Compositional Simulation

CHAPTER 5
Compositional Simulation

Introduction
Applications
PVT Behavior
Equations for Compositional Simulation
Regression of EOS Parameters
Modeling of Surface Facilities
Lab Tests
Gas Cycling in Volatile Oil and Gas Condensate
Reservoirs
Vaporizing Gas Drive (VGD)
Field Examples
Fundamentals
Heavy Fraction Characterization
Pseudoization
Regrouping Scheme of Heavy Components
Heavy Ends Characterization
Default Fluid Predictions
Pre - Regression
Component Pseudoization
Regression
Initialization Data for VIP
Run Data for VIP
Data Files

PVT Matching

Fundamentals
Heavy Fraction Characterization
Pseudoization
Regrouping Scheme of Heavy Components
Heavy Ends Characterization
Default Fluid Predictions
Pre - regression
Component Pseudoization
Regression
Initialization Data For VIP
Run Data For VI
Data Files
Heavy.dat
Heavy.eos
Default.dat
Regression.dat
Pseudo.dat
Pseudo.eos
Final.dat
Final.eos
Spe3i.dat
Spe3r.dat

1

1

Introduction

In compositional simulation of oil and gas reservoirs equation-of-state (EOS) methods are
seeing increasing usage over more traditional K-value methods for phase equilibrium calculations. It
has been found by several authors that equations-of-state are, in general, not able to accurately
predict reservoir fluid behavior using theoretical EOS parameters. It has been found that "tuning"
the EOS (by modifying the EOS parameters) is required to adequately match laboratory derived
PVT data.
In this chapter, we describe how an equation-of-state is "tuned" to match laboratory data
using nonlinear regression. We present partial justification for modification of EOS parameters and
then discuss some practical guidelines for efficient and meaningful fluid characterization. Although
we will mention component pseudoization (lumping) since it is usually part of the fluid
characterization process, the emphasis of this section will be on regression and not on pseudoization.
Most of the material presented in this text was obtained from Coats and Smart.
1

2
Applications

A compositional simulator is used in a variety of situations in which a black oil simulator
does not adequately describe the fluid behavior. Each of the applications listed requires a special
phase behavior computation and may require special laboratory analysis to determine the behavior
of the fluids. Some applications are as follows:

A) Volatile oil (
o
API changes, oil vaporizes)
B) Miscible, EOR (CO
2
, LPG)
C) Condensate (cycling, oil flows)
D) Gas cycling (injection of dry gas)
E) Injecting new fluid (N
2
)

PVT Behavior

The PVT behavior of the reservoir fluids must be well understood in order to do a competent
job of compositional simulation. We will briefly review some of the fundamentals. At any
particular time, a mixture of fluids may be in a vapor-liquid equilibrium at a particular temperature
and pressure. For most reservoirs the temperature can be considered to be constant. The variation of
the equilibrium with pressure and also with composition is also important. When the vapor and
liquid phases are in equilibrium, the molecules of each component are condensing and vaporizing at
the same rate as shown in Fig. 1. The equilibrium can be described by x, y, and z values for all the
components. A K-value can be calculated for the equilibrium for each component.

3

y = 1.0 y = mole fraction, vapor phase
i i

x
=
i
1.0
x
= mole fraction, liquid phase
i

z
= 1.0
z
= mole fraction, total
i i

i
i
i
k
=
y
x

Fig. 1

Vapor
Liquid
equilibrium

Prediction of PVT Behavior. The K-values have different patterns that can be observed. It is
common to see a log-log plot of K vs. p which shows a family of curves tending to converge to a K-
value of 1.0 at a "convergence pressure" as in Fig. 2. The behavior of this family of curves depends
on the composition as well as on temperature and pressure. For example, the amount of intermediate
hydrocarbons has a significant effect on convergence pressure.

4
Log k
Log pressure
p
k
1.0
c
1
c
2
c
3
c
4
c
5
c
n
Leaner Mixture
Lower c
2
- c
5

Fig. 2

Several methods are commonly used in compositional simulators to model the phase
behavior. K-value methods can be used with a systematic method of modifying convergence
pressures with compositions. Equation of state methods are also in common usage and have the
advantage of predicting densities as well as the molecular equilibrium.

Much research has been directed toward predicting phase behavior given temperature,
pressure and composition. However, for reservoir simulation studies laboratory composition data is
almost always available for the mixtures expected to occur in the reservoir. For that reason, it is not
as important to predict fluid behavior because the computational procedure must be modified to
match the laboratory data. All of the phase behavior methods use various parameters that can be
adjusted to accomplish the match of the laboratory data. It should be noted that a simulator using an
equation of state method can run as much as ten times slower than a simulator using the K-value
methods. This loss of computer efficiency tends to offset the advantages of the equation of state
methods. Some simulation engineers use equation of state methods to match laboratory data and
then convert to K-value methods with independent density calculation methods for the reservoir
simulation runs.

PVT Prediction Methods. The methods used for PVT prediction are as follows:

1) K-value, p
k
(convergence pressure) method.(This method is faster and can be matched with
lab data and EOS , but it needs correction for the density of oil and the density of gas)
2) Equation of State method (EOS).( It matches more accurate with the lab data and we can get
the density of oil and gas)
a) Peng-Robinson
b) Redlich-Kwong

5

Equations for Compositional Simulation

Mass Balance. The basic equations for compositional simulation are mass balance equations for
each of the components. In addition to these mass balances, it is required that saturations sum to 1.0
and the phase behavior be consistent with temperature, pressure and compositions at each cell. The
mass balance equation and other conditions applicable for component i are:

1) Mass balance equation :

( y
u )+ ( x u )= -
t
[ ( S x + S
y )]
i g g
i
o
o o
o
i g
g i
$
$ $
$

r r
(1)

vapor f l ow + l i qui d f l ow = st or age

i
i
y ,
x
= mass fractions of vapor, liquid $
$

where

2) The saturation relationship is:

S
= 1 . 0
o g
, (2)

3) Darcy's Law and Martin's Equation apply.

4) The phase behavior is:

o g o g i i
i
, , , = f(p, T, x y z ) , , (3)

Equations of State. An equation-of-state (EOS) is an equation which expresses the relationship
between pressure, temperature and volume of a gas or liquid. These equations are usually of cubic
form. Two EOS widely used in the petroleum industry today are the Peng-Robinson
2
(PR) and
Soave-Redlich-Kwong
3
(SRK) EOS.

Van der Waals Equation. One of the earliest attempts to represent the behavior of real gases by an
EOS was by van der Waals. His equation was:
(p + a / V )(V - b) = RT
M
M
2
(4)

6
where a and b are characteristics of the particular gas, R is the universal gas constant, and V
M
is the
molar volume.

Peng - Robinson Equation. Peng and Robinson later proposed the following:

[p +
a
V
(
V
+ b) + b(
V
- b)
](
V
- b) = RT
T
M M M
M
(5)

The coefficients for Eq.5 are calculated by:

T c a
=
a

(6)
C
2
C
2
C
A = 0.45724
R T
P
(7)
1/ 2 1/ 2
r
= 1 + m(1 -
T
)
(8)
b = 0.07780
RT
p
c
c
(9)

m = 0.37464 + 1.54226 - 0.26992
2

(10)

Soave-Redlich-Kwong Equation. The Soave-Redlich-Kwong equation has the form:

[p +
a
V
(
V
+ b)
](
V
- b) = RT
T
M M
M
(11)

c
2
2
c
c
a
= 0.42747
R T
p
(12)

b = 0.08664
RT
p
c
c
(13)

7
m = 0.480 + 1.574 - 0.176
2

(14)

where the Pitzer acentric factor is given by:

= - ( p + 1)
vr
log
(15)

Here p
vr
is the reduced vapor pressure, evaluated at T
r
=0.7.

These EOS contain coefficients and parameters which have theoretically derived values. For
example, the theoretical values of
a
,
b
in the PR and SRK EOS are roughly 0.4572, 0.0778 and
0.4275, 0.0866, respectively. EOS also contain parameters, called binary interaction coefficients,
which govern the interaction between all pairs of components in the mixture. These also have
theoretically or experimentally determined values.

8
Regression of EOS Parameters.

It has been found by several authors that, in general, EOS with theoretical parameter values
are incapable of accurately predicting fluid behavior of hydrocarbon mixtures commonly
encountered in oil and gas reservoirs. Flexibility in the use of the EOS can be obtained by altering
the EOS parameters or allowing them to change by component. The parameters most commonly
changed are
a
,
b
and the binary interaction coefficients. By changing these parameters, the EOS
can be forced to match laboratory measured data.

Although EOS parameters can be modified manually, they are usually determined using a
nonlinear regression algorithm. One algorithm that is used in commercial software packages is an
extension of the least-squares, linear programming (LSLP) method.
4
The linear programming
algorithm places global upper and lower limits on each regression variable v
i
. Subject to these limits,
the regression determines values of {v
i
} that minimize the objective function F defined as

9
j
F =
W d d d
j=1
n
j j jC
j
| ( - ) / |
(16)

where d
jC
and d
j
are calculated and observed values of observation j, respectively, and W
is a weighting factor.

Using the regression algorithm above and the guidelines contained in this text, it has been
found that agreement between laboratory data and regressed EOS results is generally good to
excellent.

Justification for Altering Theoretical EOS Parameters. Thus far, no justification has been given
for alteration of the theoretical values of EOS parameters. In the following sections we present
partial justification for the modification of EOS parameters for methane and the plus fraction.

Methane Fraction. The theoretical
a
and
b
values in both the PR and SRK EOS are based on the
van der Waals conditions of p/V=
2
p/V
2
=0 at the critical point. The component temperature
functions in the SRK and PR EOS essentially reflect satisfaction of pure-component density and
vapor-pressure data below critical temperature. At reservoir conditions, methane is usually well
above its critical point, where there is no theory or clear-cut guide to selection or alteration of 's.
Since the theoretical values are based on p and T conditions far removed from the p and T
conditions of interest, one could argue that the 's for methane should be determined from matching
of laboratory-derived PVT data in the p and T ranges of interest.

Plus Fraction. The primary argument for altering EOS parameters of the plus fraction is that the
plus fraction, unlike all other components, is a mixture of many components.

General Procedure for Fluid Characterization. In this section, we describe the procedure
commonly used to characterize a particular fluid for use in compositional simulation. Computer
programs (such as Integrated Technologies' EOS-PAK) are available commercially which can

10
facilitate greatly the computations required in the following steps. Some or all of the following steps
may be employed in a typical fluid characterization.
1. Decide if and how to split the plus fraction into extended fractions. The actual splitting is usually
done using the method presented by Whitson
5
. There is some debate as to whether the mole
fractions of the extended fractions should be roughly equal or decreasing (e.g., exponentially) with
increasing molecular weight.

2. Pseudoize (lump) the components into a fewer number of pseudo-components. This is done
primarily for economic reasons (fewer components results in faster run times and lower costs). The
focus of this section is not on pseudoization; however, there are references available which offer
some guidelines for optimal pseudoization.
6

3. Match the laboratory expansion data using a nonlinear regression algorithm. The regression data
set to be matched may include results from multiple fluid samples and results from tests run at
different temperatures. In general, the regression data set should include all measured laboratory
data. Any calculated data that are reported as part of a fluid analysis should not be included. The
regression variable set should be chosen such that it possesses the characteristics of an optimal
regression variable set.

4. Match the laboratory viscosity data using a nonlinear regression algorithm. The regression data
set should include all measured viscosity data. The regression variable set should include only the
critical z-factors (which affect viscosity only and do not affect the results of expansion tests).

5. At this point the EOS fluid characterization is complete, as all laboratory measured data have
been matched. However, additional adjustments to the fluid data may be required before
compositional simulation can be started. For example, it may be necessary to alter the initial fluid
composition to match the initial saturation pressure or gas-oil ratio observed in the field. In addition,
it may be necessary to determine the optimum separator conditions (only for predictive studies; for
history matching, the actual separator conditions should be used).

Practical Considerations for Minimizing Splitting of C
7+
Fraction. Coats and Smart found that
extensive splitting of the C
7+
fraction is not necessary to match laboratory data, in contrast to
previous authors. In general, none to four splits of the C
7+
fraction are all that are required. A
general, a priori guide to the extent of C
7+
splitting required can be obtained from the experimentally
observed range of C
7+
molecular weights during an expansion or multiple-contact test. The greater
the range of C
7+
molecular weights, the more C
7+
fractions are required.

In addition, the degree of splitting is also affected by the anticipated recovery process. For
example, gas cycling processes will require more splitting than depletion/waterflooding processes.

Minimizing Number of Regression Variables. Usually only five parameters -
a
and
b
for
methane and the heaviest fraction and the methane/heaviest binary interaction coefficient - are
necessary and sufficient for good matches of laboratory data. For the purposes of this discussion, the
"heaviest" fraction is the heaviest single fraction used in the EOS calculations. For example, if the

11
original plus fraction, C
7+
, is split into three components, say F
7
, F
8
, and F
9
, then F
9
is the new
heaviest fraction as defined above.
A general rule for selecting regression variables is to exclude any EOS parameter that, by
inspection, cannot affect significantly the calculated value of any of the regression data. For
example, if the fluid composition in a regression data set has a very small amount of a particular
component, then it would not make sense to include that particular component's 's or binaries as
regression variables. Conversely, it makes sense to include the EOS parameters of a component that
is predominate (such as methane) in the fluid compositions.

Regression of Surface Separation Data. Coats and Smart found that regressed EOS surface
separation results were the same regardless of whether surface separation data were included in the
regression data set. In other words, an EOS tuned with only data at reservoir conditions will, in
general, adequately predict surface separation data.

They did not address the opposite situation, i.e. whether including the surface separation data
in the regression data set could harm the match of data at reservoir conditions. They implied that
there was nothing wrong with including the separation data, but that it was not necessary. Of course,
this applies only if the surface separation data are measured in the laboratory. Surface separation
data presented in standard fluid analysis reports are commonly calculated (using K-values) and not
measured. In general, calculated data (such as surface separation data) should not be included in a
regression data set; only measured data should be included.

Characteristics of an Optimal Regression Variable Set. The characteristics of an optimal
regression variable set are:

1. The regression converges,

2. The values that the variables converge upon are realistic (reasonably close to theoretical values),

3. Deletion of any member of the variable set results in either or both of (a) significantly worse data
match and (b) unrealistic variable values, and
4. Addition of any other EOS parameter results in either or both of (a) nonconvergence and (b)
insignificantly better data matches.

Nonconvergence is dependent on both the variable set and the regression data set. It may be
possible to solve a nonconvergence problem by adding more data to the regression data set. For a
fixed regression data set, the general remedy for nonconvergence is to simply remove one of the
regression variables.

The general remedy for convergence to unrealistic variable values is to add another variable
to the regression variable set. For example, although negative values of binary interaction
coefficients may result in a good match of lab data, negative values are not reasonable. It may be
possible, by adding another regression variable, to obtain an equally good match of the lab data with
more reasonable positive interaction coefficients.

12
From personal experience, it has also been determined that (1) convergence and (2) realistic
variable values are also affected by the weighting factors assigned to the various regression data
items. It may be possible to fix problems with either or both of these characteristics by changing the
weighting factors assigned to certain data items (e.g. assigning lower weighting factors to data that
are suspect).

If a poor data match results from a regression that satisfies the above criteria for an optimal
regression variable set, then this indicates either erroneous data or inadequacy of the EOS to match
the particular fluid data (it may be difficult to distinguish between the two). It is sometimes possible
to detect erroneous data by simple mass balances on CVD data and/or surface separation data.

Example Applications. Examples of EOS regression to match laboratory data are presented by
Coats and Smart
1
and the third SPE comparative solution project, SPE 12278.
7

Modeling of Surface Facilities

Fig. 3 shows how a compositional simulator models the surface facilities as well as the reservoir
performance.
GAS
INJ .
S.T 1 2
Stage Separation
Oil Sales
Plant
Make-up Gas
Compressor
Prod.
Residue Gas
Gas Liquid
Sales
F
1
=0.0
=0.30
=0.95
=1.00 2
3
6
F
F
F
etc.

Fig. 3

The producing fluid is separated with stage separation to improve the oil recovery. The
separator calculation must be performed at the separator temperature and pressure given. The
separator liquid goes to the stock tank and is sold as oil.

The separator gas then continues to a gas liquids plant which strips the gas liquids from the
input stream. This is modeled by specifying a fraction of each component which is recovered as gas
liquids.

The remaining residue gas may be re-injected into the reservoir after compression to a
suitable injection pressure. It may be necessary to add make-up gas to the gas injection in order to
compensate for shrinkage after liquid removal.

The simulation of the surface facilities as well as the reservoir performance makes this a very
interesting project. The optimization of the operations might include runs in which the plant
configuration is modified as well as runs with the usual injection and production parameters
modified.

13
Lab Tests

Certain types of laboratory tests are usually performed for compositional problems. Before
performing compositional simulation, it is necessary to simulate laboratory tests with whatever
phase behavior and fluid property calculations which are going to be used in the simulation. We will
review some of the basic laboratory tests that are performed for gas condensate reservoirs.

Constant Composition Expansion Test. The first test is constant composition expansion,
sometimes called flash liberation. Refer to Fig. 4.

4300 4500 3500 3000 2500 1000
pressure
dew pt.
Vol. % Liq.
pressure pressure
2-phase
V
liq
GAS CONDENSATE
Constant Composition Expansion

Fig . 4

The PVT cell is expanded from above the dew point to pressures usually near the expected
abandonment pressure. Below the dew point pressure, the total volume of the two phase mixture is
measured. Sometimes the liquid volume is measured and sometimes not.

The two phase expansion curve must be matched by the simulator. Two phase
compressibility can be calculated directly from this volumetric data. This compressibility value
would be at any measured pressure for the combined liquid and gas compressibilities, at least for the
composition used.

The liquid percent curve is presented if the liquid volumes are measured in the laboratory. It
is common for the liquid to form below the dew point and then vanish again at lower pressures. This

14
would result in no residual oil saturation if this could be accomplished in the reservoir, but this test is
only applicable for constant composition.

Fig. 5 shows a phase diagram for a constant composition fluid. For this fluid, the nature of
constant composition expansion depends on the temperature of the fluid. This figure shows a bubble
point fluid at low temperatures and a fairly dry condensate which vaporizes completely during
pressure decreases at higher temperatures.

press.
temperature
1 2 3 4 5
crit.
pt.
1. black oil
2. volatile oil
3. gas condensate
(high intermediate
4. gas condendsate
5. gas
100%
80%
60%
40% 20%
0%

Fig. 5

Fig. 6 presents a family of curves which represent the pressure traverses of the various
temperatures shown in Fig. 5. For the first two temperatures, bubble points are observed. The
second temperature has greater shrinkage since it is closer to the critical point. At temperatures
above the critical temperature, the dew point elevates with increases in temperature and then
decreases with further increases in temperature. At the highest temperature, the liquid completely
vaporizes with pressure reduction.

15
% Liq.
Pressure
0
100
T
1
T
2
b.p.
b.p.
d.p. d.p. d.p.
T
3
T
4
T
5

Fig. 6

Behavior of the constant composition expansion is useful in understanding reservoir fluid
behavior. It is important to match this behavior with the simulator. However, the composition of the
reservoir fluid during depletion usually changes dramatically when two phase flow occurs.

Constant Volume Depletion Test. Fig. 7 depicts a constant volume depletion test (or simply,
depletion test). This test is normally done for a dew point fluid. This is a test which approximates
the constant reservoir volume fluid expansion.

4500
y y
y y
i i
i i
GAS CONDENSATE:
constant volume depletion
G
p
G
p
4300
dew point
3800 3800 3000 3000

Fig. 7

For this case, gas is withdrawn at pressures below the dew point, representing gas production
from the reservoir. Liquid is not taken out of the PVT cell because liquid usually does not flow in
the reservoir. Rather than "bleeding off" the reservoir gas, the gas production is taken out of the
PVT cell at discrete intervals. By performing the test in this manner the exact composition of the
equilibrium gas is known at the pressure at which the gas is withdrawn. This composition is
measured from the produced gas at that particular pressure. The liquid composition cannot be

16
measured without destroying the sample, but it can be calculated from a material balance.
Experimental errors affect the accuracy of the material balance, however.

The standard results of this test are direct measurements which are represented by three
sketches shown in Fig. 8.

% Liq.
Pressure Pressure Pressure
log y
i
G
p
cumm. gas
below D.P.
c
c
c
c
c
3
2
1
4
5

Fig. 8

The liquid percent curve can be modified to show oil saturation by accounting for the
connate water saturation. The composition of the producing gas stream is presented and can be used
to calculate the gas-oil ratio or gas-liquid ratio in the producing stream during depletion. And
finally, the cumulative gas production shows how pressure declines with gas production. These
three curves must be matched with the simulator to validate the phase behavior and fluid property
calculations. The liquid density is not measured directly, nor is the gas density. The validation of
the simulator includes the density calculations as well as the phase equilibrium calculations.

The surface separation and calculation of the gas-oil ratio must also be calculated by the
simulator during the simulation runs. This information is usually reported in the depletion
laboratory test results. These calculations are relatively simple because NGAA K-values can usually
be used at surface conditions.

17

18
Gas Cycling in Volatile Oil and Gas Condensate Reservoirs

Introduction. Gas cycling is a process used widely in volatile oil and condensate reservoirs for
primarily two reasons - for pressure maintenance, where pressures have fallen off due to depletion
and to enforce a miscible displacement process to achieve higher recoveries. In principle, the process
involves re-injection of produced gas into the reservoir so that reservoir pressures are increased and
a miscible displacement is initiated. A schematic of the gas cycling process is shown in Fig. 9.
Volatile oil and condensate reservoirs are characterized by reservoir fluids that are composed
primarily of a light oil and a rich gas. Typically these fluids are composed of 50-70 mole percent of
methane (C
1
), 20-30 mole percent of intermediates (C
2
-C
6
) and 10-20 mole percent of heavies (C
7+
).
Gravity of the stock tank oil ranges from 42
o
to 60
o
API. Characteristic gas-oil ratios range from
1200-3500 SCF/STB for volatile oil reservoirs and from 8000-30000 SCF/STB for retrograde
condensates. Approximate ranges of reservoir and fluid characteristics are summarized in Table 1
(page 20). When the reservoir pressure falls below saturation pressure, due to depletion, the reservoir
oil is in equilibrium with a rich gas. The oil tends to be heavy with high viscosities and low
mobilities. The process of gas cycling is designed to exploit the gas phase mobility compared to that
of the oil. Methane from the produced gas is recycled back into the reservoir. Sometimes, methane is
mixed with small fractions of intermediate gases before re-injection. Surface facilities required for
this include a gas plant to which part or all of the separator gas is sent, which is pressurized with a
compressor and re-injected into the reservoir.

Displacement Process. The injected lean gas miscibility displaces the in-situ rich gas toward the
producing well. The liquid generally does not move due to unfavorable mobility ratios. Also the oil
saturation can drop below the residual oil saturation. However, the oil vaporizes behind the
displacement front. The intermediates in the oil vaporize due to the contact with the injected lean
gas. This vaporization process "strips" the reservoir liquid of its intermediates which are then
transported in the gas phase to the producer. This miscible process is referred to as a multiple contact
miscible process (also called developed miscibility) as opposed to a first contact miscible process. If
the reservoir fluid is considered as being composed of three pseudo components (C
1
), intermediates
(C
2-6
) and heavies (C
7+
), a ternary diagram can be used to represent the reservoir fluid and the
displacement process at a given temperature and pressure. Two fluids are considered first contact
miscible if all possible mixtures of the two yield a single phase fluid at a given pressure and
temperature. If the path connecting the injected gas composition and the reservoir oil composition
does not pass through the two phase region, the process is termed first contact miscible. This path is
known as the dilution path and represents the composition changes as the injected gas displaces
reservoir oil. In doing so, we are neglecting the pressure change that is, of course, an essential
ingredient in making the fluids flow in the reservoir. Ternary diagrams represent constant pressure
and temperature conditions, but illustrate the mechanics of a miscible displacement process.

An Example of Compositional Simulation - Gas Cycling. A typical example of compositional
simulation is found in gas cycling performance. Fig. 9 shows that the reservoir pressure has dropped
below the dew point and an oil saturation has developed. Below the dew point the reservoir liquid
will be in equilibrium with a rich reservoir gas. During gas cycling a dry gas (mostly methane) is
injected. The dry gas miscibly displaces the rich gas toward the producing wells. The liquid
normally does not move because it is below the residual oil saturation. However, the oil does
vaporize behind the displacement front. The lighter ends of the oil vaporize because of the contact
with the dry injection gas instead of the rich reservoir gas that it previously contacted. As this
vaporization continues, the oil saturation decreases and a significant amount of the oil components
are transported to the producing wells in the gas phase.

dry gas production
GAS CYCLING
dry rich
G
O
displacement front
new equilibrium
vaporization of intermediates

Fig. 9

This gas cycling process is modeled with a compositional simulator. The displacement
mechanism is modeled simultaneously with the phase behavior.

19

20
Vaporizing Gas Drive (VGD)

If the dilution path does pass through the two phase region, the process is no longer first
contact miscible. Imagine a series of well mixed cells that represent the permeable medium in a one
dimensional displacement. The first cell initially contains crude to which we add an amount of
solvent so that the overall composition is given by M
1
. The mixture will split into two phases, a gas
G
1
and a liquid L
1
, determined by equilibrium tie lines. The gas G
1
will have a much higher mobility
than L
1
, and this phase moves preferentially into the second mixing cell to form mixture M
2
. Liquid
L
1
remains behind to mix with more pure solvent (injected gas). In the second cell mixture, M
2
splits
into gas G
2
and liquid L
2
(again determined by the equilibrium tie lines), G
2
flows into the third cell
to form mixture M
3
, and so forth. At some cell beyond the third (for this diagram), the gas phase will
no longer form two phases on mixing with the crude. From this point onward, all compositions in
the displacement will be on a straight dilution path between a crude and a point tangent to the
binodal curve. The displacement from here on will be first contact miscible, with the solvent
composition given by the point of tangency. The process has developed miscibility since the solvent
has been enriched in intermediate components to be miscible with the crude. Since the intermediates
are vaporized from crude, the process is a vaporizing gas drive. Miscibility will develop in this
process as long as the injected solvent and crude are on opposite sides of the critical tie line. A
schematic of vaporizing gas drive is shown in Fig. 9.

The injected fluid and reservoir fluids form a "mixing zone" in the reservoir, as the injected
solvent moves further into the reservoir. The point of miscibility is at the front of the mixing zone.
The key characteristics of a VGD can be summarized as:

1) Reservoir oil transfers intermediates to injected lean gas.
2) Point of miscibility is at front of the mixing zone.
3) Miscibility is achieved by progressively contacting generated gas composition with fresh oil.
Figure 9 shows a saturation profile of the injected gas as it travels from the injector toward the
producer. Note the vaporization of the intermediates, the location of the mixing zone and the point of
miscibility in the reservoir.

Reservoir Simulation Aspects. Simulation of gas cycling processes requires a compositional
model to track the phase behavior of the fluids along with modeling displacement fronts. Current
compositional models are highly sophisticated; they use both equation-of-state and K-value options
to model phase equilibria within the reservoir. A major part of such a study is phase behavior
matching, where laboratory tests on reservoir fluids are simulated by forcing equation-of-state
parameters or K-value correlations to match observed laboratory behavior.

Economic Considerations. As in all oil field operations, economic considerations dictate whether
gas cycling should be employed over other methods, i.e. waterflooding. Gas cycling involves
investment in surface facilities: gas plant to re-inject residue gas. However, in volatile oil and
condensate reservoirs, gas cycling has potential to yield higher ultimate recoveries than a waterflood
and has greater long term profit potential. Reservoir simulation is an important tool in such cases to
assess the viability of different recovery schemes.

21
Field Examples

Three field simulation studies are summarized in this section. All three reservoirs are retrograde
reservoirs. In each case, gas cycling was used for a variety of reasons. However the underlying
mechanism of displacement is the vaporizing gas drive as described earlier.

Carson Creek Field, Alberta. The Carson Creek field is a retrograde gas condensate field of
Central Alberta, discovered in 1957. The field encompasses about 18,000 acres and has two gas
pools, "A" and "B". The original gas-in-place in the two pools was about 400 BSCF. Methane
constituted 70.2 mole percent of the original reservoir fluid. Gas cycling operations began in 1962;
original gas cycling operations consisted of processing between 70 and 90 MMSCF/D raw gas to
extract about 4,000 STB/D condensate and re-injecting all residue gases into the reservoir. The
cycling also processed 5 to 10 MMSCF/D solution gas from an adjacent field and was re-injected.
Gas cycling was employed in this field for both pressure maintenance and to carry out a miscible
flood.
A compositional simulator was used to simulate gas cycling in this field. The model uses K-
values as a function of temperature, pressure and composition to model phase equilibrium. Density,
viscosity and compressibility of the fluids were treated as functions of composition and pressure.
History matches of production, injection and pressure in each well were obtained. Also, a match of
the pentanes-plus composition of each produced well stream was obtained.

Bonnie Glen D 3-A Pool Model, Alberta. The Bonnie Glen D 3-A pool, is one of the most prolific
fields in Canada. The original oil in place is estimated at 6.57 MMSTB, with an initial gas cap of
44.5 MMSCF. Since simultaneous production of gas cap during the life of the oil column could be
detrimental to oil recovery, gas cap production would normally be deferred until depletion of the oil
column. Gas cycling can however, can be carried out concurrently with continued depletion of oil
column with a beneficial effect on the overall recovery of hydrocarbon liquids. The increase in
recovery results from vaporization of retrograde liquids in the gas cap plus vaporization of lighter
components from the original oil column.

A three dimensional three phase black oil simulation study was first conducted to history
match the reservoir performance. This was followed by conversion to a compositional formulation.
The compositional formulation used K-value correlations to model phase equilibria.

Kaybob South Field , Alberta. The Kaybob South Field is one of the largest oil fields in Canada.
Initial gas in place in the field was estimated at 4,000 BSCF. The reservoir encompasses an area of
57,000 acres and is totally underlain by water (bottom water drive). Because of the possibility of
significant retrograde liquid losses upon depletion, it was decided to implement gas cycling in the
field. Reservoir simulation was used to design the gas cycling project and study its potential impact.
The design and planning of the cycling project began in 1967, without full delineation of the
reservoir limits. Several cycling patterns were simulated (involving different injection well patterns).
The effect of different cycling rates was also studied by simulating various cases. Studies were
undertaken to determine the degree of revaporization expected from cycling with residue gas.

22

Table 1

Reservoir Fluid Characteristics for
Volatile Oils and Retrograde Condensates

Fluid Classification Volatile Oils Gas Condensates
Stock tank oil color Greenish to orange Yellowish
Stock tank oil gravity
o
API 42-45 45-60
Gas-oil ratio, SCF/STB 1200-3500 8000-30000
Formation volume factor, B
o
1.7-3.0 6.0-20.0
Typical reservoir temperature,
o
F 150-300 150-300
Typical saturation pressure, Psia 3000-7500
(Bubble Point)
1500-9000
(Dew Point)

PVT MATCHING

Fundamentals:

Heavy Fraction Characterization. During the development of the application of EOS's to naturally
occuring hydrocarbon mixtures, it has become clear that insufficient description of heavier
hydrocarbons (e.g., heptanes and heavier) reduces the accuracy of PVT predictions. Volatile oil and
gas-condensate volumetric phase behavior is particularly sensitive to composition and properties of
the heaviest components. Therefore we have to develop a comprehensive method to characterize
compositional variation, which we call "molar distribution". A three-parameter gamma function was
chosen for describing molar distribution.

p(x) =
(x - h ) [(x - h) / b]
b
(a)
a-1
a
exp
(1)

p(x) is called the probability density function (or three parameter gamma function). Whereas x
corresponds to measured C
7+
molecular weight. For a =1, the distribution is exponential. Values less
than one give accelerated exponential distribution, while values greater than one yield left-skewed
distributions. Note that as a approaches infinity, the distribution becomes normal, though "folded"
at h, the minimum molecular weight included in the C
7+
fraction.

a, b, and h are parameters defining the distribution. h can be estimated accurately since it
represents the minimum molecular weight to be included in the fraction. If a is given, b is found
directly from h, a, and the measured C
7+
molecular weight, M
7+
. a can be fitted to measured molar-
and weight-distribution data, or estimated using an empirical relation.

The parameter h is defined as the minimum molecular weight expected to occur in the C
n+

fraction. That is, there is zero probability [p(x) =0] for occurrence of compounds with molecular
weight less than h. If the C
7+
is considered, then h =92 (the molecular weight of toluene) would be a
good estimate for h. Experience has shown that a good approximation of h is given by

h =14n - 6 (2)

for a C
n+
fraction. Eq. 2 is a useful empirical relation but should not be considered a restraint on the
model.
Considering the remaining two parameters in Eq. 1, a and b, a useful property of the
three-parameter gamma function is that the product ab equals the arithmetic average molecular
weight, M
n+
, minus h.

M
n+
- h =ab (3)

where M
n+
is measured directly.

23

There are several empirical correlations available for estimating a from randomly sampled
data such as fully extended molar distribution. b is easily calculated from the other variables.

b =(M
n+
- h)/a (4)

An estimate of a can be calculated using the following proposed empirical relation.

a =Y
-1
(0.5000876 +0.1648852Y - 0.0544174Y
2
) (5)

where,

Y =ln[(M
n+
- h)/m
G
] (6)

and

24
N
i
Z 1/ Z
m
= [ (
M
- h ) ]
i n+
(7)
G
i=n

Eq. 5 is valid for 0 <Y <0.5772 (i.e., a >=1.0). The variable m
G
merely represents a geometric
average molecular-weight variable. For Eq. 5 to be useful for the molar distribution problem, it is
necessary to have measured SCN (single-carbon-number) mole fractions and molecular weights
accurately. They should also constitute a full compositional analysis, preferably having a
diminishing or negligible quantity of the last fraction, N. Since such analyses are nearly nonexistent,
a set of correction tables has been developed when only partial analyses are available (the limitation
of a >=1.0 mentioned previously is also lifted in the present use of Eq. 5).

Due to the limitation of the Eq. 5, we may determine a by minimization of a error function.
The error function, E(a) , used to optimize a is defined as the sum of the squares of differences in
measured and calculated compositions and is given by

E(a) = (
Z
- Z )
N
i
2
i=n
(8)

If a molar distribution is to fit, and

E(a) = ( f - f )
N
wi
2
i=n
(9)

if a weight distribution is to fit (the more common case). Calculated values are marked with a tilde.

25
The minimization of E may proceed by a simple secant or half-interval method. Two
procedures for performing the minimization have proved useful. Reasonable limits for a are 0.5 to
3.0.
The probabilistic model is not a true physical model. One assumption is the continuous
relation between molecular weight and mole fraction. This assumption, however, along with others
implicit in its mathematical form, seems as reasonable as, for example, the assumption in distillation
(TBP) analysis that cumulative volume and boiling point have a continuous relation.
In the DESKTOP-PVT package, at least three numbers, molecular weight, specific gravity
and mole (or weight) fraction of the heavy fraction must be entered for heavy fraction
characterization.

The extended analysis can be entirely predictive or can be compared to experimental
distillation data. If distillation data of mole (or weight) fraction distillation are available, an option
can be activated to compute an optimal. This will minimize the differences between the
experimental and calculated distributions.

The gamma distribution function gives the mole (or weight) fraction and molecular weight
for each single carbon number (SCN) of the extended fractions. The calculated mole fraction and
molecular weight of the last carbon number in the extended analysis are adjusted so the computed
mole fraction and molecular weight of the heavy fraction are matched to the observed data. The
gamma function can be calculated using either a constant molecular weight interval or variable
molecular weight intervals.

A constant Watson characterization factor, K, is used for all carbon numbers in the heavy
fraction. K defines relative paraffinicity of a hydrocarbon fraction, with a typical range from 10.0
(highly aromatic) to 13.0 (highly paraffinic). If a Watson factor is entered, the input number is used
to compute the specific gravity of all carbon numbers. If a Watson factor is not entered, this number
is adjusted so the computed specific gravity of the heavy fraction is matched to the observed.

Pseudoization. The term "pseudoization" here denotes the reduction in number of components used
in EOS calculations for reservoir fluids. Pseudoization is important in reservoir calculations because
of the large number of real components(e.g. in C
7+
fraction) in reservoir fluids. Compositional model
computing times can increase significantly with number of components used.

Pseudoization must satisfy the following conditions:

1) EOS calculation will yield identical density (Z-factor) and viscosity for each pseudo
component whether performed in a) single-component mode, or b) n-component mode.

2) For all mixtures of the m pseudo components (including the original mixture z), the EOS
calculations will yield identical mixture density and viscosity whether performed in a) m-pseudo
component mode or b) n-component mode.

The first of the two pseudoization conditions is satisfied by defining pseudo component properties
as:
(l)
a
l (l)
c
2
2
(l)
c
i=1
n
j=1
n
(l)
i
(l)
j ij i j
.5
A = p( T ) / T
p
= x x (1- d )( A A
)

(10)

(l) l
b
l
c
(l)
c
j=1
n
l
j j B = p T / T
p
= x
B

(11)

(l)
c
j=1
n
(l)
j
cj
P =
x p
(12)

(l)
c
j=1
n
(l)
j cj T
=
x T
(13)

(l)
c
j=1
n
(l)
j cj v
=
x v
(14)

(l)
j=1
n
(l)
j j M
=
x M
(15)

*(l)
j=1
n
(l)
j
*
j
0.5
j
j=1
n
(l)
j
0.5
j m
=
x m

M
/ (
x M
)

(16)

The
a
and
b
appearing in Eqs. are "theoretically" universal constants, determined by
forcing the EOS to satisfy the Van der Waals conditions: (dp/dv)
Tc
and (d
2
p/dv
2
)
Tc
=0 at the critical
point. In practice, however, the
a
,
b
values are generally treated as component-dependent
functions of temperature.

The second condition of pseudoization is satisfied by Eqs. 10 through 16 and the additional
requirement that for each pair of pseudo components, the pseudo binary interaction coefficient is
given by

i=1
2
j=1
2
i j
ij (i) (j)
0.5
i=1
n
j=1
n
i j ij i j
0.5

a a
(1-
d
)(
A A
) =
x x
(1-
d
)(
A A
)

(17)

26

27
The pair of pseudo components are arbitrarily labeled components 1 and 2 on the left-hand
side of Eq. 17. {x
i
} is the n-component composition of an arbitrary mixture of the two pseudo
components, i.e.

x
i
=a
1
x
i
1
+a
2
x
i
(18)

where a
1
+a
2
=1 and 0 <a
j
<1 for j =1, 2.

The only unknown in Eq. 17, after use of Eqs. 13 and 14, is d
12
, i.e. the (pseudo) binary
interaction coefficient between pseudo components 1 and 2. This coefficient d
12
, is independent of
temperature regardless of temperature-dependence of
ai
and /or
bi
, provided, of course, that the d
ij

are independent of temperature.

Regrouping Scheme of Heavy Components. A method is proposed for estimating the number of
multiple-carbon-number(MCN) groups needed for adequate plus-fraction description, as well as
which SCN groups belong to the MCN group. It is based on Sturges rule and the observation that
the proposed distribution model is similar to a folded log-normal distribution. The number of MCN
groups, Ng, is given by

N
g
=Int [ 1 +3.3 log10(N - n)] (19)
For black-oil system, this number probably can be reduced by one.

The molecular weights separating each MCN group are taken as

M
I
=M
n
{exp[(1/N
g
)ln(M
N
/M
n
)]}
I
(20)

where M
N
is the molecular weight of the last SCN group (which may actually be a plus fraction),
and I =1, 2, .... N
g
. Molecular weights of SCN groups falling within the boundaries of these values
are included in the MCN group, I.

28
Compositional Simulation:

There are two basic steps to work on compositional simulation. One is the PVT match, the
other is the simulation of the compositional model. The following are the step-wise procedures and
the corresponding data files. DESKTOP_PVT package was used for PVT match, compositional
simulator is VIP. Data are from Third SPE Comparative Solution Project: Gas Cycling of
Retrograde Condensate Reservoirs
1
(we name this paper as SPE3).

1- Heavy Ends Characterization. This is the first step for PVT matching. The following
procedure is to be followed:
a- Invoke DESKTOP_PVT. The Peng-Robinson EOS is selected by picking CONFIG
from the main menu, then EOS submenu, then typing PR. (noted as CONFIG/EOS,
PR).
b- Turn on heavy end characterization by choosing CONFIG/Heavy, Y.
c- Enter heavy end parameters by selecting (from main menu) HEAVY/Parameter. At
least three numbers must be input to complete this step, they are: molecular weight,
specific gravity, and mole fraction of C
7+
. These numbers can be found in SPE3.
d- Specify the names and number of heavy component groups by choosing PSEUDO-
COMPONENT NO. AND NAME.
e- Select cut-off molecular weights and specify 130 for bracket 1 and 180 for bracket 2.
f- After inputting these data, you click CALCULATE option in the HEAVY menu.
g- Inspect the data by selecting HEAVY/Review.
h- Save the calculated EOS parameters by selecting HEAVY/Save EOS. Name the file
heavy.eos.
i- Save the data by choosing FILE/Save and name the file heavy.dat.

Molecular Weight = 140
Specific gravity = 0.774
Mole Fraction = 0.0659
Cut-off Molecular Weight 130 180
Names of Heavy Components: HVY1 HVY2 HVY3
Output: HEAVY.DAT, HEAVY.EOS

2- Default Fluid Predictions. At this step, all components from SPE3 should be input except
heavy components. The following procedure is to be followed:
a- Choose CONFIG/Test Type.
b- Select, Cnst Composition, Cnst Volume, and/or Swelling.
c- Choose CONFIG/Heavy, N, to turn off the heavy components option.
d- Choose COMPONENT/System, and highlight the components required (do not include
the C
7+
).
e- Load The EOS for the heavy component by choosing COMPONENT/Append EOS,
heavy.eos.
f- Enter pure component mole fraction by choosing COMPONENT/Composition.
g- Enter the data for the tests by choosing TESTS/Cnst Composition.
h- Repeat for the Cnst Volume and Swelling tests.

29
i- Run the program by Chooing RUN/Go.
j- View the data by choosing REPORT/Graphics.
k- Save the data by choosing FILE/Save. Name the file default.dat.

Components: CO
2
,N
2
,C1,C2,C3,NC4,IC4,NC5,IC5,NC6,HVY1,HVY2,HVY3
Test Types: Constant Composition Expansion, Constant Volume Depletion and
Swelling
Output: default.dat

3- Pre-regression. Before pseudoization, we make regression for EOS properties. We call this step
as pre-regression. We first need to pick regression variables, then run the program based on the
previous data file default.dat. After doing this, we save two files. One is regress.dat, and the
other is regress.eos. The updated EOS properties are included in regress.eos. The following
procedure is to be followed:

a- Choose CONFIG/Regression, Y, to turn on regression.
b- Choose (from main menu) REGRES/Variable. Choose EOS Property and enter the
number for the regression variable in the appropriate place. Then Choose Binary Coeff
and enter the number for the regression variable in the appropriate place. Then Exit.
c- Choose REGRES/Limits and change the minimum to 0.7 and maximum to 1.3.
d- Choose REGRES/Control and change the no. of iterations to 20.
e- Run the program by Choosing RUN/Go.
f- View the data by choosing REPORT/Graphics.
g- Save the calculated EOS parameters by selecting REPORT/Save EOS. Name the file
regress.eos.
h- Save the data by choosing FILE/Save and name the file regress.dat.

Components: CO
2
,N
2
,C1,C2,C3,NC4,IC4,NC5,IC5,NC6,HVY1,HVY2,HVY3

Regress Variables:

C1-OMEGAA 1
C1-OMEGAB 2
HVY1-OMEGAA 3
HVY2-OMEGAA 4
HVY3-OMEGAA 5
HVY1-OMEGAB 6
HVY2-OMEGAB 7
HVY3-OMEGAB 8
DJK C1-HVY1 9
DJK C1-HVY2 10
DJK C2-HVY3 11
Output: REGRESS.DAT, REGRESS.EOS

30
4- Component Pseudoization. This step is called the pseudoization. The pseudo components are
listed below. Once you give the names of pseudo components and components included in
them, the program is capable of calculating regrouped lab data. Test data must be reentered for
the new pseudo components. New files will be generated: pseudo.dat and pseudo.eos. The
following procedure is to be followed:

a- Choose COMPONENT/Load EOS, regress.eos. This will load the regressed EOS.
b- Turn off regression by choosing CONFIG/Regression, N.
c- Turn Pseudoization on by selecting CONFIG/Pseudoization, Y.
d- Choose PSEUDO/Pseudo Name. Give names to the new pseudo components.
e- Choose PSEUDO/Parameter. Specify the components in Pseudo-COMP P1 Lump.
Repeat for all pseudo components.
f- Choose PSEUDO/Calculate.
g- Save the calculated EOS parameters by selecting PSEUDO/Save EOS. Name the file
pseudo.eos.
h- Choose PSEUDO/Replace EOS. This replaces the active EOS.
i- Turn off pseudoization by choosing CONFIG/Pseudoization, N.
j- Reenter the test data by choosing TESTS/Const Composition. Press F2 and load the
global reference composition. Repeat for other tests.
k- Run the program by selecting RUN/Go.
l- Save the data by choosing FILE/Save and name the file pseudo.dat.

Pseudo Components:
P1 C1 +N2
P2 C2 +CO2
P3 C3 +NC4 +IC4
P4 C6 +NC5 +IC5
P5 HVY1
P6 HVY2
P7 HVY3
Test Types: Constant Composition Expansion, Constant Volume Depletion and
Swelling
Output: PSEUDO.DAT, PSEUDO.EOS

5- Regression. This is the final step for PVT match. Procedures are the same as we did in the step
3. Regression is based on the data obtained in the step 4. The main problem is the selection of
variables to be regressed in order to obtain a good match. There is no definite way to proceed.
In this case trial and error method was used to get a good match. Selecting proper variables is a
time consuming effort. The following procedure is to be followed:

a- Choose CONFIG/Regression, Y, to turn on regression.
b- Choose (from main menu), REGRES/Variable. Choose EOS Property and enter the
number for the regression variable in the appropriate place (delete old numbers if
present). Then Choose Binary Coeff and enter the number for the regression variable in
the appropriate place. Then Exit.

31
c- Choose REGRES/Limits and change the minimum to 0.7 and maximum to 1.3.
d- Choose REGRES/Control and change the no. of iterations to 20.
e- Run the program by Choosing RUN/Go.
f- View the data by choosing REPORT/Graphics.
g- Save the calculated EOS parameters by selecting REPORT/Save EOS. Name the file
final.eos.
h- Choose COMPONENT/Load EOS, final.eos. This will overwrite the old EOS.
i- Choose CONFIG/Regression, N, to turn off regression.
j- Run the program by Choosing RUN/Go.
k- Save the data by choosing FILE/Save and name the file final.dat.

Pseudo Components:
P1 C1 +N2
P2 C2 +CO2
P3 C3 +NC4 +IC4
P4 C6 +NC5 +IC5
P5 HVY1
P6 HVY2
P7 HVY3

Regress Variables:
P1-ZC 1
P2-ZC 2
P5-ZC 3
P6-ZC 4
P7-ZC 5
P1-OMEGGA 6
P1-OMEGAB 7
P5,P6,P7-OMEGAA 8
P5,P6,P7-OMEGAB 9
DJK P1-P7 10
Output: FINAL.DAT, FINAL.EOS

6- Initialization Data for Compositional Simulation. After step 5, the adjusted EOS properties
are obtained. These data are considered as the "correct values" and input in the initialization file
along with other data from SPE3. The format of initialization file is attached in the following
pages. VIP will create a spe3i.rst file when you type: "corerun" in the terminal. It should be
noted that you should check the "corefil.dat" in the current directory before you type command
"corerun."

Data Sources: SPE12278 paper, FINAL.DAT
File Name: spe3i.dat

32
7- Run Data for Compositional Simulation. This is the last step for compositional simulation.
After creating the simulation run data file-spe3r.dat and having a proper restart file-spe3i.rst
from step 6. you can type the command "execrun", VIP will automatically produce the
information specified by the user in the spe3r.dat. The format of the run data file spe3r.dat is
attached in the following pages.

Data Sources: SPE12278, Final.dat
File Name: spe3r.dat

33
Data Files

The following pages contain the data files and the output files from the VIP simulator.
Below is a list of the enclosed material.

Heavy.dat
Heavy.eos
Default.dat
Regression.dat
Regression.eos
Pseudo.dat
Pseudo.eos
Final.dat
Final.eos
Spe3i.dat
Spe3r.dat

34
(Input data file for Step 1 - Heavy Ends Characterization)
HEAVY.DAT

SPLIT
MWPLUS GPLUS ZPLUS NG MWGRP
140.00 0.7740 0.06590 3 130.00 180.00
PROPERTY CORRELATION SIMULATION
TC RIAZI-DAUBERT
PC RIAZI-DAUBERT
ACENTRIC EDMISTER
ZC RIAZI-DAUBERT
CONMWI
MWC6C7 92.00
MWINC 12.000
END

35
(Output data file for Step 1 - Heavy Ends Characterization)

HEAVY.EOS

EOS PR
COMPONENTS
HVY1 HVY2 HVY3
C
C PLUS FRACTION PSEUDO COMPONENT PROPERTIES
C
PROPERTIES
COMP MW TC PC ZC Accentric OMEGA A OMEGA B
HVY1 107.77 561.98 410.55 26239 .31504 X X
HVY2 148.05 677.63 325.18 24598 .41381 X X
HVY3 224.01 843.58 244.42 23072 .57895 X X

DJK C1
HVY1 .036113
HVY2 .041436
HVY3 .049070
ENDEOS
C
C PLUS FRACTION PSEUDO COMPONENT COMPOSITIONS
C
COMPOSITION
.034766 .019684 .011451

(Input data file for Step 2 - Default Fluid Predictions)

36

DEFAULT.DAT

EOS PR
COMPONENTS
CO2 N2 C1 C2 C3 NC4 IC4 NC5 IC5 NC6 HVY1 HVY2 HVY3
NCV 13
PROPERTIES F PSIA
COMP MW TC PC ZC ACENTRIC OMEGAA OMEGAB PCHOR
CO2 44.010 87.90 1070.9 .27420 .22250 .457236 .077796 49.60
N2 28.013 -232.40 493.0 .29100 .03720 .457236 .077796 35.00
C1 16.043 -116.60 667.8 .28900 .01260 .457236 .077796 71.00
C2 30.070 90.10 707.8 .28500 .09780 .457236 .077796 111.00
C3 44.097 206.00 616.3 .28100 .15410 .457236 .077796 151.00
NC4 58.124 305.70 550.7 .27400 .20150 .457236 .077796 191.00
IC4 58.124 275.00 529.1 .28300 .18400 .457236 .077796 191.00
NC5 72.151 385.70 488.6 .26200 .25240 .457236 .077796 231.00
IC5 72.151 369.10 490.4 .27300 .22860 .457236 .077796 231.00
NC6 86.178 453.70 436.9 .26400 .29980 .457236 .077796 271.00
HVY1 107.770 561.98 410.5 .26239 .31504 .457236 .077796 353.20
HVY2 148.050 677.63 325.2 .24598 .41381 .457236 .077796 474.02
HVY3 224.010 843.58 244.4 .23072 .57895 .457236 .077796 677.90
DJK CO2
N2 .000000
C1 .150000
C2 .150000
C3 .150000
NC4 .150000
IC4 .150000
NC5 .150000
IC5 .150000
NC6 .150000
HVY1 .150000
HVY2 .150000
HVY3 .150000
DJK N2
C1 .120000
C2 .120000
C3 .120000
NC4 .120000
IC4 .120000
NC5 .120000
IC5 .120000
NC6 .120000
HVY1 .120000
HVY2 .120000
HVY3 .120000
DJK C1
C2 .000000
C3 .000000
NC4 .020000
IC4 .020000
NC5 .020000
IC5 .020000
NC6 .025000
HVY1 .036113
HVY2 .041436
HVY3 .049070
DJ K C2
C3 .000000
NC4 .010000
IC4 .010000
NC5 .010000
IC5 .010000
NC6 .010000
HVY1 .010000

37
HVY2 .010000
HVY3 .010000
DJ K C3
NC4 .010000
IC4 .010000
NC5 .010000
IC5 .010000
NC6 .010000
HVY1 .010000
HVY2 .010000
HVY3 .010000
DJ K NC4
IC4 .000000
NC5 .000000
IC5 .000000
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJ K IC4
NC5 .000000
IC5 .000000
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJ K NC5
IC5 .000000
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJ K IC5
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJ K NC6
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJ K HVY1
HVY2 .000000
HVY3 .000000
DJ K HVY2
HVY3 .000000
HIDEAL
CO2 2.1028E+02 5.0361E+00 4.4507E-03 -1.1658E-06 1.5274E-10 -5.7829E-15
N2 -2.6164E+01 7.1489E+00 -4.9835E-04 4.4513E-07 -9.0210E-11 4.4521E-15
C1 -1.1193E+02 9.1718E+00 -4.7216E-03 6.7887E-06 -2.4493E-09 3.1208E-13
C2 -6.3147E-01 7.9649E+00 -7.5217E-04 8.7905E-06 -3.8672E-09 5.4789E-13
C3 -3.2544E+01 7.6111E+00 4.1469E-03 9.5048E-06 -4.7227E-09 7.0237E-13
NC4 4.3186E +02 5.7293E+00 1.5646E-02 3.0120E-06 -2.4420E-09 3.8132E-13
IC4 6.6831E+02 2.7132E+00 1.9460E-02 8.3832E-07 -1.8392E-09 3.1555E-13
NC5 1.9605E+03 -2.0202E-01 3.1752E-02 -6.2257E-06 5.8994E-10 -1.4225E-14
IC5 1.9930E+03 -2.2728E+00 3.3903E-02 -7.0911E-06 7.4305E-10 -2.1274E-14
NC6 -6.3694E+02 1.9744E+01 -7.0295E-03 3.9020E-05 -2.1744E-08 4.0917E-12
HVY1 0.0000E+00 4.4727E+00 4.4674E-02 -6.3579E-06 0.0000E+00 0.0000E+00
HVY2 0.0000E+00 2.7007E+00 5.8820E-02 -7.8301E-06 0.0000E+00 0.0000E+00
HVY3 0.0000E+00 4.9001E+00 8.9064E-02 -1.1580E-05 0.0000E+00 0.0000E+00
ENDEOS
C
PVTFILE
KVFILE
C
CCEXP RUN1
COMPOSITION

38
.012100 .019400 .659900 .086900 .059100 .027800 .023900
.011200 .015700 .018100 .034766 .019684 .011451
TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES VREL XLIQ ZG
6000.0 0.8045 0.0 1.129
5500.0 0.8268 0.0 1.063
5000.0 0.8530 0.0 0.9980
4500.0 0.8856 0.0 0.9930
4000.0 0.9284 0.0 0.8690
3600.0 0.9745 0.0 0.8220
3428.0 1.000 0.0 0.0
3400.0 1.004 0.009000 0.0
3350.0 1.014 0.02700 0.0
3200.0 1.047 0.08100 0.0
3000.0 1.100 0.1500 0.0
2800.0 1.164 0.0 0.0
2400.0 1.341 0.0 0.0
2000.0 1.611 0.0 0.0
1600.0 2.041 0.0 0.0
1300.0 2.554 0.0 0.0
1030.0 3.293 0.0 0.0
836.00 4.139 0.0 0.0
C
CVDEP RUN1
COMPOSITION
.012100 .019400 .659900 .086900 .059100 .027800 .023900
(Input data file for Step 2 - Default Fluid Predictions p. 1/4)

.011200 .015700 .018100 .034766 .019684 .011451
TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES 3428.0 3000.0 2400.0 1800.0 1200.0 700.00
MW HVY1 HVY3 140.00 127.00 118.00 111.00 106.00 105.00
ZGAS 0.8030 0.7980 0.8020 0.8300 0.8770 0.9240
VPROD 0.0 0.09095 0.2470 0.42026 0.59687 0.7402
SLIQ 0.0 0.1500 0.1990 0.1920 0.1710 0.1520
C
SWELL RUN1
COMPOSITION
.012100 .019400 .659900 .086900 .059100 .027800 .023900
.011200 .015700 .018100 .034766 .019684 .011451
YINJ
.000000 .000000 .946800 .052700 .000500 .000000 .000000
.000000 .000000 .000000 .000000 .000000 .000000
TEMP 200.0 F
PSIG
ZINJ VREL PSAT TYPE
0.0 1.000 3428.0 DEWPT
0.1271 1.122 3635.0 DEWPT
0.3046 1.354 4015.0 DEWPT
0.5384 1.925 4610.0 DEWPT
0.6538 2.504 4880.0 DEWPT
C
END

39
(Input data file for Step 3 - Pre-regression)

REGRESS.DAT

REGRESS
VARIABLE MIN INIT MAX
1 .8000 1.0000 1.2000
2 .8000 1.0000 1.2000
3 .8000 1.0000 1.2000
4 .8000 1.0000 1.2000
5 .8000 1.0000 1.2000
6 .8000 1.0000 1.2000
7 .8000 1.0000 1.2000
8 .8000 1.0000 1.2000
9 .8000 1.0000 1.2000
10 .8000 1.0000 1.2000
11 .8000 1.0000 1.2000
IMAX IPRINT H TOL1 TOL2 TOL3
20 1 .200000 .001000 .001000 .010000
C1 X X X X X 1 2 X
HVY1 X X X X X 3 6 X
DJK C1
HVY1 9
HVY2 10
HVY3 11
ENDREG
C
C
EOS PR
COMPONENTS
NCV 13
PROPERTIES F PSIA
CO2 44.010 87.90 1070.9 .27420 .22250 .457236 .077796 49.60
N2 28.013 -232.40 493.0 .29100 .03720 .457236 .077796 35.00
C1 16.043 -116.60 667.8 .28900 .01260 .457236 .077796 71.00
C2 30.070 90.10 707.8 .28500 .09780 .457236 .077796 111.00
C3 44.097 206.00 616.3 .28100 .15410 .457236 .077796 151.00
NC4 58.124 305.70 550.7 .27400 .20150 .457236 .077796 191.00
IC4 58.124 275.00 529.1 .28300 .18400 .457236 .077796 191.00
NC5 72.151 385.70 488.6 .26200 .25240 .457236 .077796 231.00
IC5 72.151 369.10 490.4 .27300 .22860 .457236 .077796 231.00
NC6 86.178 453.70 436.9 .26400 .29980 .457236 .077796 271.00
HVY1 107.770 561.98 410.5 .26239 .31504 .457236 .077796 353.20
HVY2 148.050 677.63 325.2 .24598 .41381 .457236 .077796 474.02
HVY3 224.010 843.58 244.4 .23072 .57895 .457236 .077796 677.90
DJK CO2
N2 .000000
C1 .150000
C2 .150000
C3 .150000
NC4 .150000
IC4 .150000
NC5 .150000
IC5 .150000
NC6 .150000
HVY1 .150000
HVY2 .150000
HVY3 .150000
DJ K N2
C1 .120000

40
C2 .120000
C3 .120000
NC4 .120000
IC4 .120000
NC5 .120000
IC5 .120000
NC6 .120000
HVY1 .120000
HVY2 .120000
HVY3 .120000
DJ K C1
C2 .000000
C3 .000000
NC4 .020000
IC4 .020000
NC5 .020000
IC5 .020000
NC6 .025000
HVY1 .036113
HVY2 .041436
HVY3 .049070
DJ K C2
C3 .000000
NC4 .010000
IC4 .010000
NC5 .010000
IC5 .010000
NC6 .010000
HVY1 .010000
HVY2 .010000
HVY3 .010000
DJ K C3
NC4 .010000
IC4 .010000
NC5 .010000
IC5 .010000
NC6 .010000
HVY1 .010000
HVY2 .010000
HVY3 .010000
DJK NC4
IC4 .000000
NC5 .000000
IC5 .000000
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJK IC4
NC5 .000000
IC5 .000000
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJK NC5
IC5 .000000
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJK IC5
NC6 .000000
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJK NC6

41
HVY1 .000000
HVY2 .000000
HVY3 .000000
DJK HVY1
HVY2 .000000
HVY3 .000000
DJK HVY2
HVY3 .000000
HIDEAL
CO2 2.1028E+02 5.0361E+00 4.4507E-03 -1.1658E-06 1.5274E-10 -5.7829E-15
N2 -2.6164E+01 7.1489E+00 -4.9835E-04 4.4513E-07 -9.0210E-11 4.4521E-15
C1 -1.1193E+02 9.1718E+00 -4.7216E-03 6.7887E-06 -2.4493E-09 3.1208E-13
C2 -6.3147E-01 7.9649E+00 -7.5217E-04 8.7905E-06 -3.8672E-09 5.4789E-13
C3 -3.2544E+01 7.6111E+00 4.1469E-03 9.5048E-06 -4.7227E-09 7.0237E-13
NC4 4.3186E+02 5.7293E+00 1.5646E-02 3.0120E-06 -2.4420E-09 3.8132E-13
IC4 6.6831E+02 2.7132E+00 1.9460E-02 8.3832E-07 -1.8392E-09 3.1555E-13
NC5 1.9605E+03 -2.0202E-01 3.1752E-02 -6.2257E-06 5.8994E-10 -1.4225E-14
IC5 1.9930E+03 -2.2728E+00 3.3903E-02 -7.0911E-06 7.4305E-10 -2.1274E-14
NC6 -6.3694E+02 1.9744E+01 -7.0295E-03 3.9020E-05 -2.1744E-08 4.0917E-12
HVY1 0.0000E+00 4.4727E+00 4.4674E-02 -6.3579E-06 0.0000E+00 0.0000E+00
HVY2 0.0000E+00 2.7007E+00 5.8820E-02 -7.8301E-06 0.0000E+00 0.0000E+00
HVY3 0.0000E+00 4.9001E+00 8.9064E-02 -1.1580E-05 0.0000E+00 0.0000E+00
ENDEOS
C
PVTFILE
KVFILE
C
CCEXP RUN1
COMPOSITION
.012100 .019400 .659900 .086900 .059100 .027800 .023900
.011200 .015700 .018100 .034766 .019684 .011451
TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES VREL XLIQ ZG
6000.0 0.8045 0.0 1.129
5500.0 0.8268 0.0 1.063
5000.0 0.8530 0.0 0.9980
4500.0 0.8856 0.0 0.9930
4000.0 0.9284 0.0 0.8690
3600.0 0.9745 0.0 0.8220
3428.0 1.000 0.0 0.0
3400.0 1.004 0.009000 0.0
3350.0 1.014 0.02700 0.0
3200.0 1.047 0.08100 0.0
3000.0 1.100 0.1500 0.0
2800.0 1.164 0.0 0.0
2400.0 1.341 0.0 0.0
2000.0 1.611 0.0 0.0
1600.0 2.041 0.0 0.0
1300.0 2.554 0.0 0.0
1030.0 3.293 0.0 0.0
836.00 4.139 0.0 0.0
C
CVDEP RUN1
COMPOSITION
.012100 .019400 .659900 .086900 .059100 .027800 .023900
.011200 .015700 .018100 .034766 .019684 .011451
TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES 3428.0 3000.0 2400.0 1800.0 1200.0 700.00
MW HVY1 HVY3 140.00 127.00 118.00 111.00 106.00 105.00
ZGAS 0.8030 0.7980 0.8020 0.8300 0.8770 0.9240
VPROD 0.0 0.09095 0.2470 0.42026 0.59687 0.7402
SLIQ 0.0 0.1500 0.1990 0.1920 0.1710 0.1520
C
SWELL RUN1
COMPOSITION

42
.012100 .019400 .659900 .086900 .059100 .027800 .023900
.011200 .015700 .018100 .034766 .019684 .011451
YINJ
.000000 .000000 .946800 .052700 .000500 .000000 .000000
.000000 .000000 .000000 .000000 .000000 .000000
TEMP 200.0 F
PSIG
ZINJ VREL PSAT TYPE
0.0 1.000 3428.0 DEWPT
0.1271 1.122 3635.0 DEWPT
0.3046 1.354 4015.0 DEWPT
0.5384 1.925 4610.0 DEWPT
0.6538 2.504 4880.0 DEWPT
C
END

43
(Output data file for Step 3 - Pre-regression)

REGRESS.EOS

EOS PR
COMPONENTS
NCV 13
PROPERTIES F PSIA
CO2 44.01 87.90 1070.90 .2742 .2225 .4572360 .0777960 49.6
N2 28.01 -232.40 493.00 .2910 .0372 .4572360 .0777960 35.0
C1 16.04 -116.60 667.80 .2890 .0126 .3776518 .0727877 71.0
C2 30.07 90.10 707.80 .2850 .0978 .4572360 .0777960 111.0
C3 44.10 206.00 616.30 .2810 .1541 .4572360 .0777960 151.0
NC4 58.12 305.70 550.70 .2740 .2015 .4572360 .0777960 191.0
IC4 58.12 275.00 529.10 .2830 .1840 .4572360 .0777960 191.0
NC5 72.15 385.70 488.60 .2620 .2524 .4572360 .0777960 231.0
IC5 72.15 369.10 490.40 .2730 .2286 .4572360 .0777960 231.0
NC6 86.18 453.70 436.90 .2640 .2998 .4572360 .0777960 271.0
HVY1 107.77 561.98 410.50 .2624 .3150 .4169631 .0933548 353.2
HVY2 148.05 677.63 325.20 .2460 .4138 .5486750 .0883512 474.0
HVY3 224.01 843.58 244.40 .2307 .5789 .5344757 .0933546 677.9
DJK N2
CO .0000000
DJK C1
CO2 .1500000
N2 .1200000
DJK C2
CO2 .1500000
N2 .1200000
C1 .0000000
DJK C3
CO2 .1500000
N2 .1200000
C1 .0000000
C2 .0000000
DJK NC4
CO2 .1500000
N2 .1200000
C1 .0200000
C2 .0100000
C3 .0100000
DJK IC4
CO2 .1500000
N2 .1200000
C1 .0200000
C2 .0100000
C3 .0100000
NC4 .0000000
DJK NC5
CO2 .1500000
N2 .1200000
C1 .0200000
C2 .0100000
C3 .0100000
NC4 .0000000
IC4 .0000000
DJ K IC5
CO2 .1500000
N2 .1200000
C1 .0200000
C2 .0100000
C3 .0100000

44
NC4 .0000000
IC4 .0000000
NC5 .0000000
DJ K NC6
CO2 .1500000
N2 .1200000
C1 .0250000
C2 .0100000
C3 .0100000
NC4 .0000000
IC4 .0000000
NC5 .0000000
IC5 .0000000
DJ K HVY1
CO2 .1500000
N2 .1200000
C1 -.1638391
C2 .0100000
C3 .0100000
NC4 .0000000
IC4 .0000000
NC5 .0000000
IC5 .0000000
NC6 .0000000
DJ K HVY2
CO2 .1500000
N2 .1200000
C1 -.1585253
C2 .0100000
C3 .0100000
NC4 .0000000
IC4 .0000000
NC5 .0000000
IC5 .0000000
NC6 .0000000
HVY1 .0000000
DJ K HVY3
CO2 .1500000
N2 .1200000
C1 .2490480
C2 .0100000
C3 .0100000
NC4 .0000000
IC4 .0000000
NC5 .0000000
IC5 .0000000
NC6 .0000000
HVY1 .0000000
HVY2 .0000000
C
HSTAR
CO2 2.10280E+02 5.03610E+00 4.45070E-03 -1.16580E-06 1.52740E-10
-5.78290E-15
N2 -2.61640E+01 7.14890E+00 -4.98350E-04 4.45130E-07 -9.02100E-11
4.45210E-15
C1 -1.11930E+02 9.17180E+00 -4.72160E-03 6.78870E-06 -2.44930E-09
3.12080E-13
C2 -6.31470E-01 7.96490E+00 -7.52170E-04 8.79050E-06 -3.86720E-09
5.47890E-13
C3 -3.25440E+01 7.61110E+00 4.14690E-03 9.50480E-06 -4.72270E-09
7.02370E-13
NC4 4.31860E+02 5.72930E+00 1.56460E-02 3.01200E-06 -2.44200E-09
3.81320E-13
IC4 6.68310E+02 2.71320E+00 1.94600E-02 8.38320E-07 -1.83920E-09
3.15550E-13
NC5 1.96050E+03 -2.02020E-01 3.17520E-02 -6.22570E-06 5.89940E-10
-1.42250E-14

45
IC5 1.99300E+03 -2.27280E+00 3.39030E-02 -7.09110E-06 7.43050E-10
-2.12740E-14
NC6 -6.36940E+02 1.97440E+01 -7.02950E-03 3.90200E-05 -2.17440E-08
4.09170E-12
HVY1 0.00000E+00 4.47270E+00 4.46740E-02 -6.35790E-06 0.00000E+00
0.00000E+00
HVY2 0.00000E+00 2.70070E+00 5.88200E-02 -7.83010E-06 0.00000E+00
0.00000E+00
HVY3 0.00000E+00 4.90010E+00 8.90640E-02 -1.15800E-05 0.00000E+00
0.00000E+00
C
ENDEOS
C
C Global Reference Component Composition
COMPOSITION
.012100 .019400 .659900 .086900 .059100 .027800 .023900 .011200
.015700 .018100 .034766 .019684 .011451

46
(Input data file for Step 4 - Component Pseudoization)

PSEUDO.DAT

EOS PR
COMPONENTS
P1 P2 P3 P4 P5 P6 P7
NCV 7
PROPERTIES F PSIA
P1 16.382 -119.91 662.8 .28883 .01330 .376528 .072862 69.97
P2 31.774 89.83 752.2 .28944 .11304 .451923 .079260 103.50
P3 50.642 245.90 581.0 .28180 .17244 .459850 .078437 169.66
P4 77.793 407.26 468.4 .26789 .26316 .460949 .078234 247.09
P5 107.770 561.98 410.5 .26240 .31500 .416963 .093355 353.20
P6 148.050 677.63 325.2 .24600 .41380 .548675 .088351 474.00
P7 224.010 843.58 244.4 .23070 .57890 .534476 .093355 677.90
DJK P1
P2 .000947
P3 .007835
P4 .021885
P5 -.160885
P6 -.155665
P7 .244672
DJK P2
P3 .003695
P4 .010541
P5 .010541
P6 .010541
P7 .010541
DJK P3
P4 .002281
P5 .002281
P6 .002281
P7 .002281
DJK P4
P5 .000000
P6 .000000
P7 .000000
DJK P5
P6 .000000
P7 .000000
DJK P6
P7 .000000
HIDEAL
P3 0.0000E+00 3.0481E+00 2.2327E-02 -3.1330E-06 0.0000E+00 0.0000E+00
P4 0.0000E+00 1.3339E+00 3.0999E-02 -4.1340E-06 0.0000E+00 0.0000E+00
P6 0.0000E+00 2.5153E+01 7.7904E-02 -1.2416E-05 0.0000E+00 0.0000E+00
ENDEOS
C
PVTFILE
KVFILE
C
CCEXP RUN1
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451

TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES VREL XLIQ ZG
6000.0 0.8045 0.0 1.129
5500.0 0.8268 0.0 1.063
5000.0 0.8530 0.0 0.9980
4500.0 0.8856 0.0 0.9930
4000.0 0.9284 0.0 0.8690

47
3600.0 0.9745 0.0 0.8220
3428.0 1.000 0.0 0.0
3400.0 1.004 0.009000 0.0
3350.0 1.014 0.02700 0.0
3200.0 1.047 0.08100 0.0
3000.0 1.100 0.1500 0.0
2800.0 1.164 0.0 0.0
2400.0 1.341 0.0 0.0
2000.0 1.611 0.0 0.0
1600.0 2.041 0.0 0.0
1300.0 2.554 0.0 0.0
1030.0 3.293 0.0 0.0
836.00 4.139 0.0 0.0
C
CVDEP RUN1
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451

TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES 3428.0 3000.0 2400.0 1800.0 1200.0 700.00
MW P5 P7 140.00 127.00 118.00 111.00 106.00 105.00
ZGAS 0.8030 0.7980 0.8020 0.8300 0.8770 0.9240
VPROD 0.0 0.09095 0.2470 0.42026 0.59687 0.7402
SLIQ 0.0 0.1500 0.1990 0.1920 0.1710 0.1520
C
SWELL RUN1
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451

YINJ
.946800 .052700 .000500 .000000 .000000 .000000 .000000

TEMP 200.0 F
PSIG
ZINJ VREL PSAT TYPE
0.0 1.000 3428.0 DEWPT
0.1271 1.122 3635.0 DEWPT
0.3046 1.354 4015.0 DEWPT
0.5384 1.925 4610.0 DEWPT
0.6538 2.504 4880.0 DEWPT
C
END

48
(Output data file for Step 4 - Component Pseudoization)

PSEUDO.EOS

C EQUATION OF STATE DATA PRODUCED BY DESKTOP-PVT FOR PSEUDO COMPONENTS
C
EOS PR
COMPONENTS
P1 P2 P3 P4 P5 P6 P7
PROPERTIES F PSIA
P1 16.38 -119.91 662.81 .2888 .0133 .3765281 .0728615 70.0
P2 31.77 89.83 752.18 .2894 .1130 .4519229 .0792598 103.5
P3 50.64 245.90 581.03 .2818 .1724 .4598499 .0784368 169.7
P4 77.79 407.26 468.43 .2679 .2632 .4609493 .0782338 247.1
P5 107.77 561.98 410.50 .2624 .3150 .4169631 .0933548 353.2
P6 148.05 677.63 325.20 .2460 .4138 .5486750 .0883512 474.0
P7 224.01 843.58 244.40 .2307 .5789 .5344757 .0933546 677.9
DJK P2
P1 .0009470
DJK P3
P1 .0078346
P2 .0036951
DJK P4
P1 .0218854
P2 .0105413
P3 .0022813
DJK P5
P1 -.1608847
P2 .0105413
P3 .0022813
P4 .0000000
DJK P6
P1 -.1556652
P2 .0105413
P3 .0022813
P4 .0000000
P5 .0000000
DJK P7
P1 .2446722
P2 .0105413
P3 .0022813
P4 .0000000
P5 .0000000
P6 .0000000
C
HSTAR
P1 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
P2 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
P3 0.00000E+00 3.04810E+00 2.23275E-02 -3.13303E-06 0.00000E+00
0.00000E+00
P4 0.00000E+00 1.33391E+00 3.09992E-02 -4.13396E-06 0.00000E+00
0.00000E+00
P5 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
P6 0.00000E+00 2.51533E+01 7.79036E-02 -1.24159E-05 0.00000E+00
0.00000E+00
P7 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
C
ENDEOS
C
C Pseudoized Component Composition
C

49
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451

50
(Input data file for Step5 - Regressio)

FINAL.DAT

REGRESS
VARIABLE MIN INIT MAX
1 .8000 1.0000 1.2000
2 .8000 1.0000 1.2000
3 .8000 1.0000 1.2000
4 .8000 1.0000 1.2000
5 .8000 1.0000 1.2000
6 .8000 1.0000 1.2000
7 .8000 1.0000 1.2000
8 .8000 1.0000 1.2000
9 .8000 1.0000 1.2000
10 .8000 1.0000 1.2000
11 .8000 1.0000 1.2000
12 .8000 1.0000 1.2000
13 .8000 1.0000 1.2000
IMAX IPRINT H TOL1 TOL2 TOL3
20 1 .200000 .001000 .001000 .010000
P1 X X X 1 X 6 7 X
P2 X X X 2 X X X X
P5 11 X X 3 X 8 9 X
P6 12 X X 4 X 8 9 X
P7 13 X X 5 X 8 9 X
DJK P1
P7 10
ENDREG
C
C
EOS PR
COMPONENTS
P1 P2 P3 P4 P5 P6 P7
NCV 7
PROPERTIES F PSIA
P1 16.382 -119.91 662.8 .28883 .01330 .376528 .072862 69.97
P2 31.774 89.83 752.2 .28944 .11304 .451923 .079260 103.50
P3 50.642 245.90 581.0 .28180 .17244 .459850 .078437 169.66
P4 77.793 407.26 468.4 .26789 .26316 .460949 .078234 247.09
P5 107.770 561.98 410.5 .26240 .31500 .416963 .093355 353.20
P6 148.050 677.63 325.2 .24600 .41380 .548675 .088351 474.00
P7 224.010 843.58 244.4 .23070 .57890 .534476 .093355 677.90
DJK P1
P2 .000947
P3 .007835
P4 .021885
P5 -.160885
P6 -.155665
P7 .244672
DJK P2
P3 .003695
P4 .010541
P5 .010541
P6 .010541
P7 .010541
DJK P3
P4 .002281
P5 .002281
P6 .002281
P7 .002281
DJK P4
P5 .000000

51
P6 .000000
P7 .000000
DJK P5
P6 .000000
P7 .000000
DJK P6
P7 .000000
HIDEAL
P3 0.0000E+00 3.0481E+00 2.2327E-02 -3.1330E-06 0.0000E+00 0.0000E+00
P4 0.0000E+00 1.3339E+00 3.0999E-02 -4.1340E-06 0.0000E+00 0.0000E+00
P6 0.0000E+00 2.5153E+01 7.7904E-02 -1.2416E-05 0.0000E+00 0.0000E+00
ENDEOS
C
PVTFILE
KVFILE
C
CCEXP RUN1
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451

TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES VREL XLIQ ZG
6000.0 0.8045 0.0 1.129
5500.0 0.8268 0.0 1.063
5000.0 0.8530 0.0 0.9980
4500.0 0.8856 0.0 0.9930
4000.0 0.9284 0.0 0.8690
3600.0 0.9745 0.0 0.8220
3428.0 1.000 0.0 0.0
3400.0 1.004 0.009000 0.0
3350.0 1.014 0.02700 0.0
3200.0 1.047 0.08100 0.0
3000.0 1.100 0.1500 0.0
2800.0 1.164 0.0 0.0
2400.0 1.341 0.0 0.0
2000.0 1.611 0.0 0.0
1600.0 2.041 0.0 0.0
1300.0 2.554 0.0 0.0
1030.0 3.293 0.0 0.0
836.00 4.139 0.0 0.0
C
CVDEP RUN1
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451
TEMP 200.0 F
DEWPT 3428.0 PSIG
PRES 3428.0 3000.0 2400.0 1800.0 1200.0 700.00
MW P5 P7 140.00 127.00 118.00 111.00 106.00 105.00
ZGAS 0.8030 0.7980 0.8020 0.8300 0.8770 0.9240
VPROD 0.0 0.09095 0.2470 0.42026 0.59687 0.7402
SLIQ 0.0 0.1500 0.1990 0.1920 0.1710 0.1520
C
SWELL RUN1
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451

YINJ
.946800 .052700 .000500 .000000 .000000 .000000 .000000

TEMP 200.0 F
PSIG
ZINJ VREL PSAT TYPE
0.0 1.000 3428.0 DEWPT
0.1271 1.122 3635.0 DEWPT
0.3046 1.354 4015.0 DEWPT
0.5384 1.925 4610.0 DEWPT

52
0.6538 2.504 4880.0 DEWPT
C
END

(Output data file for Step5 - Regression)

FINAL.EOS

EOS PR
COMPONENTS
P1 P2 P3 P4 P5 P6 P7
NCV 7
PROPERTIES F PSIA
P1 16.38 -119.91 662.80 .2888 .0133 .3528063 .0710390 70.0
P2 31.77 89.83 752.20 .2894 .1130 .4519230 .0792600 103.5
P3 50.64 245.90 581.00 .2818 .1724 .4598500 .0784370 169.7
P4 77.79 407.26 468.40 .2679 .2632 .4609490 .0782340 247.1
P5 102.03 561.98 410.50 .2624 .3150 .4192163 .0950905 353.2
P6 150.30 677.63 325.20 .2460 .4138 .5516401 .0899935 474.0
P7 233.65 843.58 244.40 .2307 .5789 .5373643 .0950905 677.9
DJK P2
P1 .0009470
DJK P3
P1 .0078350
P2 .0036950
DJK P4
P1 .0218850
P2 .0105410
P3 .0022810
DJK P5
P1 -.1608850
P2 .0105410
P3 .0022810
P4 .0000000
DJK P6
P1 -.1556650
P2 .0105410
P3 .0022810
P4 .0000000
P5 .0000000
DJK P7
P1 .2481725
P2 .0105410
P3 .0022810
P4 .0000000
P5 .0000000
P6 .0000000
C
HSTAR
P1 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
P2 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
P3 0.00000E+00 3.04810E+00 2.23270E-02 -3.13300E-06 0.00000E+00
0.00000E+00
P4 0.00000E+00 1.33390E+00 3.09990E-02 -4.13400E-06 0.00000E+00
0.00000E+00
P5 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
P6 0.00000E+00 2.51530E+01 7.79040E-02 -1.24160E-05 0.00000E+00
0.00000E+00
P7 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00 0.00000E+00
0.00000E+00
C

53
ENDEOS
C
C Global Reference Component Composition
C
COMPOSITION
.679300 .099000 .110800 .045000 .034766 .019684 .011451

54
(Input data for Step 6 - Initialization for simulation)

Spe3i.dat

C VIP-EXEC COMPOSITIONAL EXAMPLE #1
C
C DATA SET NAME: SPE3I.DAT
C
C FEATURES DEMONSTRATED:
C (1) RECTANGULAR GRID 9X9X4
C (2) GAS CONDENSATE, GAS CYCLING
C
C -----------------------------------------
C
INIT
TITLE1
SPE COMPARATIVE PROJECT 3
TITLE2
DATA SET NAME: SPE3I.DAT
TITLE3
SINGLE WELL RECTANGULAR GRID GAS CONDENSATE
C
NCOL 80
NLINES 60
C
PRINT ALL
C
C INITIAL SIMULATION DATE:
DATE 01 01 1992
C

C

NX NY NZ NCOMP
9 9 4 7
C
C CONSTANT RESERVOIR PROPERTIES
DWB BWI VW CW CR TRES TS PS
1. .9813 .3 3.0E-6 4.0E-6 200. 60. 14.7
C
C
C
NOLIST
TABLES
IEQUIL PINIT DEPTH PCWOC WOC PCGOC GOC PSAT
1 3543 7500 0 7500 0 7500 3443
C

C
SWT 1
SW KRW KROW PCWO
0.00 0.0 0.801 75.0
0.04 0.0 0.8005 71.0
0.08 0.0 0.8003 70.0
0.12 0.0 0.8001 60.0
0.16 0.00 0.80 50.0
0.2 0.002 0.65 30.0
0.24 0.01 0.513 21.0
0.28 0.02 0.4 15.5
0.32 0.033 0.315 12.0
0.36 0.049 0.25 9.20
0.40 0.066 0.196 7.00
0.44 0.090 0.15 5.30
0.48 0.119 0.112 4.20
0.52 0.15 0.082 3.40

55
0.56 0.186 0.060 2.70
0.60 0.227 0.04 2.10
0.64 0.277 0.024 1.70
0.68 0.330 0.012 1.30
0.72 0.39 0.005 1.00
0.76 0.462 0.0 0.70
0.80 0.54 0.0 0.50
0.84 0.62 0.0 0.40
0.88 0.71 0.0 0.30
0.92 0.80 0.0 0.20
0.96 0.90 0.0 0.10
1.00 1.00 0.0 0.00
C

SGT 1
SG KRG KROG PCGO
0.0 0.0 0.801 0.0
0.04 0.005 0.650 0.0
0.08 0.013 0.513 0.0
0.12 0.026 0.400 0.0
0.16 0.040 0.315 0.0
0.20 0.058 0.250 0.0
0.24 0.078 0.196 0.0
0.28 0.100 0.150 0.0
0.32 0.126 0.112 0.0
0.36 0.156 0.082 0.0
0.40 0.187 0.060 0.0
0.44 0.222 0.040 0.0
0.48 0.260 0.024 0.0
0.52 0.300 0.012 0.0
0.56 0.348 0.005 0.0
0.60 0.400 0.00 0.0
0.64 0.450 0.00 0.0
0.68 0.505 0.00 0.0
0.72 0.562 0.00 0.0
0.76 0.620 0.00 0.0
0.80 0.680 0.0 0.0
0.84 0.74 0.0 0.0
0.88 0.74 0.0 0.0
0.92 0.74 0.0 0.0
0.96 0.74 0.0 0.0
1.00 0.74 0.0 0.0
C EQUATION OF STATE DATA
C
C
EOS PR
COMPONENTS
P1 P2 P3 P4 P5 P6 P7
PROPERTIES F PSIA
COMP MW TC PC ZC ACENTRIC OMEGAA OMEGAB
P1 16.38 -119.91 662.80 .2888 .0133 .352806 .071039
P2 31.77 89.83 752.20 .2894 .1130 .451923 .079260
P3 50.64 245.90 581.00 .2818 .1724 .459850 .0784370
P4 77.79 407.26 468.40 .2679 .2532 .460949 .078234
P5 102.03 561.98 410.50 .2624 .3150 .419216 .095091
P6 150.30 677.63 325.20 .2460 .4138 .551640 .089993
P7 233.65 843.58 244.40 .23070 .57890 .537364 .095091
DJK P1
P2 .000947
P3 .007835
P4 .021885
P5 -.160885
P6 -.155665
P7 .248173
DJK P2
P3 .003695
P4 .010541

56
P5 .010541
P6 .010541
P7 .010541
DJK P3
P4 .002281
P5 .002281
P6 .002281
P7 .002281
DJK P4
P5 .000000
P6 .000000
P7 .000000
DJK P5
P6 .000000
P7 .000000
DJK P6
P7 .000000
C
ENDEOS
C
C
GASMF
.679300 .099000 .110800 .045000 .034766 .019684 .011451
OILMF
X X X X X X X
C
C
C RESERVOIR DESCRIPTION ARRAYS FOLLOW:
ARRAYS
C
DX CON
293.3
DY CON
293.3
DZ ZVAR
30 30 50 50
MDEPTH ZVAR
7330 7360 7400 7450
POR CON
.13
C
KX ZVAR
130 40 20 150
KY ZVAR
130 40 20 150
KZ ZVAR
13 4 2 15
C
RUN
STOP
END

57
(Input data for Step 7 - Simulation)
Spe3r.dat

C VIP-EXEC COMPOSITIONAL EXAMPLE #1
C
C DATA SET NAME: spe3r.DAT
C
C FEATURES DEMONSTRATED:
C (1) IMPLICIT
C (2) GAUSS
C
C -----------------------------------------
RUN
C
DIM NAMAX
300000
C
IMPLICIT
C
RESTART 0 3
C
TITLE2
DATA SET NAME: spe3r.DAT
TITLE3
SINGLE WELL - RICH GAS CONDENSATE,GAS CYCLING
C
C START CARD DEFINES WHERE DATA PROCESSING BEGINS
START
C
C USE GAUSS AS SOLUTION METHOD
GAUSS
C
C WELL DEFINITION DATA
WELL N IW JW IGC IBAT
1 7 7 1 1
2 7 1 1 1
C
C DEFINE WELL 1 AS A PRODUCER
PROD G 1
INJ G 2
YINJ 2
0.8036 0.1052 0.08 0.0112 0.0 0.0 0.0
FPERF
WELL L IW JW KH
1 3 7 7 1000
1 4 7 7 7500
2 1 1 1 3900
2 2 1 1 1200
C FLOW CHARACTERISTICS FOR WELL 1!
WI 1
1.55
C FLOW CHARACTERISTICS FOR WELL 2
WI 2
1.55
C
C PRESSURE AND RATE CONTROLS FOR WELL 1
BHP 1
500
7400
QMAX 1
6200
QMAX 2
4700
C
C SURFACE SEPARATION DATA FOR WELL 1
SEPARATOR 1

58
STAGE TEMP PRES VFRAC VDEST LFRAC LDEST
1 80. 815. 1. GAS 1. 2
2 80. 65 1. GAS 1. 3
3 60 14.6 1. GAS 1. OIL
NEWSEP 1
1
1
2500
SEPARATOR 1
STAGE TEMP PRES VFRAC VDEST LFRAC LDEST
1 80. 315 1. GAS 1. 2
2 80. 65. 1. GAS 1. 3
3 80. 14.6 1. GAS 1. OIL
C
C
C TIME STEP CONTROLS
DT -10 1. 365. 200. .2 .2 .03
C
C ITERATION CONTROLS
ITNLIM 1 8 300. .2 .2 .02
C OUTPUT CONTROLS
PRINT WELL TIME
OUTPUT SO
C
TIME 10
C SWITCH tO BLITZ
BLITZ
TIME 25
TIME 50
TIME 100
TIME 200
TIME 300
TIME 400
TIME 500
TIME 600
TIME 800
TIME 1000
TIME 1200
TIME 1500
TIME 1825
TIME 2000
TIME 2500
TIME 3000
TIME 3050
TIME 3100
TIME 3150
TIME 3200
TIME 3250
TIME 3300
TIME 3350
TIME 3400
TIME 3450
TIME 3500
TIME 3650
QMAX 2
0.0
TIME 4000
TIME 4500
TIME 5000
TIME 5500
TIME 6000
STOP
END

59
REFERENCES:

1. Coats, K.H. and Smart, G.T.: " Application of a Regression Based EOS PVT Program to
Laboratory Data," SPERE (May 1986) 277-299.

2. Peng, D.-Y. and Robinson, D.B.: "A New Two-Constant Equation of State," Ind. Eng. Chem.
Fundam. (1976) 15, 59.

3. Soave, G.: "Equilibrium Constants From a Modified Redlich Kwong Equation of State," Chem.
Eng. Sci. (1972) 27, 1197-1203.

4. Coats, K.H., Dempsey, J.R., and Henderson, J.H.: "A New Technique for Determining Reservoir
Description from Field Performance Data," SPEJ (March 1970) 66-74.

5. Whitson, C.H.: "Characterizing Hydrocarbon Plus Fractions," SPEJ (Aug. 1983) 683-94.

6. Coats, K.H.: "Simulation of Gas Condensate Reservoir Performance," JPT (Oct. 1985) 1870-86.

7. Kenyon, D.E. and Behie, A.: "Third Comparative Solution Project: Gas Cycling of Retrograde
Condensate Reservoirs," SPE 12278, JPT (August 1987), 981-997.

8. Abel, W., Jackson, R.F. and Wattenbarger, R.A.: "Simulation of a Partial Pressure Maintenance
Gas Cycling Project with a Compositional Model, Carson Creek Field, Alberta", Numerical
Simulation, SPE Reprint Series No. 11, 285-293.

9. Thompson, F.R.: "Compositional Simulation of a Gas Cycling Project, Bonnie Glen D-3A Pool,
Alberta, Canada", Numerical Simulation, SPE Reprint Series No. 11, 314-329.

10. Field, M.B., Givens, J.W. and Paxman, D.S.: "Kaybob South - Reservoir Simulation of a Gas
Cycling Project with Bottom Water Drive", Numerical Simulation, SPE Reprint Series No. 11, 294-
305.

11. Course notes, PETE 605, PETE 607 AND PETE 611, Texas A&M University.

12. Lake, L.W.: Enhanced Oil Recovery, Prentice Hall, 1989.

13. Stalkup, F.I.: Miscible Flooding Fundamentals, SPE Monograph Series, 1983.

Compositional Simulation

Uploaded by

Copyright:

Available Formats

Compositional Simulation

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Compositional Simulation

Uploaded by

Copyright:

Available Formats

CHAPTER 5

You might also like