SPE 122594

Well Test Analysis in Naturally Fractured Gas Condensate Reservoirs

Below Dew Point Pressure
Al-Baqawi, A. M. SPE, Saudi Aramco Oil Company; Al-Malki, B. H., SPE, Saudi Aramco Oil Company
Copyright 2009, Society of Petroleum Engineers

This paper was prepared for presentation at the 2009 SPE Asia Pacific Oil and Gas Conference and Exhibition held in Jakarta, Indonesia, 4–6 August 2009.

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 have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper 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 SPE copyright.

Abstract
Gas condensate reservoirs below dew point pressure have been the subject of many studies in the petroleum literature. These
studies investigated the impact of the resulting liquid condensate drop-out on well performance and concentrated mainly on
reservoirs with homogeneous behaviors.

This paper focuses on the effect of condensate drop-out in naturally fractured carbonate reservoirs with double porosity
behaviors and on the changes in well test responses due to changing oil/gas relative permeabilities. Specifically, this paper
examines the impact on the storativity ratio (ω) and the interporosity flow coefficient (λ) below the dew point pressure as the
radius of the condensate bank increases away from the wellbore. By generating well test responses under different operational
well conditions below the dew point pressure, the paper shows that λ decreases due to two-phase flow in the reservoir,
whereas ω shows very little change. The condensate bank appears as an additional skin effect, as in reservoirs with
homogeneous behavior. This skin can be removed temporarily by hydraulic fracturing.

Conclusions of the paper are verified with well test data from a carbonate formation in the Middle East and compositional
simulation models.

Introduction and background

Condensate drops out around the wellbore in gas condensate reservoirs when the well bottomhole pressure falls below the
dew-point pressure. This has a negative impact on well productivity. Many mega-reservoirs worldwide are affected, such as
the Karachaganak field in Kazakhstan, the Cupiagua field in Colombia, the North Field in Qatar, and the South Pars field in
Iran. Consequently, many studies have been conducted since the 1950’s to understand the behavior of gas condensate
reservoirs below the dew point pressure and to identify the main controlling parameters.

Reservoirs are usually depleted under isothermal conditions as shown in Figure 1. In gas condensate reservoirs, this changes
the fluid from single phase to multiphase in the near-wellbore region, creating three different mobility regions in the reservoir
as schematically represented in Figure 2. Away from the well, where the pressure is still above the dew point pressure, only
gas is present, with the initial condensate saturation (Region 3). Closer to the well, in Regions 1 and 2, the pressure is below
the dew point pressure and the condensate drops out of the gas phase. In Region 2, the condensate saturation is below a
critical value and the condensate is not mobile. In Region 1, on the other hand, the condensate saturation is above critical and
both gas and condensate are mobile and flow together into the wellbore. Regions 1 and 2 are referred to as the condensate
bank, and the decrease in gas effective permeability resulting from the existence of this condensate bank can have a
significant impact on the well performance. The condensate saturation actually decreases and gas mobility increases in the
immediate vicinity of the wellbore, due to capillary number effects (also called velocity stripping, velocity coupling, or
positive coupling), which creates a forth mobility region. These four mobility zones create a three-region composite behavior
in a well test (Gringarten et al., 2000). This composite behavior is superposed on the well reservoir behavior above the dew
point pressure (Gringarten et al. 2006).
2 SPE 122594

Figure 1: Fluid phase diagram of Isothermal depletion of Figure 2: Condensate saturation below dew point pressure flow
pressure in a gas condensate reservoir behavior in the three known regions

Figure 3: Condensate saturation profile and velocity stripping presented by Gringarten et al. (2006)

The dual porosity model was first introduced by Barenblatt et al. (1960), as low permeability and high storativity matrix
blocks surrounded by high permeability and zero-storativity fissures connected in a high permeability continuous network.
Matrix blocks do not interconnect, but produce into fissures and fissures allow fluid to flow into the well. Barenblatt, et al.’s
model was extended to finite-storativity fissures by Warren and Root (1963). Variations of the dual porosity model have been
proposed with different types of matrix-fissures interaction (Odeh, 1965; Kazemi, 1969A; Najurieta, 1980) and near-wellbore
conditions (Mavor and Cinco-Ley, 1979; Bourdet and Gringarten, 1980). Double porosity behavior is controlled by two
parameters, namely the storativity ratio ω and the interporosity flow coefficient λ. The storativity ratio ω is a measurement of
the fluid content in the fracture system:
(ϕ V c t ) f (1)
ω =
(ϕ V c t ) f +m

where f , m and f+m refer to the fissure network, the matrix blocks and the total (fissures plus matrix) system, respectively. ϕ
is the porosity, V the concentration of one medium in the total system, and ct is the total compressibility.

The interporosity flow coefficient λ characterises the communication between matrix blocks and fractures (the lower λ, the
poorer the connection). As a first approximation, λ is given by:
2 k m (2)
λ =α r w
k f

where α is a geometric parameter that depends on the geometry and surface-volume ratio of the matrix blocks, rw is the well
radius, km the matrix premeability, and kf, the fissure network permeability. In practice, the interporosity flow coefficient is
controlled by the interporosity skin, which restricts flow between matrix and fissures (Moench, 1984). In a well test, double
porosity behaviors with restricted interporosity flow exhibit derivatives characterized by a V-shaped minimum between two
same-level radial flow stabilizations at middle times (Bourdet et al., 1989).

Numerical simulations of the Warren and Root double porosity model have been discussed extensively in the literature
(Gilman, J. R., 2003). Multiphase flow requires the use of relative permeability in both matrix blocks and fractures. Relative
permeabilities in the matrix are the same as for conventional single porosity (in the absence of the relative permeability
curves generated from special core analysis, Corey’s functions can be used as shown by Liu et al., 2001). Fracture relative
permeabilities, on the other hand, is usually taken as a linear function of saturations for each phase, based on laboratory
experiments by Romm (1966), although this is only true for large fractures (Maloney and Doggett, 1977).
SPE 122594 3

Evaluation of condensate banking effects

The main goal of this study is to evaluate the effect of gas condensate banking on well test behavior in dual porosity
reservoirs, and in particular the impact on the interporosity flow coefficient (λ) and the storativity ratio (ω). These are
illustrated with test data from a hydraulically fractured wells.

The first part of the study uses compositional simulation using known well and reservoir parameters to generate well test data
for a gas condensate well in a naturally fractured reservoir under different operation conditions above and below the dew
point pressure. Results are then used to analyze well test data from vertical wells in a tight Middle East carbonate reservoir.
Further details can be found in Al-Malki (2004), Shittu (2007) and Al-Baqawi (2008).

PVT fluid model

Knowledge of the gas condensate fluid composition is required to simulate the fluid behavior below the dew point pressure as
the composition is continuously changing and to provide input parameters for well test analysis. The fluid used in this study
had been recovered at reservoir conditions with a reservoir temperature of 275°F. The PVT report included experiments on
constant volume depletion (CVD) and constant composition expansion (CCE). The fluid is a gas condensate with a maximum
liquid dropout of 9.5% at 2200 psig and a dew point pressure of 5718 psig.

To speed up compositional simulation, the heavy components were lumped into one pseudo-component of C7+. The best
match between laboratory measurements and simulated fluid properties on all the available experiments was obtained with
the Peng-Robinson three-parameter model with a dew point pressure of 5624 psig as displayed in Figure 4, Figure 5, and
Figure 6.

Figure 4: Liquid saturation match from the CCE experiment Figure 5: Relative volume CCE experiment match

Figure 6: Vapour Z-factor CCE experiment match

4 SPE 122594

Compositional Simulation
Two fine grid models were utilized. A fine grid, radial model, with 55x1x10 cells, was used to capture the details of the
condensate bank and the saturation levels (Figure 7), whereas a cartesian grid with 27x41x2 cells was used to simulate a
hydraulic fracture with LGR (Figure 8). Both models are based on the Kazemi (1969A) numerical double porosity
formulation.
Table 1 lists the well and reservoir parameters and Table 2 shows the radial cell dimensions, while the relative permeability
end point values used in Corey’s functions are listed in Table 3. Fractures are assigned a linear relative permeability -
saturation relationship.

Figure 7:Radial model for a Naturally fractured gas condensate Figure 8: Cartesian model for NFGC well with fracture
(NFGC) well

Parameter Value Number of Rings (cells) 55

Initial Reservoir pressure for gas condensate Cell radial distance (ft) 0.25, 0.1, 0.15, 0.2, 0.25, 0.30,
7326
(psia) 0.35, 0.4, 0.45, 0.7, 1.0, 1.3, 1.8,
Formation Thickness (ft) 100 2.4, 3.0, 4.0, 5.3, 7.0, 9.5, 11.5,
12.5, 15, 17, 19, 21.16, 25.15,
Average porosity (%) 11
29.88, 35.52, 42.21, 50.17, 59.62,
Absolute matrix permeability (mD) 0.1 70.86, 84.22, 100.09, 118.96,
141.38, 168.03, 199.69
Reservoir Top level (ft) 11496 237.33, 282.07, 335.23, 398.42,
Rock Compressibility (1/psi) 1.0E-5 473.51,562.76, 668.83, 794.89,
944.71, 1122.77, 1334.39,
Sigma 0.05 1585.90, 1800, 2000, 2000, 2000,
Wellbore Radius (radial model) (ft) 0.25 2000, 2000
Wellbore Radius (cartesian model) (ft) 0.06

Table 1: Simulation parameters Table 2: Radial model configuration based on Shittu (2007)

Model Swc Sgc Sorg KrgMAX KrogMAX ng no

Case1 0.24 0 0.2. 1.0 0.6 3.5 3.5

Table 3: Relative permeability end point parameters for matrix blocks

Condensate banking effects below the dew point pressure

Successive 30-day drawdowns and 3-day build ups were simulated above the dew point pressure with the radial simulation
model to verify it was capable of reproducing analytical solutions for double porosity behavior. A long 13-year production
test simulation was then simulated below the dew point pressure (Figure 9). It consisted of a successive one-year drawdowns
and one-month build ups. The simulation showed important indications on the log-plot rate normalized pseudo-pressure and
derivative due to the high production blow the dew point pressure in Figure 10. The lengthy production periods and high
condensate accumulation removes the dual porosity reservoir behavior and condensate banking shows as a skin effect. The
dual porosity behavior dose not disappear immediately as the production is below the dew point pressure but value of λ is
continuously reduced from an initial value of 0.0006 to 0.0002 throughout the simulation as the gas flow reduces in the
fracture networks due to the condensate blocking. The value of ω has been fairly stable around 0.3 which is typical for
multiphase dual porosity models.
SPE 122594 5

Figure 10: Simulation result comparison above and below the

Figure 9: Pressure response and rate history of simulation dew point pressure

A second attempt was carried out by extending Shittu (2007) empirical simulation in Figure 11 for more than 6 years to have
all the flow periods below the Pdew and the result in Figure 12 show similar simulation behavior where the fracture flow of
dual porosity behavior starts to fade away and a noticeable rise precedes the final radial flow stabilization happens after
continuous long production below the dew point.

Figure 11: Shittu (2007) extended to become completely below

dew point pressure Figure 12: Shittu(2007) simulation extension results

Velocity dependence
In the compositional simulator the following relationship to test the capillary number effects was enabled:
v μ
g g (4)
N c
=
σ

This relationship is based on Henderson et al. (2000). Where the vg is the gas velocity, µg is gas viscosity and σ is the gas-oil
surface tension. The radial model simulation was tested again for 3000 days with the North Sea lean gas condensate
parameters were obtained from Bozorgzadeh (2006), which are listed Table 4. The outcome in Figure 13 shows that capillary
number effects has no impact in gas condensate naturally fractured reservoirs and the gas condensate stabilize at the same
levels with or without capillary number effects. The condensate saturation in the matrix blocks stabilize after they reach a
certain limit in saturation and the increase continues gradually in the fractures. In addition, the condensate saturation of the
innermost radius showed that the fracture condensate saturation rises with increasing gas production in Figure 14, as the
fractures represent the easiest available spaces for continuous condensate saturation build-up.

Parameter input Gas Condensate setting for north Sea

mg/o 23.89/79.62
n1g/o 6.23/24.2
n2g/o 0
Ncbg/o 1.0E-6
Table 4: Capillary number parameters for lean gas condensate presented by Bozorgzadeh (2006)
6 SPE 122594

Figure 13: Comparison of condensate dropout with capillary Figure 14:Gas condensate distribution in fractures for
number effects in the matrix and fractures simulation result

Figure 15: Impact on λ and ω with increasing condensate

saturation Figure 16: Simulation rate dependant skin result on last BU

As the change in condensate saturation levels takes place in the matrix and fractures, the corresponding well test analysis
results Figure 15 shows that impact on λ and ω is continuous. Once the saturation level in the fracture rise to 30% after 2000
days, the condensate banking effect appears as a humping skin that removes the dual porosity reservoir behavior. The skin
effect Figure 16 agrees with Al-Malki (2004) that there is no observation of capillary number effect. In the later sections we
will note the similarities in the well test interpretations of the late build up periods of the simulation results.

Impact of induced fracture on gas condensate banking and dual porosity behavior
An initial radial model was utilized to represent a hydraulic fracture in a gas condensate well below the dew point pressure.
The radial model has two main problems in representing hydraulic fractures. First, it was not possible to accurately quantify
the fracture properties. In addition, the condensate banking effect was not noticed after long simulation.

A dual porosity cartesian model was used to represent the fracturing job in a well that produces below the dew point pressure.
The simulation was carried out for 3 years with continuous gas production increase, where selected build up periods
highlighted for analysis in Figure 17.

Figure 18 shows on the log-plot of the rate normalized pseudo-pressure and derivative. Above the dew point pressure (BU1
and BU4) show the typical fractured well response. After the simulation falls below the dew point pressure and the gas
production rates continue to increase the condensate bank effects becomes clear where it starts to appear in BU13 as a small
skin effect represented by a small hump. This skin effect rises in the subsequent build up flow periods with the increasing gas
production rate.
SPE 122594 7

Figure 17: pressure response of cartesian fracture model Figure 18: Cartesian model simulation result

All the conducted simulation behavior on the log-plot rate normalized pseudo-pressure and derivative results are in strong
agreements with Gringarten et al. (2006) analysis of fractured gas condensate wells from the Middle East. It was suggested
that the total skin effect is caused by the mechanical skin effect and it is directly proportional to the gas production rate.

Well Test data and analysis

First, three vertical wells with gas condensate dropout are investigated from a well in a naturally fractured reservoir to show
how a well behaves below the dew point pressure in different operational conditions. All the wells are selected from a
reservoir in the Middle East as the reservoir pressure is falling below the dew point pressure. These wells have been selected
for this study as the impact of gas condensate is more evident in naturally fractured reservoirs but the pressure build up test
has not been specifically designed for this paper.

This reservoir is depleting due to high production rates in many wells. This leads to high disturbance when we analyze one of
the wells at late time effects and for this reason we shall not include any pressure boundary effects in the flow behavior
interpretation. Throughout this work, the main focus will be on the build-up flow periods within the production tests as the
draw down flow periods are extremely complex to interpret due to dynamic fluid changes and phase redistribution. In a dual
porosity reservoir, the permeability thickness is actually Kfh as it is mainly attributed to the fissures permeability but in our
analysis our Kh represents the total system permeability of the fissures and the matrix blocks. This approach allows us to
keep Kh consistent when studying different pressure build up tests for a single well.

Single phase pseudo-pressure was used in analysis of the selected flow periods for each well, as we expect to observe a
change in mobility and storativity as composite behavior caused by fluid phase changes around the well. This approach will
provide the utilization of oil well test analysis equations and units as the gas formation volume factor (Bg) remains in
bbl/Mscf. It is expressed as:

⎛ μz ⎞ P (3)
P
m ( p ) = ⎜⎜ ⎟⎟ ∫ μ dp
z
n

⎝ p ⎠P P
ref
ref ( p) ( p)

For all selected flow periods the pressure and derivative response was leveled at the same radial flow stabilization on the log-
log graph of Δ m n ( P ) vs Δ t .
q

Well-2
This is a vertical well located on the upper right flank of the reservoir. It served as an offset well to monitor the impact of the
gas condensate effects in the reservoir. It has been drilled after many years of high production. This well is completely
perforated along a net-pay thickness of 67 feet with a reported average porosity of 11%. This well has been fractured before
this test took place as it penetrates a tight gas reservoir.

Model Identification
This is a relatively short production test that lasted 210 hours. The last flow period (FP), which is denoted as FP5 in Figure
19, is used as it lasts longer when analyzed on the rate normalized pressure versus time plot. It is clearly marked that the
entire test is below the dew point pressure, and the condensate banking effect was expected but it is not observed in Figure 20
as the fracture job preceded this test:
8 SPE 122594

• Near wellbore Effects: High conductivity vertical fracture with wellbore storage and no skin.
• Reservoir Behaviour: Homogenous.
• Boundary Effects: Infinite Lateral Extent.

Figure 19: Available production test for Well-2 Figure 20: Well-2 interpretation model match on flow period
(FP5)

Well test analysis interpretation results

The model selected produced the best interpretation match that was able to obtain an excellent agreement on the log-log plot,
available pressure rate history and Horner plot displayed in Figure 20. This shows that the condensate banking effect can be
removed completely with an excellent fracturing job in place, even though the well is depleted below the reservoir’s dew
point pressure. The Skin value is -6.00 showed a good implementation of a fracture on the derivative response. Wells with
low skin values of this range are due to operational treatment like fracturing or acidized jobs. The results of the interpretation
model match are listed in Table 5.

Parameter Unit Value

(pav)i psia 5656

pwf psia 3050
kh mD.ft 56
k mD 0.84
C Bbl/psi 1.7
Skin (t) - -6
r ft -
xf ft 320
ri ft 1590
Table 5: Well-2 analysis interpretation model results

The dual porosity behavior is not noticed or started to develop even after the fracturing job is completed on the well. This
could be due to the condensate banking effect that removed the dual porosity behavior as the entire reservoir region of this
well is below the dew point pressure. The test for this well is not long enough to observe clear reservoir behavior other than
the start of radial flow stabilization.

Well-A
This is a vertical well that is located at the crest of the reservoir. This well has been an active producing well for many years.
The well is fully penetrating across a net pay thickness of 30 feet with the reported average porosity of 15% from the open
hole logs. This well is surrounded by many other producing wells.

Model identification
For this well more than 10000 hours of pressure and rate history is available with two production tests highlighted on Figure
21. Both tests fall below the dew point pressure. A hydraulic fracture was conducted before the first production test and the
second test was carried out after a year and half of stable production at 12 MMscf/D. The rate validation shows that the flow
periods are coherent in Figure 22.
SPE 122594 9

During the first production test, FP16 was selected for analysis. This first test has been conducted some time after a hydraulic
fracture job which might have removed the impact of the condensate banking. In this first test the well has produced below
the dew point pressure and the start of condensate banking effect was visible with no dual porosity reservoir behavior is not
noticed. The best model that was selected for this flow period was:

Figure 22: Normalized pseudo-pressure and derivative for

Figure 21: Pressure and rate history of Well-A Well-A flow periods

• Near wellbore Effects: High conductivity vertical fracture with wellbore storage and skin.
• Reservoir Behavior: Homogenous.
• Boundary Effects: Infinite Lateral Extent.

Excellent matching agreement on the log-log, pressure-rate history, and Horner plots in Figure 23.

Figure 23:Well-A (FP16) interpretation model match Figure 24: Well-A (FP33) interpretation model match

The second production test was carried out after a year and half of production, where in FP33 a significant change in
reservoir behavior is observed in Figure 22 as the condensate banking starts to appear again as a rising skin effect and it
masks the dual porosity behavior. FP33 was not long enough to observe further changes in the reservoir behavior and late
time boundary effects. It is important to note that for this production test the slope of the derivative plot was significantly
higher than the typical ½ unit slope for a highly fractured model. Using the fractured model did not produce consistent results
with FP16. For the known conditions of this well and history the best interpretation model can be summarized as follows:

• Near wellbore Effects: Wellbore Storage and Skin.

• Reservoir Behavior: Composite behavior.
• Boundary Effects: Infinite Lateral Extent.

The well and reservoir parameters for this well is listed in Table 6. The dual porosity behavior was not seen with the
composite behavior impact in the analysis of FP33. This model shows the expected results that condensate banking have
when the wellbore pressure falls below the dew point pressure over a long period of time. The model matched very well with
the log-log, pressure-rate history test and Horner plot in Figure 24.
10 SPE 122594

Well test analysis interpretation results

The effect of gas condensate banking is very clear. Comparing the derivative shapes of FP16 and FP33 we notice that the
humping skin shape increases with constant production below the dew point pressure and the gas production rate before FP33
was increased to 16MMscf/D. The storativity ratio is estimated around 1.7 as the compressibility of the two-phase inner
region is higher than the single phase outer region and the condensate bank reaches 245 feet into the reservoir. It was not
possible to view any variations on ω and λ for this well since no dual porosity behavior is observed.
Production test Production test
Parameter Unit
# 1 (FP16) # 2 (FP 33)

(pav)i psia 7620.00 7620.00

pwf psia 2881.5 3196.1

kh mD.ft 57 57
k mD 1.9 1.9
C Bbl/psi 0.6 0.5
Skin (t) - -5 -2.4
r ft - 245
(φ ch)1/2
- - 1.7
(kh/μ)1/2 - - 0.05
Omega (ω) - - -
Lambda (λ) - - -

xf ft 77 -

ri ft 1154 1192
Table 6: Well Test analysis results for Well-A

The gas condensate bank is noticed in the skin effect where in FP33 the skin increased from -5 to -2.4. Even though the value
is negative it is not a typical value of a non damaged well in dual porosity reservoir as indicated by Gringarten (1984). This
damage is an indication of the condensate bank increase with production below Pdew.

Well-K
This well is another vertical gas condensate production well. The well is perforated through the entire net pay thickness of 66
feet. The well is producing from a single layer without any indication of cross flow from other layers. The open hole logs
showed a good reservoir porosity development of 15%. The well is located in the crest of the reservoir with many
surrounding producing wells around. Figure 25 shows the entire history available as there production tests are considered for
this well:

• 1st production test: Pre. Fracture test (FP4).

• 2nd production test: Post Fracture test (FP14).
• 3rd production test: After 1 year production test (FP22).

Figure 25 displays that this is a depleting reservoir and all the tests produced below dew point pressure as the build ups
returned above dew point pressure. The well was not produced between the first and second production test. The well
produced at a stable rate for about one year at an average gas rate of 30 MMscf/D and an average condensate rate of 720
STBD. The pressure change in this sector of the reservoir can be seen in Figure 26 as the last flow period is not coherent with
the other two tests conducted before production. This could be due to the continuous changes in the fluid properties and phase
composition around the well bore after constant high production rates in comparison with the first two tests. We shall
interpret all the three production tests separately to be able to ascertain the selected interpretation model accurately fits the
production test history, log-log and Horner plots.

Attempts to analyze the entire data as a long single history model did not produce a good match across the entire history. Not
even numerical analysis to uncover any geological structures that cause higher pressure depletion was able to give a better
match. So, the disturbance on the late time effects is due to high production in the region from surrounding wells.
SPE 122594 11

Model Identification
The first production test was conducted 1 month before the well was fractured, as the well showed a damaged wellbore
condition, probably due to the condensate banking effect around the well. The second production test has been carried out
after an acid fracturing treatment to assess the clear productivity of the well before production. The third production test takes
place one year after the production. The depletion in the reservoir is clearly obvious before and after production and the skin
effect expected to rise as the condensate saturation increases.

Figure 25: Well-K pressure and rate history

Figure 26: Well-K superposition plot shows strong pressure Figure 27: Well-K normalized pseudo-pressure and derivative
depletion with time

FP4 is conducted to obtain important well and reservoir conductions around the well to identify if it requires any maintenance
prior to full production. Part of the build-up fell below the reported dew point but no composite radial reservoir behavior is
witnessed as the existence of the condensate bank is noticed in the skin value.

The best interpretation model that can fit the 1st production test is:
• Near wellbore Effects: Wellbore Storage and Skin.
• Reservoir Behavior: Dual porosity, restricted interporosity flow.
• Boundary Effects: Infinite Lateral Extent.

In FP14, an acid fracturing treatment was done on the well to treat the damage around the wellbore. The fracturing job
showed a strong improvement in the wells performance as the skin effect is completely removed but the test was short to see
the radial flow stabilization of the total system flow.
12 SPE 122594

The best interpretation model selected to match the 2nd production test:
• Near wellbore Effects: Infinite conductivity fracture with wellbore storage and skin.
• Reservoir Behavior: Homogenous reservoir behavior.
• Boundary: Effects: Infinite Lateral Extent.

In FP22, the well has produced for about a year and the drawdown of this test is below dew point pressure, and the composite
behavior is clear now after the increase in condensate saturation levels around the wellbore. This leads to an increase in the
skin effect. This test was not long enough to see the radial flow stabilization.

The best interpretation model selected to match the condensate bank development is:
• Near wellbore Effects: Infinite conductivity fracture with wellbore storage and skin.
• Reservoir Behavior: Composite behavior.
• Boundary Effects: Infinite Lateral Extent.

Table 7 contains a summary of all the important results of all the three production tests.

Well test analysis interpretation results

The interpretation model matches are displayed in Figure 28, Figure 29, and Figure 30. The model match on FP4 provided
important parameters to understand the expected well and a reservoir conditions as Kf is 2.5 mD and the skin of -1.7. Clear
dual porosity behavior is observed on the pressure derivative plot. Even though the pressure around the wellbore falls below
dew point the dual porosity reservoir behavior is apparent. FP14 showed excellent fracturing as it removed the condensate
plugging in the fracture network around the well with a noticeable skin enhancement of -5 and the wellbore storage constant
raise to 0.05 as the volume of the fluid flowing into the well rises. Gringarten (1984) showed that this increase is a
characteristic change in fractured formations after remedial operations. In addition, the highly induced fracture has increased
the values of ω and λ, showing that connectivity between the fractures and the matrix blocks has improved. In effect, the gas
flow influx from the matrix blocks into the fractures has accelerated.

Production Production Production

Parameter Unit test # 1 test # 2 test # 3
(FP4) (FP14) (FP22)
(pav)i psia 7350 7350 7238

pwf psia 3350 5044 4638

kh mD.ft 165 165 160
k mD 2.5 2.5 2.5
C Bbl/psi 0.0206 0.0579 0.0127
Skin (t) - -1.7 -5 -4.4
r ft - - 108
(φ ch)1/2
- - - 2.3
(kh/μ)1/2 - - - 0.66
Omega (ω) - 9.0E-7 - -
Lambda (λ) - 7.6E-5 - -

xf ft - 103 86

ri ft 500 548 532

Table 7:Well Test analysis results for Well-K Figure 28: Well-K (FP4) interpretation model results
SPE 122594 13

Figure 29:Well-K (FP14) interpretation model results Figure 30: Well-K (FP22) interpretation model result

After 1 year of production, FP22 in the last production test showed stronger depletion in the pressure because of many
surrounding active producers. The observed reduction in reservoir pressure is 112 psi from the initial value. This is a result of
many surrounding production wells. It is important to note that the composite behavior is represented as a skin effect increase
where the skin raises from -5 to -4.4. This is observed as a humping skin effect before the final radial flow stabilization. It is
clear that this well begins to suffer from condensate saturation as the bank radius progressed more than 100 feet into the
reservoir. There is no indication of early stabilization before the humping skin effect that could indicate a velocity stripping
zone exists in the innermost region of the well.

The simulation patterns provides a strong agreement with the well test analysis after years of production, seen in Well-A
FP33 (after 1.5 years of production), and Well-K FP22 (after 1 year of high production). Moreover, the post fracture tests
conducted in the study for Well-2 FP5, Well-A FP16 and Well-K FP22 show the same pattern to the simulation results.

Deconvolution
The conventional well test analysis was very useful to observe the near wellbore effects and early time reservoir behavior
flow regimes but most of the build-up flow periods selected for interpretation and model matching were part of short
production tests that did not uncover the stable reservoir behavior.

The use of deconvolution becomes necessary to ensure that no other reservoir behavior is appearing or conflicting with the
analysis done earlier. Deconvolution will be conducted on each well where the longest build up flow period is selected with
the entire available production rate. Today there are various algorithms and methods of deconvolution where single or
multiple flow periods can be used together. For this study we focus our efforts on the Deconvolution algorithm of von
Schroeter et al. (2004).

Well-K
It was only possible to utilize deconvolution on Well-K to gain a better understanding of reservoir behavior. The other wells
did not produce more log cycles with Deconvolution. In FP14, the analytical well test analysis shows that the high
conductivity fracture has completely removed the skin effect. The use of deconvolution for this production test extended the
plot by another log cycle and showed that we have constant radial flow behavior and ensure no other flow behavior is present
at middle time Figure 31.

Performing Deconvolution on FP22, showed on the rate normalized log-log plot increased by two additional log cycles and a
clear effect of the condensate banking effect is observed followed by the radial flow stabilization. In Figure 32, the
Deconvolution result shows stabilization radial flow behavior.

For well-K Deconvolution assisted in confirming the flow behavior by the analytical analysis.
14 SPE 122594

Figure 31: Well-K FP14 deconvolution derivative result Figure 32: Well-K FP22 deconvolution derivative result

Conclusions
The study has looked at various vertical wells in a gas condensate banking effects under different operational conditions with
conventional well test analysis and deconvolution. The compositional simulation has confirmed the findings from the wells
test analysis. The main conclusions to highlight from this study are:

1. As the production falls below the dew point pressure the dual porosity effect dose not disappear immediately.
2. In the radial model compositional simulation, when the production condition is moving below the dew point
pressure the values of λ decrease as the condensate saturation in the fractures increase with time. ω is fairly
constant due to condensate saturation that filled the fractures. The shift in the value of λ and ω is in agreement
with the behavior trends discovered by Al-Malki (2004).
3. The condensate banking impact appears as a rising skin that masks the dual porosity behavior. This causes
fracture networks and matrix blocks around the wellbore become more filled with the condensate drop-out.
4. The capillary number did not show any effect on the saturation profile in the matrix blocks or the fractures when
applying the North Sea lean gas condensate parameters from Bozorgzadeh (2006).
5. The high conductivity fracture can treat the gas condensate banking impact below the dew point pressure but
lengthy condensate production can reduce the effective fracture and further remedial work needs to be carried
out.
6. The use of deconvolution was useful quality check tool for the well test analysis conducted on Well-k and
showed that condensate banking is a rising skin effect in a dual porosity reservoir.

Recommendations for Future works

1. It is difficult to find well test data that can fit the purpose of this study. The main recommendation is to design a
pressure build up test campaign on new wells in different regions of the reservoir and study the well and
reservoir behavior through the production history with pressure build up tests. This will allow a clearer
understanding on the scale of a well before any fracturing or treatment.
2. Obtain different test and lab data from different naturally fractured gas condensate reservoirs to be able to
confirm that the results stand for more than Middle East environments. The field test data should include the
DST report and long production tests to be able to study the impacts of gas condensate in these dual porosity
environments.
3. Overcome the rapid pressure increase in the single well model by expanding the compositional simulation either
to a reservoir sector or field model to mimic a field representative pressure profile with gas and condensate
production below dew point pressure.

Nomenclature
Ct total compressibility, psi-1
ke effective permeability, mD
kf fracture permeability, mD
km matrix permeability, mD
SPE 122594 15

Greek
νg gas velocity, fts-1
α geometric parameter for fracture network, ft-2
λ interporosity flow coefficient, dimensionless
Φ porosity
ω storativity ratio, dimensionless
μ viscosity

Abbreviations
bbl barrel unit
BU build-up
CCE Constant Composition Expansion
CVD Constant Volume Depletion
DD drawdown
EOS Equation of State
F Fahrenheit
FP Flow Period
ft foot
Kh Permeability Thickness
MMscf/D Million standard cubic feet pre Day
NFGC Naturally Fractured Gas Condensate
P pressure
Psia pound square inch – Absolute
Psig pound square inch – Gauge
PR-3P Pang-Robinson three-parameter
PVT Pressure/Volume/Temperature
Mscf Thousand standard cubic feet
mD Mili Darcy
STB Stock Tank Barrel

Subscripts
Dew At dew point
f fracture/fissure
g gas
i initial condition
m matrix
o oil
r relative
t total
w water

References
1. Al-Baqawi, A. M.: “Well Test Analysis In Naturally Fractured Gas Condensate Reservoirs Below Dew Point Pressure” MSc Thesis,
Centre for Petroleum Studies, Imperial College London, UK, 2008
2. Al-Malki, B.: “Well Test Analysis in A Dual Porosity Gas-Condensate reservoir,” MSc Thesis, Centre for Petroleum Studies, Imperial
College London, UK, 2004
3. Barenblatt, G. E., Zheltov, I. P. and Kochina, I. N.: “Basic Concepts in the Theory of Homogeneous Liquids in Fisured Rocks,” J.
Appl. Math. Mech. (USSR, 1960), 24 (5), 1286-1303
4. Bourdet, D.: Well Test Analysis: The use of Advanced Interpretation Models, Handbook of Petroleum Exploration on Production
Series, 3, first edition, Elsevier Science B. V., Amsterdam, The Netherlands (2002)
5. Bourdet, D. Ayoub, J. A. and Pirard, Y. M., 1989. Use of Pressure Derivative in Well-Test Interpretation. SPEFE, June, 293-302.
6. Bourdet D. P. and Gringarten, A. C.: "Determination of Fissure Volume and Block Size in Fractured Reservoirs by Type Curve
Analysis," paper SPE 9293, presented at the 55th Annual Technical Conference and Exhibition of SPE, Dallas, Tex., Sept. 21-24,
1980. 37.
7. Bozorgzadeh, M.: “Characterization and Determination of Gas Condensate Dynamics from Pressure Transient Data and Fluid PVT
Properties,” PhD dissertation, Centre for Petroleum Studies, Imperial College London, UK, March 2006
8. Bozorgzadeh, M. and Gringarten, A. C.: “New Estimate for the Radius of a Condensate Bank from Well Test Data Using Dry Gas
Pseudo-Pressure,” paper SPE 89904, presented at the 2004 SPE Annual Technical Conference and Exhibition, Houston, Texas,
September 26-29
9. Cinco-Ley, H., Samaniego-V., F. and Kuchuk, F.: “The Pressure Transient Behavior for Naturally Fractured Reservoirs with Multiple
Block Size,” paper SPE 14168, presented at the 1985 60th Annual Fall Meeting, Las Vegas, Nevada, September 22-26
10. Gilman, J. R., 2003: “Practical Aspects of Simulation of Fractured Reservoirs”, Stanford-Leoben International Forum on
Reservoir Simulation, January 23-27.
11. Gringarten, A. C.: “Interpretation of Tests in Fissured and Multilayered Reservoirs with Double-Porosity behavior: Theory and
Practice” JPT (April 1984) 549-564. SPE 10044-PA.
12. Gringarten, A. C.: “How To Recognize “Double Porosity” Systems From Well Tests,” JPT (June 1987) 631-633. SPE
16437.Gringarten, A. C., Bozorgzadeh, M., Daungkaew, S. and Hashemi, A.: “Well Test Analysis in Lean Gas Condensate
Reservoirs: Theory and Practice,” paper SPE 100993, presented at the 2006 SPE Russian Oil and Gas Technical Conference and
Exhibition held in Moscow, Russia, October 3-6
16 SPE 122594

13. Gringarten, A. C.: “MSc Petroleum Engineering Notes on Well Test Analysis and Flow in Porous Media,” Centre for Petroleum
Studies, Imperial College London, UK, 2007-2008
14. Gringarten, A. C., Al-Lamki, A., Daungkaew, S., Mott, R. and Whittle, T. M.: “Well Test Analysis in Gas-Condensate Reservoirs,”
paper SPE 62920, presented at the 2000 SPE Annual Technical Conference and Exhibition, Dallas, October 1-4
15. Henderson, G. D., Danesh, A., Peden, J. M., Tehrani, D. H. and Al-Shaidi, S.: “Measurement and Correlation of Gas Condensate
Relative Permeability by the Steady-State Method,” SPEJ (June 1996), 191-202
16. Kazemi, H.: “Pressure Transient Analysis of Naturally Fractured Reservoirs with Uniform Fracture Distribution,” SPEJ (December,
1969A), vol. 9, no. 4, pp. 451-462
17. Liu, J. S., Wilkins, J. R., Al-Qahtani, M. Y. and Al-Awami, A. A.: “Modelling a Rich Gas Condensate Reservoir With Composition
Grading and Faults,” paper SPE 68178, presented at the 2001 SPE Middle East Oil Show, Bahrain, March 17-20
18. McCain, W.D.J.: "Properties of Petroleum Fluids", Penwell Books, Penwell Publishing Co, 1990.
19. Maloney and Doggett, 1977:”Multiphase Flow in Fractures”, DOE Contract DE-AC22-94PC91008, Report SCA9730.
20. Moench, A. F.: “Double-Porosity Models for a Fissured Groundwater Reservoir With Fracture Skin,” Water Resources Res. (July
1984), vol. 20, no. 7, 831-846
21. Najurieta, H.L., 1980. A Theory for Pressure Transient Analysis in Naturally Fractured Reservoirs. J. Pet. Tech., July, 1241
22. Odeh, A.S., 1965. Unsteady-State Behavior of Naturally Fractured Reservoirs. SPEJ, March, 60-64; Trans. AIME, 234.
23. Poon, D. C. C.: “Pressure Transient Analysis of a Composite Reservoir With Uniform Fracture Distribution,” paper SPE 13384, 1984.
24. Shittu.: “Well Test Analysis in Naturally Fractured Gas/Condensate reservoirs” MSc Thesis, Centre for Petroleum Studies, Imperial
College London, UK, 2007
25. Romm, E.: "Filtrasionnie Svoistsa Teschinovatich Porod (Flow Phenomena in Fractured Rocks)" (in Russian), Nedra,
Moscow, (1966). (English translation, W. R. Blake, Bartlesville, OK, 1972)
26. Schlumberger: Eclipse 300 Simulation Software Manuals 2007.1, 2007
27. Tiab, D. and Donaldson, E.: "Petrophysics: Theory and Practice of Measuring Reservoir Rock and Fluid Transport properties”
second edition, Elsevier Books, Gulf Professional Publishing Co (2004) 489
28. von Schroeter, T., Hollaender, F. and Gringarten, A. C.: “Deconvolution of Well Test Data as a Non-linear Total Least Square
Problem, SPEJ (December 2004), vol. 9, no. 4, pp. 375-390, paper SPE 77688-PA
29. Warren, J. E. and Root, P. J. :”The Behavior of Naturally Fractured reservoirs,” paper SPE 426, presented at the 1962 Fall Meeting of
the Society of Petroleum Engineers in Los Angeles, October 7-10