SPE 93986

Evaluation of Two-Phase IPR Correlations for Horizontal Wells

R. Kamkom and D. Zhu, Texas A&M U.

Copyright 2005, Society of Petroleum Engineers

other correlations reviewed in this paper.
This paper was prepared for presentation at the 2005 SPE Production and Operations The modified Vogel correlation for two-phase flow
Symposium held in Oklahoma City, OK, U.S.A., 17 – 19 April 2005.
horizontal wells is simple compared with the other models,
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal submitted by the author(s). Contents of the paper, as
and works as well as the other models at the normal flow
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to condition. It can be used to estimate horizontal well
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 performance conveniently.
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 Introduction
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
words; illustrations may not be copied. The proposal must contain conspicuous To predict inflow performance in a vertical well with
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.
analytical models, correlations have to be used to avoid the
complexity caused by relative permeabilities. The most
Abstract commonly used models are Vogel’s correlation1 and
To predict well performance for horizontal wells, several Fetkovitch’s correlation2. Inflow performance relationships of
analytical models have been developed to generate inflow horizontal wells are different than the ones of vertical wells.
performance relationships (IPR) for single-phase wells. The two most pronounced factors are; first, the flow
Similar to vertical wells, analytical solutions of IPR in two- streamline is a combination of radial flow and linear flow with
phase flow horizontal wells are unavailable because of the linear flow dominating; and second, the inflow performance
complexity of relative permeability and the variable phase depends not only on horizontal permeability, but also on
distribution in the reservoir. Correlations, lead by Vogel’s vertical permeability; and therefore the anisotrop ratio of the
equation, have been used for two-phase IPR calculations for reservoir becomes important when modeling a horizontal well
vertical wells, and the methods have been adopted for performance. These yield additional difficulty to obtain
horizontal well applications. analytical models of horizontal well inflow performance.
This paper reviews several two-phase correlations for Today, long horizontal wells and multilateral wells are more
horizontal wells (Bendakhlia and Aziz’s equation, Cheng’s and more widely used. In such wells, wellbore hydrodynamics
equation, and Retnanto and Economides’ equation). Instead of plays an important role in well performance. Using single-
using the original formula of Vogel’s equation, these models phase inflow relationship to predict two-phase flow well
include the effects of reservoir recovery factor or fluid bubble- performance can result in significant deviation in flow rate and
point pressure on IPR behaviors. We also presented a pressure distribution in the wellbore, and deliver misleading
modified Vogel’s correlation for two-phase horizontal IPR information for well performance management.
following Kabir’s work. The modified correlation uses Babu There have been two-phase correlations published before
and Odeh’s single-phase horizontal well model to calculate the to predict the potential productivity in two-phase flow
absolute open flow potential, qo,max, and then uses the original horizontal wells. We review several two-phase IPR
Vogel’s correlation to generate the relationship between flow correlations for horizontal well in this paper. The models
rate and pressure. The reservoir anisotropy is considered in reviewed in this paper were all following Vogel’s approach for
Babu and Odeh’s model for horizontal well flow. The results two-phase vertical wells, and modifying the parameters in the
of the modified Vogel’s equation are compared with the correlations to capture the effects of fluid properties and
reviewed two-phase correlations, and also checked against the reservoir recovery factor on the inflow performance. We will
results from a reservoir simulator (Eclipse™). It is found that compare and discuss these models with the original Vogel’s
the modified Vogel’s equation results in close predictions of formula and reservoir simulation results.
two-phase inflow performance for horizontal wells at normal
recovery factors compared with the reservoir simulation Two-Phase Inflow Models for Horizontal Wells
results. At low recovery factor, it deviates from the simulation In 1968, Vogel presented the empirical equation to estimate
results more than some other correlations reviewed. For low two-phase inflow performance relationships in vertical wells.
bubble-point pressure fluids, flow conditions in the reservoir This empirical equation is given as
2
are more likely to be single-phase rather than two-phase, and qo ⎛ p wf ⎞ ⎛p ⎞
gradually tranforms from single-phase to two-phase flow as = 1 − 0.2⎜⎜ ⎟ − 0.8⎜ wf
⎟ ⎜ p
⎟
⎟ (1)
qo,max ⎝ p ⎠ ⎝ ⎠
the reservoir pressure declines. The modified Vogel’s inflow
equation does not capture this transient period; neither do the
2 SPE 93986

2
where p wf and p are flowing bottomhole pressure and average qo ⎛ p wf ⎞ ⎛p ⎞
= 0.9885 + 0.2055⎜⎜ ⎟ − 1.1818⎜ wf
⎟ ⎜ p
⎟
⎟ (3)
reservoir pressure respectively. qo, max is the oil production qo,max ⎝ p ⎠ ⎝ ⎠
rate at the maximum drawdown or at zero bottomhole flowing
pressure for single-phase oil flow. This equation was obtained Cheng’s reservoir simulation model used a rectangular
from curve fitting of dimensionless inflow performance bounded reservoir with a slanted or horizontal well at the
relationship (the plot of dimensionless oil flow rate, middle of the reservoir. The reservoir was assumed
qo q o,max , versus dimensionless pressure, p wf p ). The data homogeneous and isotropic with constant water saturation.
Initially, there were oil and immobile water in the reservoir.
of dimensionless IPR were generated by a computer program6
The well was produced at pseudo-steady state condition.
for solution gas drive reservoirs. The reservoir model was a Cheng’s model kept the same exponents for the pressure ratio
cylindrical bounded reservoir with a fully penetrated well at
the middle of the reservoir. The reservoir was homogeneous ( )
terms p wf p as Vogel’s correlation, but modified the
and isotropic with constant water saturation. The effects of constants in the equation to fit the reservoir simulation results
gravity and the compressibility of formation and reservoir better.
fluid were neglected. To simulate two-phase flow conditions,
the initial pressure was set equal to or below the bubble-point Retnanto and Economides Correlation5. Retnanto and
pressure of the reservoir fluid. By running the reservoir Economides presented a model to estimate the two-phase
simulation model, Vogel obtained the data and plotted IPR inflow performance in horizontal wells in 1998. They used the
curves at different constant recovery factors. These plots numerical simulator, VIP8, to generate the reservoir inflow
showed the same trend of IPR behavior at different recovery performance for horizontal and multilateral wells. The model
factors. Vogel extended the study over a wide range of was run over an extended range of fluid and reservoir
conditions and obtained the empirical relationship (Eq. 1). properties, and the dimensionless IPR curves were created at
The correlation has been used widely and successfully to the simulated conditions. By applying nonlinear regression
estimate two-phase inflow performance relationship in vertical techniques to dimensionless IPR curves, the empirical
wells. equation that fitted the simulation results was given by

Bendakhlia and Aziz Correlation3. Bendakhlia and Aziz n

qo ⎛ p wf ⎞ ⎛p ⎞
followed Vogel’s approach to develop a model for horizontal = 1 − 0.25⎜⎜ ⎟ − 0.75⎜ wf
⎟ ⎜ p
⎟
⎟ (4)
wells. They presented the empirical equation to calculate two- qo,max ⎝ p ⎠ ⎝ ⎠
phase flow in horizontal wells as
where,
n
⎛ ⎛ p wf ⎞ ⎛p ⎞
2⎞

= ⎜⎜1 − V ⎜⎜ ⎟ − (1 − V )⎜ wf ⎟
qo
⎟ ⎛ ⎛ p ⎞ ⎞⎟
2

( )
⎟ ⎜ p ⎟ ⎟⎟ (2) ⎛ p ⎞
qo,max ⎜ ⎜
n = ⎜ − 0.27 + 1.46⎜⎜ ⎟ − 0.96⎜ ⎟ −3
⎝ ⎝ p ⎠ ⎝ ⎠ ⎠ ⎜ p ⎟ ⎜ p ⎟ ⎟⎟ 4 + 1.66 × 10 pb
⎝ ⎝ b ⎠ ⎝ b ⎠ ⎠
They modified Vogel’s equation by replacing the constants, (5)
0.2 and 0.8, and the exponent in Eq. 1 with parameters, V and
n. These parameters, V and n, as shown in Fig. 1, are functions Retnanto and Economides’ correlation modified both the
of reservoir recovery factor. This equation was obtained from exponents and the constants in the original Vogel’s equation to
curve fitting of simulated dimensionless IPR curves. The address the issue of effects of fluid properties and reservoir
simulated data were generated by a three-phase black-oil conditions on two-phase inflow performance. Eq. 5 shows that
simulator, IMAX. The reservoir simulation used a box-shaped n is a function of the bubble-point pressure of the reservoir
drainage area with a horizontal well in the middle of the fluid. In the normal range of bubble-point pressure, as the
reservoir. The well was fully penetrated through the reservoir. bubble-point pressure increases, the exponent n increases,
The reservoir was assumed homogeneous and isotropic. which implies that non-linear behavior of two-phase flow
Capillary pressure was neglected. The initial pressure was set would be more pronounced. However, this model has a
at the bubble point pressure. The simulation model was run limitation of the range that can be applied. Table 1 lists the
with a wide range of fluid properties and reservoir conditions.
The method of obtaining dimensionless IPR curves was
( )
value of n as a function of the ratio of p / p b . It shows that

similar to Vogel’s algorithm. The dimensionless IPR curves at low p / pb ratio, n could be less than one, and in some cases,
were different at each recovery factor because V and n are it even becomes negative. As consequence, the predicted
dependents of reservoir recovery factors. dimensionless IPR is meaningless in these cases. Fig. 2 shows
the dimensionless IPR curves with n less than one or negative.
Cheng’s Correlation4. Cheng generated an equation to
calculate well deliverability for slanted wells. His simulation Kabir’s work9. Kabir presented a method to estimate the
data were obtained from the NIPER simulation model7 for absolute open flow potential or q o,max for horizontal wells. To
slanted and horizontal wells. His empirical equation was calculate q o,max , he differentiated the dimensionless inflow
presented as,
SPE 93986 3

performance relationship and calculated the q o,max in term of reviewed in this paper. The reservoir simulator used in the
productivity index. The differentiation of Vogel’s Equation study is ECLIPSE13. The reservoir model for the simulation is
(Eq. 1), Bendakhlia and Aziz’ Equation (Eq. 2) and Cheng’s a box-shaped reservoir with a 2000-foot length in the x-
direction, a 2000-foot width in the y-direction, and a 150-foot
Equation (Eq. 3) with respect to p wf are given as
height in the z-direction (Fig. 3). The horizontal well is in the
middle of the reservoir. The wellbore is 1000 feet long and has
⎛ p wf ⎞ a wellbore radius of 0.25 feet. The reservoir is homogeneous
= q o,max ⎜ 0.2 + 1.6 2 ⎟
dq o 1
−
()
(6) and anisotropic. The reservoir porosity is 25%, and the
dp wf ⎜ p ⎟
⎝ p ⎠ permeability is 100 md in the x- and y- direction, and 10 md in
the z- direction. The rock compressibility is 4x10-5 psi -1. The
n −1 initial reservoir pressures used in the study are 3815 psi and
⎡ ⎛ p wf ⎞ ⎛ p wf ⎞ ⎤
2
5600 psi, which are also set as the bubble-point pressure of the
dq o
− ⎢
= q o,max n 1 − V ⎜ ⎟ − (1 − V )⎜ ⎟ ⎥
dp wf ⎢ ⎜ p ⎟ ⎜ p ⎟ ⎥ reservoir fluid.
⎣ ⎝ ⎠ ⎝ ⎠ ⎦ The IPR curves were generated from simulation results at
(7)
different recovery factors. For each recovery factor, the
⎡ 1 p wf ⎤
× ⎢V + 2(1 − V ) ⎥ simulation was run with pressure constraint to develop one set
⎢ p
⎣ p
2⎥
⎦() of IPR curves. Then the data were presented in a
dimensionless form. The pressure was normalized by the

⎛ ⎞
( )
average reservoir pressure p wf p , and the flow rate was
−
dq o 1 p wf
= q o,max ⎜ − 0.2055 + (2 × 1.1818) 2 ⎟ normalized by maximum flow rate at zero pwf (q o q o,max ) .
()
(8)
dp wf ⎜ p ⎟
⎝ p ⎠
Results and Discussions
where (− dq o dp wf ) is defined as productivity index, J , Figures 4 and 5 show the simulation results of IPR curves and
dimensionless IPR curves at different recovery factors
which has the maximum value when p wf is equal to p . Thus, respectively. From Fig. 4 we can see that for the specified
the maximum productivity index for each model can be case, 10% recovery factor is about the upper limit of recovery
written as factor to retain the constant flowing bottomhole pressure
constraint. The flow rate at 10% recovery factor is much lower
compared with the other curves. The dimensionless IPR
⎛ 1⎞
J = q o,max ⎜⎜1.8 ⎟⎟ (9) curves (Fig. 5) show that the IPR curves with different
⎝ p⎠ recovery factors follow the same trend. However, at 10%
recovery factor, the dimensionless curve is slightly deviated
J =0 (for n ≠ 1) (10)
from the other curves. This is caused by producing in a highly
depleted reservoir.
The reviewed correlations, modified Vogel’s, Bendakhlia
⎛ 1⎞ and Aziz’s, Cheng’s, and Retnanto and Economides’ are then
J = q o,max ⎜⎜ 2.1581 ⎟⎟ (11) compared against the simulation results. The dimensionless
⎝ p⎠
IPR curves from the simulator and correlations at 1%, 5% and
10% recovery factor are presented in Figs. 6, 7 and 8
Obviously, this approach cannot be applied to Bendakhlia and respectively.
Aziz’ model (Eq. 7). Kabir suggested that q o,max can be Figure 6 shows the results of dimensionless IPR at 1%
evaluated by using Eq. 9 where J is calculated from analytic recovery factor. From the plot we can see that Bendakhlia &
expression for single-phase flow in horizontal wells such as Aziz’s model gives the closest results to the simulated data.
Joshi10, Kuchuk11 and Babu and Odeh12. This model yields close estimation of flow rate as a function
Following Kabir’s suggestion, Vogel’s equation is of pressure until the dimensionless pressure is about 0.70.
modified for horizontal wells application. Babu and Odeh’s Beyond this point, the model estimates higher oil production
model is used to calculate qo,max, and then Eq. 1 is used to rates than the simulation result, like all of the other
generate IPR curves for two-phase flow horizontal wells. The correlations. At the dimensionless pressure of 0.5, the
advantage of modified Vogel’s model for horizontal wells is dimensionless oil flow rate from the simulated data is 0.59.
that it is simple, and it requires the least information to use the The calculated dimensionless rates are 0.65, 0.70, 0.74 and
model. Notice that the modified Vogel’s model does not 0.79 obtained from Bendakhlia and Aziz’s, Vogel’s, Retnanta
explicitely consider the effect of recovery factor or bubble- and Economides’ and Cheng’s correlation respectively.
point pressure on IPR behavior. Bendakhlia and Aziz’s correlation, as the closest correlation,
deviated 10% from the simulation result. Cheng’s correlation
Reservoir Simulation Model results in a 35% difference which is the highest compared with
To compare the correlations discussed in this paper, we used a other correlations. Vogel’s correlation has 18% difference, and
reservoir simulation model to generate dimensionless IPR Retnanto and Economides’ correlation is 26% deviated.
results for horizontal wells to evaluate the correlations As the recovery factor increases, Vogel’s correlation
gives better result than the other correlations. Fig. 7 shows the
4 SPE 93986

dimensionless IPR curves from the correlations and the Conclusions

simulation. When the dimensionless pressure is above 0.85, Several two-phase IPR correlations for horizontal wells were
Vogel’s correlation is almost the same as the simulated curve. reviewed, and their accuracies were evaluated by comparing
Comparing the result at dimensionless pressure of 0.5, the the calculated results from the correlations against the
simulation results for dimensionless rate is 0.62, and the rate reservoir simulation data. From this study, it is found that
from the Vogel’s correlation is 0.70, resulting in a difference 1. Different correlation works better at different
of 12%. The differences from the simulation result at this conditions. Bendakhlia and Aziz’s correlation yields
point are 18%, 22% and 27% for Bendakhlia and Aziz’s, the closest results to the simulation result at high or
Cheng’s, and Retnanta and Economides’ correlation very low reservoir recovery factors, while the
respectively. modified Vogel’s equation does so at the recovery
At 10% recovery factor (Fig. 8), Cheng’s correlation factors in between.
gives the closest result to the simulated curve at high 2. Cheng’s equation can be used when the reservoir
dimensionless pressures (>0.5). The deviation of Cheng’s recovery factor is high, and deviates from the
correlation increases as the dimensionless pressure decreases. simulation results as high as 35% at low recovery
At low dimensionless pressure, Bendakhlia and Aziz’s factors.
correlation gives the closest curves to the simulated curve. 3. At high bubble-point pressure, all of the reviewed
However, the curve obtained from Vogel’s correlation is fairly correlations deviate from the simulation results more
close to Bendakhlia and Aziz’s correlation, and yields a better significantly than at low bubble-point pressure.
result at very low dimensionless pressure. Table 2 shows the 4. In general, the modified Vogel’s model can be used to
summarized results of the differences between the correlations estimate the two-phase inflow performance in
and the simulation at dimensionless pressure of 0.5. It shows horizontal wells with a reasonable accuracy.
from Table 2 that either Vogel’s correlation or Bendakhlia and
Aziz’s correlation work well in general. When recovery factor Nomenclature
is high, Cheng’s correlation should be used. J = productivity index, STB/psi
Bubble-point pressure of fluids is one of the parameters pb = bubble point pressure, psi
that affect the IPR behavior of a two-phase flow system. The
comparison of the dimensionless IPR curves at a bubble-point p = average reservoir pressure, psi
pressure of 5600 psi is shown in Figs. 10 and 11 for 5% and p wf = flowing bottomhole pressure, psi
10% recovery factor respectively. These plots reveal that at qo = oil flow rate, STB
high bubble-point pressure, all of the correlations reviewed
deviate more significantly from the simulation result q o , max = absolute open flow potential, STB
compared with at the low bubble-point pressure results. V, n = parameters in Bendakhlia and Aziz model
Vogel’s correlation would be the closest one compared with
the other correlations, and should be used to predict inflow References
performance of horizontal wells. At high recovery factor, the 1. Vogel, J. V.: “Inflow Performance Relationships for Solution-
correlations should be used with caution. Gas Drive Wells,” Journal of Petroleum Technology, (January,
All of the reviewed correlations for two-phase inflow 1968), 83-92.
2. Fetkovich, M.J.: “The Isochronal Testing of Oil Wells,” Paper
performance relationships for horizontal wells have the similar
SPE 4529, presented at the 1973 SPE Annual Fall Meeting, Las
formula with different constants and/or exponent in the Vegas, NV, Sept. 30- Oct 3.
equations. For Bendakhlia and Aziz’ model, the parameter n 3. Bendakhlia, H., and Aziz, K.: “Inflow Performance
varies from 0.9 to 1.4, and V varies from 0.1 to 0.45 (Fig. 1). Relationships for Solution-Gas Drive Horizontal Wells,” Paper
Although these parameters change as the recovery factor SPE 19823, presented at the 64th Annual Technical Conference,
changes, Figs. 7 and 8 show that Bendakhlia and Aziz’ San Antonio, Texas, October 8-11, 1989.
correlation gives very close results to Vogel’s correlation, 4. Cheng, A.M.: “Inflow Performance Relationships for Solution-
which does not consider the effect of recovery factor, Gas-Drive Slanted/Horizontal Wells,” Paper SPE 20720,
presented at the 65th Annual Technical Conference, New
especially at high recovery factors. Cheng’s model deviates Orleans, Louisiana, September 23-26, 1990.
the most compared with the simulated results at low recovery 5. Retnanto, Albertus, and Economides, J. Michael: “Inflow
factors. In general, the model over-estimates the oil production Performance Relationships of Horizontal and Multibranched
rate at low recovery factors. The deviation is corrected in Wells in a Solution-Gas-Drive Reservoir,” Paper SPE 50659,
certain dimensionless pressure range at higher recovery presented at the 1998 European Petroleum Conference, Hague,
factors. Retnanto and Economides’ model considered the Netherlands, October 20-22.
effect of bubble-point pressure on IPR. However, the model is 6. Weller, W.E.: “Flowing and Gas-Lift Well Performance,”
limited to certain ranges of the bubble point pressure. It also Journal of Petroleum Technology, (February, 1966), 240-246.
shows relatively higher deviation from the simulation results 7. Cheng, M.M.: “Simulation of Production from Wells with
Horizontal/Slanted Laterals,” Department of Energy Report No.
for the conditions studied in this paper, compared with the NIPER-328 (Revised, October, 1988).
other reviewed models. Among the reviewed correlations, 8. VIP Manual Version 3.31, Western Atlas Software, Houston
modified Vogel’s correlation is convenient to use. Even (1996).
without additional consideration of fluid and reservoir 9. Kabir, C.S.: “Inflow Performance of Slanted and Horizontal
properties, it gives reasonable results compared to the Wells in solution-Gas-Drive Reservoir,” Paper SPE 24056,
simulation data at the conditions studied. presented at the 1992 SPE western regional meeting,
SPE 93986 5

Bakersfield, California, March 30-April 1.

10. Joshi, S.D.: “Augmentation of Well Productivity Using Slanted
and Horizontal Wells,” Journal of Petroleum Technology, (June
1988), 729-739.
11. Kuchuk, F.J. et al.: “Pressure Transient Analysis and Inflow
Performance for Horizontal Wells,” Paper SPE 18300, presented
at the 1988 SPE Annual Technical Conference and Exhibition,
Houston, Texas, October 2-5.
12. Babu, D.K. and Odeh, Aziz S.: “Productivity of a Horizontal
Well,” SPE Reservoir Engineering, (November 1989), 417-421.
13. Eclipse 100 Version 2004A_1 User Manual, Schlumberger-
GeoQuest.

Table 1. Evaluation of parameter n

pb=1000 pb =2000 pb =3000 pb =4000 pb =5000

pr /pb n n n n n

0.00 -1.5282 -1.9764 -2.4246 -2.8728 -3.3210

0.05 -1.1286 -1.4596 -1.7906 -2.1216 -2.4526
0.10 -0.7562 -0.9780 -1.1997 -1.4215 -1.6433
0.15 -0.4109 -0.5314 -0.6519 -0.7725 -0.8930 Figure 1: Parameter V and n versus recovery factor
0.20 -0.0928 -0.1200 -0.1473 -0.1745 -0.2017 Bendakhlia & Aziz2

0.25 0.1981 0.2562 0.3143 0.3724 0.4305

0.30 0.4619 0.5973 0.7328 0.8682 1.0037
0.35 0.6984 0.9033 1.1081 1.3130 1.5178
1
0.40 0.9079 1.1741 1.4404 1.7067 1.9729
0.45 1.0901 1.4098 1.7295 2.0493 2.3690 0.9

0.50 1.2452 1.6104 1.9756 2.3408 2.7060 0.8

0.55 1.3731 1.7758 2.1785 2.5813 2.9840
0.7
0.60 1.4739 1.9061 2.3384 2.7707 3.2029
0.65 1.5474 2.0013 2.4551 2.9090 3.3628 0.6

pwf / pr
0.70 1.5939 2.0613 2.5288 2.9962 3.4637
n=1 0.5
0.75 1.6131 2.0862 2.5593 3.0324 3.5055 n = 0.5
0.80 1.6052 2.0760 2.5467 3.0175 3.4883 n = -0.1 0.4

0.85 1.5701 2.0306 2.4911 2.9515 3.4120 n = -0.3

0.3
0.90 1.5078 1.9500 2.3923 2.8345 3.2767
0.2
0.95 1.4184 1.8344 2.2504 2.6664 3.0824
1.00 1.3018 1.6836 2.0654 2.4472 2.8290 0.1

0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Table 2. Deviation of correlations from simulation
qo/qo,max
(% difference at dimensionless pressure =0.5)

Recover Vogel Bendakhlia Retnanto Cheng

Factor & Aziz &Economides
Figure 2: Dimensionless IPR Curve of Retnanto and Economides’
1% 18 10 26 35
Model at different n
5% 12 18 22 27
10% 6 5 15 6
6 SPE 93986

2000 ft 0.9
Z
Y
0.8

0.7
X
0.6
1000 ft

pwf / pr
2000 ft 0.5 1%

0.4 3%

0.3 5%

0.2 8%
150 ft 10%
0.1

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
qo /qo,max

Figure 3: The horizontal model for simulation data Figure 5: The dimensionless IPR curves of 1%, 3%, 5%, 8% and
10% recovery factors

4000

1%
3500
3%
3000
5% 1

2500 8% 0.9
pwf (psi)

10% 0.8
2000
0.7
1500
0.6
pwf /pr

1000
0.5

500 0.4

0.3 Eclipse
0
Vogel
0 10000 20000 30000 40000 50000 60000 0.2 Bendakhlia and Aziz
qo (STB/DAY) Cheng
0.1 Retnanto and Economides
Figure 4: The IPR curves of 1%, 3%, 5%, 8% and 10% recovery 0
factors 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
qo/qo,max

Figure 6: The dimensionless IPR curves of 1% recovery factor

SPE 93986 7

1 1

0.9 0.9

0.8 0.8

0.7 0.7

0.6 0.6
pwf /pr

pwf / pr
0.5 0.5
1%
0.4 0.4
3%
0.3 Eclipse 0.3 5%
Vogel
8%
0.2 Bendakhlia and Aziz 0.2
Cheng 10%
0.1 Retnanto and Economides 0.1 Vogel

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
qo/qo,max qo /qo,max

Figure 7: The dimensionless IPR curves of 5% recovery factor Figure 9: The comparison of dimensionless IPR curves of Vogel
with different recovery factors
1

0.9

0.8

0.7 1

0.6 0.9
pwf /pr

0.5 0.8

0.4 0.7

0.3 Eclipse 0.6

Vogel
pwf /pr

0.2 Bendakhlia and Aziz 0.5

Cheng
0.1 Retnanto and Economides 0.4
Eclipse
0 0.3
Vogel
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0.2 Bendakhlia and Aziz
qo/qo,max Cheng
0.1 Retnanto and Economides
Figure 8: The dimensionless IPR curves of 10% recovery factor 0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
qo/qo,max

Figure 10: The comparison of dimensionless IPR curve of

Vogel with 5% recovery factor at 5600 and 3815 psi bubble
point pressure
8 SPE 93986

0.9

0.8

0.7

0.6
pwf /pr

0.5

0.4

0.3 Eclipse
Vogel
0.2 Bendakhlia and Aziz
Cheng
0.1 Retnanto and Economides

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
qo/qo,max

Figure 11: The comparison of dimensionless IPR curve of

Vogel with 10% recovery factor at 5600 and 3815 psi
bubble point pressure