You are on page 1of 8

Society of Petroleum Engineers

SPE 71572

Transient Well Index for Numerical Well Test Analysis


Rosalind A. Archer, SPE, Texas A&M University, and Tabiat T. Yildiz, SPE, Texas A&M University

Copyright 2001, Society of Petroleum Engineers, Inc. grids (e.g. 15 cells by 15 cells, 167 feet by 167 feet). When
This paper was prepared for presentation at the SPE Annual Technical Conference the same simulations were performed using the Peaceman
and Exhibition held in New Orleans, Louisiana, 30 September - 3 October 2001. well index, the simulated pressure transients showed sig-
This paper was selected for presentation by an SPE Program Committee follow- nificant artifact wellbore storage, especially in low perme-
ing review of information contained in an abstract submitted by the author(s).
Contents of the paper, as presented, have not been reviewed by the Society of ability reservoirs.
Petroleum Engineers and are subject to correction by the author(s). The material,
as presented, does not necessarily reflect any position of the Society of Petroleum
Engineers, its officers, or members. Papers presented at SPE meetings are subject
to publication review by Editorial Committees of the Society of Petroleum Engi- Background
neers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Peaceman1 initially proposed the use of a well index in
Engineers is prohibited. Permission to reproduce in print is restricted to an ab-
stract of not more than 300 words; illustrations may not be copied. The abstract reservoir simulators. His later work2− 6 made provision
must contain conspicuous acknowledgment of where and by whom the paper was for a variety of reservoir and well geometries. Babu et al.7
presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836,
U.S.A., fax 01-972-952-9435. presented a well index relationship for horizontal wells and
wells at arbitrary locations within grids. Mochizuki8 con-
sidered well indices for arbitrarily inclined wells. Chen et
Abstract al.9, 10 developed a pseudo-skin factor to accurately pre-
Incorporating well test data into integrated reservoir char- dict well pressure and productivity for wells with various
acterizations may require the well test to be modeled in a inclinations. Sharpe and Ramesh11 modified the Peace-
reservoir simulator that also forward simulates other dy- man well model, restoring its validity for problems with
namic data being matched. Care is required to ensure the non-uniform local grid refinement. Sharpe and Ramesh
simulator is not introducing unnecessary numerical arti- also introduced a modified Peaceman well index suitable
facts. Standard finite-difference reservoir simulators use for modeling problems dominated by vertical flow such as
the Peaceman well index, which is based on the solution gas and water coning. Ding et al.12 presented a new rep-
to single-phase, steady-state, incompressible flow. For a resentation of wells in reservoir simulators which is par-
pressure transient test the assumptions of steady-state, in- ticularly applicable to non-uniform grids and leads to im-
compressible flow are not applicable. However instead of provements in the calculation of the productivity index.
modifying the well index it is common to account for this Wan et al.13 compared horizontal well performance simu-
by using highly refined grids around wells and making very lated by using a uniform coarse grid, a uniform fine grid,
careful choices of timestep sizes to model well tests. and a non-uniform fine grid using Peaceman’s well index
This work presents a new well index formulation that model. Wan et al. found that when a coarse grid was
allows well tests to be simulated accurately in finite- used to simulate a partially penetrating horizontal well,
difference simulators using uniform, relatively coarse grids, the flow rate of the well was under-predicted. None of the
without the problem of artifact wellbore storage (early works cited (except a brief mention by Peaceman1) specif-
time unit slope on the pressure derivative only) that oth- ically address the problem of fully transient flow such as
erwise occurs. The well index model computes the average the flow occuring in pressure transient tests.
pressure in the well block directly from the analytical so- Integrated reservoir characterizations frequently re-
lution for infinite acting radial flow. This approach can quire reservoir simulation models to accurately reproduce
also be applied to non-square gridblocks and anisotropic dynamic data such as pressure transient tests, production
reservoirs. data and long term pressure measurements. When well
Well tests simulated using finite-difference simulation tests are interpreted in isolation the measured data are
with the proposed transient well index and the Peaceman compared to analytically generated models to determine
well index are compared to analytical well test solutions. the reservoir properties. The catalog of models well test
The simulated well tests using the proposed transient well software has available for this process is limited to simple
index closely follow analytical solutions - even on coarse geometries such as square and circular reservoirs.
2 TRANSIENT WELL INDEX FOR NUMERICAL WELL TEST ANALYSIS SPE 71572

Conversely, finite-difference simulators can handle any Inflow into the well is controlled by the difference be-
geometry but cannot always accurately reproduce known tween the average block pressure and the wellbore pres-
pressure transient solutions because of limitations in the sure. We propose a model in which the grid block pres-
well index models available and discretization errors caused sure is taken to be the analytically determined average
by gridding and timestepping. This study proposes a tran- pressure. Therefore:
sient well index model so that pressure transient tests can
be accurately simulated in reservoir simulators for both q = W I(p̄B − pwf ) (8)
well test analysis and integrated reservoir characterization
purposes. If the simulator replicates the analytical solution then
The simulated well test examples presented in this study the average well block pressure is:
are for square shaped reservoirs. In these cases the sim-
ulated pressure transients (Eclipse14) can be compared  xB
D  yB
D  t2D  2 
1 r
4 1
2
1
2
t1D 2 Ei − 4tDD dtD dxD dyD
with analytical solutions generated by well test software
p̄B = (9)
(Saphir15). The well index models proposed in this study B B −
(xD yD 2 )(t2 − t1 )
πrwD D D
will allow engineers to simulate pressure transient tests
accurately in arbitrarily shaped reservoirs. where t1D is the dimensionless time at the start of the
timestep being considered, t2D is the dimensionless time
at the end of the timestep, xBD is the dimensionless length
Theory B
of the well block in the x-direction and yD is the dimen-
Peaceman Well Index Model Peaceman1 derived the sionless length of the well block in the y-direction.
following well known equivalent wellbore radius expres- The integrals in Eq. 9 are not of forms that can be
sion:  π computed analytically. In this study they were performed
ro
= exp − ≈ 0.208 (1) numerically using Mathematica17. Once the average block
∆x 2
pressure for a time step has been computed the well index
For non-square grid blocks, with any aspect ratio, the can then be calculated:
equivalent wellbore radius for an isotropic permeability is
given by: kh
WI = (p̄B − pD,wf ) (10)
1
ro = 0.14(∆x2 + ∆y 2 ) 2 (2) 141.2µB D

For an anisotropic permeability and non-square grid where pD,wf is the dimensionless flowing wellbore pres-
blocks, the equivalent wellbore radius is given by: sure generated from the analytical solution given in Eq. 4.
 1/2 Extension to Anisotropic Reservoirs The formula-
(ky /kx )1/2 ∆x2 + (kx /ky )1/2 ∆y 2
ro = 0.28 (3) tion available Eq. 9 and Eq. 10 is for an isotropic reser-
(ky /kx )1/4 + (kx /ky )1/4 voir. To incorporate anisotropic permeability distances are
rescaled in the following manner:
Direct Calculation of Average Block Pressure As
an alternative to computing ro and then using it to de-   14
ky
fine the well index the proposed well index formulation C= (11)
assumes the well block pressure is representative of the kx
average well block pressure16. This is computed directly

from the solution for infinite acting radial flow18 (Eq 4). 
rD = (xD × C)2 + (yD /C)2 (12)
This treatment allows fully transient flow (and compress-
ible fluids) to be handled. The model is derived in terms 
of dimensionless pressure, distance and time given in Eq. kx ky t
5 to Eq. 7: tD = 2
(13)
φµct rw
   
1 r2 1 2
rD
pD (rD , tD ) = Ei − D2 (4) 
pD (rD , tD ) = Ei − (14)
2 4tD 2 4tD
where yielding:
kh
pD = (pi − pwf ) (5) xB ×C yB  2 
141.2qBµ  D  D  t2D 1 r
4 1
2 2×C
1 t1D 2 Ei − 4tDD dtD dxD dyD
0.000264kt p̄B = (15)
tD = (6) (xB B 2 2 1
D yD − πrwD )(tD − tD )
2
φµct rw
r Once p̄B is calculated the well index is calculated in
rD = (7) the same manner as before.
rw
SPE 71572 ROSALIND A. ARCHER and TABIAT T. YILDIZ 3

Results was presented in Table 2. Six well tests were simulated


Square Grid Blocks and Isotropic Permeability To to test the performance of the transient well index. The
compare the proposed transient well index to the Peace- results are shown in Fig. 7 to Fig. 10.
man well index a numerical simulation study was per- These models used permeabilities of 1 md, 10 md, and
formed. The reservoir is a square homogenous reservoir 100 md in the x and y-direction and gridding of 51 × 17
containing single-phase oil, and the well is located at the cells, 51 × 51 cells and 51 × 153 cells. The results show
center of the reservoir. The distance from the well to each that well tests simulated using the transient well index
boundary is 1,225 ft. There are no wellbore storage or skin match the corresponding analytical solutions much bet-
effects in the simulation model. The simulation model was ter than the well tests simulated using the Peaceman well
run for a total producing period of 1000 hours, at a con- index.
stant rate of 100 STB/D of oil. The four gridding varia-
tions used for this areal model are shown in Table 1. In Well Test Analysis of Simulated Well Tests Each
each case the grid is uniform and square. Petrophysical simulated well test was imported into a well test analysis
and fluid parameters are given in Table 2. package and analyzed as if it were data. Example anal-
The time stepping scheme for the simulation study was yses are given in Table 8 to Table 11. The permeability
arranged so that the time increased logarithmically i.e. ti estimates that were determined from the simulated well
= 10i hours, where i = -4, -3.9, 3.8,...,2.9, 3. Three dif- tests using the proposed well index are very close to the
ferent permeability values 1 md, 10 md, and 100 md, were true reservoir permeability. Also, the simulated well tests
used to observe the effect of permeability on the perfor- using the proposed well index do not show artifact well-
mance of the well index. Results for some of the gridding bore storage effects. The permeability used in the
 well test
and permeability combinations are shown in Fig. 1 to Fig. analysis software was isotropic with a value of kx ky .
6. These results show that the proposed transient well
index approach allows simulated well tests to accurately Sensitivity to Grid Block Size and Permeability
reproduce both drawdown and pressure derivative at early The performance of the Peaceman well index and the per-
time - even on coarse grids (such as 15 by 15 cells). formance of the proposed well index were sensitive to both
the grid block size and the reservoir permeability. The ar-
tifact wellbore storage was more pronounced in low per-
Well Test Analysis of Simulated Well Tests Each
meability reservoirs and coarsely gridded models.
simulated well test was imported into a well test anal-
ysis package (Saphir15) and analyzed as if it were data.
Examples of the reservoir properties obtained from these Pressure Derivative Accuracy All the simulated well
analyses are given in Table 3 to Table 6. In the well tests tests exhibited a “bump” in the in level of the pressure
simulated using the Peaceman well index no single ana- derivative during the infinite acting radial flow period. The
lytical solution could be properly matched to all the data. position of this bump moves later in time as the grid size
The well test interpretations showed storage coefficients of increases and the permeability decreases. To analyze this
0.012 - 0.6 STB/psi. The permeability estimates deter- effect further, the radius of investigation was calculated18:
mined from the simulated well tests using the proposed
well index and the Peaceman well index, are close to the kt
rinv = 0.03 (16)
true reservoir permeability. However the simulated well φµct
tests using the proposed well index do not show artifact
wellbore storage effects that occur when the Peaceman well Table 12 shows the radius of investigation when the
index is used. calculated pressure derivative first became greater than
the true pressure derivative. The radius of investigation
Sensitivity to Grid Block Size and Permeability was between 64 and 72 % of the grid cell size in every
The performance of the Peaceman well index was sensitive case. The bump artifact begins with a period where the
to both the grid block size and the reservoir permeability. simulator slightly underpredicts the pressure derivative.
Larger grid blocks and lower permeability values gave rise During this time the well test is sampling the grid block it
to greater artifact wellbore storage effects. The perfor- is located in. The presence of the bump makes it appear
mance of the proposed transient well index was insensitive as if the simulated well test is sensing the boundary of this
to gridding and permeability. grid block in some fashion.
To gain further insight into this problem a history match-
Simulation Study using Non-Square Grid Blocks ing study was performed in which the well index was man-
and Anisotropic Permeability To compare the pro- ually changed at every time step of a simulation run in
posed transient well index to the Peaceman well index, a order to match an analytical well test solution. The well
numerical a second simulation study was performed. The indices required are shown in Fig. 11. The well index val-
grid parameters used are shown in Table 7. The basic ues are non-monotonic and the curve in Fig. 11 shows a
reservoir parameters used in this model were the same pa- dip in the same region that the well tests simulated in this
rameters as used in the previous simulation study, which study possess a bump. For a well index model to perfectly
4 TRANSIENT WELL INDEX FOR NUMERICAL WELL TEST ANALYSIS SPE 71572

represent the behavior of the pressure derivative a formu- Acknowledgments


lation must be derived which is capable of reproducing the The second author would like to thank the Turkish Petroleum
non-monotonic behavior in Fig. 11. The transient well in- Corporation and Texas A&M University for their financial
dex proposed in Eq. 10 can only produce monotonic well support.
indices.
Nomenclature
Time Dominated by Artifact Wellbore Storage All
ct Total compressibility, psi− 1
the well tests simulated using the Peaceman well index
Ei Exponential integral function
showed some period of time in which they were dominated
h Reservoir thickness, ft
by artifact wellbore storage. This was evaluated in terms
k Permeability, md
of the radius of investigation.
kx Permeability in the x-direction, md
The time and corresponding radius of investigation at ky Permeability in the y-direction, md
which the artifact wellbore storage effect was completed in p̄ Average pressure, psi
the 12 isotropic examples is shown in Table 13. In every po Simulator well-block pressure, psi
case rinv /∆x is less than 1.2. Therefore at the end of pwf Wellbore pressure, psi
artifact wellbore storage: q Flow rate, BBL/D
rinv Radius of investigation, ft
rinv 0.03 kt ro Equivalent well-block radius, ft
= = 1.2 (17) rw Wellbore radius, ft
∆x ∆x φµct t Time, hours
W I Well Index, (BBL/D)/psi
therefore
 
φµct Greek
t = 1600(∆x)2 (18) ∆x Grid size in the x-direction, ft
k
∆y Grid size in the y-direction, ft
Eq. 18 gives guidelines for when the Peaceman well φ Porosity
index is applicable and when a transient well index should µ Viscosity, cp
be considered.
Subscripts
D Dimensionless
Conclusions
Superscripts
In this study we presented the development and applica- B Block
tion of a transient well index model, which could be used
to simulate pressure transient tests accurately in finite-
difference simulators. References
- The transient well index can be readily incorporated 1. Peaceman, D.W.:“Interpretation of Well - Block
into a commercial simulator by over-riding the default cal- Pressures in Numerical Reservoir Simulation” SPE
culations commercial simulators perform and it is accurate 6893, SPE - AIME 52nd Annual Fall Technical
even on coarse grids. Conference and Exhibition, Denver, Colorado, Oct.
1977.
- The transient well index removes the requirement for
local grid refinment (to achieve accurate pressures) and it 2. Peaceman, D.W.: “Interpretation of Well - Block
is computationally efficient. Pressures in Numerical Reservoir Simulation with
- Comparison of simulated well tests to analytical solu- Nonsquare Grid Blocks and Anisotropic Permeabil-
tions showed the transient well index approach enables well ity” SPE 10528, SPEJ (June 1983), 531-569.
tests to be more accurately simulated in finite-difference 3. Peaceman, D.W.: “Interpretation of Well - Block
simulators. The results are not affected by artifact well- Pressures in Numerical Reservoir Simulation - Part
bore storage. 3: Some Additional Well Geometries” SPE 16976,
- When simulations were performed using the Peace- 62nd SPE Annual Technical Conference and Exhi-
man well index the simulated pressure transients showed bition, Dallas, Texas, Sept. 1987.
significant artifact wellbore storage especially in low per- 4. Peaceman, D.W.: “Representation of a Horizon-
meability cases, and in cases with large grid cells. tal Well in Numerical Reservoir Simulation,” SPE
- Though the transient well index approach is formu- 21217, 11th SPE Symposium on Reservoir Simula-
lated for fully completed vertical wells it could be gener- tion, Anaheim, California, Feb. 1991.
alised to other cases where analytical well test solutions 5. Peaceman, D.W.: “Further Discussion of Produc-
are available. These solutions would replace Eq. 4. The tivity of a Horizontal Well,” SPERE (February
rest of the well index derivation would remain the same. 1991), 149 - 150.
SPE 71572 ROSALIND A. ARCHER and TABIAT T. YILDIZ 5

6. Peaceman, D.W.: “A New Method for Represent-


ing Multiple Wells with Arbitrary Rates in Numer- Table 1 - Gridding Study
ical Reservoir Simulation,” SPE 29120, 13th SPE Number of cells ∆x = ∆y, ft
Symposium on Reservoir Simulation, San Antonio, 15 × 15 166.667
Texas, Feb. 1995. 25 × 25 100
7. Babu, D.K., Odeh, A.S., Al-Khalifa, A.A. and 51 × 51 49.0196
McCann, R.C.: “The Relation Between Wellblock 101 × 101 24.2575
and Wellbore Pressure in Numerical Simulation of
Horizontal Wells - General Formulas for Arbitrary
Well Locations in Grids,” SPE 20161, unsolicited
manuscript, 1989.
8. Mochizuki, S.: “Well Productivity for Arbitrarily
Inclined Well,” SPE 29133, presented at the 13th
SPE Symposium on Reservoir Simulation, San An-
tonio, Texas, Feb. 1995.
9. Chen, G., Tehrani, D.H., Peden, J.M.: “Calcula- Table 2 - Reservoir Parameters
tion of Well Productivity in a Reservoir Simulator Initial reservoir pressure 3000 psi
(1),” SPE 29121, 13th SPE Symposium on Reser- Wellbore radius 0.3 ft
voir Simulation, San Antonio, Texas, Feb. 1995. Porosity 0.2 fraction
10. Chen, G., Tehrani, D.H., Peden, J.M.: “Calculation Oil viscosity 0.7 cp
of Well Productivity in a Reservoir Simulator (2),” Total compressibility 1.0e-05 psi− 1
SPE 29932, International Meeting on Petroleum Production period 100 hours
Engineering, Beijing, China, Nov. 1995. Oil flow rate 100 BBL/D
11. Sharpe, H.N., Ramesh, B.A.: “Development and Reservoir thickness 60 ft
Validation of a Modified Well Model Equation for Reservoir width 2500 ft
Nonuniform Grids with Application to Horizon- Reservoir length 2500 ft
tal Well and Coning Problems,” SPE 24896 67nd
SPE Annual Technical Conference and Exhibition,
Washington, D.C., Oct. 1992.
12. Ding, Y., Renard, G., Aristegui, C.: “New Repre-
sentation of Wells in Numerical Reservoir Simula-
tion,” SPE 25248, 12th SPE Symposium on Reser-
voir Simulation, New Orleans, Louisiana, Feb. 28 -
March 3, 1993.
13. Wan, J., Penmatcha, V.R., Arbabi, S., Aziz, K.: SI Metric Conversion Factors
“Effects of Grid Systems on Predicting Horizon- cp × 1.0∗ E-03 = P a · s
tal Well Productivity,” SPE 46228, presented at f t × 3.048∗ E-01 = m
the SPE Western Regional Meeting and Exhibition, md × 9.869233 E-04 = µm2
Bakersfield, Texas, May 1998. psi × 6.894757 E+00 = kP a

14. “Eclipse 100”, Ver. 99A, Schlumberger GeoQuest Conversion factor is exact
Technology Company, 1997.
15. “Saphir Welltest Interpretation Software”, Ver.
2.20, Kappa Engineering, 1996.
16. Yildiz, T.T.: “Development and Application of a
Transient Well Index,” M.S. Thesis, Texas A&M
University, College Station, Texas, May 2001.
17. “Mathematica”, Ver. 3.0, Wolfram Research, 1996.
18. Horne, R.N.: “Modern Well Test Analysis - A Table 3 - Results using the proposed well index and
Computer Aided Approach”, Palo Alto, California, the Peaceman well index (k=100md and 101 × 101
1995. gridding)
True Peaceman Transient
Value Well Index Well Index
C, STB/psi 0.0 0.01255 0.0
S 0 -0.0264 -0.23
k, md 100 96.2 99.3
6 TRANSIENT WELL INDEX FOR NUMERICAL WELL TEST ANALYSIS SPE 71572

Table 4 - Results using the proposed well index and the Table 9 - Results using the proposed well index and
Peaceman well index (k=100md and 15 × 15 gridding) the Peaceman well index (kx =1md, ky =10md, 51 × 51
gridding)
True Peaceman Transient
Value Well Index Well Index True Peaceman Transient
C, STB/psi 0.0 0.6 0.0 Value Well Index Well Index
S 0 -0.0112 -0.0264 C, STB/psi 0 0.05 0
k, md 100 96.2 96.2 S 0 -0.0012 -0.026
k, md 3.16 3.04 2.95

Table 5 - Results using the proposed well index and the


Peaceman well index (k=1md and 101 × 101 gridding) Table 10 - Results using the proposed well index and
the Peaceman well index (kx =10md, ky =100md, 51 ×
True Peaceman Transient 153 gridding)
Value Well Index Well Index
C, STB/psi 0.0 0.012 0.0 True Peaceman Transient
S 0 0.0192 -0.15 Value Well Index Well Index
k, md 1.0 0.96 0.93 C, STB/psi 0 0.02 0
S 0 -0.0012 -0.0253
k, md 31.62 3.04 27.7

Table 6 - Results using the proposed well index and the


Peaceman well index (k=1md and 15 × 15 gridding)
Table 11 - Results using the proposed well index and
True Peaceman Transient the Peaceman well index (kx =10md, ky =1md, 51 × 17
Value Well Index Well Index gridding)
C, STB/psi 0.6 0.6 0.0
S 0 -0.0568 0.00365 True Peaceman Transient
k, md 1.0 0.96 0.99 Value Well Index Well Index
C, STB/psi 0 0.15 0
S 0 -0.0012 0.0196
k, md 3.16 3.04 2.95

Table 7 - Gridding Study, non-square grid blocks


Number of cells ∆x, ft ∆y, ft
15 × 153 49.0196 16.3398 Table 12 - Radius of investigation when “bump” artifact
51 × 51 49.0196 49.0196 appears
51 × 17 49.0196 147.0588
Gridding ∆x, ft k, md rinv , ft
101 × 101 24.7525 100 16.0
101 × 101 24.7525 10 16.0
101 × 101 24.7525 1 16.0
Table 8 - Results using the proposed well index and the 51 × 51 49.0196 100 34.0
Peaceman well index (kx =100md, ky =100md, 51 × 153 51 × 51 49.0196 10 34.0
gridding) 51 × 51 49.0196 1 34.0
25 × 25 100 100 71.7
True Peaceman Transient 25 × 25 100 10 71.7
Value Well Index Well Index 25 × 25 100 1 71.7
C, STB/psi 0 0.02 0 15 × 15 166.667 100 113.4
S 0 0.004 -0.25 15 × 15 166.667 10 113.4
k, md 100.0 96.2 93.3 15 × 15 166.667 1 113.4
SPE 71572 ROSALIND A. ARCHER and TABIAT T. YILDIZ 7

104

Table 13 - Time and radius of investivation when artifact Analytical Solution


103 Analytical Solution
wellbore storage is complete a
b
This Work
This Work
c

Pressure Drop, Pressure Derivative, psi


Peaceman Well Index
d Peaceman Well Index
aaa aaa aaa
ccc aaa
ccc aaa
ccc cca
ccaa
k, md Gridding t, hr ∆x, ft rinv , ft rinv /∆x 102 aaa
aa cc
aa cc aaa
ccc aaa
ccc ccc ccc
aa cc
aa cc aa cc
aa cc
aa aa aa
cc cc
aa cc cc
100 101 × 101 0.01 24.7525 25.35 1.024 aaaa
aa cc
ccc
a a a a aa
ccccccccc

100 51 × 51 0.04 49.0196 50.70 1.034 b b bb bbbb b db bb


d ddddd db db
db
db
10
bbbbbbbbbbbb bb b db d d d dd ddd
bb bbbbb bbbbbbbbbbbbbbbbbb b db
100 25 × 25 0.2 100 113.39 1.134 d d
ddddddddddddddddddddddd

dd
100 15 × 15 0.4 166.667 160.35 0.962 d
dd
d
10 101 × 101 0.1 24.7525 25.35 1.024 1
d
d
d
d
10 51 × 51 0.5 49.0196 56.69 1.156 d
d
d

10 25 × 25 1.5 100 98.19 0.982


10-1
10 15 × 15 4.0 166.667 160.35 0.962
1 101 × 101 1.0 24.7525 25.35 1.024
1 51 × 51 4.0 49.0196 50.70 1.034 10-2
10-3 10-2 10-1 1 10 102 103
1 25 × 25 15 100 98.19 0.982 Production Time, hr
1 15 × 15 40 166.667 160.35 0.962
Fig. 3 - Diagnostic plot, k=10md, 101 by 101 gridding

104
104

Analytical Solution
103 Analytical Solution
103 a This Work
b This Work
c
Pressure Drop, Pressure Derivative, psi
Analytical Solution Peaceman Well Index
Pressure Drop, Pressure Derivative, psi

Analytical Solution d Peaceman Well Index


a aaa cc
aaa ccc aa
b
This Work
aa cc aaa ccc
aa ccc aaa ccc
aaa ccc
This Work 102 aa aa aa cc
ccc aa cc
aa cc
102 c ccccccccccccccccccccccccccc aa aa aa cccc
Peaceman Well Index
aa aa aa aa aa aa
d a a a a aa
aaaa
Peaceman Well Index
a
acacc db
aacac bd db
aaacc
aaaccc aaccaacacacacc bd ddbbb
b b bb dddb db dd
b bb db db
aaaaa aaaaaa
ccc aaaccc
ccc
aaaccc
aaaccc
b
d
bd 10
bbbbbbbbbbbbbbbbbbbbbbbbbbbbb bb b db d d d ddbbb bbbbb db db
ddddddd
10
aaaaaa ccccccccc
ccaa
ccaa bd dd
ccaa
aaaa cccccc
ccaa
ccaa bd
dd
aaaaaaccc
ccccccccc bdbd dd
aaaa bd d
bd d
b
bdd d
d
bbbbdbbbbbbdd
ddddd bbdd bdbd 1 d
1
bbbbbbbbbbbbbbbbdddd bbddddd bbbbbbbbbbbbbbbbbbbdbd
bbbbbddddddddddddddddddd
d
d
dd d
d
dd d
d
dd d
d
d 10-1 dd
10-1 d d
d d
d d
d d
d d
d dd
d
d
d d
10-2
10-2
10-3 10-2 10-1 1 10 102 103
10-4 10-3 10-2 10-1 1 10 102 103 Production Time, hr
Production Time, hr

Fig. 1 - Diagnostic plot, k=100md, 101 by 101 gridding Fig. 4 - Diagnostic plot, k=10md, 15 by 15 gridding

104 104

aaa aaa aaa aaa


ccc aa
ccc
103 103 aa cc
aa cc aaa
aa cc ccc ccc ccc
aa cc
aa cc
aa cc
aa cc
aa cc
aa cc
cccccccccccccccccc a cccc aa cc
aa aa cc
Analytical Solution aa a a a
Pressure Drop, Pressure Derivative, psi

aaaa
Pressure Drop, Pressure Derivative, psi

Analytical Solution aaa


a This Work aaa
b aa bb bbbb bb bb
This Work
b b b b d d dddddb db dd
2
10 dddbbbbbb
bb bbb bbbbbbbbbbbbbbbb bb b d d d bbbbbbbbbbbbb
102 c dddddddddddddddddd
Peaceman Well Index
d Peaceman Well Index d d
d d
ca
ca ca dd
a ca ca db d
aaa ccc aaa cc
aaa ccc aa caa c ca ca c db d
d
aa aa aa
cc ccc aaa ccc
aaa ccc db d
10 aa
aa cc cc cc db 10
aa aa aaccc d
aa aa aa aa cccc
ccccccccccccccccc
db
db d
aa aa aa aa db d
d
db d
db d
db d
db d Analytical Solution
db d
ddbbb
b b bb dddb dd
b b dd db db d Analytical Solution
bbbbb db db
1
1
bbbbbbbbbbbbbbbbbbbb b b db d d d bb ddd
bbb ddddd
d
d a This Work
d d b
dd
This Work
d c
dd d
d
Peaceman Well Index
dd Peaceman Well Index
d
d 10-1
10-1 dd
d
d
d
d
d
d
d d
10-2 d 10-2
10-3 10-2 10-1 1 10 102 103 10-3 10-2 10-1 1 10 102 103
Production Time, hr Production Time, hr

Fig. 2 - Diagnostic plot, k=100md, 15 by 15 gridding Fig. 5 - Diagnostic plot, k=1md, 101 by 101 gridding
8 TRANSIENT WELL INDEX FOR NUMERICAL WELL TEST ANALYSIS SPE 71572

104 104

Analytical Solution
aa ccc
aa cc
aa cc aaa ca
aaa ccc 103 Analytical Solution
103 aa ccaa cc a
aa aa aa ccc
aa aa aa aa cccc
ccccccccccccccccccccccccccccccccccccc This Work
b
a a aa aa aa aa aa c
This Work
a

Pressure Drop, Pressure Derivative, psi


aaa
Pressure Drop, Pressure Derivative, psi

Peaceman Well Index


aaaa d
aaa Peaceman Well Index
aaa
aa bb bbb b bb
b b b db d dd dddb db dd
2 102
10 ddbbb
dddb
bb bbb bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb bb d d d aa caa aa
c ca c c
aa cc
d d
aaa
aa cc
aa cc aaa aaa
ccc
aaa
ccc
aaa
ccc ccaaa cc cc
d aa cc
aa cc
aa cc
aa cc ccc
dd aa aa
cccaa
ccc aa cc
aa cc db
d a aa cccc
cccccccc aa b
d aaaaa bd
10 d
d 10 b
d db d
d b
d
d db db
d db db
d bbbb bbbbbbb bb ddd db db
d
d bbbb bbb bbbb bbbb bbb d d d d dd ddddddbb bbb bbbbbbbbb
ddddddddd d db db
d ddd
d
d Analytical Solution ddd
d Analytical Solution ddd
ddd
1 1
d a d
d This Work d
d
d b This Work dd
d c dd
dd
d Peaceman Well Index
d d Peaceman Well Index
d
d
10-1 d 10-1
d
d
d
d
d
dd
d
10-2 dd 10-2
10-3 10-2 10-1 1 10 102 103 10-3 10-2 10-1 1 10 102 103
Production Time, hr Production Time, hr

Fig. 6 - Diagnostic plot, k=1md, 15 by 15 gridding Fig. 9 - Diagnostic plot, kx =10md, ky =100md, 51 by
104
153 gridding
104

103
103
Analytical Solution
Pressure Drop, Pressure Derivative, psi

Analytical Solution aaa


aa cc aaa aaa
ccc aaa
ccc c
a Pressure Drop, Pressure Derivative, psi aa aa aa cc
aa cc ccc
This Work
cccccccccccccccccccccccccccc aa aa aa ccc
aa aa cccc
ccccc ccc
b This Work aa aa aa aa
102 c a aa aa aa
Peaceman Well Index aaaaa
d 102 aaaa
Peaceman Well Index aaa
ca
ca ca db
a ca ca db bbbb bbbb bbbbbbbbbb bbbb db db db
bb dddd
ddddd
cccaa ccaa ccaa c ca ca c db bbbbbbbbbbb bbbbbbbbbb bbb bbb bbbb d d d dd
aaa aaa
cccaaa
cccaaa
cccaaa db ddd
10 ccaaa
ccaa
ccaa
ccaa
aaa
ccc ccc ccc db ddd
ccaa
ccaa ccaa
ccaa db d d
aa aa ccc
cccc ccaa
aa aa db
db 10 d ddd
db dd
db dd
db d dd
b
bd dd
db d
1 bbbb bb bbb bbb d
bbbbbbbbb
dd dd bb bbbb
ddddd ddd bbbbbbbbbbbbbbbb
dddddddddddddddddddd b db db dd Analytical Solution
ddd d Analytical Solution
dd 1 d a
d
dd d b
This Work
d dd d
d This Work
d d c Peaceman Well Index
d d
d Peaceman Well Index
10-1 d
d
10-1 dd
d
d
d
d
d
10-2
10-3 10-2 10-1 1 10 102 103 10-2
Production Time, hr 10-3 10-2 10-1 1 10 102 103
Production Time, hr
Fig. 7 - Diagnostic plot, kx =100md, ky =100md, 51 by
Fig. 10 - Diagnostic plot, kx =100md, ky =1md, 51 by
153 gridding
17 gridding
104 35

103
30
aaa
aa cc aaa aaa aaa
ccc
aaa
ccc ccaa
c
Pressure Drop, Pressure Derivative, psi

aa cc
aa cc
aa cc
aa cc ccc ccc This Work
aa cc aa cc
aa cc Peaceman Well Index
ccccccccccccccccccccc aa aa aa aa
cccc ccc
aa aa aa
aaaaa
102 aaaa
aaa
Well Index, STB/psi

bb bb bbbbbbbbbb bb bb ddd db db
bbbbbbbbbbb bbbbbbb bb bb bb bbbbbb db db db
dd ddddddd bb bbbb 25
dddd ddddddddd
dddd
10 d dd
dd
dd
d
d
d
d Analytical Solution 20
d
d Analytical Solution
1 d a
d This Work
d b
d This Work
d c
d Peaceman Well Index
d d
d Peaceman Well Index
d
d 15
10-1 d
d

10-2 10
10-3 10-2 10-1 1 10 102 103 10-4 10-3 10-2 10-1 1 10 102 103
Production Time, hr Time, hr

Fig. 8 - Diagnostic plot, kx =1md, ky =1md, 51 by 51 Fig. 11 - History matched well index and Peaceman well
gridding index, k=100md, 101 by 101 gridding