You are on page 1of 17

SPE 100993

Well Test Analysis in Lean Gas Condensate Reservoirs: Theory and Practice
A.C. Gringarten, M. Bozorgzadeh, S. Daungkaew,Ð and A. Hashemi,ÐÐ SPE, Imperial College, London

Copyright 2006, Society of Petroleum Engineers


verify the fluid behavior. It is shown that, in addition to the
This paper was prepared for presentation at the 2006 SPE Russian Oil and Gas Technical usual well test analysis results, it is possible to obtain
Conference and Exhibition held in Moscow, Russia, 3–6 October 2006.
parameters required for reservoir simulation and well
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
productivity forecasting, such as gas relative permeabilities at
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to the end point, critical oil saturation, and the base capillary
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 number.
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 Introduction
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.
Gas condensate reservoirs are becoming more common as
deeper depths are being targeted in the exploration for oil and
gas. The behaviors of such systems are complex and are still
Abstract not fully understood, especially in the near-wellbore region.
Gas condensate reservoirs exhibit a complex behavior when Well tests, in particular, are difficult to interpret.
wells are produced below the dew point, due to the existence A discussion of the state-of-the-art in gas condensate well
of a two-fluid system, reservoir gas and liquid condensate. test interpretation was published in 2000 by Gringarten et al1.
Different mobility zones develop around the wellbore with an extensive review of the related literature. To
corresponding respectively to the original gas in place (away summarize, a characteristic of gas condensate production is
from the well), the condensate drop-out, and capillarity the creation of a condensate bank when the bottomhole
number effects (close to the well). Condensate drop-out causes pressure drops below the dew point2 pressure. This reduces
a non-reversible reduction in well productivity, which is the gas relative permeability3 around the well and leads to a
compensated in part by capillarity number effects. loss of well productivity4-7, with some wells even ceasing
All these effects can be identified and quantified from well production completely due to condensate loading in the
test data. Tests in condensate reservoirs, however, tend to be wellbore5. This “condensate banking” effect, however, is
difficult to interpret. Build-up and/or drawdown data are compensated by “velocity stripping” which increases the gas
usually dominated by wellbore phase redistribution effects and mobility in the immediate vicinity of the wellbore8. “Velocity”
the main analysis challenge is to distinguish between reservoir or “viscous” stripping (also called “positive coupling”)9-13
effects, boundary effects, fluid behavior and wellbore phase occurs at high capillarity number, a dimensionless parameter
redistribution perturbations. that represents a ratio of viscous to capillary forces14,15:
The paper compares theoretical well test behaviors in vertical vμ
Nc = (1)
and horizontal wells as obtained from compositional σ
simulation with actual behaviors selected from more than
twenty different gas condensate reservoirs. An interpretation In Eq.1, ν represents the velocity, μ the viscosity, and σ the
methodology is described, which uses time-lapse analyses, interfacial tension (IFT). High capillary numbers are obtained
deconvolution and different analytical and numerical tools to for high flowrate8 or low interfacial tension16.
identify the probable causes of the pressure data behavior: Consequently, four regions develop around the wellbore
two-region and three-region analytical composite models to with different liquid saturations7,8,17,18. Away from the well, an
represent the various mobility zones around the wellbore; a outer region, still above the dew point pressure, contains gas
voronoi-grid numerical simulator to represent discontinuous with the initial liquid saturation; next, there is an intermediate
boundaries; a multilayered analytical simulator to account for region with a rapid increase in liquid saturation and a
the geological description and a compositional simulator to corresponding decrease in gas relative permeability. Liquid in
that region is immobile. Closer to the well, a region forms
where the liquid saturation reaches a critical value2, and the
Ð
Now with SCHLUMBERGER Data & Consulting Services effluent travels as a two-phase fluid with constant composition
ÐÐ
Now with NIOC (the condensate deposited as pressure decreases is equal to that
2 SPE 100993

flown towards the well). The fourth region is in the immediate In this paper, we use a normalized single phase pseudo-
vicinity of the well and is characterized by a decrease of the pressure23:
p
liquid saturation and an increase in gas relative ⎛ μZ ⎞ p

mn ( p) = ⎜⎜ ⎟⎟ dp (3)
permeability4,19 at low interfacial tensions or high rates (Fig.
⎝ p ⎠ p p μ p Z ( p)
( )
1). ref ref

instead of pseudo-pressure, because of its convenience: units


Pressure < pdew P > pdew are the same as for pressure, and equations for well test
analysis are the same as for oil, with the gas formation volume
Condensate bank region
factor in Bbl/Mscf.
0.1
Condensate saturation (So)

Mobile oil zone


The Gringarten et al.’s paper1 provided the first well test
Immobile
oil zone
0.08 Velocity stripping
zone
Original So evidence in the literature of the existence of the velocity
stripping zone. Previous well test publications had only
0.06 reported the existence of a condensate bank7,24-27 (zones 2 to 3
in Fig. 1) as a two-region radial composite behavior (curve b
0.04
1 2 4
in Fig. 2).
3

0.02
Critical So start of The objective of this paper is to discuss the impact of the
two-phase
zone condensate bank on well test data, to document further
0 examples of complex well test behaviors from a number of
0.1 1 10 100 1000 lean gas condensate sandstone reservoirs and to describe an
Radial distance from the well (ft) interpretation methodology to identify the probable causes of
these behaviors. This methodology, which combines time-
lapse analyses, deconvolution and the use of different
Fig. 1: Condensate saturation profile with condensate drop-out analytical and numerical tools, provides parameters required
and velocity stripping
for reservoir simulation and well productivity forecasting in
addition to the usual well test analysis results.
The existence of capillary number effects has been inferred
from numerical studies8, field production data8,9,20 and
laboratory experiments16,21. Gringarten et al.1 showed that the Dynamic Behavior of the Condensate Bank
corresponding near-wellbore region, the second and third As in all complex well test situations, it is helpful to use
regions, and the fourth region appeared as three different simulation to identify key features that can then be recognized
mobility zones, exhibiting a three-region radial composite on actual data. The following summarizes the main findings
behavior in a well test (curve a in Fig. 2), when the analysis on the dynamic behavior of the condensate bank from
was made in terms of a single-phase or dry-gas pseudo- numerical compositional simulation12.
pressure function, also called real gas potential22:
p
p Drawdown versus build up condensate saturation
m( p ) = 2 ∫ dp (2)
p
μ ( p ) Z ( p ) ref
distribution
Fig. 3 presents a pressure and rate history for a lean gas
where Z is the real gas compressibility factor. condensate well in a sandstone reservoir of infinite extent with
homogeneous behavior, and the corresponding simulated
103 condensate saturation distributions in the reservoir for both
drawdowns and build up’s. The first drawdown, Dd1, is below
mn(p) change and derivative (psi)

the dew point pressure and has a duration which is long


102
H ig h c o n d e n s a te enough so that a condensate bank has time to fully develop
s a tu r a tio n
(b ) (liq u id d r o p -o u t) (zones 1 to 3 in Fig. 1). It is followed by a build up where the
(a ) final pressure is above the dew point pressure. The next
G a s w it h lo w e r drawdown, Dd2, has a much lower rate and the pressure
10 c o n d e n s a te s a tu r a tio n
( C a p illa r y n u m b e r e ff e c t) remains above the dew point pressure. Subsequent drawdowns
G a s w ith in itia l
C o n d e n s a te
are at increasing rates and below the dew point pressure,
s a tu r a tio n whereas all build up’s end above.
1
1 0 -3 1 0 -2 1 0 -1 1 10 102 103
Fig. 3 shows that the saturation distributions in drawdowns
E la p s e d tim e ( h o u rs ) and build up’s are only slightly different. This is because,
Fig. 2: Schematic of pressure and derivative composite during a shut-in, the high accumulation of condensate mass
behaviors: (a) three-region composite; (b) two-region composite
1 near the well prevents the re-vaporization of the condensate
that should occur due to the increase in pressure (hysteresis
effect)19. The condensate saturation may even be higher than
SPE 100993 3

during the preceding drawdown because excessive while the pseudo-pressure shows a skin effect that corresponds
accumulation of condensate and migration of the heavy to the wellbore skin. Dd3 and Dd4 are the first drawdowns
components towards the well changes the fluid composition where the pressure drops below the dew point pressure. The
and makes the fluid to behave as a black oil28 (the gas may derivatives in the following build up’s, Bu3 and Bu4
dissolve into the oil, thus increasing the condensate saturation respectively, still exhibit a homogeneous behavior, but the
in the vicinity of the well). Some re-vaporization does occur, corresponding pressures display higher total skins, which
but only near the outer edge of the two-phase region, and only include the wellbore skin and the effect of the condensate drop
if the preceding drawdown has a high production rate. The out. This is because the condensate saturation around the well
leaner the gas and the lower the production time, production has not yet reached the critical saturation and only zones 3 and
rate and critical saturation, the smaller the saturation profile 4 (Fig. 1) exist. Only the gas phase is produced and the
difference between a drawdown and the subsequent build-up29. immobile condensate appears as an additional skin effect.

4000
4000 Dew
3443 Point

Dd1
Dew point pressure

Pressure (psia)

Dd2

Dd12
3000 Pressure

Dd3

Dd4

Dd14
3000
Pressure (psia)

2000

Dd10

Dd11
Bu10
Bu11
Bu12
40

Bu4
Bu3

Gas Rate (MMscf/D)


Bu1
Flow rate (MMscf/D)
18 1000
2000 Dd1 Dd5
30
14
0
Dd4 20
10
1000
Dd3 6 10
Dd2
Bu4
Bu2

Bu5
Bu3
Bu1

2 0
0 0 0 100 200 300
0 50 100 150 200 250 300
Time (days)
Time
Rate normalized Δmn(p) and Derivative (psi)

(Days) 10 4

0.25
Dd1 15MMscf/D for 200 days
Dd5 15MMscf/D for 10 days
0.20
10 3
Dd4 10MMscf/D for 10 days Bu10
Bu12 27.5 MMscf/D
5 MMscf/D
0.15 End of Dd1 (200 days) Condensate bank
End of Bu1 (210 days) Bu11 30 MMscf/D
So

End of Dd2 (220 days)


10 2
End of Bu2 (230 days) Reservoir effective permeability
0.10 Dd2 End of Dd3 (240 days) Bu3
2.5MMscf/D for 10 days
End of Bu3 (250 days) 7.5 MMscf/D
End of Dd4 (260 days) Bu1
0.05
Bu4
End of Bu4 (270 days) 2.5 MMscf/D 10 MMscf/D
Dd3 2.5MMscf/D for 10 days End of Dd5 (280 days)
10
10 -5 10 -4 10 -3 10 -2 10 -1 1 10 10 2 10 3
End of Bu5 (290 days)
Elapsed time, hrs
0.00
10-1 1 10 102 103 104
Fig. 4: Lean gas condensate well test behavior with increasing
30
and decreasing flow rates
Radial distance from well (ft)
Fig. 3: Lean gas condensate saturation distribution with As production time increases, the condensate saturation
29
increasing and decreasing flow rates
reaches a critical value and the condensate becomes mobile.
Homogeneous versus composite fluid behavior Zones 1 and 2 (Fig. 1) develop while the saturation at the
Fig. 4 shows another pressure and rate history for a similar wellbore increases further (up to a maximum value). This
well-reservoir configuration, and a rate-normalized log-log yields a composite behavior and a further increase in total skin
plot of the simulated normalized pseudo-pressure and (Bu10 and Bu11).
derivative curves of some of the build up’s. The pressure in
the first two drawdowns, Dd1 and Dd2, remains above the Initially, capillary number effects create a three-region
dew point pressure, whereas the pressure in all subsequent composite well test behavior, with three derivative
drawdowns, from Dd5 to Dd14, is mainly below. As there is stabilizations (Fig. 5). The first derivative radial flow
no depletion in the reservoir, the pressure at the end of all the stabilization line develops as soon as oil becomes mobile in
build up’s is above the dew point pressure. The log-log the reservoir because the gas effective permeability in the
derivative for the build up following the first drawdown, Bu1, near-wellbore region is greater than that in the condensate
exhibits a homogeneous behavior, as expected from a dry gas, bank due to capillary number effects. As production continues,
4 SPE 100993

the oil saturation increases. Consequently, the gas effective Condensate re-vaporization with decreasing gas rate
permeability decrease in the two-phase zone. This Figs. 3 and 4 also show the effects of decreasing gas flow
permeability reduces faster in the velocity stripping zone than rates. In Fig. 3, lowering the rate maintains the pressure above
elsewhere in the reservoir. Eventually, the first stabilization the dew point pressure in Dd2 and causes the condensate
line disappears when the near-wellbore gas effective saturation in the near-wellbore region to decrease, as part of
permeability becomes less than that in the mobile zone, in the condensate accumulated near the wellbore is recovered. As
which case a two-zone radial composite behavior develops, soon as the bottomhole pressure drops again below the dew
with only two stabilizations on the derivative (after 50 days point pressure in a subsequent flow period (Dd4), oil
production in Fig. 5)30. saturation increases near the wellbore due to renewed
condensate deposit. The bank radius (i.e. the start of the two-
Condensate Saturation (So)

0.12 50 days phase zone) remains approximately constant in this particular


0.10
45 days example. Fig. 4 shows the corresponding impact on the
20 days pressure derivative: as the flow period Dd12 has a much lower
0.08
5 days rate than Dd11, the condensate saturation decreases around the
0.06
wellbore, and consequently, the following build up, Bu12, has
a lower condensate stabilization level and a lower total skin
0.04 effect. In a multirate exploration test, where the pressure
during successive drawdowns may fall below or remain above
0.02
the dew point pressure, depending on the flow rate, the skin
0 due to the condensate may appear and disappear accordingly.
0.1 1 10 100

Radial distance from well (ft)


Well test analysis with a composite model
10 3
mn(p) Change and Derivative (psi)

50 days Field examples of condensate banking with capillary number


45 days
20 days
effects are shown in Fig. 6, where the normalized pseudo-
5 days pressures from Eq. 3 and derivatives are plotted versus elapsed
time on a log-log graph. For demonstration purpose, data have
been shifted vertically so that the derivatives coincide in the
10 2
2nd condensate bank area.
stabilisation
Rate normalized Δmn(p) and derivative (psi)

1st
3rd stabilisation
stabilisation
G erm any
10
10-5 10-4 10-3 10-2 10-1 1 10 102 103 High condensate
saturation
Elapsed time (hrs) (liquid drop-out)

Fig. 5: Impact of capillary number effects on saturation profiles Gas with lower
and log-log derivative shapes (10 day build-ups following Croatia condensate saturation North Sea
30 (Capillary num ber effect)
drawdowns with various production durations at 15MMscf/D) Algeria
Gas with initial
The two-region composite behavior yields two stabilizations Condensate
saturation
on the derivative. The last one corresponds to the gas mobility
in the portion of the reservoir still above the dew point, i.e. it Elapsed tim e (hours)
represents the reservoir effective permeability. The first one, Fig. 6: Examples of well test derivatives from different lean gas
at a higher level, corresponds to the lower gas relative condensate reservoirs showing condensate drop-out and velocity
mobility in the condensate bank. The level of this first stripping
stabilization depends on the condensate saturation in Zone 2
The single-phase pseudo-pressure is normally used for well
(Fig. 1) and therefore increases as the condensate saturation
test analysis in dry gas reservoirs, in an attempt to linearize the
increases until the condensate saturation reaches its maximum
diffusivity equation. Using it for gas condensate reservoirs
value. This maximum condensate saturation is slightly higher
amounts to considering the gas as the dominant fluid, and the
at high production rates than at low production rates, for the
condensate as a fluid heterogeneity. When the bottomhole
same cumulative production. As production time increases at
pressure drops below the dew point pressure, this creates a
constant rate, the maximum condensate saturation at the
fluid-induced composite behavior, as illustrated in Fig. 2, on
wellbore and the first derivative stabilization level no longer
top of any possible geology-induced heterogeneous behavior
change, but the bank radius increases.
and/or complex well behavior, as in hydraulically fractured or
horizontal wells (Fig. 7).
SPE 100993 5

Fractured well59
0.40
Assuming the durations of the analyzable flow periods are
Increasing production
time below the dew
Radial flow 0.30 long enough so that all the features of the composite behavior
point pressure have had time to develop, interpretation of gas condensate
Rate normalized Δmn(p) and derivative (psi)

at constant rate
0.20
Increasing production pressure transient data that exhibit a two- or a three-region
time below the dew
0.10
point pressure radial composite behavior yields the effective reservoir
½ slope

Condensate Saturation
at constant rate

¼ slope
Above the dew
point pressure
permeability and the total skin effect from the final derivative
10-4 10-3 10-2 10-1 1 10 102 103 104 105
0

10-2 10-1 1 10 102 103 104


radial flow stabilization, and the mobility or permeability
Elapsed time (hours) Distance perpendicular to the fracture (ft) ratios between the various regions from the respective
104

Horizontal well60
0.3
Increasing production derivative radial flow stabilizations. The radius of each
time below the
Increasing production time below the
dew point pressure
mobility zone, on the other hand, cannot be un-coupled from
dew point pressure at increasing rate
103 0.2
at increasing rate
the storativity ratio between that zone and the next because
½ slope
analytical well test analysis methods can only account for one
102
0.1 set of PVT data. An independent estimate of the condensate
Above the dew Radial bank storativity is therefore required to calculate the bank
10
point pressure flow
0
radius, the contribution of the bank to the skin effect, and the
10-1 102 103 104

wellbore skin effect. Bozorgzadeh and Gringarten29 have


10-4 10-3 10-2 10-1 1 10 102 103 104 1 10

Elapsed time (hours) Distance perpendicular to the well (ft)


Fig. 7: Composite behaviors due to condensate banking in a shown that, for the analysis of build up data, the total
31 32
hydraulically fractured well and in a horizontal well obtained compressibility of the two-phase region could be obtained
from compositional simulation from the wellbore flowing pressure at the time of shut-in and
the corresponding saturation profile. They demonstrated on
The diffusivity equation in a gas condensate reservoir can be both computer-generated and real data that the calculation
further linearized33 with a two-phase pseudo-pressure procedure they proposed (Fig. 9) ensures that the condensate
function: bank characteristics obtained from well test analysis were
p consistent with those from compositional simulation.
k k
m2ϕ ( p) = ∫ ( rg + ro )dp (4)
p
μ B
g g μ B
o o ref
1. Construct an equation of state (EOS) model to predict the
This amounts to converting the two-phase fluid into a single actual reservoir fluid properties
fluid equivalent in the two-phase flow regions (Regions 1 to 3
in Fig. 1). As a result, the fluid-induced composite behavior 2. Tune the EOS parameters with reservoir fluid
experimental data, using the critical properties of the pseudo-
obtained with single-phase pseudo-pressures no longer exists components as variables in the tuning process.
and any heterogeneous behavior must be due to the geology,
not to the fluid. In the case of a reservoir with homogeneous 3. Generate live oil and wet gas PVT tables as a function of
behavior and infinite extent, a single derivative stabilization is pressure over the pressure range of the test, using the tuned
EOS and Whitson and Trop’s procedures 52 .
therefore obtained, which yields the absolute permeability.
The skin effect obtained from the corresponding analysis 4. Calculate the total compressibility for the two-phase region
represents the wellbore skin (Fig. 8). Calculation of the at the pressure and time of shut-in, using the PVT table
integral in Eq. 4, however, is challenging, because it requires generated in Step 3:
the knowledge of the gas phase relative permeability as a ⎛ Sg ⎡ − dBg ⎛ dRv ⎛ B0 − Rs Bg ⎞ ⎞ ⎤ ⎞
function of pressure. ⎜ ⎢ + ⎜⎜ ⎜⎜ ⎟⎟ ⎟ ⎥ + ⎟
⎜ Bg ⎟
⎢⎣ dp ⎝ dP ⎝ 1 − Rs Rv ⎠ ⎠ ⎥⎦ ⎟
ctc = (1 − S w ).⎜ ⎟ + S w c w + cr
⎜S ⎡ − dBo ⎛ dRs ⎛ Bg − Rv Bo ⎞ ⎞⎤ ⎟
T w o -p h a s e S in g le -p h a s e ⎜ o ⎢ + ⎜⎜ ⎜⎜ ⎟⎟ ⎟⎥ ⎟
k rg Δmn(p) and derivative (psi)

⎜ Bo ⎟
⎝ dP ⎝ 1 − Rs Rv ⎠ ⎠⎦⎥ ⎟⎠
Δmn,2φ(p) and derivative (psi)

max

⎝ ⎣⎢ dp
Sw 5. Calculate the total compressibility for the outer gas region
(where the reservoir pressure is above the dew point pressure)
at the average reservoir pressure
St
6. Calculate the storativity ratio using the total
Sw compressibilities from Steps 4 and 5:
ct1
G a s w ith in itia l
(φ h ct )1 / 2 =
ct 2
c o n d e n s a te
s a tu ra tio n
Fig. 9: Procedure to calculate the total compressibility in the
E la p s e d tim e (h o u rs ) 29
condensate bank
Fig. 8: Single-phase vs. two-phase pseudo-pressure formulation
(single-phase pseudo-pressures are multiplied by k rg
max
to yield a final Table 1 emphasizes the dependency of the condensate bank
radius on the storativity ratio. Experience indicates that the
derivative stabilization that corresponds to the absolute permeability
instead of the effective permeability) storativity ratios are almost always underestimated by
6 SPE 100993

interpreters, resulting in an overestimation of the condensate experimental data, using single-phase and two-phase pseudo-
bank radius (by up to one order of magnitude). The impact on pressures simultaneously. It is also possible to obtain the base
the wellbore skin, on the other hand, is small. capillary number (i.e., the minimum value required to see
capillary number effects) using single-phase gas pseudo-
(ϕcth)bank/reservoir
Outer zone Total skin Wellbore pressures. The gas relative permeability at near-wellbore
radius (ft) effect skin effect saturation and at the initial liquid saturation, and the absolute
0.3 289 20 5.7 permeability are thecontrolling parameters for predicting well
1 103 20 6.7 productivity in gas condensate reservoirs. The trial and error
4.9 (correct value) 30 20 7.9 procedure, schematically described in Fig. 11, has been
Table 1: Relationship between storativity ratio, zone radius and wellbore skin
effect for a North Sea well29
validated with field data and has been found to be accurate
enough to be used for forecasting well productivity in gas
One particularity of gas condensate well tests below the dew condensate reservoir34.
point pressure is that the wellbore skin effect may increase,
decrease (Fig. 10) or remain constant as the gas rate increases, Well test analysis using compositional simulation
instead of increasing with rate, the expected behavior above
The conventional well test analysis approach described above
the dew point pressure. This reflects the balance between the
provides a series of snapshots of the well-reservoir-fluid
positive impact on productivity of the capillary number effect,
characteristics at specific times from the start of production. It
and the negative impact of inertia (non-Darcy or turbulent
does not allow predicting how the system will evolve in the
flow).
future. This can only be achieved by compositional
30
simulation. Compositional simulation also provides a
verification of the results from conventional well test analysis,
25 and in particular of the condensate bank radius. Using the
analytical well test interpretation results as inputs, the
Capillary number
20
effects dominate compositional model must provide a reasonable match not
Wellbore skin

over inertia only on the pressure-rate history and the producing GOR, but
15 also on the log–log plot of pseudo-pressure and derivatives of
the main build-up’s and drawdowns.
10 In general, a radial local grid is used with logarithmically
Above the dew Below the dew increasing sizes away from the well to permit evaluation of the
point pressure point pressure
5 near-wellbore gas condensate behavior in enough detail. High
resolution time steps must be used, which provide linear
0
pressure gradients and smooth saturation profiles on a semi-
0 5 10 15 20 25 log scale.
Gas rate (MMscf/D) In order to ensure the reliability of the compositional
Fig. 10: Wellbore skin effect vs. rate simulation, it is essential to have a fluid model that behaves as
the actual reservoir fluid, within the applicable pressure range.
Therefore, a proper characterization of the most representative
START
fluid sample is mandatory. Often, however, there is no fluid
Conventional Build-up
- sample from the well under study or the available samples are
analysis pressure data
not representative and a PVT fluid sample from a different
Estimation of
pseudo-relative
m
Guess krg @ S wi Single - phase well, which had been correctly recombined, must be selected.
pseudo -pressure
permeabilities
i = 0 The heavier fractions of the fluid samples (i.e. C7+) must be
m
krg @ Swi keff @ Swi Mobility ratio of Base lumped into fewer pseudo-components in order to decrease
capillary
bank to reservoir
number CPU time. An equation of state must be selected to predict the
actual reservoir fluids properties and its parameters tuned by
(k abs )i
regressing on the critical properties of the plus fraction until a
k eff / (k abs )i+1 good match was obtained between predicted values and
Pseudo- k r
≠1
observed data, such as dew point pressure, fluid density and
(k abs )i+1
(k abs )i+1
Two-phase
pseudo -
viscosity, fluid volume and composition, and liquid shrinkage
(k abs )i pressure
during CVD and CCE experiments. These variables are
=1
END
selected for regression because the properties of the plus
30
fraction were less accurate and not well defined.
Fig. 11: Procedure for estimating relative permeability data . Relative permeability curves should be selected from a set
In addition to the usual results from conventional analysis, it is used in the actual full-field compositional simulation based on
possible to estimate the relative permeabilities and the the reported fluid connate water saturation. The end point of
absolute permeability from well test analysis in lieu of the gas relative permeability must be adjusted in order to
SPE 100993 7

obtain a gas effective permeability consistent with that from Incorrect or inconsistent rate history
conventional well test analysis. When special core analysis is The issues more specifically associated with rates are
not available, the relative permeability characteristics can illustrated in Fig. 13. They are of two types: uncertainties in
estimated using Corey function12. Sensitivity runs must be the values and incomplete histories.
made on the Corey parameters to define a consistent set of Rates measured at the separator during a well test have
relative permeability curves that provides both calculated oil typically errors between 10 and 20%. In production wells,
rates equal to the measured oil rate at a specific gas rate, and a rates are often allocated, with the allocation calibrated at
good match with the flowing bottomhole pressure. regular interval, and may be incorrect. Even the latest
Capillary number and inertia effects must be included in the multiphase flowmeters provide noisy data which must be
simulation. The Forchheimer62 parameter β that defines inertia smoothed with a moving average algorithm, thus introducing
can be obtained from Geertsma’s correlation36, whereas additional errors. The rates must therefore be validated and
capillary numbers can be obtained from a number of made consistent between flow periods, otherwise errors in
correlations37-42. These require coefficients that must be rates will be transmitted to the calculated permeability values.
determined experimentally or from correlations29. This is usually done by plotting rate-normalized derivatives
Once the capillary number parameters are obtained, well skins together on a log-log graph and verifying that they share the
must be adjusted by trial and error to obtain a good match for same stabilization during radial flow: if there is an inconsistent
all the drawdown periods on the pressure history plot. flowrate in any period, the radial flow derivative stabilization
An example of results from compositional simulation analysis for that period would be different from that for other periods.
is shown in Fig.1229. Flowrate adjustments should be limited to the uncertainty in
the rates (10-20%). When the main radial flow stabilization
6000 16
Data has not been reached or does not exist, a different flow regime
5000
could be used for adjustment, with the caveat that such a flow
Pressure (psia)

15
GOR(Mscf/D)

4000
Simulation with
<10%
regime may be affected by the rate history and therefore have
capillary number 14
3000
different characteristics in different flow periods. In a gas
2000
13 GOR
condensate well test, for instance, the radial flow stabilization
1000 Simulation without Measured
Calculated
corresponding to the condensate bank could be used if the
capillary number
0 12 reservoir stabilization has not been reached (a common
0.0 0.4 0.8 1.2 1.6 2.0 2.4 0 2 4 6 8 10 12 14
Elapsed time (days) 10000 Elapsed time (days) occurrence, as discussed later). In that case, great caution must
103 0.1 be exercised in adjusting the rates, as the level of the bank
Δmn(p) and derivative (psi)

Simulation with capillary number


Condensatesaturation

1 2 3 4
and no wellbore storage 0.08 stabilization varies with the condensate saturation and
(Numerical)
r2=38 ft

0.06 Bu6 therefore with the rate. Such an adjustment has been
102 Data
performed in the left hand side of Fig. 1344.
(Numerical)

0.04
Dd5
r1=5 ft

r2=30 ft
0.02 (WTA) Corrected but uncomplete rates Complete rates
Dd5 Effective permeability
10 0
10-4 10-3 10-2 10-1 102 6000
1 10 10-1 1 10 102
Raw rates Corrected rates
Pressure (psia)

R=38 ft
5000
Elapsed time (hours) Radial distance (ft) (Numerical)
4000 100

Rate (MMscf/D)
Fig. 12: Verification of conventional well test analysis with 3000 80
29
compositional simulation 2000 60

1000 40

20
Interpretation Challenges 0

0
0 400 800 1200 1600 0 2000 4000 6000 8000 10000
Data Time from start of production (hrs) Time from start of production (hrs)
104
Well test interpretation consists of solving an inverse problem,
Rate normalized Δmn(p)
and derivative (psi)

the solution of which is non-unique by definition43. The only 103


Bu392 Bu392
way to decrease the non-uniqueness is to increase the amount Bu499 Bu499
102
of data one is working with. These include more pressure and
Bu322 Bu322
rate data (ideally, all the pressure and rate information on the 10
Bu456 Bu310 Bu456
well), and all possible interpreted information on the reservoir 1
Bu310

(geology, seismics and logs), the fluid (PVT) and the well 10-2 10-1 1 10 102 10-2 10-1 1 10 102

Elapsed time (hours) Elapsed time (hours)


(drilling and completion) and on the way the test has been run
(sequence of operations). The more complex the environment 44
Fig. 13: Impact of rate history on derivative shapes
of the test, the more information is required. Tests in gas
condensate reservoirs below the dew point pressure are
probably at the high end of the difficulty scale, as they
combine the complexities of the geology and the well with the
complexity of the fluid.
8 SPE 100993

Incomplete rate history Lack of final radial flow derivative stabilization


The entire production history is required in order to obtain a The interpretation results described earlier can only be
correct analysis, especially for production well tests. obtained if all the flow regimes required for analysis can be
Truncating the rate history by ignoring its early part or identified on the data. Fig. 6 points to a very common problem
simplifying it with an “equivalent” Horner time45 (cumulative in condensate gas well test analysis, namely, the derivatives
production divided by last rate) may distort the derivative46 have not reached their final stabilizations. This is mostly the
and lead to erroneous interpretations. This is illustrated in Fig. case in production tests, as illustrated in Fig. 15.
13: the derivatives calculated on the left hand side with only 104
1 second Belo w dew point pressure
the rates corresponding to the pressure measurements are

Rate normalized Δmn(p) and derivative (psi)


pressure
different from those on the right hand side, which include all pressure
acquisition Above dew point
the rates from the start of production. As a rule, the more
recent the changes in production rates, the more detailed the
103
rate history must be. Describing accurately the rate history Condensate
over a period corresponding to the last 40% of the cumulative bank
production of the well, and using the equivalent Horner time
to represent the first 60% provides a correct derivative46. Production test
102

Phase redistribution in the wellbore


Wellbore phase redistribution is a recurrent problem in gas DST Reservoir
Effective
condensate reservoir well tests. It occurs whenever gas and 24 hour flow period permeability
condensate flow in different directions in the wellbore1. This 10
would not happen in a drawdown with a gas rate high enough 10-5 10-3 10-1 10 103 105 107
Elapsed time (hours)
to lift the condensate droplets to the surface, but it would most
Fig. 15: Influence of the size of the condensate bank on the onset
likely be present in a subsequent build-up or a subsequent of the final derivative stabilization
drawdown period at a lower rate20. Similarly, phase
redistribution can take place during low flowrate drawdowns, If we consider a typical test, a maximum build up duration of
in which case there would be no phase redistribution in a 24 hour may be enough to see a radial flow stabilization on the
following build-up or lower rate drawdown20. Wellbore phase derivative if the pressure remains above the dew point
redistribution creates an increase in wellbore storage during pressure, and even in a DST below the dew point pressure,
both drawdown or build-up periods and may dominate the test. when the condensate bank has just started to form. In a
Figure 14 is a rate-normalized log-log plot of drawdown data production test, on the other hand, the radius of the condensate
for a North Sea lean gas condensate well1. Phase re- bank may be such that the derivative is still only showing the
distribution dominates the drawdown data and is more stabilization due to the bank at the end of the 24 hour period.
pronounced and lasts longer for low flow rates (Flow periods Seeing the final derivative stabilization would possibly take
14, 7 and 15). The highest rate drawdown, FP 17, is less ten times as long.
affected and only at very early times, which makes it
analyzable. Its derivative is similar to the build-up ones, Flow 104
DST well B
periods 8, 18 and 21.
Rate normalized Δmn(p) and derivative (psi)

10 3
FP 8 Bu 103
(DD 45 M Scf/D)
Rate normalized Δmn(p) and derivative (psi)

10 2 102
FP 21 Bu FP 15
(DD 38 MScf/D) (DD 47 M Scf/D)

FP 17 10
10 (DD 57.5 M Scf/D)

FP 14 Effective permeability from core


(DD 38 M Scf/D)
1
FP 7 FP 18 Bu
(DD 45 M Scf/D) (DD 57.5 M Scf/D) 10-3 10-2 10-1 1 10 102
North Sea
1 Elapsed time (hours)
10 -3 10 -2 10 -1 1 10 10 2 10 3
Fig. 16: Final derivative stabilization from a DST in a North Sea
Elapsed time (hours) 44
well: well test effective reservoir vs. core permeability
Fig. 14: Example of increasing wellbore storage due to phase
1
redistribution in the wellbore during low rate drawdowns The interpretation challenge, when there is only a single
derivative stabilization on the derivative, is to decide whether
SPE 100993 9

that stabilization corresponds to the condensate bank or to the normalized to a unit rate. In Fig. 18, the downward trend in the
reservoir. In sandstone reservoirs, our experience indicates deconvolved derivative suggests that the derivative
that the effective permeability from the final derivative stabilization represents the condensate bank mobility rather
stabilization is in good agreement with the arithmetic average than the reservoir effective permeability. The deconvolved
core permeability44 (Fig. 16). The single derivative derivative does not match the data at the end because it is an
stabilization should therefore be compared with the average over the deconvolved period, and therefore the
stabilization level corresponding to the core permeability: if it downward trend represents the average of the condensate front
is above, it is likely to represent the condensate bank. locations. The final stabilization does not appear in the
104 deconvolved derivative because the test was too short.
Bu456 Bu392

Fluid versus reservoir behavior


mn(p) change and derivative (psi)

103
The derivative shapes are often ambiguous and can due either
Bu499 to the fluid or to geological features. Forward modeling with
? analytical or numerical models must be used to identify
102
possible causes.

1 6 /2 6 -B 1 0
Condensate mobility
PLT 0.2
WIRE.AT10_1
OHMM 20 kh ϕ Sw
PLT.FLOW_1 WIRE.PEF_1 WIRE.AT20_1

C o re

TOPS.ZONE_1
0 MSCFT/DAY 50000 0 B/E 10 0.2 OHMM 20

Layer
CONTACTS
CORE.PLT_1 WIRE.DT_1 WIRE.AT30_1

FACIES

TVDSS

PERFS
0 2082 140 US/F 40 0.2 OHMM 20
WIRE.GR_REF_1 WIRE.VSHTH_1 WIRE.NPHI_1 WIRE.AT60_1 WIRE.KAHCOMPN_1 CORE.PORBEST_1 CORE.SWC_1

BED

FEET
0 API 150 0 V/V 1 WIRE.PAYAUTO_2 0.45 V/V -0.15 0.2 OHMM 20 0.01 MD 1000 0.3 % 0 1 V/V 0

10 Core permeability 6
WIRE.CALI_1
IN 16 0
WIRE.VSHGR_1
V/V 1
0
WIRE.NETAUTO_2
0
6

6 1.95
WIRE.RHOB_1
G/C3 2.95 0.2
WIRE.RT_1
OHMM 20
CORE.KAHBEST_1
0.01 MD 1000 0.3
WIRE.PHIT_1
V/V 0 1
WIRE.SW_1
V/V 0
S k in h (ft) kh kv
12860

Bu322 Bu310 12870


86
10
12880

12886
12880

84
11
12896
Z50
12900 97
82

Production well A 13

12910

80
15 12920
1 0 82 13 0 .0 6
12926

12933

78
13 12940

12946 12947

1
10-2 10-1 1 10 102
12960
GAS
232
2 c lo s e d 24 0 .1 4 0 .0 7
12977
12980

3 c lo s e d 19 25 12
Elapsed time (hours)
66
37
13000
4 0 to 4 0 20 12 7
13014

Fig. 17: Final derivative stabilization from a production test in a 62


22

13037
13020

5 c lo s e d 20 19 0 .0 5
44 13042
60
13040
Z45

6 0 30 2 0 .1
North Sea well: condensate mobility vs. core permeability
7 198
13049

58
17
13060

13066
56
9
13075

13081 13080
13082

This is the conclusion reached from Fig. 17 on the test shown


54
11
13092

52
11 13100
13105

104 13108

50 WATER

in Fig. 12. A second conclusion is that the end of Bu310,


12 63
Δmn(p) and derivative (psi)

13120 13120
48
10
13130

13140

which falls below the bank stabilization, must correspond to


the transition between the bank and the reservoir 103

stabilizations, reflecting the advancement of the condensate


40
bank.
10 2
0
1E+01
I n c r e a s in g S 4 ( 0 ,5 , 1 0 ,2 0 ,4 0 )
Actual Normalised Derivative
10
1 0 -2 1 0 -1 1 10 102 103 104 106
Rate Normalised Derivative (psi)

E la p s e d t im e ( h o u r s )
1E+00 Fig. 19: Multilayer behavior in a production test in a North Sea
44
well

In Fig. 17, the question mark,”?”, refers to the derivative data


between the end of wellbore storage and skin effects and the
1E-01 start of the condensate bank stabilization. These data could be
part of the condensate bank, represent spherical flow or
Deconvolved Derivative correspond to the reservoir layering. An analytical
multilayered simulator48 was used to test the various
1E-02
possibilities (Fig. 19). The reservoir includes three, non-
1E-04 1E-03 1E-02 1E-01 1E+00 1E+01 1E+02 1E+03 communicating, layers, labeled (1), (3-4-5) and 6 in the upper
Elapsed time (hrs) part of Fig. 19. The middle layer, (3-4-5), is only perforated in
Figure 18: Deconvolved derivative suggesting the existence of a (4). If the skin coefficient for every perforated layer is the
43
condensate bank same (say, zero), the multilayer derivative displays a double
permeability behavior, with a U-shaped minimum
Another helpful tool is deconvolution47, particularly if core corresponding to the V-shape found in double porosity
permeability is not available. Deconvolution transforms behavior49. As the contrast in skin increases, however, with the
variable rate pressure data into a constant rate initial skin factor in layer (4) increasing to 5, 10, 20 and 40, the
drawdown with a duration equal to the total duration of the minimum disappears and the early part of the derivative shape
test, and yields directly the corresponding pressure derivative, tends toward that found in commingled layers. This leads to
10 SPE 100993

the conclusion that the “?” shape in Fig. 16 is due to layering, Another example of an ambiguous derivative shape that must
not to the condensate bank. be resolved through forward simulation is shown in Fig. 20.
104 The derivative shapes of the build up’s, Bu3 from well 4 in the
log-log plot at the top, and Bu3, Bu7 and Bu9 from well E in
the log-log graph at the bottom, are very similar. In well 4, the
103 Bank next build up, Bu6, shows clearly a 3-region composite
behavior due to condensate banking. It can therefore be
Rate Normalised Δmn(p) & derivative (psi)

Bu3 concluded that Bu3 also shows the bank and more specifically,
102
Bu6 the capillary number zone (zone 1 in Fig. 1).
The core permeability for well E suggests that the derivatives
D S T w e ll 4 data correspond to the condensate bank. The upward trend at
10 the end of the derivatives, confirmed by deconvolution, could
1 0 -2 1 0 -1 1 10 102
therefore be due to the bank or to the discontinuous faults
E la p s e d tim e (h rs )
identified by seismic (Fig. 21). The effect of the faults is tested
104
in Fig. 21 with a voranoid grid simulator50, for a single phase
Bu9 D e c o n v o lv e d
d e riv a tiv e gas above the dew point pressure. The resulting derivative
clearly indicates that the upward trend is due to the fault, with
103 Bu7
B a n k o r b o u n d a rie s ?
the derivative going directly from the condensate bank into the
fault without reaching the reservoir effective permeability
Bu3
stabilization.
102

E ffe c tiv e p e rm e a b ility fro m c o re Such a situation is very common in gas condensate well tests
D S T w e ll E and is due to the growth of the condensate bank. Another
10
1 0 -2 1 0 -1 1 10 102 example is shown in Fig. 22. The build up labeled FP16 has a
E la p s e d tim e (h o u rs )
very different derivative from that of the build up FP50. FP16
44
Fig. 20: Example of ambiguous derivative shape is at the start of production and the condensate bank has just
begun to form. Its derivative exhibits a downward trend which
represents the transition between the stabilization due to the
condensate bank and that corresponding to the reservoir
effective permeability (represented by the core permeability
4550 ft line). The derivative in FP50, on the other hand, is mostly flat
13800 ft
except for an upward trend at the end which is confirmed by
1180 ft
13300 ft
9660 ft
deconvolution. Voronoid grid simulation50 shows that that
10000 ft trend is due to the faults of limited extent around the well. As
the radius of the condensate bank increases, the behavior
180 ft 9110 ft becomes less dominated by the bank and more by the
1090 ft boundaries.
6000 80 103
FP16 FP50
Δmn(p) and derivative (psi)

5800 70
Dew point
5600 60
pressure
Pressure (psia)

Gas Rate (MMscf/D)

Deconvolution of all the build-ups


5400 50

5200 40 102
Numerical simulation with core arithmetic average permeability (with bank)
Numerical simulation with core arithmetic average permeability (w/o bank) 5000 30
FP50
Rate Normalised Δmn(p) & derivative (psi)

4800 20 FP16
10 4 Core
No-flow boundary of limited extent 4600 10
permeability
Deconvolution of FP16
4400 0 10
20 40 60 80 100 120 140 160 180 200 220 240 260 280 10-2 10 -1
1 10 102
103 104
Time from start of test (hrs) Elapsed time (hours)
10 3 Bu3 103
Deconvolution
Δmn(p) and derivative (psi)

500
Simulation with core
arithmetic average
10 2 Increasing permeability
102 bank
-500
radius
FP50
Condensate FP16
10 mobility Effective permeability from core -1500
Core
permeability
10 -3 10-2 10 -1 1 10 10 2 103 10 4 10 5 10
-1500 -500 500 1500 10-2 10-1 1 10 102 103 104
Elapsed time (hours) Length [ft] vs Length [ft] Elapsed time (hours)
Fig. 21: Verification of boundary effects with a voranoi grid Fig. 21: Example of changing well test behavior due to the growth
44
simulator of the condensate bank and the presence of boundaries
44
SPE 100993 11

Finally, Fig. 22 shows a derivative in a gas condensate DST capillary number effect, the total skin decreases to become
slightly below the dew point pressure, which exhibits the equal to 7 when k1=k3, i.e. for the theoretical case where all
characteristics of a condensate bank, although it was not clear condensate has been removed around the well. The other log-
from PVT data whether a condensate bank could have formed log plots in Fig. 23 show the effects of r1 and r2 and of the
or not. The ambiguity was solved by deconvolution: the bank size (r2 – r1). They are negligible compared to the impact
deconvolved derivative indicates a homogeneous behavior and of k1. The main conclusion is that the condensate saturation
channel boundaries, with the derivative shape due to the need to be decreased only in the immediate vicinity of the well
derivative calculation algorithm (the multirate derivative to improve productivity significantly.
differs from the drawdown derivative43 because of the
St is independent of the distance to a condensate
previous rate history). 10000
St decreases significantly as k1 increases
10000
bank of constant width
r1= 100 r2 = 250 St 28 St 19
18
Rate Normalised Δmn(p) & derivative (psi)

10 12 18
7 r1= 10 r2 = 160
1000 1000
k1= 1/3 k3 = k2 r1= 100 r2 = 250

nm(p) Change and Derivative (psi)


k2
r1= 1000 r2 = 1150
Sw=5
1 100 k3 100
Condensate bank? k1= ½ k3 = 3/2 k2
k1 = k 3 = 3 k2
k1= 2/3 k3 = 2 k2 k1= 1/2 k3 = 3/2 k2
10
10

10 -1 0.001 0.01 0.1 1 10 100 1000 0.001 0.01 0.1 1 10 100 1000 10000

St increases slightly as the condensate doughnut St decreases slightly as the stripping zone radius
outer radius increases increases
10000 10000

St 18 - 21 St 25
Reservoir effective 18
permeability? r1= 1 r2 = 1150
10 2 r1= 100 r2 = 550
r1= 30 r2 = 1150
1000 1000
r1= 100 r2 = 1150 r1= 100 r2 = 1150
Channel?
10 3 100 100

r1= 500 r2 = 1150


10 -4 10 -3 10-2 10-1 1 10 10 2 103
Elapsed time (hours) k1= 1/2 k3 = 3/2 k2
r1= 100 r2 = 250
k1= 1/2 k3 = 3/2 k2 r1= 1000 r2 = 1150
10 10
Rate Normalised Δmn(p) & derivative (psi)

10 0.001 0.01 0.1 1 10 100 1000 10000 0.001 0.01 0.1 1 10 100 1000 10000

Drawdown derivative type curve Elapsed time (hrs)


Pressure
Figure 23: Impact of the condensate bank saturation, size and
1 Deconvolved derivative location on the total skin effect

Several techniques have been used or proposed to increase


10 -1 Derivative productivity. They include solvent injection, gas cycling,
Shape due to algorithm acidising and hydraulically fracturing of vertical wells before
for calculating the or after the development of the condensate bank, and drilling
10 2 multirate derivative
horizontal wells instead of vertical wells.
Multirate derivative type curve In the solvent injection technique, a solvent is pumped into the
10 3 formation in order to alter the wettability of the reservoir rocks
10 -4 10 -3 10-2 10-1 1 10 10 2 10 3 near the wellbore from strongly liquid-wet to intermediate gas
Elapsed time (hours) wet51. The results from laboratory tests show that wettability
Figure 22: Deconvolved derivative proving the derivative shape is
due to the pressure derivative calculation algorithm, not to a significantly affects both critical condensate saturation and gas
condensate bank
43
phase relative permeability52. This method however, is still at
the research stage and has not yet been used at reservoir scale.
Productivity impairment remediation In gas cycling, the condensate liquid is removed from the
produced (wet) gas, usually in a gasoline plant, and the
Fig. 23 illustrates with a simple example the importance on the residue, or dry gas is returned to the reservoir through
total skin effect of the condensate saturation in the condensate injection wells. The injected gas maintains reservoir pressure
bank and in the velocity stripping zone. The four log-log plots and retards retrograde condensation. Although this technique
included in Fig. 23 all show a 3-region composite behavior appears to be an ideal solution to the retrograde condensate
due to condensate banking. k3 is the reservoir effective problem, it is often less attractive due to practical limitations
permeability, k2 the gas relative permeability in the condensate and/or economical considerations. Gas cycling usually
bank, and k1 the gas relative permeability in the velocity requires additional expenditures for drilling additional wells,
stripping zone (respectively, zones 4, 2+3, and 1 in Fig. 1). r1 liquid recovery plants and compression systems53.
is the inner radius of the condensate bank (the outer radius of
the stripping zone) and r2 is the outer radius. The wellbore The most common ways of improving productivity are still
mechanical skin effect is equal to 5. The maximum total skin acidification, fracturing, and horizontal wells, all aiming at
effect is obtained when there is no capillary number effect decrease or delay the pressure drawdowns, and therefore the
(k1=k2). Its value, 28, depends on the condensate saturation in condensate saturation, in order to increase or maintain gas
the condensate bank, i.e. on k2. As k1 increases with the well deliverability. In the following, we present examples
12 SPE 100993

showing the impact of these various techniques on well test Fig. 25 shows a log-plot of normalized pseudo-pressure and
behavior. derivatives for a North Sea horizontal well54. The pressure and
rate data, shown in Fig. 26, include a DST (build up’s FP29
Fig. 24 shows a log-log plot with build up’s before (FP3 and and 38), mostly above the dew point pressure, and two
10) and after (FP26) a matrix acidification, performed because production tests below the dew point pressure (respectively,
a high skin damage had been observed during the test. build up’s FP 48-50 and FP62-65). The derivatives clearly
Deconvolution and the permeability from cores suggest that show the existence and growth of a condensate bank, which
only the condensate bank is seen on the derivatives before the appears as a composite behavior superimposed on the
acid job, whereas the transition between condensate bank and horizontal well behavior, as predicted by forward composition
reservoir is seen afterwards. The post-acid data exhibit a high simulation (Fig. 7). There is, however, no published analytical
wellbore storage, which hides the condensate bank derivative solution for such a model and the growth of the condensate
stabilization and makes it impossible to decide whether the bank must be handled through an increase in the total skin in
condensate saturation has been decreased or not. The total skin conventional analysis. Deconvolution indicates the presence of
effect, however, has not decreased, which would imply that parallel faults, which is consistent with the seismic
acidification has not been effective in improving productivity information. Here again, the derivative would go directly from
impairment due to the condensate bank. the condensate bank to the boundaries, never reaching the
final radial flow stabilization identified from cores.
104
4000 40
Deconvolution FP 29 38 48 50 62 65

Total Rate (MMscf/D)


Dew point pressure 30

Pressure (psia)
mn(p) change and derivative (psi)

3000
103
20

2000
10

Before acid After acid 1000 0


0 10 20 30 40 50 2544 2570 7602 7630
102
Elapsed time (hrs) Elapsed time (hrs) Elapsed time (hrs)
FP3
FP10
4000 40

Total Rate (MMscf/D)


1st production 2nd production
Condensate mobility DST
FP26 test test
Pressure (psia)

3000 30
10 Dew point pressure
2000 20

Effective permeability from core 1000 10


1
10-2 10-1 1 10 102 0 0
0 1000 2000 3000 4000 5000 6000 7000 8000
Elapsed time (hours) Elapsed time (hrs)
Figure 24: Impact of acidizing on the condensate bank saturation, Figure 26: Pressure and rate history for a North Sea horizontal
44 54
and size well

103 0.35

0.30 1st production


Reservoir Boundary at 700 ft
mn(p) Change and Derivative (psi)

test
FP50
5
P6

Condensate saturation

0.25 FP48
dF

2nd
e

FP65 production
olv
nv

test
0.20
co
De

FP65
102 2nd production
0.15 test

Core permeability 0.10


FP48
FP50
FP38
1st FP29 0.05 DST
production DST FP29
test
FP38
10
0
10-3 10-2 10-1 1 10 102 103 104 1 10 100 1000 10000
Elapsed time (hrs) Radial distance perpendicular to the well from the heel, ft
Figure 25: Log-log plot of normalized pseudo-pressure and Figure 27: Condensate saturation distribution for a North Sea
54 54
derivative for a North Sea horizontal well horizontal well
SPE 100993 13

As no analytical composite model is available for horizontal derivatives, below the dew point pressure, show a growing
wells, the only way to characterize the condensate bank is condensate bank, which appears as a composite behavior
through compositional simulation, which provides the superimposed on the fractured well behavior, as predicted by
condensate saturation distribution in the reservoir (Fig. 27). forward composition simulation31 (Fig. 7). Although there is
This in turn yields the bank radius, the contribution of the no published solution for such a model either, a composite
bank to the total skin effect, and the wellbore skin coefficient. model for wellbore storage and skin with CDe2S <0.5 (which is
The condensate saturation around the well in the final build up a characteristic of an infinite conductivity fracture)55 can be
of the second production test, FP65, is less than that in the first used. The gas rate is lower in the second production test,
production test. Therefore, the contribution of the condensate which triggers re-vaporization and a smaller condensate bank.
bank to the total skin must be less. Yet, the total skin is higher The total skin effect increase must therefore be due to an
(Fig. 25). This means that the wellbore mechanical skin has increase in the mechanical skin, as in the previous example.
increased between the first and the second production test. In Deconvolution indicates a reservoir of infinite extent.
this particular example, the condensate bank reaches the
boundary in the second production test, confirming that later Not all fractured wells behave as in Fig. 28. Fig. 29 shows a
derivatives would go directly from the condensate bank to the pressure and rate history for a well in a multilayered reservoir
boundaries, never reaching the final radial flow stabilization. in the North Sea56. The pressure was below the dew point
pressure at all times. The wells productivity declined gradually
8000
over time and a remedial massive hydraulic fracturing
DST 2nd production operation was carried out which saw a sixty percent
Pressure (psia)

FP12 test improvement in the well productivity index.


6000
Dew point pressure 5000
FP66 FP79
4000 FP7 1st production 4000
FP50
test Pressure (psia)

Remedial Hydraulic Fracture


Gas Rate (MMscf/D)

40 3000
2000

FP30 30 2000
0
20 1000 20
FP28 FP59 FP70

Gas Rate (MMscf/D)


10 0

0 10
0 200 400 600 800 1000 1200 1400
Elapsed time (days)

104 0
Rate Normalised Δmn(p) and Derivative (psi)

FP50 0 2000 4000 6000 8000


Elapsed time (hrs)
Condensate bank
Rate Normalised Δmn(p) and Derivative (psi)

103 104
FP30
FP12

102

Condensate bank

ive
FP59

at
103
riv
10 Deconvolved de
FP7
derivative
d
ve
ol

FP28
nv

Model, above the dew point pressure FP70


co

1
De

10-3 10-2 10-1 1 10 102 103 104 105 FP66

Elapsed time (hrs) FP79


102
Figure 28: Log-log plot of normalized pseudo-pressure and
Phase redistribution Core permeability
derivative for a hydraulically fractured well in the Middle East
10-2 10-1 1 10 102 103 104
Elapsed time (hrs)
Fig. 28 shows a log-plot of normalized pseudo-pressure and
Figure 29: Log-log plot of normalized pseudo-pressure and
derivatives for a fracture well in a low permeability Middle derivative for a hydraulically fractured well in the North Sea
56

East reservoir. In this example again, the pressure and rate


data include a DST (build up’s FP 7 and 12, with FP7 above The data in the log-log graph of Fig. 29 exhibit a wellbore
the dew point pressure), and two production tests below the storage and skin two-region composite behavior even after
dew point pressure (build up’s FP30 and FP50, respectively). fracturing, instead of the fractured well composite behavior of
FP7 exhibits a normal fractured well behavior, represented by Fig. 28. This is due to the layering of the reservoir and is not
the “model above the dew point pressure”. All the other unusual in the North Sea. As a result, the fracture
14 SPE 100993

characteristics cannot be determined. The derivative radial This yields an increase in the condensate bank size and a
flow stabilization coincides with that corresponding to the decrease in well productivity.
core permeability, as expected. The deconvolved derivative
suggests the existence of multiple boundaries. Whereas fractured vertical wells and horizontal wells increase
The behavior of the outer bank radius is hidden by phase productivity in dry gas systems, their performance is even
redistribution, so it is not clear whether it is growing or not. better in gas-condensate reservoirs below the dew point, where
On the other hand, the derivative stabilization level they decrease pressure drawdowns and condensate blockage
corresponding to the condensate bank, and therefore, the compared to a vertical well. Fig. 31 shows that they are
condensate saturation in the bank, decreases after fracturing in equally effective in improving productivity in gas-condensate
the build up’s FP59, FP66 and FP70 from the level before frac reservoirs below the dew point. The optimum choice, when
(FP28), and then increases again in FP79. This is paralleled by both are technically feasible, can only be made from economic
a decrease then an increase in the total skin. considerations57.

100

% increase in total gas production over vertical well


90 LH (ft)

80 1000

70 xf=300 ft

xf =200 800
xf =50 xf =10
60 0 ft ft
ft
50

Before frac 600


50 hours after frac 40

30

20

400
10

0
0 1 2 3 4 5 6 7

Production Time, years


Figure 31: Comparison of relative increase in cumulative gas
production for horizontal and fractured vertical wells compared to
57
100 hours after frac 500 hours after frac that from a vertical well below the dew point

Summary of Results
The results presented in this paper cam be summarized as
follows:
1. Condensate is deposited around the well when the
bottomhole pressure drops below the dew point during
production.
2. The corresponding impediment to flow is compensated by
1000 hours after frac 20000 hours after frac capillary number effects.
31
3. Condensate deposit and capillary number effects yield a
Figure 30: Evolution of condensate bank after fracturing two- or three region composite well test behavior when
single phase pseudo-pressures are used for analysis.
Such a behavior matches the compositional simulation results
4. The final derivative stabilization corresponding to the
shown in Fig. 3031: a non-fractured well is produced until the
reservoir effective permeability in the composite behavior
well bottomhole flowing pressure drops below the dew point
is usually not reached in production tests. The
pressure and a condensate bank forms around the wellbore. A
stabilization seen on the derivative is likely to represent
hydraulic fracture is then created with a facture half-length
the condensate bank mobility.
extending beyond the condensate bank and the well is
5. The reservoir effective permeability is consistent with
produced again at the same rate as before frac. The bottomhole
core permeability in sandstone reservoirs. The core
pressure is above the dew point pressure initially, and no
permeability can be used to distinguish between
additional condensate is deposited in the reservoir. Instead, the
condensate bank and reservoir mobility if only a single
existing condensate is produced to the surface, decreasing the
stabilization is seen on the derivative.
condensate saturation and the size of the condensate bank. As
6. The derivative stabilization corresponding to the
time increases, the bottomhole pressure fall again below the
mobility of the condensate bank varies with the
dew point pressure and condensate drops out in the reservoir.
SPE 100993 15

condensate saturation and therefore the rate. At ctc total compressibility in the condensate bank (psi-1)
constant rate, its level increases with time until a c compressibility (psi-1)
maximum level is reached. h reservoir thickness (ft)
7. The condensate bank decreases in size and saturation k permeability (mD)
when the production rate decreases. Lh horizontal well length (ft)
8. The condensate saturation distribution in a build up is m(p) single phase pseudo-pressure (psi)
approximately the same as that as the end of the preceding mn(p) normalized single phase pseudo-pressure (psi)
drawdown. m2ϕ(p) two-phase pseudo-pressure (psi)
9. Wellbore phase redistribution may dominate the entire p pressure (psi)
test. pref reference pressure (psi)
10. It is often difficult to distinguish condensate bank effects Pdew dew point pressure (psi)
from layering, boundary or derivative calculation effects. r1 condensate bank inner radius (ft)
A series of tools must be used for identification, including r2 condensate bank outer radius (ft)
conventional well test analysis, deconvolution, forward Rs solution gas /oil ratio
modeling with analytical and numerical models, and Rv dissolved oil/gas ratio
compositional simulation. Sw wellbore skin effect; water saturation
11. Behavior often changes with time as the condensate bank So condensate saturation, fraction
grows and reaches the boundaries. Successive drawdowns Sg gas saturation, fraction
and build up’s must be analyzed together to understand St total skin effect
these changes (time-lapse well test analysis). xf fracture half length (ft)
12. Capillary numbers often compensate for inertia effects. Z gas deviation factor
As a result, the wellbore skin may increase, decrease or
remain constant as the gas rate increases. Greek
13. Calculating the bank outer radius requires to know the φ porosity, fraction
bank total compressibility, which is greater than the gas σ interfacial tension, (lb/ft)
compressibility above the dew point pressure. μ viscosity (cp)
14. Pseudo-relative permeabilities, absolute permeability and ν interstitial velocity, (ft/sec)
base capillary number can be estimated using single-
phase and two-phase pseudo-pressures together. Abbreviation
15. Fracturing vertical wells and drilling horizontal wells is Bu build up
equally effective for improving productivity in gas- CCE constant composition expansion
condensate reservoirs below the dew point. CVD constant volume depletion
Dd drawdown
DST drill stem test
Acknowledgement EOS equation of state
Portions of this study were conducted at Imperial College FP flow period
London by Manijeh Bozorgzadeh30, Saifon Daungkaew44 and GOR gas-oil ratio
Abdolnabi Hashemi32, Olalekan Aluko56 and Tariq Baig58 in IFT interfacial tension
partial fulfillment of post-graduate degree requirements. The PVT pressure-volume-temperature
authors gratefully acknowledge the financial support of this WTA well test analysis
research by the members of the Imperial College Joint
Industry Project (JIP) on Well test Analysis in Gas Condensate Subscripts
and Volatile Oil Reservoirs: the UK Department of Trade and Abs absolute
Industry, Anadarko, Burlington Resources, BHP Billinton, Eff effective
Britannia Operator Ltd, ConocoPhillips, Gaz de France and F formation
Total. They are also indebted to software vendors for allowing G gas
them access to the software products required for this work, i initial
respectively Kappa Engineering (SaphirTM), Paradigm o oil
Geotechnology B.V (InterpretTM 2005) and Schlumberger r relative
(EclipseTM 300 and PVTiTM). Dr Daugkaew and Dr Hashemi ref reference
further acknowledge partial financial support from Britannia t total
Operator Ltd. and the Royal Thai Government, and from w water
NIOC, respectively.
Superscripts
Nomenclature Max maximum
B formation volume factor
(reservoir volume/standard volume)
16 SPE 100993

References 16. Danesh A. S., Dandekar. A. Y., Todd A. C., Sakar R.: “ A
Modified Scaling Law and Parachor Method Approach for
1. Gringarten, A. C., Al-Lamki, A., Daungkaew, S., Mott, R., and Improved Prediction of Interfacial Tension of Gas Condensate
Whittle, T.: “Well Test Analysis in Gas-Condensate Reservoirs,” System,” SPE22710, paper presented at the 66th Annual Technical
SPE 62920, paper presented at the SPE Annual Technical Conference and Exhibition of the Society of Petroleum Engineer
Conference and Exhibition held in Dallas, Texas, October 1-4, held in Dallas, TX, October 6-9, 1991.
2000. 17. Kalaydjian, F. J-M., Bourbiaux, B. J., Lambard, J-M., 1996, “
2. Muskat, M.: “Physical Principle of Oil Production,” McGraw- Predicting Gas-Condensate Reservoir Performance: How flow
Hill Book Co., Inc., New York, 1949, pp. 793. parameters are altered when approaching Production Wells.”,
3. Kniazeff, V. J., and Naville, S. A.: “Two-Phase Flow of Volatile paper SPE 36715 presented at the 1996 SPE Annual Conference
Hydrocarbons,” Society of Petroleum Engineering Journal and Exhibition, Colorado, 6-9 October, 1996.
(March 1965), pp. 37. 18. Ali, J. K., McGauley, P. J., and Wilson, C. J.: “ Experimental
4. Fussell, D. D.: “Single-Well Performance Predictions for Gas Studies and Modelling of Gas Condensate Flow Near the
Condensate Reservoirs”, Journal of Petroleum Engineering SPE Wellbore,” SPE39053, paper presented at the Fifth Latin
4072 (July 1973), pp. 860-870. American and Caribbean Petroleum Engineering Conference and
5. Barnum, R. S., Brinkman, F. P., Richardson, T. W., and Spillette, Exhibition, Brazil, 30 August - 3 September, 1997.
A. G.: “Gas Condensate Reservoir Behaviour: Productivity and 19. Economides, M. J., Dehghani, K., Ogbe, D. O., and Ostermann,
Recovery Reduction Due to Condensation,” SPE30767, paper R. D.: “Hysteresis Effects for Gas Condensate Wells Undergoing
presented at the SPE Annual Technical Conference and Build-up Tests below the Dew Point Pressure,” SPE16748, paper
Exhibition, Texas, October 22-25, 1995. presented at the 62nd Annual Technical Conference and
6. Afidick, D., Kaczorowski, N. J., and Bette, S.: “Production Exhibition of the Society of Petroleum Engineers, Texas,
Performance of a Retrograde Gas Reservoir: A Case Study of September 27-30, 1987.
Arun Field,” SPE 28749, paper presented at the SPE Asia Pacific 20. Behrenbruch, P. and Kozma, G.: “Interpretation of Results From
Oil and Gas Conference, Australia, November 7-10, 1994. Well Testing Gas-Condensate Reservoirs: Comparison of Theory
7. Favang, ∅., Whitson, C. H.: “Modelling Gas Condensate Well and Field Cases,” SPE13185, paper presented at the 59th Annual
Deliverability,” SPE30714, paper presented at the SPE Annual Technical Conference and Exhibition held in Houston, Texas,
Technical Conference and Exhibition, Texas, October 22-25, September 16-19, 1984.
1995. 21. Mott, R., Cable, A., and Spearing, M.: “ A New Method of
8. Gondouin, M., Iffly, R. and Husson, J.: “An Attempt to Predict Measuring Relative Permeabilities for Calculating Gas-
the Time Dependence of Well deliverability in Gas-Condensate Condensate Well Deliverability,” SPE56484, paper presented at
Fields,” Society of Petroleum Engineering Journal (June 1967), the 1999 SPE Annual Conference and Exhibition, Texas, 3-6
pp. 112-124. October, 1999.
9. Boom, W., Wit, K., Schulte, A. M., Oedai, S., Zeelenberg, J. P. 22. Al-Hussainy, R., Ramey, H. J. Jr. and Crawford, P. B., "The Flow
W., and Maas, J. G.: “Experimental Evidence for Improved of Real Gases through Porous Media", J. Pet. Tech., May 1966,
Condensate Mobility at Near-Wellbore Flow Conditions,” pp. 624-636.
SPE30766, paper presented at the 70th Annual Technical 23. Meunier, D. F., Kabir, C. S., and Wittmann, M. J., "Gas Well Test
Conference and Exhibition of the Society of Petroleum Engineers, Analysis: Use of Normalised Pressure and Time Functions",
Dallas, TX, U.S.A., October 22-25, 1995. SPEFE , Dec. 1987, p. 629.
10. Henderson, G. D., Danesh, A., Tehrani, D. H., Al-Shaidi, S., and 24. Chu, W.C., and Shank, G.D.: “A New Model for Fractured Well
Peden, J. M., “Measurement and Correlation of Gas-condensate in Radial Composite Reservoir,” SPE20579, paper presented at
Relative Permeability by the Steady-State Method”, SPE30770, the 65th Annual Technical Conference and Exhibition, Los
SPE Annual Technical Conference & Exhibition, Dallas, October Angles, September 23-26, 1990.
1995. 25. Al-Shaidi, S. M.:“Modelling of Gas-Condensate Flow in
11. Henderson, G.D., Danesh, A., Tehrani, D.H., Al-Kharusi, B.: Reservoir at Near Wellbore Conditions,” Thesis Submitted for the
"The Relative Significance of Positive Coupling and Inertial Degree of Doctor of Philosophy in Petroleum Engineering,
Effects on Gas Condensate Relative Permeabilities at High Department of Petroleum Engineering, Heriot-Watt University,
Velocity," paper SPE 62933 presented at the 2000 SPE Annual UK, August, 1997.
Technical Conference and Exhibition, Dallas, Texas, 1-4 Oct. 26. Eilerts, C. K., Sumner, E. F., and Potts, N.L.: “Intergration of
12. Ali J. K., McGauley, P. J., and Wilson, C. J.: "The Effects of High Prtial Differential Equation for Transient Radial Flow of Gas-
Velocity Flow and PVT Changes Near Wellbore on Condensate Condensate Fluids in Porous Structures,” Society of Petroleum
Well Performance," paper SPE 38923 presented at the 1997 SPE Engineering Journal (June, 1965), pp. 141-152.
Annual Technical Conference and Exhibition, 5-8 Oct. 27. El-Banbi A. H., McCain W. D. Jr., and Semmebeck M. E.:
13. Blom, S. M. P., Hagoort, J., and Soetekouw, D. P. N.: “Relative “Investigation of Well Productivity in Gas-Condensate
Permeability at Near-Critical Conditions,” SPE 38935, paper Reservoirs,” SPE59773, paper presented at the 2000 SPE/CERI
presented at the SPE Annual Technical Conference and Exhibition Gas Technology Symposium held in Calgary, Alberta Canada,
held in Antonio, Texas, March 5-8, 1997. April 3-5, 2000.
14. 14 Moore, T.F. and Slobod, R.L.:”Displacement of Oil by Water- 28. Vo, D. T., Jones, J. R., and Raghavan, J. R., "Performance
Effect of Wettability, Rate, and Viscosity On Recovery,” SPE Predictions for Gas Condensate Reservoirs", SPE Formation
paper 502, presented at the 30th Annual Fall Meeting of the Evaluation, 1989, vol. 4, no. 4, December.
Petroleum Branch of the AIME, New Orleans, La., Oct. 2-5, 29. Bozorgzadeh, M., and Gringarten, A. C., "New Estimate for the
1955. Radius of a Condensate Bank from Well Test Data Using Dry Gas
15. Bardon, C. and Longeron, D.G.:”Influence of very low interfacial Pseudo-Pressure", paper SPE 89904, presented at the SPE
tension on relative permeability,” SPEJ (Oct. 1980)391-401. Technical Conference and Exhibition held in Houston, Texas,
SPE 100993 17

U.S.A., 26–29 September 2004. 45. Horner, D. R.: "Pressure Build-ups in Wells", Proc., Third World
30. Bozorgzadeh, M.:”Characterisation and Determination of Gas Pet. Cong., E. J. Brill, Leiden (1951) II, 503-521. Also, Reprint
Condensate Dynamics from Pressure Transient Data and Fluid Series, No. 9 — Pressure Analysis Methods, Society of Petroleum
PVT Properties”, PhD thesis, Centre for Petroleum Studies, Engineers of AIME, Dallas ( 1967) 25-43.
Imperial College London, UK, Feb. 2006. 46. Daungkaew, S., Hollaender, F., and Gringarten A. C.: “Frequently
31. Baig, T., Droegemueller, U. and Gringarten, A.C.:”Productivity Asked Questions in Well Test Analysis,” SPE 63077, paper
Assessment of Fractured and Non-Fractured Wells in a presentatedt the 2000 SPE Annual Technical Conference and
Lean/Intermediate Low Permeability Gas Condensate Reservoir, “ Exhibition held in Dallas, Texas, October 1-4, 2000.
paper SPE 93136 presented at the 14th Europec Biennial 47. von Schroeter, T., Hollaender, F., Gringarten, A.:"Deconvolution
Conference held in Madrid, Spain, 13-16 June 2005. of Well Test Data as a Nonlinear Total Least Square Problem,"
32. Hashemi, A.:”Evaluation of Horizontal Gas-Condensate Wells SPE Journal (Dec., 2004)375-390.
Using Pressure Transient Analysis and Compositional 48. Bidaux, P., Whittle, T. M., Coveney, P. J. and Gringarten, A. C.:
Simulation," PhD Thesis, Centre for Petroleum Studies, Imperial "Analysis of Pressure and Rate Transient Data From Wells in
College London, UK, Feb. 2006. Multilayered Reservoirs,” SPE paper 24679 presented at the 67th
33. Raghavan, R.:"Well test Analysis for Multiphase Flow," paper Annual Technical Conference and Exhibition of SPE,
SPE 14098 presented at the 1986 SPE International Meeting on Washington, DC, Oct. 4-7, 1992. Implemented in InterpretTM
Petroleum Engineering, Beijing, China, 17-20 March. 2005, well test analysis software from Paradigm Geotechnology
34. Bozorgzadeh, M. and Gringarten, A.C. "Application of Build-Up B.V.
Transient Pressure Analysis to Well Deliverability Forecasting in 49. Bourdet, D.:”pressure Behavior of Layered Reservoirs with
Gas Condensate Reservoirs Using Single-Phase and Two-Phase Crossflow”, paper SPE 13628, presented at the Annual California
Pseudo-Pressures", paper SPE 94018, presented at the 14th SPE Regional Meeting, Bakersfield, California, march 27-29, 1985.
Europec Technical Conference and Exhibition held in Madrid, 50. Saphir, well test analysis software from Kappa Engineering
Spain, 13-16 June 2005. 51. Tang, T., and Firoozabadi, A.:" Relative Permeability
35. Forchheimer, P.: "Wasserbewegung durch Boden" ZVD1, (1901) Modification in Gas-Liquid Systems Through Wettability
(45) 1781. Alteration to Intermediate Gas-Wetting" SPE paper 62934, Proc.
36. Geertsma, J.: "Estimating the Coefficient of Inertial Resistance in 2000 SPE Ann. Tech. Conf. Exh., Soc. Pet. Eng., Dallas (2000).
Fluid Flow through Porous Media," SPEJ (Oct. 1974) 445. 52. Fahes, M. and Firoozabadi, A.: “Wettability Alteration ro
37. Henderson, G.D., Danesh, A., Tehrani, D.H. and Al-Kharusi, B.: Intermediate Gas-Wetting in Gas/Condensate Reservoirs at High
"Generating Reliable Gas Condensate Relative Permeability Data Temperatures,” paper SPE 96184 presented at the 2005 SPE
Used to Develop a Correlation with Capillary Number," Journal ATCE, Dallas, Texas,9-12 October.
of Petroleum Science and Engineering (2000) (25), 79. 53. Marokane, D., Logmo-Ngog, A.B., Sakar, R.: “Applicability of
38. App, J.F. and Mohanty, M.: "Gas and Condensate Relative Timely Gas Ingection in Gas Condensate Fields to Improve Well
Permeability at Near Critical Conditions: Capillary and Reynolds Productivity,” paper SPE 75147 presented at the SPE/DOE
Number Dependence," Journal of Petroleum Science and Improved Oil Recovery Symposium, Tulsa, Oklahoma13-17 April
Engineering , (2002) (36) 111-126. 2002.
39. Henderson, G. D., Danesh, A., Tehrani, D. H. and Al-Kharusi, B.: 54. Hashemi, A., Nicolas, L. and Gringarten, A.C.:“Well Test
"Effect of Positive Rate Sensitivity and Inertia on Gas Condensate Analysis of Horizontal wells in Gas-Condensate Reservoirs”,
Relative Permeability at High Velocity," Petroleum Geoscience, paper SPE 89905 presented at the 2004 SPE ATCE, Houston,
(2001) (7) 45-50. Texas, 26-29 September 2004; SPE Reservoir Engineering and
40. Jamiolahmady, M., Henderson, G. D., Danesh, A., Tehrani, D.H. Evaluation Journal, Feb. 2006.
and Al-Kharusi, B.: "Variation of Gas-Condensate Relative 55. Gringarten, A. C., Bourdet, D. P., Landel, P. A., and Kniazeff, V.
Permeability with Production Rate at Near Wellbore Conditions," J.: "A Comparison Between Different Skin and Wellbore Storage
paper SPE 83960 presented at the 2003 Offshore Europe, Type-curves for Early-time Transient Analysis," paper SPE 8025
Aberdeen, UK, 2-5 Sep. presented at the Fifty-fourth Annual Fall Technical Conference
41. Mott, R., Cable, A. and Spearing, M.: "Measurements and and Exhibition of SPE, Las Vegas, Nevada, Sept. 23-26, 1979.
Simulation of Inertial and High Capillary Number Flow 56. Aluko, O.:”Well Test Analysis of a North Sea Gas Condensate
Phenomena in Gas-Condensate Relative Permeability," paper SPE Reservoir,” MSc thesis, Centre for Petroleum Studies, Imperial
62932 presented at the 2000 the SPE 75th Annual Technical College London, UK, Sept. 2003.
Conference and Exhibition of the Society of Petroleum Engineers 57. Abdolnabi Hashemi and Alain C. Gringarten: “Comparison of
held in Dallas, TX, 1-4 Oct. Well Productivity between Vertical, Horizontal and Hydraulically
42. Mott, R.: "Engineering Calculations of Gas-Condensate-Well Fractured Reservoirs in Gas Condensate Reservoirs” paper SPE
Productivity," SPEREE (Oct.2003) 298-306. 94178 presented at the 14th Europec Biennial Conference, Madrid,
43. Gringarten, A.C.:“From Straight-lines to Deconvolution: the Spain, 13-16 June 2005.
Evolution of the State-of-the-Art in Well Test Analysis,” paper 58. Baig, M.T.A.:”Productivity Assessment of Fractured and Non-
SPE 102079 presented at the 2006 SPE Annual Technical Fractured Gas Condensate Wells,” MSc thesis, Centre for
Conference and Exhibition held in San Antonio, Texas, U.S.A., Petroleum Studies, Imperial College London, UK, Sept. 2003.
24–27 September 2006.
44. Daungkaew, S.:" New development in Well Test Analysis," PhD
Thesis, Centre for Petroleum Studies, Imperial College London,
UK, October 2002.

You might also like