SPE 99647
Simulation of Experimental Gas-Recycling Experiments in Fractured
Gas/Condensate Reservoirs
S.R. Shadizadeh, Petroleum U. of Technology, and D. Rashtchian and S. Moradi, Sharif U. of Technology
Copyright 2006, Society of Petroleum Engineers
This paper was prepared for presentation at the 2006 SPE Gas Technology Symposium held
in Calgary, Alberta, Canada, 15–17 May 2006.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Abstract
Gas recycling experiments were conducted on a gas
condensate sample in a fractured model. Experimental Model
was then numerically simulated with compositional simulator.
The objectives were to investigate effects of gas recycling on
condensate recovery from a naturally fractured reservoir under
complete pressure maintenance process.
In laboratory scale we could accurately analyse behavior of
in-place and injected gas in fracture and matrix blocks and
understand how the production mechanisms affect ultimate
recovery in this reservoir model. In reservoir scale, due to high
heterogeneity and lack of required data such an accurate
investigation is impossible.
In the simulated model after history matching of
cumulative gas and condensate production and composition
variation history of produced gas, some sensitivity analysis
were performed. Simulation results show that gas recycling
increases liquid recovery more than 2.5 times for the reservoir
if the fracture model behaves as dual permeability fractured
reservoir. If model behavior supposed to be dual porosity,
condensate recovery decreases dramatically due to early break
through of injected gas via fractures. Cyclic injection-
production and decreasing the production rate are the factors
directing the process to higher ultimate recoveries. Nitrogen
injection effect has also been established showing less
recovery than dry gas injection case.
Finally, lacks of data for a complete compositional
simulation of such models are mentioned. If the data be
prepared prior to performing the experiment, results could be
up-scaled to the reservoir with higher degree of assurance.
Introduction
Gas recycling in gas condensate reservoirs has been
recommended for several years as an optimum production
scenario of increasing condensate recovery. For fractured
media, the process is more complex due to early break through
of injected gas in fracture network, and activation of
production mechanisms like diffusion and gravity drainage of
condensates.
Experimental gas recycling is most often used to predict
recycling performance of the reservoir. With a complete rock
and fluid analysis before starting the experiment, laboratory
results can be up-scaled to design the recycling procedure in
the reservoir. By preparing most similarity between laboratory
model and reservoir conditions, probable errors would be
minimized during up-scaling.
Laboratory gas recycling experiments are time and cost
consuming; also in reservoir scale we face high heterogeneity
and large scales and couldn't perform accurate investigation on
recycling process. Numerical simulation of laboratory models
prepares accurate investigation of recycling performance to
select best conditions of pressure maintenance process for a
reservoir. Most important factors are the injection and
production gas rates, composition of gas to be injected, and
when is the best time to start and finish the recycling to
achieve most economic conditions.
Gas condensate behavior is highly sensitive to the heavy
fraction composition of the fluid; so the most probable errors
occur in fluid sampling and phase behavior modeling.
Reducing deviations as will be discussed would yield in
accurate up- scaling of the experiment to the reservoir.
Experimental setup and procedure
Six sandstone core plugs stacked vertically in a core-
holder. Horizontal and vertical spaces were considered
between cores and core holder as horizontal and vertical
fractures (Fig.1). Average porosity of matrix was 21% and
average permeability of matrix and fracture were measured to
be 409 md and 550 darcy respectively. (See properties of rock
samples in table 1.)
Gas and liquid samples were taken from surface separators
and recombined with liquid gas ratio of 20 bbl/MMScf to
represent the reservoir fluid. Compositions of reservoir gas
and dry gas are shown in tables 2, 3 respectively.
Cores were washed in the core holder with toluene and
methanol. Vacuum then imposed on the system under 180°F
temperature to evaporate solvents and exhaust them out of
pore spaces. Prepared reservoir gas then injected into the core-
holder under pressure of 4500 psi and 180°F . Core-holder
closed-in for two weeks for uniformity of fluid distribution in
matrix blocks and fractures. To ensure uniformity, reverse gas
2 SPE 99647
cycling was then performed; 300 cc/hr of reservoir gas was
injected from bottom and same produced from top of core-
holder. Liquid gas ratio was measured during gas cycling until
became steady in 20 bbl/MMScf.
Natural gas production began in 4500 psi and 180 °F with
rate of 1000 cc/hr (standard conditions) and continued until
core holder pressure decreased to 3550 psi (15 days).
Complete pressure maintenance applied by dry gas injection of
2.9 cc/hr (@ 3550 psi and 180°F) from top and 1000 cc/hr of
gas production (standard conditions). Injection and production
were not continuous and core-holder was shut-in at nights
during recycling period. Composition of produced gas samples
was determined by gas chromatography. Recycling continued
until producing gas composition became nearly the same as
injection gas (after 1.3 pore volume of dry gas injection). After
58 days of recycling, core-holder depleted naturally in 2000
cc/hr.Increased production rate was because core-holder was
mostly occupied with methane.
Experimental results show over 80% of condensates was
recovered.
Simulation of experimental model
A compositional simulation was performed using available
physical properties of experiment setup and history of the
recycling process as uncertainities in this scale are lower
compared with actual reservoir scale.
Phase behavior of reservoir gas was modeled by
WINPROP module of CMG package using Peng-Robinson
EOS. The unique measured property for the current gas
sample was gas dew point pressure; for better tuning of EOS,
CCE&CVD tests of similar samples from same reservoir were
used. To ensure correct tuning of EOS describing phase
behavior of reservoir gas, dew point pressure weighted higher
than other tuning parameters.
Gas chromatography analysis of gas sample described the
composition up to C28. Pseudo-components were defined to
minimize numerical dispersion and unnecessary calculations
during simulation run (coupled phase and flow differential
equations muse be solved simultaneously for each block at any
time steps assigned).
In order to minimize final EOS tuning error, C7+ heavy
fraction was splitted into three psedo-components. Adjustable
parameters were critical properties and volume shift of special
components and molecular weight and specific gravity of
heavy pseudo-components.
There are some potential errors during phase behavior
modeling of gas condensate samples that need to be aware of
and take the appropriate actions to minimize them as follows:
- Improper well conditioning: well clean up time should
be carefully planned to ensure that all contaminants are
removed and hydrocarbon withdrawal is minimized prior to
sampling. It is essential that the well production is stable prior
to and during sampling.
For ideal subsurface or surface PVT sampling, it is best to
have small perforation intervals. Hydrocarbon columns are
known to have compositional variations with depth. Short
perforation intervals allow these variations to be captured
during sampling. If the perforations are over a large pay
interval, the fluid sample will represent the most mobile fluid.
If the tubing is too large, greater drawdown is necessary to
maintain stable flow. Greater drawdown means more chance
you will be sampling below dew point. If drawdown is
reduced, flow may become unstable.
- Sample contamination: Even a small amount of
contamination, especially in the heptanes plus, can cause
significant errors in the lab evaluated PVT properties.
Contamination from oil based mud is particularly difficult to
detect. Common sources of contamination are dirty sampling
lines and dirty containers. Another fairly common and easily
preventable source of error is faulty metering. Two common
problems with conventional separator sampling and
recombination are liquid carry over into the gas stream and gas
carry under into the liquid stream.
- Inadequate or poor PVT analysis: The combination of
small volumes of liquid with large volumes of gas leads to
problems in obtaining accurate PVT measurements on gas
condensate fluids. Measurements of small liquid volumes and
dew point pressure have a particularly high uncertainty. Large
variations can occur between results from different samples,
different labs or even different lab technicians. This is
especially true for liquid rich condensates where fluids may be
at conditions close to the critical point. Specially designed
laboratory testing can greatly improve the accuracy of PVT
measurements. Material balance calculations can be performed
on CVD and CCE experiments to investigate potential data
problems.
- Inappropriate fluid model: Various equations of state
(EOS) models are available. The choice of which model to use
depends on fluid type and condition. The Peng- Robinson EOS
is preferred for highly volatile oils or liquids rich gas
condensates nearer the critical point. Sometimes not all the
data can be matched. The data to be matched should be listed
in order of data quality and importance. It is often prohibitive
to use fully compositional simulation for full field simulation.
Modified black oil models can be used with reasonable
confidence to model gas condensate fields under depletion or
recycling above the dew point. For modeling of gas
condensate fields cycling below the dew point, fully
compositional simulation with a minimum number of about
six components is required.
Matrix and fracture permeabilities were measured and
considered constant in both horizontal and vertical directions.
Relative permeabilities were used from similar
experiments performed on the same rock types as SCAL core
analysis wasn't performed on used core plugs.
As explained in experimental procedure, gas recycling
wasn't continuous; so average injection and production rates
were defined in a manner to match measured cumulative
production gas values.
After matching pressure, gas production, liquid recovery
and composition variation of produced gas histories, effects of
production rate changes, dual porosity or dual permeability
behavior of matrix-fracture system, nitrogen injection and
cyclic production on ultimate condensate recovery were
investigated.
SPE 99647
Results and discussion
Results of the simulation are presented in two parts;
history matching and performed sensitivity analysis.
Fig. 2 shows composition variation of methane vs. pore
volume of injected gas. It is evident that after 1.3 pore volume
of dry gas injection, methane mole percent in the produced gas
becomes nearly the same as injection gas. Figures 3, 4, 5, 6
show history matching for C2, C3, C4 and CO2 mole percents
respectively.
Pressure history of the process during three production
periods (first depletion, pressure maintenance and final
depletion) is shown in fig. 7. Simulation results have
acceptable match on experimental data in all mentioned cases.
Fig. 8 represents effect of recycling on condensate
recovery for dual permeability behavior of model. Results
show more than 2.5 times increase in recovery comparing with
complete natural depletion of core holder. Due to good matrix
permeability, dual permeability behavior is more probable
than dual porosity model. Block to block flow prepares larger
contact area and time between injected gas and condensates in
pore volume yielding in later break through of dry gas and
higher recovery.
Fig. 9 compares ultimate recovery for two cases whether
the matrix fracture behaves as dual porosity or dual
permeability model. If the system is modeled to be dual
porosity, even with cyclic injection process, early
breakthrough of injection gas yields in smaller recovery factor
than the dual permeability case.
Effect of cyclic injection-production for dual porosity
model is investigated in fig. 10. Due to static contact of
injected gas and in-place fluids during shut-in period,
molecular diffusion, revaporization and gravity drainage
mechanisms would become stronger and increase ultimate
recovery by delaying injection gas break through to production
well.
Nitrogen injection has always mentioned as alternative
pressure maintenance in gas condensate reservoirs. In
laboratory scale, mixing of nitrogen with reservoir gas causes
more condensation because the mixture dew point pressure is
higher than initial gas dew point. Lower revaporization and
hydrocarbon carrying capacity of nitrogen (compared with dry
gas) also makes dry gas injection more efficient. Results are
shown in fig. 11.
Increasing gas production rate yields in reduction of
ultimate recovery because of performing earlier injection gas
break through via fractures (fig.12).
Conclusions
Based on results of this study, following conclusions can
be drawn:
1. Effects of rate changes on condensate recovery in
homogeneous and fractured reservoirs are not the same. In
homogeneous reservoirs increasing dry gas injection rate
improves condensate recovery due to improvements in relative
permeability values. In fractured media, earlier breakthrough
of injected gas due to high injection rate decreases ultimate
condensate recovery.
2. Fractured reservoir behavior has important effect on
condensate recovery; in dual porosity model early break
through of injected gas via fractures reduces condensate
3
recovery. In case of dual permeability model behavior, block
to block flow increases contact between injected gas and
liquids and delays breakthrough yielding in higher recoveries.
3. For the actual reservoir (from which the fluid sample
was taken), where the matrix blocks are carbonate, lower
matrix permeability reduces block to block flows. As a result,
in reservoir scale we face lower recoveries than what is
referred in experimental data (in laboratory model, system
behavior is more likely dual permeability comparing with dual
porosity behavior of actual reservoir).
4. Recovery increases with cyclic injection-production
scenario. Performing static contact between injected gas and
reservoir condensates by means of shutting in injection and
production wells activates diffusion and revaporization
mechanisms and increases ultimate condensate recovery by
delaying injection gas breakthrough.
5. In laboratory scale, nitrogen injection is less efficient
than dry gas injection. In actual reservoir, negligible
dispersion of N2 reduces negative effects of nitrogen mixing
with reservoir gas and tends the recovery to the ideal dry gas
injection.
References
1. Broad, T.J., Varotsis, N., Pasadakis, N., "The Compositional
Characterization of Gas Condensate Fluids - A Review Featuring the
Impact of the Analysis Data Quality on the Accuracy of Equation of
State Based PVT Predictions", SPE Paper No. 68231 Prepared for
Presentation at the 2001 SPE Middle East Oil Show Held in Bahrain,
17.20 March, 2001.
2. Diamond, L., Rondon, C.R., "Simulation studies of Gas-
Injection into Gas- Condensate Reservoirs", SPE Paper No. 21060
Prepared for Presentation at the 1990 SPE
Latin American Petroleum Engineering Conference held in Rio
de Janeiro, 14-19 October, 1990.
3. Fontaine, T., Fekete, "Four Reasons Why Your Gas
Condensate Fluid Model Is Probably Wrong", Australia Pty Ltd,
1998.
4. Goldthorpe, W.H., "Simulation of Gas Injection Processes in
Gas-Condensate Reservoirs Using a Binary Pseudo-Component
Representation", SPE Paper No.19470 Prepared at the SPE Asia-
Pacific Conference Held in Sydney, Australia, 13-15 September
1989.
5. Sanger, P.J., Hagoort ,J., "Recovery of Gas-Condensate By
Nitrogen Injection Compared with Methane Injection", SPE Paper
No. 30785 Prepared for Presentation at the SPE Annual Technical
Conference & Exhibition Held in Dallas, TX, U.S.A., 22-25
October,1995.
6. Sajjadian, V.A., Danesh A., Tehrani D.H., "Laboratory Study
of Gravity Drainage Mechanism in Fractured Carbonate Reservoir-
Reinfiltration", SPE Paper No. 54003, 1999.
4 SPE 99647

#6
Height Diameter Bulk
φ = 21.8 % 5 mm Core no.
(cm) (cm) Volume
k = 422 md 1 23.5 6.3383 742.486
2 42 6.315 1315.485
3 3939 6.35 1263.721
4 30.8 6.33 969.277
5 18 6.3315 566.729
6 28 6.322 878.94
#5
Table 1. Dimensions of core plugs.
φ = 23.4 %
k = 390 md

#4 Component Mole Fraction

N2 0.296
CO2 2.234
C1 83.987
φ = 19.44 % C2 6.109
k = 408 md C3 2.891
I-C4 0.546
N-C4 1.06
I-C5 0.444
N-C5 0.494
C6 0.726
C7 0.081
C8 0.196
#3 C9 0.237
C10 0.19
C11 0.135
C12 0.099
φ = 20.37 % C13 0.081
k = 409 md C14 0.059
C15 0.046
C16 0.028
C17 0.008
C18 0.019
C19 0.011
C20 0.007
C21 0.004
C22 0.003
C23 0.002
#2 C24 0.001
C25 0.003
C26 0.001
C27 0.001
φ = 20.37 % C28 0.001
k = 409 md
0.095 mm Table 2. Composition of gas condensate sample.

Component Mole Fraction

N2 0.509
#1
C1 98.718
C2 0.586
φ = 21.54 % C3 0.1
k = 414 md I-C4 0.012
N-C4 0.04
I-C5 0.011
N-C5 0.024

Fig.1. Experiment setup. Table 3. Composition of dry injection gas.

SPE 99647 5

100
Simulation 10
Simulation
Experimental
Experimental Data
8
Methane Mole%

Ethane mole%
6
90

80 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
Pore volume gas injected Pore volume gas injected

Fig. 2. Methane cmposition in produced gas during recycling Fig. 3. Ethane composition in produced gas during recycling
period period

5 4
Simulation Simulation
Experimental Data Experimental Data
4
3
Propane Mole%

Butane Mole%

3
2
2

1
1

0 0
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
Pore volume gas injected Pore volume gas injected

Fig. 4. Propane cmposition in produced gas during recycling Fig. 5. Butane cmposition in produced gas during recycling
period period

4 5000
Simulation
Simulation
Experimental Data
Experimental Data
4000
3
Pressure (Psi)
CO2 Mole%

3000
2
2000

1
1000

0 0
0 20 40 60 80 100 120 140 0 200000 400000 600000 800000
Pore volume gas injected Cumulative gas production (cc)

Fig. 6. Carbon dioxide cmposition in produced gas during Fig. 7. Pressure history matching
recycling period
6 SPE 99647

100 100
Simulation-Dual Permeability

Cumulative Condensate Recovery %

GasRecycling-Dual Permeability
Cumulative Condensate Recovery %

Natural Depletion Simulation-Dual Porosity

Experimental Data Experimental Data
80 80

60 60

40 40

20 20

0 0
0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500
Time (hr) Time (hr)

Fig. 8. Effect of gas recycling on condensate recovery. Fig. 9. Effect of fractured reservoir behavior on condensate
recovery.

50
Dual Porosity-Continuous Injection 100
Cumulative Condensate Recovery %

Dry Gas Injection

45 Cumulative Condensate Recovery % 90 Nitrogen Injection
Dual Porosity-Cyclic Injection
40 Natural Depletion
80
35
70
30
60
25
50
20
40
15
30
10 20
5 10
0 0
0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500
Time (hr) Time (hr)

Fig. 10. Cyclic and continuous injection-production. Fig. 11. Comparison between nitrogen and dry gas injection.

100
300cc/hr= ‫يلوت يبد‬
‫ زاگ د‬cc/hr
Cumulative Condensate Recovery %

Gas Rate= 300

90
‫تاع الطا‬ ‫ﻩاگشيامزﺁ‬
Natural ‫ي‬
Depletion
80
500cc/hr= ‫يلوت يبد‬
Gas Rate= ‫ زاگ د‬cc/hr
500
70
60
50
40
30
20
10
0
0 500 1000 1500 2000 2500
Time (hr)

Fig. 12. Effect of gas rate on ultimate recovery.