative temperature dependence as suggested by Mathias and Copeman (1983) may be
applied
(T) = (1+ C1 1 Tr + C2 1 Tr
) + C (1
2
Tr ) 2 , Tr < 1
M&C is not used default in PVTsim, but is it possible for the user to change temperature
dependence from classical to M&C and to enter M&C coefficients (C1, C2 and C3) when these
are not given in the PVTsim database. The M&C coefficients used in PVTsim are from Dahl
(1991).
RT
a
V b (V + c )(V + b + 2c )
V =V c
The b parameter in the Peneloux equation
follows
~
b= b c
The parameter c can be regarded as a volume translation parameter, and it is given by the
following equation
c = c + c (T 288.15)
where T is the temperature in K. The parameter c is the temperature independent volume
correction and c the temperature dependent volume correction. Per default the temperature
dependent volume correction c is set to zero unless for C+ pseudocomponents. In general the
temperature independent Peneloux volume correction for defined organics and other organics
is found from the following expression
c' = 0.40768
RTc
(0.29441 ZRA )
Pc
correlation and the Peneloux molar volume for the same temperature, where the Peneloux
volume is found assuming c=c.
PR/PR78 Equation
The PR/PR78 equations both take the form
P=
RT
a(T)
V b V(V + b ) + b(V b )
where
a(T) = ac
a c = a
(T)
R 2Tc2
Pc
T 0.5
(T) = 1 + m1
Tc
R Tc
b = b
Pc
where
a = 0.45724
b = 0.07780
The parameter m is for the PR equation found from
m = 0.37464 + 1.54226
 0.26992
(1.48503 0.164423
+ 0.01666
The MathiasCopeman temperature dependence presented in the SRK section may also be
applied with both the PengRobinson equation and the PengRobinson 78 equation.
RT
a(T)
V b (V + c )(V + 2c + b ) + (b + c )(V b )
where c is a temperature dependent constant as presented in the SRK section. In general the
temperature independent Peneloux volume correction for defined organics and other organics
is found from
c' = 0.50033
RTc
(0.25969 Z RA )
Pc
where ZRA is defined as for the Peneloux modification of the SRK equation. For other
components c is found as explained in the SRK section, which also explains how to determine
the temperature dependent term c.
a = z i z ja ij
i =1 j =1
b = z i bi
i
c = z i ci
i
The parameter kij is a binary interaction coefficient, which by default is zero for hydrocarbonhydrocarbon interactions and different from zero for interactions between a hydrocarbon and a
nonhydrocarbon and between unlike pairs of nonhydrocarbons.
The greater part of the interaction coefficients in the PVTsim database has been found in Knapp
et al. (1982). The interaction coefficients between hydrogen and other components have been
found using the Tccorrelation of Tsonopoulos and Heidman (1986). It is given by
BX3
k ij =A +
1 + X3
T 50
X = cj
;
1000 Tcj
50 Tcj 1000
where Tcj is the critical temperature of the component interacting with hydrogen.
The values of A and B are given in Tsonopoulos and Heidman (1986) for the SRK and PR/PR78
equations of state
SRK
PR/PR78
A
0.0067
0.0736
B
0.63376
0.58984
Furthermore the option exists to calculate interaction parameters from critical volumes using the
following equation (Chueh and Prausnitz, 1967)
1
1
2 V3 V3
ci
cj
k ij =1
1
1
Vci3 + Vcj3
where
is specific for the selected equation of state. For SRK and PR the values for
are
SRK : = ln2
PR : =
2 + 1
ln
2 2 2 1
1
G E is the excess Gibbs energy at infinite pressure. G E is found using a modified NRTL mixing
rule
N
G E N
= zi
RT i = l
j =1
N
ji
b jz jexp( ji ji )
b z exp(
k =1
k k
ki ki
is a nonrandomness parameter, i.e. a parameter for taking into account that the mol
where
fraction of molecules of type i around a molecule of type j may deviate from the overall mol
fraction of molecules of type i in the mixture. When
is zero, the mixture is completely
random. The parameter is defined by the following expression
ji =
g ji g ii
RT
where gji is an energy parameter characteristic of the ji interaction. In PVTsim the gparameters
are temperature dependent and given by the expression (Pedersen et al., 2001)
gji gii = (gji gii) + T (gji gii)
The parameter b entering into the expression for
is the bparameter of the equation of state.
The classical mixing rule is used for the bparameter.
The local composition of a binary pair that can be described using the classical mixing rule, will
not deviate from the overall composition, i.e.
should be chosen equal to zero. By further
selecting the following expressions for the interaction energy parameters
g ii =
ai
bi
g ji = 2
bi b j
bi + b j
(g
g jj ) (1 k ij )
0.5
ii
the H&V mixing rule reduces to the classical one. When the H&V mixing rule is used, the latter
expressions are therefore used for gij and gii of binary pairs not requiring the advanced mixing
rule. This gives a continuous description of both hydrocarbons and aqueous components. The
H&V mixing rule can for PVTsim version 13 and onwards be used both with the SRK and PR
equations of state.
((P/n )
i T, V, n j
RT/V dV lnZ
where ni is the number of mols of type i. Subsequently when the SRK equation is used, the
following relation can be derived for the fugacity coefficient
N
0.5
ln i = (b i /b )(Z 1) lnZ + ln [V/(V b )] + a/b RT b i /b 2 z i (1 k ij )(a i a j ) /a ln[(V + b ) / V]
j=1
With two phases present, the phase compositions are related to the total composition as follows
xi =
zi
1+ (K i 1)
yi =
K i zi
1 + (K i 1)
For details on how to determine the number of phases and on how to determine the amounts of
each phase, the P/T flash section should be consulted.
References
Chueh, P.L., and Prausnitz, J.M., VaporLiquid Equilibrium at High Pressures: Calculation of
Partial Molar Volumes in NonPolar Liquid Mixtures, AIChE J 6, 13, 1967, pp. 1099.
Dahl, S., Phase Equilibria for Mixtures Containing Gases and Electrolytes, Ph.D. thesis,
Department of Chemical Engineering, Technical University of Denmark, 1991.
Huron, M.J. and Vidal, J., New Mixing Rules in Simple Equations of State for Representing
Vaporliquid Equilibria of Strongly NonIdeal Mixtures, Fluid Phase Equilibria 3, 1979, p. 255.
Knapp H.R., Doring, R., Oellrich, L., Plocker, U., and Prausnitz, J.M., VaporLiquid Equilibria
for Mixtures of Low Boiling Substances, Chem. Data. Ser., Vol. VI, 1982, DECHEMA.
Mathias, P.M. and Copeman, T.W., Extension of the PengRobinson Equation of State to
Complex Mixtures: Evaluation of the various Forms of the Local Composition Concept, Fluid
Phase Equilibria 13, 1983, pp. 91108.
Pedersen, K.S., Milter, J., and Rasmussen, C.P., Mutual Solubility of Water and Reservoir
Fluids at High Temperatures and Pressures, Experimental and Simulated Phase Equilibrium
Data, Fluid Phase Equilibria 189, 2001, pp. 8597.
Pedersen, K.S., Milter, J. and Srensen, H., Cubic Equations of State Applied to HT/HP and
Highly Aromatic Fluids, SPE 77385, SPE ATCE in San Antonio, Tx, September 29October 2,
2002.
Peneloux, A., Rauzy, E. and Frze, R., A Consistent Correlation for RedlichKwongSoave
Volumes, Fluid Phase Equilibria, 8, 1982, pp. 723.
Peng, D.Y. and Robinson, D.B., A New TwoConstant Equation of State, Ind. Eng. Chem.
Fundam., 15, 1976, pp. 5964.
Peng, D.Y., and Robinson, D.B., The Characterization of the Heptanes and Heavier Fractions
for the GPA PengRobinson Programs, GPA Research Report RR28, 1978.
Pitzer, K. S., Volumetric and Thermodynamic Properties of Fluids. I., Theoretical Basis and
Virial Coefficients, J. Am. Chem. Soc. 77, 1955, 3427.
Reid, R.C., Prausnitz, J.M. and Sherwood, J. K., The Properties of Gases and Liquids
McGrawHill, NewYork 1977.
Soave, G., Equilibrium Constants From a Modified RedlichKwong Equation of State, Chem.
Eng. Sci. 27, 1972, 1197.
Srensen, H., Pedersen, K.S. and Christensen, P.L., "Modeling of Gas Solubility in
Brine", Organic Geochemistry 33, 2002, pp. 35642.
Tsonopoulos, C., and Heidman, J.L., HighPressure VaporLiquid Equilibria with Cubic
Equations of State, Fluid Phase Equilibria 29, 1986, pp. 391414.
Characterization of Heavy
Hydrocarbons
Classes of Components
Naturally occurring oil and gas condensate mixtures consist of three classes of components
Defined Components
These are per default N2, CO2, H2S, C1, C2, C3, iC4, nC4, iC5 and C6 in PVTsim. C6 is in PVTsim
considered to be pure nC6.
C7+ Fractions
Each C7+ fraction contains hydrocarbons with boiling points within a given temperature interval.
Carbon number fraction n consists of the components with a boiling point between that of nCn1 +
0.5C/0.9F and nCn + 0.5C/0.9F. The C7 fraction for example consists of the components with
a boiling point between those of nC6 + 0.5C/0.9F and nC7 + 0.5C/0.9F . For the C7+fractions
the density at standard conditions (1 atm/14.969 psi and 15C/59F) and the molecular weight
must be input.
The Plus Fraction
The plus fraction consists of the components, which are too heavy to be separated in individual
C7+fractions. The average molecular weight and the density must be known.
Properties of C7+Fractions
PVTsim supports two different characterization procedures

Tc, Pc and
Tc = c1 + c2 1n M + c3 M + c4/M
lnPc = d1 + d2 d5 + d3/M + d4/M2
m = e1 + e2 M + e3 + e4 M2
where m is defined in the Equation of State section and the coefficients are given in the tables
below.
Standard characterization  SRK (Pedersen et al., 1989b and 1992)
Subindex/
1
2
3
4
Coefficient
c
1.6312 x 102 8.6052 x 10
4.3475 x 101 1.8774 x 103
1
d
1.3408 x 10
2.5019
2.0846 x 102 3.9872 x 103
e
7.4310 x 101 4.8122 x 103 9.6707 x 103 3.7184 x 106
Subindex/
Coefficient
c
d
e
Subindex/
Coefficient
c
d
e
Subindex/
Coefficient
c
d
e
7.3404 x 10
7.2846 x 102
3.7377 x 101
9.7356 x 10
2.1881
5.4927 x 103
6.1874 x 101
1.6391 x 102
1.1793 x 102
2.0593 x 103
4.0434 x 103
4.9305 x 106
3.04143 102
3.05081
4.96902101
3.80073 103
1.27154 104
5.24300106
3.26725102
2.68058
1.89723101
5.23447101
5.3227410
7.42901103
5.77248101
2.04507102
3.28795102
1.77498103
9.45434103
7.36151106
5
1.0
5
1/3
5
0.25
5
0.25
M is in g/mol, is in g/cm3, Tc is in K and Pc in atm. The correlations are the same with and
without volume correction.
Estimation of the molar distribution, i.e. mol fraction versus carbon number.
Estimation of the density distribution, i.e. the density versus carbon number.
Estimation of the boiling point distribution, i.e. boiling point versus carbon number
Estimation of the molecular weight distribution, i.e. molecular weight versus carbon number.
Calculation of Tc, Pc and of the resulting pseudocomponents.
0.04609
in g/cm3.
where TB is in K and
n=
1 + 2I
1 I
0.9182
TB is the boiling point in K and the liquid density at atmospheric conditions in g/cm3. Based on
the refractive index, the density and the molecular weight the PNA distribution (in mol%) can be
estimated as described by Nes and Westerns (1951)
v = 2.51 (n 1.4750) 
+ 0.8510
for v > 0
for v < 0
for w > 0
for w < 0
PVTsim default uses a weight based lumping where each lumped pseudocomponent contains
approximately the same weight amount and where Tc, Pc and of the individual carbon number
fractions and found as weight mean average values of Tc, Pc and of the individual carbon
number fractions. If the kth pseudocomponent contains the carbon number fractions M to L, its
Tc, Pc and will be found from the relations
L
z i M i Tci
Tck = i =ML
zi Mi
i=M
z i M i Pci
Pck = i =ML
zi Mi
i=M
z i M i i
ck = i = ML
zi Mi
i=M
where zi is the mol fraction and Mi the molecular weight of carbon number fraction i. The weight
based procedure ensures that all hydrocarbon segments of the C7+ fraction are given equal
importance.
ab Grouping
This represents a lumping scheme, which minimizes the variation in the equation of state
parameters a and b within a pseudocomponent (Lomeland and Harstad, 1994). The terms used in
the following are further explained in the Equation of State section. The aparameter may for a
pure component i be written
T
a i = a ci 1 + m i m i
Tci
or
a i = a1i 1 a 2i T
where
a1i = a ci (1 + mi )
mi
(1 + mi ) Tci
a 2i =
The expression for the parameter a of an N component mixture may similarly be rewritten to
a a a a (a + a )
a N N
= z i z j 1i 1j 1i 1j 2i 2j + a1i a1ja 2i a 2j (1 k ij )
T i =1 j=1
T
T
For pseudocomponent k comprising the carbon number fractions from Ln to Un the average a1
and a2 parameters are calculated by
z i z j a 1i a 1j (1 k ij )
Un
a lk2 =
Un
i = L n j= L n
Un z
i
i=Ln
z i z j a 1i (a 2i + a 2j )a 1j (1 k ij )
Un
2a 2k a 1k2 =
Un
i = L n j= L n
Un z
i
i=Ln
bk =
zi bi
i =Ln
Un
zi
i=Ln
a a a a
S = 1i lk + 2i 2k
n = Ls i = L n
a 1i a 2i
N pc
Un
bi bk
+
bi
by varying Ln and Un. Ls is lowest carbon number considered for grouping and Npc is the final
number of pseudocomponents. The parameters n, Tcn and Pcn are found by backcalculation
using the following formulas
b a 1k a 2k
a
bk
mk =
mk
Tck =
1 + m k a 2k
RT
Pck = b ck
bk
where m is a second order polynomial in as defined in the Equation of State section. In case
nonzero binary interaction coefficients are used for the hydrocarbonhydrocarbon interactions,
the binary interaction coefficient between pseudocomponent n and m is determined from the
formula
Un
k nm =
Um
z z M M k
i = L n j= L m
Un
Um
i=Ln
j= L m
ij
Mn Mm zi z j
where the pseudocomponent m comprises the carbon number fractions for Lm to Um and
are the average molecular weights of pseudocomponents m and n, respectively. For
and
interactions with methane the following correction term is to be added to the binary interaction
parameters calculated from the above formula
(M
Mm )
where
Mn Mm
n
Delumping
In compositional reservoir simulations it is desirable to use as few components as possible in
order to minimize the computation time. This is accomplished by component lumping. Not only
C7+ components but also some of the defined components will usually have to be lumped. In
subsequent process simulations it may be desirable to reestablish all the defined components and
possibly also to increase the number of C7+ pseudocomponents. This may in PVTsim be
accomplished by use of the Delumping Option. A lumped component consisting of defined
components is split into its constituents. The relative molar amounts of the individual
components are assumed to be the same as in the original composition before lumping. The C7+
pseudocomponents of the lumped fluid are possibly split to cover smaller carbon number ranges.
To start with the C7+ pseudocomponent containing the largest weight fraction is split into two
new pseudocomponents of approximately equal weight amounts. Next the pseudocomponent
which now contains the largest weight amount is split into two and so on until the number of C7+
pseudocomponents equals that specified.
It is possible to adjust the gas/oil ratio of the delumped composition to match that of the lumped
composition.
Tciunique =
Wgt( j)z T
j j
i ci
j =1
NFL
Wgt( j)z
j=1
NFL
Pciunique =
j
i
Wgt ( j)z P
j j
i ci
j=1
NFL
Wgt ( j)z
j=1
NFL
imix =
Wgt( j)z
j =1
NFL
j
i
Wgt( j)z
j =1
j
i
j
i
z iunique =
Wgt( j)z
j =1
NFL
j
i
Wgt( j)
j =1
Wgt ( j)z i M i
NFL
unique
i
j=1
Wgt ( j)z i
NFl
j=1
This composition is now treated like an ordinary composition to be lumped into pseudocomponents. The lumping determines the carbon number ranges to be contained in each pseudocomponent, and Tc, Pc and of each pseudocomponent. The properties of the lumped
composition are assumed to apply for all the individual compositions. If the kth pseudocomponent contain the carbon number fractions M to L, the mol fraction of this pseudocomponent in the jth composition will be
L
z kj = z ij
i=M
References
Katz, D.L. and Firoozabadi, A., Predicting Phase Behavior of Condensate/CrudeOil Systems
Using Methane Interaction Coefficients, J. Pet. Technol. 20, 1998, pp. 16491655.
Lomeland F. and Harstad, O., Simplifying the Task of Grouping Components in Compositional
Reservoir Simulation, SPE paper 27581, presented at the European Petroleum Computer
Conference in Aberdeen, U.K., 1517 March, 1997.
Nes, K. and Westerns, H.A., van, Aspects of the Constitution of Mineral Oils, Elsevier, New
York, 1951.
Pedersen, K.S., Thomassen, P. and Fredenslund, Aa., Thermodynamics of Petroleum Mixtures
Containing Heavy Hydrocarbons. 3. Efficient Flash Calculation Procedures Using the SRK
Equation of State, Ind. Eng. Chem. Process Des. Dev. 24, 1985, pp. 948954.
Pedersen, K.S. , Fredenslund, Aa. and Thomassen, P., Properties of Oils and Natural Gases,
Gulf Publishing Inc., Houston, 1989a.
Pedersen, K.S., Thomassen, P. and Fredenslund, Aa., Advances in Thermodynamics 1, 1989b,
137.
Pedersen, K.S., Blilie, A. and Meisingset, K.K., "PVT Calculations of Petroleum Reservoir
Fluids Using Measured and Estimated Compositional Data for the Plus Fraction", Ind. Eng.
Chem. Res. 31, 1992, pp. 924932.
Pedersen, K.S., Milter, J. and Srensen, H., Cubic Equations of State Applied to HT/HP and
Highly Aromatic Fluids, SPE 77385, SPE ATCE in San Antonio, Tx, September 29October 2,
2002.
Riazi, M.R. and Daubert, T.E., Prediction of the Composition of Petroleum Fractions, Ind.
Eng. Chem. Process Des. Dev. 19, 1980, pp. 289294.
SRK
SRKPeneloux
SRKPeneloux(T)
PR
PRPeneloux
PRPeneloux(T)
PR78
PR78Peneloux
PR78Peneloux(T)
where (T) means that the Peneloux volume translation parameter is temperature dependent.
Enthalpy
The enthalpy, H, is calculated as the sum of two contributions, the ideal gas enthalpy and residual
enthalpy, Hres
N
H = z i H idi + H res
i =l
where N is the number of components, zi is the mol fraction of component i in the phase
considered and
H idi =
Tres
Cidpi dT
CF
)104
)107
C4 = 0
where
CF = ((12.8 K)(10K)/(10 ))2
and K is the Watson characterization factor defined as
K = TB1/3 /SG
TB is the normal boiling point in R and SG the specific gravity, which is approximately equal to
the liquid density in g/cm3.
For hydrocarbons with a molecular weight above 300,
factors below 0.1 are replaced by = 0.1.
is replaced by 1.0 if
< 1. Acentric
The residual term of H is derived from the equation of state using the following general
thermodynamic relation
H res = RT 2
ln
T
where is the fugacity coefficient of the mixture and the derivative is for a constant
composition.
Internal Energy
The internal energy, U, is calculated as U = H PV. Where H is the enthalpy, P the pressure and
V the molar volume.
Entropy
The entropy is calculated as the sum of two contributions, the ideal gas entropy and residual
entropy
N
S = z iSidi + Sres
i =1
S =
id
i
Tref
Cidpi
P
dT Tln
R ln z i
T
Pref
H ref
R ln
T
Heat Capacity
The heat capacity at constant pressure is calculated from
H
CP =
T P
and the heat capacity at constant volume from
V P
C V = C P T
T P T V
where the derivatives are evaluated using the equation of state. H is the enthalpy, T the
temperature, P the pressure and V the molar volume.
JouleThomson Coefficient
The JouleThomson coefficient is defined as the pressure derivative of the temperature for
constant enthalpy. It is derived as follows
1 H
T
jT = =
Cp P T
P H
Velocity of sound
The velocity of sound is derived as
u sonic =
V
V
P
=
MW V S
MW
CP P T
CV T V V P
where M is the molecular weight and the derivatives are evaluated using the equation of state.
References
Kesler, M.G. and Lee, B.I., Improve Prediction of Enthalpy of Fractions, Hydrocarbon
Processing 55, 1976, 153.
Reid, R.C., Prausnitz, J. M. and Sherwood, J.K., The Properties of Gases and Liquids.
McGrawHill, NewYork 1977.
Transport Properties
Transport Properties
Viscosity
Corresponding States Method
The viscosity calculations in PVTsim are default based on the corresponding states principle in
the form suggested by Pedersen et al. (1984, 1987).
The idea behind the corresponding states principle is that the relation between the reduced
viscosity
r = f (Pr , Tr )
If the function f is known for one component (a reference component) within the group it is
possible to calculate the viscosity at any (P,T) for any other component within the group. The
viscosity of component x at (P,T) is for example found as follows
T
x (Pr , Tr ) = cx
Tco
1/6
2/3
1/2
Pcx M x
PPco TTco
,
o
Pco M o
Pcx Tcx
1/6
(P
c, mix
/Pco )
2/3
(M
/M o ) ( mix / o ) o (Po , To )
1/2
mix
where
Po =
TTco o
Tc, mix mix
The critical temperature and the critical molar volume for unlike pairs of molecules (i and j) are
found using the below formulas
To =
TTco o
Tc, mix mix
1 1/3
Vci + Vcj1/3
8
The critical molar volume of component i may be related to the critical temperature and the
critical pressure as follows
Vci =
RZ ci Tci
Pci
where Zci is the compressibility factor of component i at the critical point. Assuming that Zc is a
constant independent of component, the expression for Vcij may be rewritten to
Vcij =
1
constant
8
Tc, mix =
z z T
i =1 j =1
N N
cij
Vcij
z z V
i =1 j =1
cij
where zi and zj are mol fractions of components i and j, respectively and N the number of
components. This expression may be rewritten to
3
T 1/3 T 1/3
1/2
cj
z i z j ci + [Tci Tcj ]
i =1 j=1
Pci
Pcj
Tc,mix =
3
T 1/3 T 1/3
N N
cj
z i z j ci +
i =1 j=1
Pci
Pcj
N N
For the critical pressure of a mixture, Pc,mix, the following relation is used
Pc,mix = constant Tc,mix / Vc,mix
Where Vc,mix is found as follows
N
T 1/3 T 1/3
8 z i z j ci + cj Tci Tcj
P
Pci
i =1 j =1
cj
Pc, mix =
2
1/3 1/3
N N
1/3
Tcj
Tci
z i z j +
Pci
i =1 j=1
Pcj
1/2
The applied mixing rules are those recommended by Mo and Gubbins (1976).
The mixture molecular weight is found as follows
2.303
2.303
M mix =1.30410 4 M w M n
and
where
respectively.
N
i =1
j=1
)+ M
M w = z i M i2 / z i M i
N
M n = zi Mi
i =1
The constants 1.304104 and 2.303 in the above equation are the main tuning parameters applied
when performing regression on data with the CSP model. The viscosity correction factors
referred to in the regression output are multiplication factors for these coefficients, i.e. with a
default value of 1. The first viscosity correction factor is multiplied onto the coefficient
1.304104 while the second viscosity correction factor is multiplied onto the exponent 2.303.
This is further described in the section about Regression to Experimental PVT Data.
The parameter
is defined as
TT PP
o co , co
Tc, mix Pc, mix
r =
co
The reference viscosity correlation is based on the methane viscosity model of Hanley et al.
(1975)
' (, T ) = o (T ) + 1 (T ) + ' (, T )
are functions defined in the above reference. The methane density is found
where
using the BWRequation in the form suggested by McCarty (1974). In the dense liquid region
this expression is mainly governed by the term
j3
j
j
0.5
' (, T ) = exp ( j1 + j4 / T ) exp 0.1 j2 + 2/3
+ j5 + 6 + 72 1.0
T
T T
In the work of Hanley (1975) the coefficients j1 j7 have the following values (viscosities in P)
j1 = 10.3506
j2 = 17.5716
j3 = 3019.39
j4 = 188.730
j5 = 0.0429036
j6 = 145.290
j7 = 6127.68
is given by
= ( c )/ c
The presented viscosity calculation method presents some problems when methane is in a solid
form at its reference state. This is the case when the reduced temperature is below approximately
0.4. This problem is overcome by replacing
Fredenslund (1987))
k3
k k
0.5
' ' (, T ) = exp (k1 + k 4 /T ) exp 0.1 k 2 + 2/3
+ k 5 + 6 + 72 1.0
T
T T
with
k1 = 9.74602
k2 = 18.0834
k3= 4126.66
k4 = 44.6055
k5 = 0.9676544
k6 = 81.8134
k7= 15649.9
Continuity between viscosities above and below the freezing point of methane is secured by
as a fourth term in the viscosity expression
introducing
HTAN + 1
2
F2 =
1 HTAN
2
HTAN =
exp(T ) exp( T )
exp(T ) + exp( T )
with
T = T TF
where TF is the freezing point of methane.
LohrenzBrayClark (LBC) Method
[( ) +10 ]
*
4 1/4
=a1 + a 2 r + a 3 2r + a 4 3r + a 5 4r
where
a1 = 0.10230
a2 = 0.023364
a3 = 0.058533
a4 = 0.040758
a5 = 0.0093324
* is the lowpressure gas mixture viscosity. is the viscosityreducing parameter, which for a
mixture is given by the following expression:
= z i Tci
i=1
1/6
zM
i=1 i i
1/2
zP
i =1 i ci
2/3
where N is the number of components in the mixture and zi the mol fraction of component i.
The critical density,
c = (Vc )
= (z i Vci )
i =1
For C7+ fractions the critical volume in ft3/lb mol is found from
Vc = 21.573 + 0.015122 M 27.656
+ 0.070615 M
In this expression, M is the molecular weight and the liquid density in g/cm3. For defined
components literature values are used for the critical volumes.
If the composition has been entered in characterized form and densities are not available, the
critical volume is calculated from a correlation of Riedel (1954)
Vc =
RTc
[3.72 + 0.26( c 7.0)]1
Pc
Tb
lnPc
c = 0.90761.0 + c
Tb
Tc
If the normal boiling point is not available, the critical volume is calculated from the following
correlation (Reid et al., 1977)
Vc =
* =
z MW
i =1
N
*
i i
1/2
i
z MW
i =1
1/2
i
The following expressions (Stiel and Thodos, 1961) are used for the dilute gas viscosity of the
individual components,
1 0.94
Tri , Tri < 1.5
i
1
5/8
*i =17.7810 5 (4.58 Tri 1.67 ) , Tri > 1.5
*i = 34 10 5
where
i =
is given by
Tci1/6
M 1/2
Pci2/3
i
When performing tuning on the LBC viscosity model, either the critical volumes, the coefficients
a1a5 or both may be selected as tuning parameters. The ability to tune the coefficients makes the
LBC model extremely flexible, but if no data are available the CSP model generally provides
better predictions.
For fluids containing solid wax particles, a nonNewtonian viscosity model may be applied as is
described in the Wax section.
Thermal Conductivity
The thermal conductivity is defined as the proportionality constant,
(Fouriers law)
dT
q =
dX
where q is the heat flow per unit area and (dT/dX) is the temperature gradient in the direction of
the heat flow.
The thermal conductivity is in PVTsim calculated using a corresponding states principle
(Christensen and Fredenslund (1980) and Pedersen and Fredenslund (1987)).
According to the corresponding states theory, the thermal conductivity can be found from the
expression
r = f (Pr , Tr )
where f is the same function for a group of substances obeying the corresponding states principle.
For the reduced thermal conductivity, r, the following equation is used
r (P, T ) = (P, T ) / [ Tc1/6 Pc2/3 M 1/2
Using simple corresponding states theory, the thermal conductivity of component x at the
temperature T and the pressure P may be found from the following equation
x (P, T ) = (Tcx / Tco )
1/6
(P
cx
/ Pco )
2/3
(M
/M o )
1/2
o (Po , To )
where Po = PPco/Pcx and To = TTco/Tcx and o is the thermal conductivity of the reference
substance at the temperature To and pressure Po. As is the case for viscosity, methane is used as
reference substance. However some corrections must be introduced as compared with the simple
corresponding states principle. The thermal conductivity of polyatomic substances (Hanley
(1976)) can be separated into two contributions, one due to transport of translational energy and
one due to transport of internal energy
=
tr
int
PVTsim uses the modification of Christensen and Fredenslund (1980), which only applies the
corresponding states theory to the translational term. A term int,mix is used to correct for the
deviations from the simple corresponding states model. The final expression for calculation of
the thermal conductivity of a mixture at the temperature, T, and the pressure, P, is the following
mix (P, T ) = (Tc,mix / Tco )
mix
1/6
(P
c, mix
/ Pco )
2/3
(M
mix
/ Mo )
1/2
where
T
P
To = T/ c,mix mix and Po = P/ c,mix mix
Tco co
Pco o
The mixture molecular weight Mmix is found from ChapmanEnskog theory as described by Mo
and Gubbins (1976)
M mix =
][
1
1/4
1/3
1/3
1/2
z i z j (1/M i +1/M j ) (Tci / Tcj ) / (Tci /Pci ) + (Tcj /Pcj )
i
j
16
]]
2 2
1/3
Tc,mix
Pc,4/3mix
where z are mol fractions and i and j component indices. The internal energy contributions to the
thermal conductivity, int,o (reference substance) and int,mix (mixture) are both given by
int = 1.18653 i (C idp 2.5R )f ( r )/M
i =1 j=1
0.5
which ensures that components having small values, i.e. small molecules, are attributed more
importance than those having larger values. The smaller molecules are more mobile than the
larger ones. They thereby contribute relatively more to the transfer of energy than the larger ones.
The calculation of the thermal conductivity of the reference substance, methane, is based on a
model of Hanley et al. (1975), which has the form
(, T ) = o (T ) + 1 (T ) + ' (, T ) + c (, T )
In the dense liquid region the major contribution to this expression comes from
, which
l3
l
l
0.5
' ' (, T ) = exp(l1 + l 4 /T )exp 0.1 l 2 + 2/3
+ l 5 + 6 + 7 1.0
T
T T2
where
l1= 8.55109
l2= 12.5539
l3= 1020.85
l4= 238.394
l5= 1.31563
l6= 72.5759
l7= 1411.60
1/4 = ( L [P ] i x i v [P ] i y i )
i =1
L and
are the molar densities in mol/cm3 (the density divided by the molecular weight) of
the oil and gas phases, respectively and xi and yi are the mol fractions of component i in the oil
and gas phases. The Parachors of the defined components have fixed values. The Parachor of a
C7+ component is calculated from the following expression
[P]
= 59.3 + 2.34 M i
where Mi is the molecular weight of the component. The phase densities are calculated using the
equation of state.
References
Christensen, P.L. and Fredenslund Aa., A Corresponding States Model for the Thermal
Conductivity of Gases and Liquids, Chem. Eng. Sci. 35, 1980, p. 871.
Hanley, H.J.M., McCarty, R.D. and Haynes, W.M., Equation for the Viscosity and Thermal
Conductivity of the Individual Gases, Cryogenics 15, 1975, 413.
Hanley, H.J.M., Prediction of the Viscosity and Thermal Conductivity Coefficients of
Mixtures, Cryogenics 16, 1976, p. 643.
Herning, F. and Zippener, L., Calculation of the Viscosity of Technical Gas Mixtures from the
Viscosity of the Individual Gases, Gas u. Wasserfach 79, 1936, pp. 6973.
Lohrenz, J., Bray, B.G. and Clark, C.R., Calculating Viscosities of Reservoir Fluids from Their
Compositions, J. Pet. Technol., Oct. 1964, pp. 11711176.
McCarty, R.D., A Modified BenedictWebbRubin Equation of State for Methane Using Recent
Experimental Data, Cryogenics 14, 1974, 276.
Mo, K.C. and Gubbins, K.E., Conformal Solution Theory for Viscosity and Thermal
Conductivity of Mixtures, Mol. Phys. 31, 1976, 825.
Pedersen, K.S., Fredenslund, Aa., Christensen, P.L. and Thomassen, P., Viscosity of Crude
Oils, Chem. Eng. Sci. 39, 1984, 1011.
Pedersen, K.S. and Fredenslund, Aa., An Improved Corresponding States Model for the
Prediction of Oil and Gas Viscosities and Thermal Conductivities, Chem. Eng. Sci. 42, 1987,
182.
Reid, R. C. and Sherwood, T. K., "The Properties of Gases and Liquids", 2nd ed. Chap 2,
McGrawHill, New York, 1966.
Reidel L., Chem. Ing. Tech., 26, 1954, 83
Stiel, L. I. and Thodos, G., The Viscosity of NonPolar Gases at Normal Pressures, AIChE J. 7,
1961, pp. 611615.
Weinaug, C.F. and Katz, D.L., Surface Tensions of MethanePropane Mixtures, Ind. Eng.
Chem. 35, 1943, pp. 239246.
PVT Experiments
PVT Experiments
PVTsim may be used to simulate the most commonly performed PVTexperiments. A
description of these experiments has been given by Pedersen et al. (1984, 1989).
PVT experiments are carried out with reference to standard conditions that may be specified in
PVTsim. Default values are default 1 atm/14.696 psi and 15C/59F. The results tabulated in a
simulation of a PVT experiment are explained in the following.
For oil systems the primary output for each pressure stage comprises
Relative volume
V/Vb  V is the actual volume and Vb is bubble point or saturation point volume.
Compressibility (only for pressures above the saturation point)
co =
1 V
V P T
Y=
Psat P
V
P t 1
Vsat
For gas condensate systems the primary output for each pressure stage comprises
Rel Vol
Liq Vol
Z Factor
Differential Depletion
This experiment is only carried out for oil mixtures. The reservoir fluid is kept in a cell at the
reservoir temperature. The experiment is usually started at the saturation pressure. The pressure
is reduced stepwise and all the liberated gas is displaced and flashed to standard conditions. This
procedure is repeated 610 times. The end point is measured at standard conditions.
The primary output for each pressure stage comprises
Oil FVF
Rsd
Gas FVF
Gas Gravity
Separator Experiments
Separators in Series
Sometimes the separator GOR is seen reported as the standard volume of gas divided by the
separator oil volume (oil volume at actual stage). The latter GOR can be converted into that
reported by PVTsim by dividing it by FVF.
Separator Test with Recirculation
PVTsim has an option for simulating a special two stage separator setup. The gas stream from
the first separator is separated in a second separator. The liquid stream from the second separator
is mixed into the feed to the first separator. The product streams are the liquid stream from the
first separator and the gas stream from the second separator.
Viscosity Experiment
A viscosity experiment is performed at the reservoir temperature. The pressure is reduced in
steps as in a differential depletion experiment. At each step the phase viscosities are recorded.
Swelling Experiment
When gas is injected into a reservoir containing undersaturated oil, the gas may dissolve in the
oil. The volume of the oil increases, which is also called swelling. A swelling test experiment
may simulate this process. The cell initially contains reservoir oil. A known molar amount of a
gas is added at a constant temperature. The saturation pressure of the swollen mixture and the
volume at the saturation point divided by the volume of the original reservoir oil are recorded.
More gas is added. The new saturation pressure and saturation point volume are recorded and so
on. The primary output consists of
Mol%
GOR
Sat P
Swollen volume
Density
It is further indicated in the output whether the saturation point is a bubble point (Pb) or a dew
point (Pd).
References
Pedersen, K.S., Thomassen, P. and Fredenslund, Aa., Thermodynamics of Petroleum Mixtures
Containing Heavy Hydrocarbons. 3. Efficient Flash Calculation Procedures Using the SRK
Equation of State, Ind. Eng. Chem. Process Des. Dev. 24, 1985, pp. 948954.
Pedersen, KS., Fredenslund Aa. and Thomassen, P., Properties of Oils and Natural Gases, Gulf
Publishing Company, Houston, 1989.
Isothermal reservoirs
Reservoirs with a vertical temperature gradient.
In the first case temperature is assumed constant over the entire fluid coloumn and the
compositional variations with depth are assumed only to originate from the effect of gravitational
forces. In the second case temperature varies over the fluid coloumn and compositional variations
with depth are affected both by the temperature gradient and by the gravitational forces.
Isothermal case
For an isothermal system the chemical potentials,
height h0 are related as follows
( )
(h ) h 0 = M w,i g h h 0
M stands for molecular weight and g is the gravitational acceleration. The chemical potential is
related to the fugacity through the following relation
i = RT ln f i
where T is the temperature.
The fugacities of component i in height h and in height h0 are therefore related through
ln f ln f
h
i
ho
i
M w,i g h h 0
RT
f i = i z i P
which gives the following relation between the fugacity coefficients of component i in height h
and in height h0
ln ( ih z ih P h ) ln ih z ih P h
0
) = M gRT(h h )
0
wi
This equation is valid for any component i. For a system with N components there are N such
equations. The mol fractions of the components must sum to 1.0 giving one additional equation
N
z
i =1
=1
If the pressure
ln i,SRK ln i,PEN =
ci P
RT
where c is the volume translations term. In a usual phase equilibrium calculation the temperature
and pressure are the same throughout the system and the term on the right hand side of the
equation cancels. This is not the case in a calculation of the compositional variations with depth.
The pressure changes with depth and this change is related to the fluid density for which different
results are obtained with the SRK and PR Peneloux equations. The SRK and PR Peneloux
equations are both presented in the Equation of State section.
~ ~
H
H T
RTln( ih z ih P h ) RTln( ih0 z ih0 P h0 ) = M i g(h h 0 ) M i i
; i = 1, N
M
M
T
i
Relative to the isothermal expression by Schulte, an additional term including the effect of the
temperature gradient T has been added. The term furthermore contains average molecular
weight M, component molecular weight Mi and partial molar enthalpies H and Hi.
It follows that a proper determination of partial molar enthalpies is the key to obtaining
reasonable predictions with the model. In typical process simulations it is appropriate to work
with enthalpy differences since overall composition is normally constant, and the reference state
therefore the same in all cases. This assumption cannot be applied to the present problem.
Instead, absolute enthalpies with a unique reference state must be used.
In PVTsim enthalpies are normally calculated relative to an ideal gas at 273.15 K/0C/32F and
the same composition. Absolute enthalpies, being the sum of an ideal gas contribution and a
residual term are obtained as follows
H abs = H PVTsim + H ig273.15K = H res + (H ig H ig273.15K ) + H ig273.15K
A simultaneous parameter fit has been carried out to thermal diffusion data for the mixtures C1C3 and C1nC4. The enthalpy of C1 as ideal gas at 273.15 K/0C/32F was assumed to be zero.
The data were fitted using
N2 :
H ig273.15
= 1.00 M i
R
H ig273.15
CO 2 :
= 1.70 M i
R
C2 :
H ig273.15
= 3.93 M i
R
H ig273.15
C3 :
= 15.8 M i
R
C4 :
H ig273.15
= 7.07 M i
R
C5 :
H ig273.15
= 37.3 M i
R
C6 :
H ig273.15
= 48.4 M i
R
For all other components the reference ideal gas enthalpy is assumed to be
H ig273.15 = 50 M i
When tuning to experimental data is performed, two multiplication factors are used, one for the
ideal gas reference of the defined components and a second one for the ideal gas reference
enthalpy of the C7+ components.
References
Haase, R., Borgmann, H.W., Dcker, K. H. and Lee, W. P., "Thermodiffusion im kritischen
Verdampfungsgebiet Binrer Systeme", Z. Naturforch. 26 a, 1971, pp. 12241227.
Schulte, A.M., Compositional Variations within a Hydrocarbon Column due to Gravity, paper
SPE 9235 presented at the 1980 SPE Annual Technical Conference and Exhibition Dallas, Sept.
2124, 1980.
Pedersen, K.S. and Lindeloff, N., Simulations of Compositional Gradients in Hydrocarbon
Reservoirs Under the Influence of a Temperature Gradient, SPE Paper 84364, to be presented at
the SPE ATCE in Denver, 58 October, 2003.
Rutherford, W.M. and Roof, J.G., "Thermal diffusion in methane nbutane mixtures in the critical
region", J. Phys. Chem. 63, 1959, pp. 15061511.
Regression to Experimental
Data
Experimental data
The two tables below show the type of PVTdata to which regression may be performed.
Oil mixtures.
Sat.
Point
*)
x
Saturation
Point
Bo
GOR (Rs)
Rel. volume
Compressibility
YFactor
Oil density
Z factor Gas
Two phase Z
factor
Liquid volume %
Gas Gravity
Bg
Mole % removed
Oil viscosity
Gas viscosity
*)
May also be critical point.
CME
Dif. Dep.
Separator
Viscosity
Swelling
CVD
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Sat. points
*)
Saturation
x
Point
Z factor gas
Two phase
Z factor
Rel volume
Liq vol%
Bo
GOR
Gas density
Oil density
Gas gravity
Mol%
removed
Oil viscosity
Gas
viscosity
*)
May also be critical point.
CME
x
CVD
x
x
x
x
x
Separator
x
Viscosity
X
x
x
x
x
x
x
x
x
x
x
NOBS
j=1
rj
w
j
where NOBS is the number of experimental observations used in the regression, wi is the weight
factor for the jth observation, and rj is the jth residual
rj =
OBSexp OBScalc
OBSexp
where OBS stands for the observed value and the subindices exp and calc stand for experimental
and calculated, respectively. For liquid dropout curves from a constant mass expansion and
constant volume depletion experiment, a constant is added to all OBSvalues. This constant
equals the maximum liquid dropout divided by 3 and is added to reduce the weight assigned to
data points with small liquid dropout relative to data points with larger liquid dropouts. The
weight factor, wj, and the user specified weight, WOBS to be assigned to the jth observation are
interrelated as follows
WOBS=
1
w 2j
If the user has allowed the plus molecular weight to be adjusted, an initial regression calculation
is performed where the plus molecular weights are adjusted to give the best possible match of the
measured saturation points. The molecular weight of the plus fraction is used as regression
parameter because there is usually an experimental uncertainty of 510% on the experimental
determination of this quantity. Furthermore even small changes in the molecular weight of the
plus fraction may have a major influence on the calculated saturation point. When modifying the
molecular weight of the plus fraction, the weight composition is kept constant while the molar
composition is recalculated. The weight composition is the one actually measured and is
accordingly kept constant.
Secondly a regression is performed where the coefficients in the Tc, Pc and m correlations
presented in the Characterization section are treated as regression parameters. The default
number of regression parameters is
NPAR = 1 + ln (NDAT)
Where NDAT is the number of experimental data points not considering viscosity data. The
maximum number of regression parameters is 10. The NPAR regression parameters are selected
in the following order (Christensen, 1999)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Coefficient c2 in Tc correlation.
Coefficient d2 in Pc correlation.
Peneloux volume shift parameter.
Coefficient c3 in Tc correlation.
Coefficient d3 in Pc correlation.
Coefficient e2 in m correlation.
Coefficient e3 in m correlation.
Coefficient c4 in Tc correlation.
Coefficient d4 in Pc correlation.
Coefficient e4 in m correlation.
In each iteration the parameters c1, d1 and e1 are recalculated to give the same Tc, Pc and m of a
component with a molecular weight of 94 and a density of 0.745 g/cm3 as is obtained with the
standard coefficients. This is done to ensure that Tc, Pc and m of the lower C7+ fractions are
assigned properties, which are physically meaningful. The user therefore has no control of the
parameters c1, d1 and e1 in the regression input menu.
The user may modify the default selection of regression parameters, but the number of regression
parameters must not exceed the number of experimental data points.
Regression to Viscosity Data
The regression parameters depend on which viscosity correlation is used. With the corresponding
states model the assumed mixture molecular weight is found from the following equation
VISC2
VISC2
M w,mix = VISC1 M W
MW
+ M W,n
,w
,n
With no regression VISC1 = 1.304 x 104 and VISC2 = 2.303. During a regression to viscosity
data, VISC1 and VISC2 are multiplied by two viscosity correction factors to give the best
possible agreement with the experimental viscosity data. The optimum viscosity correction
factors may be viewed in the Char Options menu accessed from the composition input menu.
With the LBC viscosity correlation three regression options exist. The default one is to let the
regression determine a unique correction factor to be multiplied with the critical volumes of the
pseudocomponents. It is further possible to determine optimum values of the five coefficients a1
a5 in the LBC correlation. A third option is to combine the Vc and a1 a5 regression.
The optimum viscosity correction factors and/or the optimum values of a1 a5 may be viewed in
the Char Options menu accessed from the composition input menu.
Tc
Pc
VPEN
(volume shift parameter)
Vc
a
b
kij (binary interaction parameter)
The mentioned properties are all defined in the Equation of State section. A total of up to 15
regression parameters may be specified. The number of experimental data points must be at least
as high as the number of regression parameters. One regression parameter may consist of for
example Tc of one specific component or it may consist of the Tcs of a number of consecutive
components in the component list. In the latter case the Tcs of all these components will be
adjusted equally.
The critical volume only affects the viscosities and only if the LBC correlation has been specified
(see Transport Property section)
With the LBC viscosity model it is further possible to regress on the coefficients a1 a5.
For the binary interaction parameters it is possible to specify single pairs of components for
which the binary interaction parameters are to be adjusted. Alternatively one may specify a
component triangle. The binary interaction parameters for each component pair contained in this
triangle will in that case be adjusted equally.
The user may specify a maximum allowed adjustment for each parameter.
Regression Algorithm
The minimization algorithm used in the parameter regression is a Marquardt algorithm
(Marquardt, 1963).
References
Christensen, P.L., Regression to Experimental PVT Data, Journal of Canadian Petroleum
Technology 38. 1999, pp. 19.
Marquardt, D.W., SIAM J 11 1963, 431441.
(FCMMP) is the lowest pressure at which the reservoir oil and the injection gas are miscible in
all ratios. The multiple contact minimum miscibility pressure (MCMMP or just MMP) is the
lowest pressure at which the oil and the gas phases resulting from a multicontact process
(vaporizing or condensing) between a reservoir oil and an injection gas are miscible in all ratios.
The MMP module of PVTsim permits calculation of FCMMP as well as MCMMP. The FCMMP
may be calculated by tracking the saturation pressure as a function of oil/gas mixing ratio. The
highest saturation point pressure located during the tracking procedure equals FCMMP.
Two different procedures are used in PVTsim for calculating vaporizing and condensing
MCMMP's. One is based on an extension of the procedures behind a ternary diagram (e.g.
Stalkup, 1984) to multicomponent mixtures.
With the second procedure the miscibility process for a vaporizing gas drive is simulated as a
continuous addition of reservoir oil at constant pressure and temperature to a cell initially
containing injection gas. Any oil phase formed in the cell is continuously removed.
The initial situation, where the cell contains pure injection gas corresponds to a situation just
before the gas enters into the well. Starting the continuous addition of oil corresponds to the
situation where the gas moves into the reservoir where it becomes saturated with oil. Further
addition of oil to the now saturated gas causes an oil phase to be formed in the cell. The removal
of this oil from the cell corresponds to the saturated gas moving in the reservoir leaving the oil
behind. It is seen that the experiment corresponds to a cell moving with the gas front in the
reservoir.
Simulating a condensing gas drive, the cell initially contains reservoir oil. Injection gas is added
and any formed gas is removed.
A set of differential mass balances and algebraic equilibrium relations of the following form
describe the vaporizing drive
dv i =(f i o i )dt ; i = 1,2,..., N
oi L vi V
i = i ; i = 1,2,..., N
O
V
N
i =1
i =1
O = oi ; V = vi
fi and oi are respectively the rates of addition to and removal of component i from the cell in
mols/time. vi is the number of mols of component i in the cell.
Since the feed rate and feed composition are constant, integration over an arbitrary time step
gives
v i (t + t ) = v i (t ) + f i t tt + t o i dt ; i = 1,2,...., N
Using the trapezoidal rule for evaluation of the integral one gets
v i (t + t ) = v(t ) + f i t
o i (t ) + o i (t + t )
t ; i =1,2,...., N
2
The compositions of the first equilibrium phases (t = 0) are known from the calculation of
mols of reservoir oil to a gas at equilibrium causes
FCMMP. Adding
mols of oil to be
formed, and the initial relative flow rate of oil, / , can be determined manipulating the
equilibrium relations. Successive solution with a chosen time step
simulates the miscibility
process. If an additional time step at some point does not result in a two phase solution,
miscibility has been established. If on the other hand the end result is constant compositions,
miscibility cannot be obtained at the specified conditions. Simulations at different pressures
determine MCMMP as the lowest pressure at which miscibility is obtained.
(1)
+ (1 ) x
(2)
= y
(2)
+ (1 ) x
and
(1)
The succession of N1 intersecting key tie lines can then be written as
x ij+1 (1 j ) + y ij j x ij (1 j ) y ij+1 j = 0
i = 1,..., N 1 , j = 1,..., N 2
where i are the component number and j the tie line number.
The first and last tie lines in the sequence are specified by the tielines through the original oil
and the injection gas respectively. Following the above nomenclature, these two compositions are
specified as follows:
z ioil = x ij=1 (1 oil ) + y ij=1 oil
j= nc 1
(1 inj ) + y ij=N1 inj
z inj
i = xi
i = 1,..., N 1
The above mentioned equations are solved subject to the usual phase equilibrium and mass
balance constraints
x ij iL y ij iV = 0
N
x
i =1
j
i
i = 1,..., N , j = 1,..., N 1
y ij = 0
All of this may be rearranged to a set of nonlinear equations to be solved for the coplanarity
parameters ( , ) and the phase compositions. A more thorough description may be found in
Jessen et al. (1998)
References
Jensen, F. and Michelsen, M.L., Calculation of First Contact and Multiple Contact Minimum
Miscibility Pressure In Situ 14, 1990, pp. 117.
Jessen, K.; Michelsen, M.L. and Stenby, E.H.: Effective Algorithm for Calculation of Minimum
Miscibility Pressure, SPE Paper 50632, 1998.
Stalkup, F.I., Miscible Displacement, Monograph Volume 8, H.L. Doherty Series, Society of
Petroleum Engineers, 1984.
Wang, Y., and Orr, F.M., Calculation of Minimum Miscibility Pressure, SPE paper 39683,
1998.
Unit Operations
Unit Operations
Compressor
PVTsim supports two compressor options
The two options differ in the way the compression path is corrected for isentropic efficiency.
The isentropic efficiency,
=V
, is defined as
dP
dH
where V is the molar volume, P the pressure and H the enthalpy. From the general
thermodynamics relation
dH = VdP+TdS
where S is the entropy it can be seen that
=1 for
=0 and that
dH
=V
dP s
meaning that the definition of the efficiency can be rewritten to
dH
dP s (dH)S
=
=
dH
dH
dP
Neglecting variations in efficiency along the compression path, one arrives at the classical
definition of the efficiency
(H) s
H
where (
)S is the enthalpy change of a compression following an isentropic path (=reversible
adiabatic compression) and
is the enthalpy change of the real compression (adiabatic but
partly irreversible).
The difference between the two compressor options is illustrated in the below figure.
The dashed line illustrates a compression path following the classical definition of isentropic
efficiency. Initially an isentropic path is followed from inlet pressure Pin to outlet pressure Pout.
The corresponding enthalpy change is (
)S. The outlet enthalpy is determined by dividing the
isentropic enthalpy change by the efficiency. The Pout pressure line is followed to the outlet
enthalpy meaning that the efficiency is determined by the slope of the Pout curve.
Pout
P..
P..
P2
P1
Pin
Schematic HSdiagram.
The dotted line shows a compression path of an almost constant efficiency (polytropic
as
compression). The compression path is divided into small Psegments each of the size
illustrated by the dotted line in the figure. Each segment is simulated as an isentropic
compression with the pressure increase
. The corresponding enthalpy change (
)S is
derived. The actual enthalpy change,
=(
)S/ , and
determine the conditions in the
next point on the compression path including the volume.
The sequence of calculations is the following
X=
T V
1
V T P
Y=
P V
V P T
HEAD =
WORK
g mf
where WORK is the total work done by the compressor on the fluid, g the gravitational
acceleration and mf the flow rate of the fluid through the compresor.
As can be seen from the above equation, the unit of HEAD is m or ft depending on selected unit.
HEAD therefore expreses the vertical lift height corresponding to the total work done by the
compressor on the fluid.
Expander
The input is inlet pressure and temperature and outlet pressure. An efficiency can be specified
which is 1.0 by default. For an efficiency of 1 the expansion process is assumed to be isentropic
(constant entropy (S)). In general the efficiency is defined as
H
(H )s
where
change.
Cooler
Input consists of inlet and outlet temperature and pressure. The outlet pressure is entered as a
pressure drop, which is zero by default. The cooling capacity is calculated which is the enthalpy
to be removed from the flowing stream per time unit.
Heater
Input consists of inlet and outlet temperature and pressure. The outlet pressure is entered as a
pressure drop, which is zero by default. The heating capacity is calculated which is the enthalpy
to be transferred to the flowing stream per time unit.
Pump
Input consists of inlet temperature and pressure and outlet pressure. A thermal efficiency can be
specified, which is defined through the relation
=
(Vout + Vin )P
H
where Vout is the outlet volume, Vin the inlet volume and
the pumping.
Valve
The outlet temperature is found by assuming that there is no enthalpy change by the passage of
the valve.
Separator
Input consists of inlet temperature and pressure for which a PTflash calculation is performed.
References
Schultz, J. M., "The Polytropic Analysis of Centrifugal Compressors", Journal of Engineering for
Power, January 1962, pp. 6982.
Hydrate Formation
Hydrates consist of geometric lattices of water molecules containing cavities occupied by lighter
hydrocarbons or other light gaseous components (for example nitrogen or carbon dioxide).
Hydrates may be formed where the mentioned components are in contact with water at
temperatures below approximately 35C/95F. Using the hydrate module in PVTsim it is
possible to calculate the conditions at which hydrates may form and in what quantities.
Calculations concerning the effect of the most commonly applied liquid hydrate inhibitors may
be performed, and the inhibiting effect of dissolved salts in the water phase is also accounted for.
The hydrate phase equilibrium calculations considers the phases
Gas
Oil
Aqueous
Ice
Hydrates of structures I, II and H
Solid salts.
Types of Hydrates
PVTsim considers three different types of hydrate lattices, structures I, II and H. Each type of
lattice contains a number of smaller and a number of larger cavities. In a stable hydrate,
components called guest molecules occupy either a fraction or all of these cavities.
Structures I and II hydrates can only accommodate molecules of a rather modest size and
appropriate geometry. The table below indicates which of the components in the PVTsim
component database may enter into the cavities of hydrate structures I and II. The cavities may
contain just one type of molecules or they may contain molecules of different chemical species.
Component
N2
CO2
sI  Small
Cavities
+
+
sI  Large
cavities
+
+
sII  Small
cavities
+
+
sII  Large
cavities
+
+
H2S
O2
Ar
C1
C2
C3
iC4
nC4
2,2dimC3
cC5
cC6
Benzene
+
+
+
+

+
+
+
+
+

+
+
+
+

+
+
+
+
+
+
+
+
+
+
+
+
The last four components in the above table are designated structure II heavy hydrate formers
(HHF). Neglecting these components (Danesh et al. (1993), Tohidi et al. (1996), Tohidi et al.,
(1997)) show that may lead to malpredictions of the hydrate formation temperatures of heavy
reservoir oil mixtures by more than 2C/3.6F.
Structure H consists of three different cavity sizes. These are in PVTsim modeled as just two
cavity sizes, a small/medium one and a huge one. The huge cavity can accommodate molecules
containing from 5 to 8 carbon atoms. The small/medium sized molecules will usually be
accommodated with N2 or C1. The below table gives an overview of structure H formers
considered in PVTsim.
Component
Methane
Nitrogen
Isopentane
Neohexane
2,3Dimethylbutane
2,2,3Trimethylbutane
3,3Dimethylpentane
Methylcyclopentane
1,2Dimethylcyclohexane
Cis1,2Dimethylcyclohexane
Ethylcyclopentane
Cyclooctane
Small/Medium Cavities
+
+

Huge Cavities
+
+
+
+
+
+
+
+
+
+
Hydrate Model
Hydrates are formed when the hydrate state is energetically favorable as compared to a pure
water state (fluid water or ice). The transformation from a pure water state to a hydrate state can
be regarded as consisting of two steps
1.
2.
where , and H are used to identify each of the three states considered. The  state is purely
hypothetical and only considered to facilitate the hydrate calculations. Which state is
energetically favorable depends on which state has the lowest chemical potential. The difference
between the chemical potential of water in the hydrate state (H) and in a pure water state ( ) can
be expressed as
H = ( H )+ ( )
can be regarded as the stabilizing effect on the
The first term on the right hand side
hydrate lattice caused by the adsorption of the gas molecules. This latter effect depends on the
tendency of the molecules to enter into the cavities of the hydrate lattice. This tendency is in
PVTsim expressed using a simple adsorption model. The difference between the chemical
potential of water in the empty and in the filled hydrate lattice is calculated as follows
)= R T
NCAV
i =l
v i ln 1 YKi
K =1
where i is the number of cavities of type i and YKi denotes the possibility that a cavity i is
occupied by a gas molecule of type K. NCAV is the number of cavities per unit cell in the hydrate
lattice and N is the number of components present, which may enter into a cavity in the hydrate
lattice. The probability YKi is calculated using the Langmuir adsorption theory
N
YKi = C Ki f K / 1 + C ji f j
j=l
where fK is the fugacity of component K. CKi is the temperature dependent adsorption constant
specific for the cavity of type i and for component K. The adsorption constant accounts for the
waterhydrate forming component interactions in the hydrate lattice. The adsorption constant C is
calculated from the following expression (Munck et al., 1988)
is equal to the difference between the chemical potentials of water in the empty
hydrate lattice (the state) and water in the form of liquid or ice (the state). An expression for
this difference in chemical potentials can be derived using the following thermodynamic relation
H
V
d
dT +
dP
=
2
RT
RT
RT
 and
P V
(T, P ) (T0 , P0 ) T H
=
=
dP
dT +
2
T0 RT
P0 R T
RT
RT
RT0
T + T0
2
If the reference pressure, P0, is chosen to be equal to be zero, the above equation can be rewritten
to
P V
(T, P ) (T0 , P0 ) T H
=
=
dP
dT +
2
T0 RT
P0 R T
RT
RT
RT0
 and the
H(T ) = C p dT
T
T0
Unit
J/mol
Structure I
1264
Structure II
883
Structure H
1187.33
H 0 (liq)
J/mol
4858
5201
5162.43
H 0 (ice)
J/mol
1151
808
846.57
V0 (liq)
cm /mol
4.6
5.0
5.45
V0 (ice)
cm3/mol
3.0
3.4
3.85
C p (liq)
J/mol/K
39.16
39.16
39.16
between
At those conditions the hydrate state and the liquid or solid water states are equally favorable. To
the left of the hydrate curve
H < 0
and some of the water will at equilibrium be in a hydrate form. Whether this is a structure I or a
structure II hydrate depends on which of the two structures has the lower chemical potential in
the presence of the actual gas components as potential guest molecules. To the right of the
hydrate curve
H > 0
i.e. at equilibrium at those conditions no hydrate can exist and the water will be in the form of
either liquid or ice.
j,
subscript K is a component index, j a phase index, stands for phase fraction and N for
number of components.
3. Using the selected equation of state and the calculated compositions (xKj), the fugacities of all
components in all the phases except the hydrate and pure solid phases are calculated, i.e. (fKj,
K = 1,2,N, j hyd and pure solid).
4. Based on these fugacities (fKj, K = 1,2..,N, j hyd and pure solid), mixture fugacities
are calculated. For the nonwater components, a mixture fugacity is
calculated as the molar average of the fugacities of the given component in the present
hydrocarbon phases. For water the mixture fugacity is set equal to the fugacity of water in the
water phase.
5. The fugacities of the components present in the hydrate phase are calculated using
where
from
where w stands for water and
empty hydrate lattice.
6. The hydrate compositions are calculated using the expression
is found
refers to the
j,
K = 1,2,,N, j =
Calculation of Fugacities
Fluid Phases
To use the flash calculation procedure outlined above, expressions must be available for the
fugacity of component i in each phase to be considered. The fugacity of component i in a solution
is given by the following expression
fi = i x i P
where
For the fluid phases, is calculated from the selected equation of state. See Equation of State
section for details. Fugacities calculated with PR will be slightly different from those calculated
with SRK.
Hydrate Phases
The fugacities of the various components in the hydrate phases are calculated as described by
Michelsen (1991)
Water:
N (1 )
N 0
ln f wH ln f w = v i ln 0
+ v 2 ln
v1
v2
Other Hydrate Formers:
f kH =
Nk
N 0 C k 2 ( + k (1 ))
In these equations
f w = fugacity of water in empty hydrate lattice
k = Ck1/Ck2
The determination of and N0 follows the procedure described by Michelsen. As the fluid phase
fugacities vary with the equation of state choice, the hydrate model parameters are equation of
state specific in order to ensure comparable model performance for both SRK and PR.
Ice
The fugacity (in atm) of ice is calculated from the following expression
273.15
273.15 0.0390 P
f ice = 2.064 1
4.710 ln
+
T
T T + 273.15
Salts
The fugacities of a salt in pure solid form is assumed to be equal to the fugacity of the mentioned
salt in saturated liquid solution in water. The solubilities in mol salts per mol water are found
from the following expressions (with T in C)
Sodium Chloride, NaCl
Mol salt
= 0.108 + 0.000125 T
Mol water
T < 50C :
Mol salt
= 0.145 + 0.00355 T
Mol water
T 50C :
Mol salt
= 0.313
Mol water
T < 20C :
Mol salt
= 0.712 + 0.00705 T
Mol water
T 20C :
Mol salt
= 0.964 + 0.0174 T
Mol water
T < 6C :
Mol salt
= 0.0248 + 0.00143 T
Mol water
50oC T :
Mol salt
= 0.272 + 0.006 T
Mol water
Mol salt
= 0.572
Mol water
The remaining salts in the database are assigned the solubility of CaCl2, if they consist of 3 ions
and the solubility of NaCl, if they consist of a number of ions different from 3.
References
Danesh, A., Tohidi, B., Burgass, R.W., and Todd, A.C., "Benzene Can Form Gas Hydrates",
Trans. IChemE, Vol. 71 (Part A), pp. 457459, July, 1993.
Erickson, D.D., Development of a Natural Gas Hydrate Prediction Computer Program, M. Sc.
thesis, Colorado School of Mines, 1983.
Madsen, J., Pedersen, K.S. and Michelsen, M.L., Modeling of Structure H Hydrates using a
Langmuir Adsorption Model, Ind. Eng. Chem. Res., 39, 2000, pp. 11111114.
Michelsen, M.L., Calculation of Multiphase Equilibrium in Ideal Solutions, SEP 8802, The
Department of Chemical Engineering, The Technical University of Denmark, 1988.
Michelsen, M.L., Calculation of Hydrate fugacities , Chem. Eng. Sci. 46, 1991, 11921193.
Munck, J., SkjoldJrgensen S. and Rasmussen, P., Computations of the Formation of Gas
Hydrates, Chem. Eng. Sci. 43, 1988, 26612672.
Rasmussen, C.P. and Pedersen, K.S., Challenges in Modeling of Gas Hydrate Phase Equilibria,
4th International Conference on Gas Hydrates Yokohama Japan, May 19  23, 2002.
Tohidi, B., Danesh, A., Burgass, R.W., and Todd, A.C., Equilibrium Data and Thermodynamic
Modelling of Cyclohexane Gas Hydrates, Chem. Eng. Sci., Vol. 51, No. 1, pp. 159163, 1996.
Tohidi, B., Danesh, A., Todd, A.C., Burgass, R.W., stergaard, K.K., "Equilibrium Data and
Thermodynamic Modelling of Cyclopentane and Neopentane Hydrates", Fluid Phase Equilibria
138, pp. 241250, 1997.
f iL = f iS
When a cubic equation of state is used for the liquid phase it is practical to express the liquid
phase fugacities in terms of fugacity coefficients
f iL =x iL iL P
is the liquid phase mol fraction of component i,
the liquid phase
In this expression
fugacity coefficient of component i and P the pressure. For an ideal solid phase mixture, the solid
phase fugacity of component i can be expressed as
f iS = x Si f ioS
the solid standard state fugacity
where x Si is the solid phase mol fraction of component i, and
of component i. The solid standard state fugacity is related to the liquid standard state fugacity as
f ioS (Pref )
G = RT ln oL
(
)
f
P
ref
i
f
i
is the molar change in Gibbs free energy associated with the transition of pure
where
component i from solid to liquid form at the temperature of the system. To calculate
following general thermodynamic relation is used
the
G = H TS
where
may be expressed as
G if = H if + TSif
is the enthalpy and
the entropy of fusion of component i at the normal melting
where
point. Again neglecting any differences between the liquid and solid state heat capacities, the
entropy of fusion may be expressed as follows in terms of the enthalpy of fusion
Sif =
H if
Tif
where is the melting temperature of component i. The following expression may now be
derived for the solid standard state fugacity of component i
H if
f ioS = f ioL (Pref )exp
RT
T
1 f
Ti
Vi (P Pref )
RT
where
is the difference between the solid and liquid phase molar volumes. Based on
experimental observations of Templin (1956), the difference
i between the solid and liquid
phase molar volumes of component i is assumed to be 10% of the liquid molar volume, i.e. the
solidification process is assumed to be associated with a 10% volume decrease.
The liquid standard state fugacity of component i may be expressed as follows
f ioL = ioL P
is the liquid phase fugacity coefficient of pure i at the system temperature and
where
pressure. This leads to
H if
f ioS = ioL (Pref ) P exp
RT
T
1 f
Ti
Vi (P Pref )
RT
The following expression may now be derived for the solid phase fugacity of component i in a
mixture
H i f
f i =xiS ioL (Pref ) P exp
RT
T
1 f
Ti
Vi (P Pref )
RT
is found using an equation of state on pure i at the temperature of the system and the
reference pressure.
and
The wax model is based on the assumption that a wax phase primarily consists of nparaffins.
The user may input the nparaffin content contained in each C7+ fraction. Otherwise the following
expression is used to estimate the mol fraction,
i
i
z = z 1 (A + B M i )
P
s
i
tot
i
In this expression Mi is the molecular weight in g/mol and i the density in g/cm3 at standard
conditions (atmospheric pressure and 15 oC) of pseudocomponent i. A, B and C are constants of
the following values
A = 1.0744
B = 6.584 x 104
C = 0.1915
iP is the densities (g/cm3) at standard conditions of a normal paraffin with the same molecular
weight as pseudocomponent i. The following expression is used for the paraffinic density.
iP = 0.3915 + 0.0675 ln M i
For a (hypothetical) pseudocomponent for which
will be equal to
meaning that
all the components contained in that particular pseudocomponent are able to enter into a wax
phase. In general
The wax forming and the nonwax forming fractions of the C20+ pseudocomponents are assigned
different critical pressures. The critical pressure of the wax forming fraction of each pseudocomponent is found from
P
P = Pci i
i
3.46
s
ci
Pci equals the critical pressure of pseudocomponent i determined using the characterization
1 FracinoS
=
Pci
PcinoS
) + (Frac )
2
S 2
i
S
ci
2FracinoS
PcinoS PciS
where S and noS are indices used respectively for the wax forming and the nonwax forming
fractions (Frac) of pseudocomponent i. By using this relation the contribution to the equation of
state aparameter of pseudocomponent i divided into two will be the same as that of the pseudocomponent as a whole.
For the wax forming C7+ components, the following expressions proposed by Won (1986) are
used to find the melting temperature and enthalpy of melting
Tif = 374.5 + 0.02617 M i
20172
Mi
H if = 0.1426 M i Tif
The division of each C7+component into a potentially wax forming component and a component,
which cannot form wax, implies that it is necessary to work with twice the number of C7+components as in other PVTsim modules. The equation of state parameters of the wax forming
and the nonwax forming parts of a pseudocomponent are equal, but the wax model parameters
differ. Presence of nonwax forming components in the wax phase is avoided by assigning these
components a fugacity coefficient of exp(50) in the wax phase independent of temperature and
pressure.
When tuning to an experimentally determined wax content or to an experimental wax
appearance. The wax forming fraction of each pseudocomponent is adjusted to match the
experimental data.
4
E wax F wax
= liq exp(D wax )1 +
+
dv x
dv x
dy
dy
Having performed regression of the viscosity model to experimental data, the modified model
can be applied in the wax module by entering the multiplication factors for the coefficients D, E
and F. For the original model these multiplication factors should all have values of 1. Viscosity
values at different T and P can then be calculated by specifying the P,T grid of interest.
Wax Inhibitors
Wax inhibitors are often added to oils being transported in subsea pipelines with the purposes of
decreasing the apparent viscosity of the oil. In PVTsim the wax inhibitor effect is modeled as a
depression of the melting temperature of wax molecules within a given range of molecular
weights (Pedersen and Rnningsen, 2003). The range of affected molecular weights and the
depression of the melting temperature may be estimated by entering viscosity data for the oil
with and without wax inhibitor and running a viscosity tuning to this data material.
References
Pedersen, K.S., Prediction of Cloud Point Temperatures and Amount of Wax Precipitation,
Production & Facilities, February 1995, pp. 4649.
Pedersen, K.S. and Rnningsen, H.P., Effect of Precipitated Wax on Viscosity A Model for
Predicting NonNewtonian Viscosity of Crude Oils, Energy & Fuels, 14, 2000, pp. 4351.
Pedersen, K.S. and Rnningsen, H.P., Influence of Wax Inhibitors on Wax Appearance
Temperature, Pour Point, and Viscosity of Waxy Crude Oils, Energy & Fuels 17, 2003, pp. 321328.
Rnningsen, H. P., Smme, B. and Pedersen, K.S., An Improved Thermodynamic Model for
Wax Precipitation; Experimental Foundation and Application, presented at 8th international
conference on Multiphase 97, Cannes, France, 1820 June, 1997.
Templin, R.D., Coefficient of Volume Expansion for Petroleum Waxes and Pure nParaffins,
Ind. Eng. Chem., 48, 1956, pp. 154161.
Won, K.W., Thermodynamics for SolidLiquidVapor Equilibria: Wax Phase Formation from
Heavy Hydrocarbon Mixtures, Fluid Phase Equilibria 30, 1986, pp. 265279.
Asphaltenes
Asphaltenes
Asphaltene precipitation is in PVTsim considered as a that can be described by equilibrium
thermodynamics. An equation of state is used for all phases including the asphaltene phase.
By default the aromatic fraction of the C50+ component is considered to be asphaltenes (Rydahl et
al, 1997). The user may enter an experimental weight content of asphaltenes in the oil from a
flash to standard conditions. This will change the cut point between asphaltenic and nonasphaltenic aromatics from C50 to a carbon number that will make the total asphaltene content
agree with that measured. In asphaltene simulations pseudocomponents containing asphaltenes
are split into an asphaltene and nonasphaltene component. Having completed an asphaltene
simulation, selecting Split Pseudos will maintain the split fluid.
In contrast to most other calculation options in PVTsim, the asphaltene module should not be
considered a priori predictive. Being a liquidliquid equilibrium the oilasphaltene phase split is
extremely sensitive to changes in model parameters. Consequently the asphaltene module should
be considered a correlation tool rather than a predictive model. It is strongly recommended that
an experimental asphaltene onset P,T point is used to tune the model before further calculations
are made. Having tuned the model to a single data point, the model will in general correlate the
remaining part of the asphaltene precipitation envelope quite well.
where Tci is the critical temperature of pseudocomponent i before being split. The critical
pressure
equation
1 FracinoA
=
Pci
PcinoA
) + (Frac )
2
A 2
i
A
ci
2FracinoA
PcinoA PciA
while the acentric factor of the nonasphaltene forming fraction of pseudocomponent i is found
from
i = FracinoA inoA + FraciA iA
The binary interaction parameters between asphaltene components and C1C9 hydrocarbons are
by default assumed to be 0.017 where binary interaction parameters of zero are default used for
all other hydrocarbonhydrocarbon interactions. Tuning the model to an experimental point may
either be accomplished by tuning the asphaltene Tc and Pc or by tuning the asphaltene content in
the oil.
References
Rydahl, A., Pedersen, K.S. and Hjermstad, H.P., Modeling of Live Oil Asphaltene
Precipitation, AIChE Spring National Meeting March 913, 1997, Houston, TX, USA.
H2S Simulations
H2S Simulations
The H2S module of PVTsim is based on the same PTflash as is used in many of the other
modules. What makes this module different is the way H2S is treated in the aqueous phase. The
dissociation of H2S is considered.
H2S HS + H+
The degree of dissociation is determined by the pH
[ ]
pH = log10 H +
and pK
pK 1 = log10
[HS ][H ]
H 2S
pK1 is calculated using considerations based on chemical reaction equilibria. This gives
approximately the following temperature dependence
pK1 = 7.2617 0.01086(T 273.15)
where T is the temperature in K. From the knowledge of the amount of dissolved H2S on
molecular form, pH and pK1 it is straightforward to calculate [HS].
In principle the following equilibrium should also be considered
HS S + H+
Its pK value defined by the following expression
pK 2 = log10
[H ][S ]
[HS ]
+
is however of the order 1314, meaning that the second order dissociation for all practical
purposes can be neglected. It is therefore not considered in the H2S module.
Dynalog
Prosper/Mbal
Multiphase meter interface if license does not give access to multiflash options.
The options treating water as pure water calculates the physical properties and transport
properties using a separate thermodynamics instead of an EOS. In the OLGA2000 interface the
water property routines are used in cases where no hydrate inhibitors are present. This is also an
option in the Property Generator.
The thermodynamic properties of pure water are calculated using an equation for Helmholtz free
energy developed by Keyes et al. (1968)
= 0 (T ) + RT [ln + Q( , T )]
where
=
=
=
R =
and
0 (T ) = C1 + C2T + C3T 2 + (C4 + C5T ) ln T
i l
Q( , T ) = Aij ( a ) + e E ( A9,1 + A10,1 )
i =l
8
7
j 2
o l
+ ( c ) ( a ) Aij ( b ) + e Ep (A9 j + A10 j )
i =1
j=2
where
a = 0.634 g/cm3
b = 1.0 g/cm3
a = 2.5 K1
c = 1.544912 K1
E = 4.8 cm3/g
CI
1855.3865
3.278642
.00037903
46.174
1.02117
1
29.492937
132.13917
274.64632
360.93828
342.18431
244.50042
155.18535
5.9728487
410.30848
416.05860
2
5.1985860
7.779182
33.301902
16.254622
177.31074
127.48742
137.46153
155.97836
337.31180
209.88866
3
6.8335354
26.149751
65.326396
26.181978
0
0
0
0
137.46618
733.96848
j
4
01564104
0.72546108
9.2734289
4.3125840
0
0
0
0
6.7874983
10.401717
5
6.3972405
26.409282
47.740374
56.323130
0
0
0
0
136.87317
645.81880
6
3.9661401
15.453061
29.142470
29.568796
0
0
0
0
79.847970
399.17570
7
0.69048554
2.7407416
5.1028070
3.9636085
0
0
0
0
13.0411253
71.531353
1000
2 Q
= 2
= R
P = 2
1 + Q +
T
The pure water density, , is obtained from this equation by iteration. The enthalpy, H, the
entropy, S, and the heat capacity at constant pressure, Cp, are obtained from the following
relations
P
( )
H =
+
Q
d 0
1000 R
Q
+ 0 T
+
1+ Q +
dT
S =
= R
T
Q d 0
ln + Q
dT
H H
Cp =
T T
P
T
P
T
Viscosity
Four different expressions (Meyer et al. (1967) and Schmidt (1969)) are used to calculate the
pure water viscosity. Which expression to use depends on the actual pressure and temperature. In
two of the four expressions an expression enters for the viscosity, , at atmospheric pressure
(=0.1 MN/m2) valid for 373.15 K/100C/212F < T < 973.15 K/700C/1292F
T
b2 + b3 10 6
Tc
1 = b1
Region 1:
Psat < P < 80 MN/m2 and 273.15 K/0C/32F < T < 573.15 K/300C/572F
=10 6 a1 1+
Psat
Pc
a2
a4 a5 10
Tc
(T / Tc ) a3
where Tc and Pc are the critical temperature and pressure, respectively and
critical point.
Region 2:
0.1 MN/m2 < P < Psat and 373.15 K/100C/212F < T < 573.15 K/300C/572F
= 1 10 6 10
Region 3:
6
c1 c2 c3 10
c
Tc
0.1 MN/m2 < P < 80 MN/m2 and 648.15 K/375C/707F < T < 1073.15 K/800C/1472F
3
6
= 1 10 + d 3 + d 2
10 6
d1
Region 4:
Otherwise
= 1 +
10Y
0.0192
where
Y = C5kX4 + C4kX3 + C3kX2 + C2kX + C1k
X = log10
c
The parameter k is equal to 1 when / <= 4/3.14 and equal to 2 when
following coefficients are used in the viscosity equations
a1
a2
a3
a4
a5
b1
b2
b3
c1
c2
c3
d1
d2
d3
241.4
0.3828209486
0.2162830218
0.1498693949
0.4711880117
263.4511
0.4219836243
80.4
586.1198738
1204.753943
0.4219836243
111.3564669
67.32080129
3.205147019
C1k
C2k
C3k
C4k
C5k
6.4556581
1.3949436
0.30259083
0.10960682
0.015230031
For k = 1
For k = 2
C1k
C2k
C3k
C4k
C5k
6.4608381
1.6163321
0.07097705
13.938
30.119832
D2
T 273.15
where Psat is in MN/m2 and T in K. The coefficient, Di, are given in the table below.
Coefficients of vapor pressure correlation.
I
1
2
3
4
5
6
7
Di
2.9304370
2309.5789
.34522497 x 101
.13621289 x 103
.25878044 x 106
.24709162 x 109
.95937646 x 1013
Thermal conductivity
Six different expressions (Meyer et al. (1967), Schmidt (1969) and Sengers and Keyes (1971))
are used to calculate the pure water thermal conductivity (in W/cm/K). Which expression to use
depends on the actual pressure and temperature. The following expression for the thermal
conductivity, 1, at atmospheric pressure (=0.1 MN/m2) and 373.15 K/100C/212F < T <
973.15 K/700C/1292F enters into two of the six expressions
1
where
t = T 273.15
Region 1:
Psat < P < 55 MN/m2 and 273.15 K/0C/32F < T < 623.15K/350C/662F
P Psat
Pc
= S1 +
where
P Psat
S 2 +
Pc
S 3 10 2
T
S1 = ai
i =0
Tc
4
T
S 2 = bi
i =0
Tc
3
T
S3 = ci
i =0
Tc
3
Region 5:
When P,T is not in region 1 and P (in MN/m2) and T (in K) are in one of the following ranges

= 1 +10Y
where
Y = C5kX4 + C4kX3 + C3kX2 + C2kX + C1k
and
X = log10
c
k = 1 for
<= 2.5
k = 2 for
> 2.5
0.5786154
1.4574646404
0.17006978
0.1334805
0.032783991
C1k
C2k
C3k
0.70859254
0.94131399
0.064264434
for k = 2
C4k
C5k
1.85363188
1.98065901
Region 3:
When P,T is not in regions 1 or 5 but in one of the following ranges (P in MN/m2 and T in K)
45 < P and 723.15 K/450C/842F < T < 823.15 K/550C/1022F

where
T
Pbound = Pc ei
i =0
Tc
2
P
T
T
d 32 exp 9d 33 1
A
P
T
Pc
Tc
Tc
=
d 35 d 36 exp d 33 1
+
7
12
P
T
Pc
Tc
1 Bd 31
1+ d 34
Pc
Tc
P
A = a31 + a32
Pc
1.63
P
b31
Pc
B=
3.26
P
1 + b32
Pc
1.5
P
c31 + c32
P
C= c
c33
B
Region 4:
When P,T is not in region 1, 3 or 5 but in one of the following ranges (P in MN/m2 and T in K)

T 8
P
8
= a4i k i + c40 b4i k i
Tc i = 0
Pc
i =0
where
k = 100
The solution for
is iterative.
Region 6:
When P,T is not in region 1, 3, 4 or 5 and in one of the following ranges
15 MN/m2 < P and T > 633.15 K/360C/680F
14 MN/m2 < P and T > 618.15 K/345C/653F
= 0.01 0.2
+ v1
c
where
v1 = 1.76 x 102 + 5.87 x 105 t + 1.04 x 107 t2 4.51 x 1011 x t3
Region 2:
Otherwise
2.1482 x 1014 2
x 10 5
4.2
t
The following coefficients are used in the equations for thermal conductivity
a0
a1
a2
a3
a4
a31
a32
a40
0.92247
6.728934102
10.11230521
6.996953832
2.31606251
0.01012472978
0.05141900883
1.365350409
a41
a42
a43
a44
a45
a46
a47
a48
b0
b1
b2
b3
b31
b32
b40
b41
b42
b43
b44
b45
b46
b47
b48
c0
c1
c2
c3
c31
c32
c33
c40
d31
d32
d33
d34
d35
d36
e0
e1
e2
4.802941449
23.60292291
51.44066584
38.86072609
33.47617334
101.0369288
101.2258396
45.69066893
0.20954276
1.320227345
2.485904388
1.517081933
6.637426916 x 105
1.388806409
1.514476538
19.58487269
113.6782784
327.0035653
397.3645617
96.82365169
703.0682926
542.9942625
 85.66878481
0.08104183147
0.4513858027
0.8057261332
0.4668315566
3.388557894 x 105
576.8
0.206
1.017179024
2.100200454 x 106
23.94
3.458
13.6323539
0.0136
7.8526 x 103
50.60225796
105.6677634
55.96905687
The surface tension of liquid water (in mN/m) is calculated from the following formula
1.256
T
= 235.81
Tc
1 0.625 1 T
T
The interfacial tension, , between a water phase and a hydrocarbon phase (gas or oil) is
calculated from the following expression (Firoozabadi and Ramey, 1988)
1/ 4 =
a1 b1
Tr0.3125
where:
= w HC
In this equation
HC
phase. The values of the constants a1 and b1 are given in the below table as a function of
Values of the constants a1 and b1 with
(g/cm3)
< 0.2
0.2  0.5
0.5
b1
0.94716
0.76852
0.62590
Tr is a pseudoreduced temperature for the hydrocarbon phase. It equals the temperature divided
by a molar average of the critical temperatures of the individual hydrocarbon phase components.
Salt Water Density
The density of a water phase with dissolved salts is calculated using a correlation suggested by
Numbere et al. (1977)
s
 1 =CS [7.65 x 103 1.09 107 P + CS (2.16 x 105 + 1.74 x 109 P)
w
(1.07 x 105 3.24 x 1010 P)T + (3.76 x 108 1.0 x 1012 P)T2
where s is the salt water density, w the density of salt free water at the same T and P, Cs is the
salt concentration in weight%, T the temperature in oF and P the pressure in psia.
Salt Water Viscosity
The viscosity of a water phase with dissolved salts is calculated using a correlation suggested by
Numbere et al. (1977)
s
1= 1.87 10 3 C s0.5 + 2.18 10 4 C s2.5 + (T 0.5 1.35 x10 2 T )(2.76 10 3 C s 3.44 10 4 C s1.5 )
w
where s is the salt water viscosity, w the viscosity of pure water at the same T and P, Cs the
salt concentration in weight% and T the temperature in F.
Viscosity of WaterInhibitor Mixtures
The viscosities of mixtures of water, methanol and/or mono ethylene glycol (MEG) are
calculated from the viscosities of the pure fluids using appropriate mixing rules.
Methanol
The viscosity of saturated liquid methanol can be calculated from the following equation (Alder,
1966)
ln
= A + B/T + CT + DT2
where
A = 2.687 x 10
B = 1.150 x 103
C = 1.875 x 101
D = 5.211 x 104
Mono Ethylene Glycol (MEG)
The viscosity of saturated liquid mono ethylene glycol can be calculated from the following
equation (Alder, 1966)
ln
= A + B/T
where
A = 7.811
B = 3.143 x 103
DiEthylene Glycol (DEG)
The viscosity of saturated liquid diethylene glycol is calculated from the following equation
(van Velzen et al., 1972)
1 1
log10 = B
T T0
where
The viscosity of saturated liquid triethylene glycol is calculated from the following equation
(van Velzen et al., 1972)
1 1
log10 = B
T T0
where
Saturation Pressures
To be able to determine the pressures corresponding to the above inhibitor viscosities the pure
component vapor pressures are needed. The vapor pressures are determined from the Antoine
equation
ln P sat = A
B
T +C
where the vapor pressure, Psat, is in atm, the temperature, T, in Kelvin and A, B and C are
constants for which values are given in the below table.
Antoine constants for methanol (MEOH), mono ethylene glycol (MEG), diethylene glycol
(DEG), triethylene glycol (TEG) and water. T is in K and P is in atm.
Component
MeOH
MEG
DEG
TEG
Water
A
11.9542
13.6168
10.3993
0.0784
11.6703
B
3626.55
6022.18
4122.52
8699.44
3816.44
C
34.290
28.250
122.50
2.2040
46.1300
The effect of pressure on the pure component liquid viscosity is calculated using the following
formula (Lucas, 1981)
1 + D (Pr / 2.118)A
=
SL
1 + C Pr
where
= viscosity of liquid at actual temperature and pressure
SL = viscosity of saturated liquid at current T
Pr = (P Psat)/Pc
= acentric factor
 1.0513))
D = (0.3257 /
)  0.208616
+ 44.1706
 84.8291
+ 96.1209
 59.8127
15.6719
Pc is the critical pressure and Tr the reduced temperature, T/Tc, where Tc is the critical
temperature.
Viscosity Mixing Rules
Mixture viscosities are calculated using the following relation (Grunberg and Nissan, 1949)
ln mix = zi ln i +
i
z z G
i
>
ij
where zi and zj are the mol fractions of component i and j, respectively and Gij is a binary
interaction parameter, which is a function of the components i and j as well as the temperature.
The following temperature dependence is assumed
Gij (T ) = 1 (1 Gij ( 298))
573 T
275
Other glycols
Other glycols are assigned the properties of that of the above glycols that is cloest in molecular
weight.
+ 0.0002497 t
where
r=
( ) r =100
r , h = 1+
1.19 w
r =100
2.5
( ) r =100
w, h = 1+
1.19 h
r =100
2.5
and
=100
1.19 1 r
0 .4
r) is
, if w < Inv
, if w > Inv
is used to calculate
is calculated as a function of
References
Alder, B.J., Prediction of Transport Properties of Dense Gases and Liquids, UCRL 14891T,
University of California, Berkeley, California, May 1966.
Firoozabadi, A. and Ramey, H.J., Journal of Canadian Petroleum Technology 27, 1988, pp. 4148.
Grunberg, L. and Nissan, A.H., Nature 164, 1949, 799.
Keyes, F.G., Keenan, J.H., Hill, P.G. and Moore, J.G., A Fundamental Equation for Liquid and
Vapor Water, presented at the Seventh International Conference on the Properties of Steam,
Tokyo, Japan, Sept. 1968.
Lucas, K., Chem. Ing. Tech. 53, 1981, 959.
Meyer, C.A., McClintock, R.B., Silverstri, G.J. and Spencer, R.C., Jr., Thermodynamic and
Transport Properties of Steam, 1967 ASME Steam Tables, Second Ed., ASME, 1967.
Numbere, D., Bringham, W.E. and Standing, M.B., Correlations for Physical Properties of
Petroleum Reservoir Brines, Work Carried out under US Contract E (043) 1265, Energy
Research & Development Administration, 1977.
Pal, R. and Rhodes, E., "Viscosity/Concentration Relationships for Emulsions", J. Rheology,
33(7), 1989, 1021.
Rnningsen, H.P., Conditions for Predicting Viscosity of W/O Emulsions based on North Sea
Crude Oils, SPE Paper 28968, presented at the SPE International Symposium on Oilfield
Chemistry, San Antonio, Texas, US, February 1417, 1995.
Schmidt, E., Properties of Water and Steam in SIUnits, SpringerVerlag, New York, Inc.
1969.
Sengers, J.V. and Keyes, P.H., Scaling of the Thermal Conductivity Near the GasLiquid
Critical Point, Tech. Rep. 71061, University of Maryland, 1971.
Thomson, G.H. Brabst, K.R. and Hankinson, R.W., AIChE J. 28, 1982, 671.
van Velzen, D., Cordozo, R.L. and Langekamp, H., Ind. Eng. Chem. Fundam. 11, 1972, 20.
A water analysis, including the concentrations (mg/l) of the inorganic ions Na+, K+, Ca++,
Mg++, Ba++, Sr++, Fe++, Cl, SO4, of organic acid and the alkalinity.
Since the major part of the organic acid pool is acetic acid and since the remaining part behaves
similar to acetic acid, the organic acid pool is taken to be acetic acid.
The alkalinity is defined in terms of the charge balance. If the charge balance is rearranged with
all pHdependent contributions on one side of the equality sign and all pHindependent species
on the other, the alkalinity appears, i.e. the alkalinity is the sum of contributions to the charge
balance from the pHindependent species. Therefore the alkalinity has the advantage of
remaining constant during pH changes.
The calculation of the scale precipitation is based on solubility products and equilibrium
constants. In the calculation, the nonideal nature of the water phase is taken into account.
Thermodynamic equilibria
The thermodynamic equilibria considered are
Acidequilibria
H2O(l) H+ + OHH2O(l) + CO2(aq) H+ + HCO3HCO3 H+ + CO3HA(aq) H+ + AH2S(aq) = H+ + HS
H OH
+
K H 2O = mH + mOH
K CO2 1 =
a H 2O ( l )
mH + m HCO
H HCO
mCO2
CO ( aq ) a H O (l )
K CO2 , 2 =
mH + mCO H + CO
3
HCO
mHCO
3
K HA =
mH + mA H + A
KH2S =
mHA( aq ) HA( aq )
mH + mHS H + HS
mH 2 S ( aq ) H 2 S ( aq )
K FeS =
mFe + + mHS Fe ++ HS
mH +
B
E
+ C ln T + DT + 2
T
T
K CO2 ,1
A
820.433
B
50275.5
C
126.8339
1000D
140.273
E
3879660
KCO2 , 2
248.419
11862.4
38.92561
74.8996
1297999
K HA
KH 2S
10.937
16.112
0
0
0
0
0
0
0
0
K CaSO4
11.6592
2234.4
48.2309
K CaSO 4 2 H 2 O 815.978
26309.0
138.361
167.863
18.6143
K BaSO4
208.839
13084.5
32.4716
9.58318
2.58594
K SrSO4
89.6687
4033.3
16.0305
1.34671
31402.1
K FeCO3
21.804
56.448
16.8397
0.02298
K FeS
8.3102
K CaCO3
395.448
6461.5
71.558
180.28
24847
Ref.:
Haarberg
(1989)
Haarberg
(1989)
stvold
(1998)
Haarberg
(1989)
Haarberg
(1989)
Haarberg
(1989)
Haarberg
(1989)
stvold
(1998)
stvold
(1998)
Haarberg
(1989)
142613.6
+ 4229.195 log10 T 9.7384 T + 0.0129638 T 2
T
1.1506810 5 T 3 + 4.602 10 9 T 4 8908.483
log10 K H 2O (T ) =
Where
is the partial molar compressibility change of the reaction,
is the partial molar
volume change of the reaction and R is the universal gas constant.
for the sulfate
precipitation reactions is expressed by a third degree polynomial
 103
= a + bt + ct2 + dt3
Where t is the temperature in oC. The coefficients a, b, c and d for each of the sulfate
precipitation reactions are listed in the below table
Coefficient in compressibility change expression for sulfate mineral precipitation reactions.
in cm3 /mol/bar.
Units: t in oC and
a
17.54
17.83
16.13
17.83
BaSO4
SrSO4
CaSO4
CaSO42H2O
100b
1.159
1.159
0.944
1.543
1000c
17.77
17.77
16.52
16.01
106d
17.06
17.06
16.71
16.84
A
343.6
306.9
282.3
263.8
B
1.746
1.574
1.438
1.358
1000C
2.567
2.394
2.222
2.077
D
11.9
20
21.7
21.7
E
4
8.2
9.8
9.8
The partial molar volume changes of both of the acid equilibria of CO2 are (Haarberg, 1989)
ln M = z M2 F + ma (2 BMa + ZC Ma ) + mc 2 Mc + ma Mca +
a
c
a
m m
a ' a >a '
a'
Maa '
+ zM
m m C
c
ca
m m
c ' c >c '
c'
cc ' X '
+ zX
m m C
c
ca
for the anions. c denotes a cation species, whereas a denotes an anion species. m is the molality
(mols/kg solvent) and I is the ionic strength (mols/kg solvent)
I=
1
2
mi z i
2 i
z is the charge of the ion considered in the unit of elementary units. ijk is a model parameter
that is assigned to each cationcationanion triplet and to each cationanionanion triplet. The
remaining quantities in the activity coefficient equations are
I 1/ 2
2
+ ln 1 + bI 1 / 2
F = A
1/ 2
b
1 + bI
m m
c ' c >c '
c ' cc '
) + m m B'
ca
+ ma ma 'aa '
a ' a >a '
1
e2
1/ 2
A = (2N 0 d w )
3
4 0 DkT
3/ 2
N0 is the Avogadro number, dw is the water density, e is the elementary charge, D is the dielectric
constant of water and k is the Boltzman constant.
(0)
(1)
( 2)
BMX = MX
+ MX
g (1I 1 / 2 )+ MX
g ( 2 I 1 / 2 )
where
g (x ) =
(0)
2{1 (1 + x ) exp ( x )}
x2
(1)
ij , ij and ij
(2)
pair. 1 and 2 are constants, with 1 = 2 kg1/2 mol1/2 and 2 =12 kg1/2mol1/2. However, for
pairs of ions with charge +2 and 2, respectively, the value for 1 is 1.2 kg1/2mol1/2.
Further
Z = mi z i
i
CMX =
CMX
2 zM z X
1/ 2
ij = s ij + E ij (I ) + I E ij (I )
ij = s ij + E ij (I )
ij is a model parameter assigned to each cationcation pair and to each anionanion pair and
ij is an electrostatic term
ij =
zi z j
1
1
J (xij ) J (xii ) J (x jj )
4I
2
2
where
xij = 6 zi z j A I 1/ 2
Also the Pitzer model describes the activity of the water in terms of the osmotic coefficient
( 1) mi =
i
m m
c ' c >c '
c'
2 A I 3 / 2
1 + bI
1/ 2
+ mc ma Bca
+ ZCca +
c
where
(0)
(1)
( 2)
BMX
= MX
+ MX
exp( 1I 1 / 2 ) MX
exp( 2 I 1 / 2 )
and the relation between the osmotic coefficient and the activity of the water is
ln aH 2 O = M H 2 O mi
i
K+
0.12980
0.04810
0.00000
0.01070
0.12880
Mg++
0.00000
0.35090
0.21500
0.32900
0.00000
Ca++
0.17470
0.30530
0.20000
1.49800
0.40000
K+
0.32000
0.21870
1.10230
0.04780
1.43300
Mg++
0.00000
1.65100
3.36360
0.60720
0.00000
Ca++
0.23030
1.70800
3.19730
7.89900
5.30000
Sr++
0.00000
0.28370
0.20000
0.00000
0.00000
Sr++
0.00000
1.62600
3.19730
0.00000
0.00000
Ba++
0.17175
0.26280
0.20000
0.00000
0.00000
Ba++
1.20000
1.49630
3.19730
0.00000
0.00000
Fe++
0.00000
0.44790
4.70500
0.00000
1.91900
Fe++
0.00000
2.04300
17.00000
14.76000
5.13400
(2 ) parameters at 25C
H+
Na+
OH
0.00000 0.00000
Cl0.00000 0.00000
SO4
0.00000 0.00000
HCO3 0.00000 0.00000
CO30.00000 0.00000
HS
0.00000 0.00000
K+
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
C parameters at 25C
H+
Na+
OH0.00000
0.00410
Cl
0.00080
0.00127
SO4
0.04380
0.00571
HCO3 0.00000
0.00000
CO3
0.00000
0.00520
HSS
parameters at 25C
H+
H0.00000
Na+
0.03600
+
K
0.00500
++
Mg
0.10000
Ca++
0.06120
++
Sr
0.06500
Ba++
0.00000
OHClSO4HCO3CO3
Mg++
0.00000
0.00000
32.74000
0.00000
0.00000
0.00000
K+
0.00410
0.00079
0.01880
0.00000
0.00050
Mg++
0.00000
0.00651
0.02797
0.00000
0.00000
Ca++
0.00000
0.00000
54.24000
0.00000
879.20000
0.00000
Ca++
0.00000
0.00215
0.00000
0.00000
0.00000
Sr++
0.00000
0.00000
54.24000
0.00000
0.00000
0.00000
Sr++
0.00000
0.00089
0.00000
0.00000
0.00000
Ba++
0.00000
0.00000
54.24000
0.00000
0.00000
0.00000
Ba++
0.00000
0.01938
0.00000
0.00000
0.00000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
Na+
K+
Mg++
Ca++
Sr++
Ba++
0.00000
0.01200
0.07000
0.07000
0.05100
0.06700
0.00000
0.00000
0.03200
0.00000
0.00000
0.00000
0.00700
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
Cl
SO4
HCO3
CO3
0.00000
0.08900
0.00000
OH0.00000
0.05000
0.01300
0.00000
0.10000
0.00000
0.02000 0.00000
0.03590 0.01000
0.05300 0.02000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
parameters at 25C
Na+
K+
Mg++
Ca++
Sr++
Ba++
0.00000
0.00180
0.01200
0.00700
0.00210
0.01200
0.00000
0.02200
0.02500
0.00000
0.00000
0.00000
0.01200
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
OH
ClSO4HCO3CO3Cation 1 fixed as Mg++
Na+
K+
Mg++
Ca++
Sr++
Ba++
0.00000
0.01000
0.01500
0.05500
0.00000
0.00000
0.00000
0.04800
0.00000
0.00000
0.00000
0.00000
0.02400
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
Na+
K+
Mg++
Ca++
Sr++
Ba++
0.00000
0.00300
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
Na+
K+
Mg++
Ca++
Sr++
Ba++
0.00000
0.00300
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
Cl
SO4
HCO3
CO3
0.00000
0.00140
0.01430
0.00000
0.00000
0.00500
0.00500
0.00000
0.00000
0.00000
OH0.00000
0.00600
0.00900
0.00000
0.01700
OH0.00000
0.00800
0.05000
0.00000
0.01000
OH
Cl
SO4
HCO3
CO3
0.00000
0.00000
0.00000
0.02400
0.00000
0.00000
0.00900
0.00000
0.03600
0.00000
SO4
HCO3
Cl
CO3
OH
ClSO4HCO3CO3
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00400
0.09600
0.00000
OH0.00000
0.02500
0.00000
0.00000
0.00000
0.00000
0.16100
0.00000
0.00000
0.00000
0.00000
Cl
SO4
HCO3
CO3
0.00000
0.01800
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
ijk
X (T ) = X (298.15) +
2
X
(T 298.15)+ 1 X2 (T 298.15)2
T
2 T
Due to the appearance of Na and Cl in many systems, Pitzer et al. (1984) have developed a more
sophisticated description of the temperature dependence of the parameters for these species. Also
a pressure dependence is included in the description. The functional form is for temperatures in K
Q1
+ Q2 + Q3 P + Q4 ln (T ) + (Q5 + Q6 P )T
T
Q + Q10 P Q11 + Q12 P
+ (Q7 + Q8 P )T 2 + 9
+
T 227
680 T
X (T ) =
and
H+
0.00000
0.18133
0.00000
0.00000
0.00000
Na+
0.01879
0.007159
0.16313
0.10000
0.17900
K+
0.00000
0.03579
0.09475
0.10000
0.11000
Mg++
0.000000
0.05311
0.00730
0.00000
0.00000
x 100.
Ca++
0.00000
0.02124
0.00000
0.00000
0.00000
Sr++
0.00000
0.02493
0.00000
0.00000
0.00000
Ba++
0.00000
0.06410
0.00000
0.00000
0.00000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
Ca++
0.00000
0.00057
0.00000
0.00000
0.00000
Sr++
0.00000
0.00621
0.00000
0.00000
0.00000
Ba++
0.00000
0.00000
0.00000
0.00000
0.00000
Ca++
0.00000
0.36820
5.46000
0.00000
0.00000
Sr++
0.00000
0.20490
5.46000
0.00000
0.00000
Ba++
0.00000
0.32000
5.46000
0.00000
0.00000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
x 100.
Mg++
0.00000
0.00074
0.00901
0.00000
0.00000
Ca++
0.00000
0.00232
0.00000
0.00000
0.00000
Sr++
0.00000
0.05000
0.00000
0.00000
0.00000
Ba++
0.00000
0.00000
0.00000
0.00000
0.00000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
Mg++
0.00000
0.00000
0.06100
0.00000
0.00000
Ca++
0.00000
0.00000
0.51600
0.00000
0.00000
Sr++
0.00000
0.00000
0.51600
0.00000
0.00000
Ba++
0.00000
0.00000
0.51600
0.00000
0.00000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
Mg++
0.00000
0.00000
0.01300
0.00000
0.00000
Ca++
0.00000
0.00000
0.00000
0.00000
0.00000
Sr++
0.00000
0.00000
0.00000
0.00000
0.00000
Ba++
0.00000
0.00000
0.00000
0.00000
0.00000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
Ca++
Sr++
Ba++
Fe++
OHClSO4HCO3CO3HS
H+
0.00000
0.00376
0.00000
0.00000
0.00000
Na+
0.00003
0.00150
0.00115
0.00192
0.00263
K+
0.00000
0.00025
0.00008
0.00000
0.00102
Mg++
0.00000
0.00038
0.00094
0.00000
0.00000
x 100.
Mg++
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
OHClSO4HCO3CO3HS
0.00000
0.00590
0.00000
0.00000
0.00000
0.00790
0.01050
0.36300
0.00000
0.00000
0.00000
0.00400
0.00625
0.00000
0.00000
0.00000
0.01990
0.02950
0.00000
0.00000
0.00000
0.01300
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.01540
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
x 100.
Mg++
0.00000
0.00018
0.00010
0.00000
0.00000
Ca++
0.00000
0.00005
0.00000
0.00000
0.00000
Sr++
0.00000
0.00000
0.00000
0.00000
0.00000
Ba++
0.00000
0.00000
0.00000
0.00000
0.00000
Fe++
0.00000
0.00000
0.00000
0.00000
0.00000
Temperature coefficients in expression for temperature dependence of the Pitzer parameters for
NaCl
NaCl(1)
NaCl(0)
C NaCl
Q1
6.1084589
6.5684518102
1.1931966102
1
1
Q2
2.48691295010
4.830932710
4.0217793101
Q3
0
5.381275267105
2.2902837105
Q4
4.4640952
0
7.5354649104
Q5
1.4068095103
1531767295104
1.110991383102
7
Q6
0
2.65733990610
9.0550901108
Q7
0
5309012889106
1.53860082108
Q8
0
8.6340233251010
8.692661011
Q9
1.579365943
4.2345814
3.53104136101
Q10
0
0.0022022820790103
4.3314252104
Q11
9.706578079
0
9.187145529102
Q12
0
2.686039622102
5.190477104
The coefficients correspond to units of pressure and temperature in bars and Kelvin, respectively.
Reference: Pitzer (1984)
Calculation procedure
The amount of minerals that precipitate from a specified aqueous solution is evaluated by
calculating the amount of ions that stay in solution when equilibrium has established. This
amount is given as the solution to the system of thermodynamic equilibrium constant equations.
Only the solubility products of the salts precipitating, need be fulfilled. Solving the system of
equations is an iterative process
The thermodynamic equilibrium constants are calculated for the specified solution at the
specified set of conditions, pressure and temperature.
The stoichiometric equilibrium constants are calculated from the thermodynamic ones and
from the activity coefficients.
The ratio of CO2(aq) to H2S(aq) is calculated. This determines if any of the ferrous iron
minerals FeCO3 and FeS will precipitate. Only one can precipitate, since both H2S and CO2
are fixed in concentration, and then the Fe++ concentration cannot fulfil both solubility
products at the same time.
The equilibrium in the acid/base reactions is determined without considering the precipitation
reactions. The convergence criterion is that the charge balance must be fulfilled.
The ion product of the iron mineral identified at a previous step is checked against the
solubility product. If the solubility product is exceeded, the amount of precipitate of the iron
mineral is determined. The convergence criterion in this iteration is the charge balance.
Precipitation of calcium carbonate is not included in the calculation.
The ion product of calcium carbonate is checked against its solubility product. If the
solubility product is exceeded, simultaneous precipitation of calcium carbonate and the iron
mineral is calculated. A double loop iteration is applied. The inner loop: With a given
amount of ferrous iron mineral precipitation (which comes from the outer loop), the amount
of calcium carbonate precipitate is determined. During the calcium carbonate precipitation,
the sulfate precipitate is influenced since some Ca++ is removed from the solution. The state
in the sulfate system is therefore corrected in each of these inner loop iterations. In the inner
loop, the charge balance is used to check for convergence. The outer loop: The iteration
variable is the amount of ferrous iron mineral precipitate. Convergence is achieved when the
ion product of the ferrous mineral matches the thermodynamic solubility product.
The resulting amount of each precipitate is compared to that of the previous iteration. If the
weighted sum of relative changes in the amounts of precipitates exceeds 106, then all activity
coefficients are recalculated from Pitzers activity coefficient model for electrolytes. The
procedure is then repeated from the 3rd step.
References
Atkinson, A. and Mecik, M., The Chemistry of Scale Prediction, Journal of Petroleum Science
and Engineering 17 (1997) pp. 113121.
Haarberg, T. Mineral Deposition During Oil Recovery, Ph.D. Thesis, Department of Inorganic
Chemistry, Trondheim, Norway (1989).
Haarberg, T., Jakobsen, J.E., and stvold, T., The effect of Ferrous Iron on Mineral Scaling
During Oil Recovery, Acta Chemica Scandinavia 44 (1990) pp. 907915.
Kaasa, B. and stvold, T., Prediction of pH and Mineral Scaling in Waters with Varying Ionic
Strength Containing CO2 and H2S for 0<T(oC)<200 and 1<p(bar)<500 Presented at the
conference Advances in Solving Oilfield Scaling held January 28 and 29, 1998 in Aberdeen,
Scotland.
Pitzer, K.S., Thermodynamics of Electrolytes I. Theoretical basis and general equations,
Journal of Physical Chemistry 77 (1973) pp. 268277.
Pitzer, K.S., Thermodynamics of Electrolytes V. Effects of HigherOrder Electrostatic Terms,
Journal of Solution Chemistry 4 (1975) pp. 249265.
Pitzer, K.S., Theory: Ion Interaction Approach. Activity Coefficients in Electrolyte Solutions,
Book by Pytkowicz, R.M., pp. 157208, CRC Press, Boca Raton, Florida (1979).
Pitzer, K.S., Peiper, J.C. and Busey, R.H., Thermodynamic Properties of Aqueous Sodium
Chloride Solutions. Journal of Physical Chemistry 13 (1984) pp. 1102.
Pitzer, K.S., Theoretical Considerations of Solubility with Emphasis on Mixed Aqueous
Electrolytes, Pure and Applied Chemistry 58 (1986) pp. 15991610.
Pitzer, K.S., Thermodynamics 3. edition, McGrawHill, Inc. (1995).
The equation states that under the above assumptions, the temperature Tx at a given position x can
be calculated based on the mass flowrate , the heat capacities Cp, the pipeline diameter D, and
the overall heat transfer coefficient Utot. Tamb is the ambient temperature, while Tin is the fluid
temperature at the pipeline inlet. This expression may be exploited to optimize the discretization
of the pipeline by assigning section lengths in such a way that the temperature only declines a
predefined amount in each section. This results in short section lengths near the inlet, while
sections are longer further down the pipeline where the temperature changes less.
Energy balance
The principles of the algorithm are illustrated in the below figure. The inlet conditions to a
section, such as mass flowrate, temperature, pressure and composition are known. Also, the
pipeline specifications such as insulation and temperature of the surroundings are known. This
allows the program to calculate heat loss from the pipeline, enthalpy of the exiting fluid, and
pipewall temperatures. A steady state flow model, OLGAS 2000, calculates pressure drop, flow
regime, and liquid holdup, based on information about phase equilibria and viscosity passed from
the thermodynamic models in PVTsim. Subsequently the wax model in PVTsim is used to
calculate wax concentrations, which are needed to determine deposition on the walls of the
pipeline. Knowing pressure, enthalpy, and feed composition at the outlet of the section, an
integrated waxPH flash is used to calculate the temperature and phase compositions. These
values are then used as inlet conditions for the next section. This proceeds until the calculation
has been completed for the entire pipeline in the current timestep. Subsequent timesteps are
calculated similarly, the only change from one timestep to the next being that the pipeline
diameter and insulation have changed due to a layer of deposited wax on the pipewall.
Tamb
Ti, Pi
To , P o
mi
Hi
Ho
Q=UAT
The structure of the algorithm, as described above, can be summarized by the following four
points that are further illustrated in the above figure
Heat balance, HO = Hi (Q + W)
Pressure drop and flow regime, OLGAS 2000 PO
Wax flash at wall and deposition
PHwax flash, (PO, HO) TO
The energy balance determines the total fluid enthalpy at the outlet of the section based on the
enthalpy of the fluid entering into the section. This enthalpy is known from the last section. For
the first section, the inlet enthalpy is obtained from a PTflash. The enthalpy of the fluid exiting
the section depends on the amount of heat transferred through the pipeline walls (Q) and the
work done (W) due to changes in elevation. The work term, which becomes significant for
instance in a riser, is calculated from the following equation.
W = bulk g h
In this equation
is the average bulk fluid density, g is the gravitational acceleration and h
the elevation change.
ln i
NLAY
r
1
1
+ i i1,i1 +
U in = rin1
rin hin i =1 k
rout hout
In this equation, the heat transfer coefficient is referred to the inner radius of the pipeline rin. ki1,i
is the thermal conductivity of the layer between the radii ri1 and ri. Deposited wax is included as
an additional layer at radius rwax = rin xwax, where xwax is the deposit layer thickness. hin and hout
are the inside film heat transfer coefficient and outside film heat transfer coefficient, respectively.
For a more detailed description of this, please refer to example 9.61 of the text by Bird et al.
(1960).
hin D
k
where k is the thermal conductivity of the fluid and D is the inside diameter of the pipeline. The
Nusselt number has been related to the Reynolds and Prandtl numbers through different
correlations depending on flow regime. Four sets of correlations are available, of which the
SiederTate and DittusBlter are probably the most popular and the Petukov/ESDU
(recommended set) the most reliable and well documented.
SiederTate
N Re > 10
N Nu = 0.027 N N
0.8
Re
1/ 3
Pr
0.8
Re
b
w
1/ 3
Pr
0.25
6 10 5 b
1
1.8
N Re
w
0.25
N Re > 2300 :
1/ 3
DittusBlter
N Re > 10
N Nu = 0.023 N N
0.8
Re
0.3
Pr
b
w
0.25
6 10 5 b
1/ 3
N Re < 2300 :
N Nu = max 0.184( N Gr N Pr ) , 3.66
4
0.8
Re
0.25
0.3
Pr
PetukovGnielinski
2300 < N Re :
(N Re 1000)N Pr
8
N Nu =
2/3
1+12.7
N Pr 1
8
where =
N Re > 2300 :
D 2 / 3 b
1+
L w
0.25
1
(1.82 log N Re 1.64)2
0.0677 N re N Pr
N Nu = 3.657 +
1 + 0.1N Pr N Re
4/3
L
0.3
D
0.25
b
w
b
( f / 2 )N Re N Pr
N Nu =
1/ 2
2/3
1.07 + 12.7( f / 2 ) (N Pr 1) w
f=
N Re < 2300 :
0.25
1
2
4(1.82 log N Re 1.64)
1/ 3
0 .7
)]
3
1/ 3
0.25
= 1.33
N Re
6000
Depending on the insulation of the pipeline, the film heat transfer coefficient may strongly affect
the wax deposition calculation through the temperature gradient over the laminar film layer
inside the pipeline. For that reason, the choice of correlation is important.
f=
64
N Re
1/ 3
10 6
4
+
turb : f = 0.0055 1+ 2 10
D N Re
S wet = 2 r
sin
HOL =
2
Here HOL is the holdup, is the angle corresponding to the part of the pipeline circumference
swept by liquid and Swet is the wetted perimeter corresponding to that angle. The wetted
perimeter is corrected for presence of an aqueous phase based on the phase volume fractions.
phases and an aqueous phase. The wax deposition can only take place from the hydrocarbon
wetted part of the inner pipewall.
Wax deposition
Wax deposition from the oil phase is always considered. Furthermore it is optional whether or
not wax deposition from the gas phase should be considered. The wax deposition mechanisms
considered for the gas and oil phases are molecular diffusion and shear dispersion.
The volume rate of wax deposited by molecular diffusion for a given wax forming component i is
calculated from the relation
Vol
diff
wax
NWAX
i =1
where is the molar concentration of wax component i in the bulk phase and is the molar
concentration of wax component i in the phase at the wall. Swet is the fraction of the perimeter
wetted by the current phase. NWAX is the number of wax components, Mi the molecular weight
and i the density of wax component i. L is the length of the pipeline section and r the current
inner pipeline radius considering wax deposition.
The thickness of the laminar film layer
(1913) expression
= 58 D N Re8
where is a user defined thickness correction factor. The allowed values of are between 0 and
100. The introduction of provides the user with the possibility of tuning a predicted thickness
of a wax layer to experimental data, since a very narrow film layer will result in an increase in
wax deposition and vice versa.
The diffusion coefficient, Di of the wax forming component is calculated from a correlation by
Hayduk and Minhas (1982).
Di = 13.3 10
where
12
1.47
10.2
M w , wax ,i
wax ,i
0.791
M w,wax ,i
,
wax
i
0.71
are between 0
and 100. The introduction of provides the user with yet another possibility of tuning a
predicted wax layer thickness to experimental data, since a large diffusion coefficient for a given
wax component will result in an increased deposition of that particular component and vice
versa.
For systems with a large oil fraction, it is generally expected that deposition is dominated by oil
phase deposition to an extent where contributions from the gas phase are negligible. For rich
gases and lean condensate systems, it may however be of interest to include contributions from
the gas as well. The model considers wax deposition from the gas phase as results of both
molecular diffusion and shear dispersion. The same assumptions are used as for the oil phase.
Whether wax deposition from the gas phase should be considered or not is selected on the
Simulation Options menu.
Shear dispersion accounts for deposition of wax already precipitated in the bulk phase. The
volume rate of wax deposited from shear dispersion is estimated from the following correlation
of Burger et al. (1981)
shear
Volwax
=
k *cwallA
wax
where k* is a shear deposition rate constant, Cwall is the volume fraction of deposited wax in the
oil in the bulk, is the shear rate at the wall, A is the surface area available for deposition and
wax is the average density of the wax precipitated in the bulk phase. The shear dispersion
mechanism is often assumed to be negligible as compared with molecular diffusion (Brown et al.
(1993) and Hamouda (1995)). Therefore the allowed values of k* is set to [0;0.0001 g/cm2] or
[0;0.025 lb/ft2] or [0;0.001 kg/m2].
Boost pressure
It is possible to specify a pressure increase or boost pressure at the entrance of each user
specified segment. The boost pressure may originate from a pump or a compressor, which is
located between two sections.
Porosity
The porosity of the deposited wax is understood as the space between the wax crystals occupied
by captured oil. This porosity is reported to be quite significant in many cases (70%) and to
depend on the shear rate. The program has the possibility of treating the porosity as a constant or
to depend linearly on shear rate. The expression used is:
=A +B
In this expression, is the porosity and the shear rate. The constants A and B are determined
from two input data points of shear rate and corresponding porosity. If a constant porosity is to be
used, A = 0 and B is the constant porosity value.
Boundary conditions
By boundary conditions is understood the fluid inlet specifications to the pipeline. This includes
pressure, temperature, flow rate and fluid composition. One or more boundary conditions may be
changed during the simulation at specified timesteps. In case the inlet composition is to be
changed.
Mass Sources
A mass source in this context is understood as a side stream to the pipeline. Mass sources may be
defined to enter in a specified segment inlet in a given timestep. Mass sources cannot be
specified to enter in the first segment. In this case a change of boundary conditions may be
specified instead. Temperature and flow rate of the source are specified. The pressure in the
source is assumed to be equal to that of the fluid at the current position in the pipeline. The fluid
composition for the source is specified by referring to a fluid in the current fluid database. It is
possible to change conditions for the source in a later timestep, or to change the composition of
that source. The source composition is mixed into the main pipeline stream, and a PHflash
determines the phase distribution and temperature of the mixed stream. This is done by first
determining the enthalpy of the source through a PTflash and then determine the mixture
enthalpy based on the molar flow rates. Fluids entered as sources must be characterized to the
same pseudocomponents as the original fluid in the simulation.
References
Bendiksen, K.H., Maines, D., Moe, R., Nuland, S.: SPE 19451, (1991), SPE Production
Engineering, May, pp. 171180.
Bird, R.B., Steward, W.E., Lightfoot, E.N., Transport Phenomena, (1960), Wiley, NY. pp. 28628.
Blasius, H., Das hnlichkeitsgesetz bei Reibungsvorgngen in Flssigkeiten, Forch. Ver. Deut.
Ing. 131, 1913.
Brown, T.S., Niesen, V.G. and Erickson, D.P., Measurement and Prediction of Kinetics of
Paraffin Deposition, SPE 26548, 68th Annual Technical Conference and Exhibition of SPE
Houston, Tx, 36 October, 1993.
Burger, E.D., Perkins, T.K. and Striegler, I.H., Studies of Wax Deposition in the Trans Alaska
Pipeline, Journal of Petroleum Technology, June 1981, 19751086.
ESDU 93018 and 92003: Forced convection heat transfer in straight tubes, ESDU 1993.
Hamouda, A., An Approach for Simulation of Paraffin Deposition in Pipelines as a Function of
Flow Characteristics with a Reference to Teeside Oil Pipeline, SPE 28966 (1995), presented at
SPE Int. Symposium on Oilfield Chemistry, San Antonio, 1417 February 1995.
Hayduk, W. and Minhas, B.S., Correlations for Predictions of Molecular Diffusivities in
Liquids, The Canadian Journal of Chemical Engineering, 60, 1982, pp. 295299.
Lindeloff, N. and Krejbjerg, K., Compositional Simulation of Wax Deposition in Pipelines:
Examples of Application, Presented at Multiphase 01, Cannes, France, June 1315, 2001.
Lindeloff, N. and Krejbjerg, K., A Compositional Model Simulating Wax Deposition in
Pipeline Systems, Energy & Fuels, 16, pp. 887891, 2002.
Szilas, A.P.: Production and Transport of Oil and Gas, part B, 2. Ed. Developments in
Petroleum Science, 18B, (1986), Elsevier, Amsterdam.
Cleaning Procedure
In order to use the Mud module, the following compositional data are needed
Composition of contaminated reservoir fluid. It is customary to analyze to either C7+, C10+,
C20+, or C36+.
Composition of base oil contaminate. It will usually consist of components in the carbon
number range C11 C30.
Weight% contaminate in stock tank oil (optional for compositions to C36+)
The cleaning procedure will differ depending on the extent of the compositional analysis
Reservoir fluids to C7+ or C10+
The base oil contaminate will seldom contain components lighter than C11. With a composition to
C7+ or C10+ all base oil contaminate will be contained in the plus fraction of the contaminated
reservoir fluid. The base oil affects molar amount, density and molecular weights of the plus
fraction. The weight% contaminate in the oil from a flash of the contaminated reservoir fluid to
standard conditions is required input.
After the contraction of the contaminated reservoir fluid composition the cleaning procedure is
the same as for a C7+ or a C10+ composition.
Reservoir fluids to C36+
With a composition to C36+ the carbon number fraction C7C10 will usually be free of
contamination and the same will be the case for the fractions C30C36. This allows the degree of
contamination to be estimated.
For a clean reservoir fluid PVTsim assumes the following relation between the mol fraction (z)
of C7+ fractions and carbon number i.
ln zi = A + B CN i
A and B are estimated by a fit to mol%s for C7+ mol fractions against carbon number.
The above relation will not apply for fractions contaminated by base oil, but it will still be true
for uncontaminated C7+ fractions. A and B may be determined by a linear fit to zi versus CNi,
where i stands for uncontaminated C7+ fractions. Using A and B, the mol fractions of the
remaining C7+ fractions in the uncontaminated fluid may be estimated. The remaining molar
amount of each carbon number fraction is assumed to originate from the base oil, which enables
the composition of the contaminate to be estimated. The estimated base oil composition will not
necessarily be identical to the input composition.
Chapter 1
Regression on the characterized contaminated fluid is also an option, in which case the same
regression parameters are used as with ordinary regression for characterized fluids. To allow the
program to identify the mud components in the contaminated fluids, the characterized mud must
be saved in the database prior to the regression and selected as mud comtaminate in the Clean for
Mud menu. The result of the regression is a cleaned, tuned and characterized reservoir fluid
composition.
Chapter 1