Sei sulla pagina 1di 14

SPE

Society of Petroleum Enginee/'s

SPE 23444

Transient-Pressure Analysis for an Interference Slug Test


L. Chu and A.S. Grader, Pennsylvania State U.
SPE Members

Copyright 1991, Society of Petroleum Engineers. Inc.

This paper was prepared for presentation at the SPE Eastern Regional Meeting held in Lexington. Kentucky, October 22-25,1991.

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are SUbject to correction by the author(s). The material. as presented, does not necessarily ref!ect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to pUblication review by Editorial Committees of the SOCiety
of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Librarian, SPE, P.O. Box 833836, Richardson. TX 75083·3836 U.S.A. Telex, 730989 SPEDAL.

ABSTRACT provide this straight line yield a 45 degrees line in the log-log
This paper presents a new general solution for an interference classical presentation of the slug test, denoted as the Ferris and
slug test in a two-region composite system with both active well and Knowles 45 degrees line.
observation well having wellbore storage and skin. The solution The effect of a finite wellbore with wellbore storage is addressed
technique, which employs the Laplace transformation and addition by Carslaw and Jaege? [1960] and more related to aquifer flow by
theorem of Bessel function, is new and different from past works. As Cooper et af [1967]. Cooper et af present the classical semi-log
a reduced form of the general solution, slug test interference analysis diagram of the residual head as it declines from a unit value to zero
in a homogeneous domain is readily obtained by letting the properties for various wellbore storage values. Ramey et als [1975] include in
in both regions be identical. The paper presents the characteristic the slug test formulation wellbore skin, and present a correlating
pressure responses for interference slug tests, as well as some slug test parameter, Coe2S , that forms the basis for their late-time log-log
design criteria. Dimensionless correlation parameters and parametric presentation of the slug test. From the type curves presented by
analyses are discussed, Ramey et als it is dear that as the dimensionless pressure declines to
An interpretation technique for a slug test in a homogeneous values of 0.01 or less, the slug test response does not depend on
system is presented.. Two type curves are presented along with two wellbore storage and is identical to the Ferris and Knowles 45 degrees
correlation groups. The first correlation group, (Co/Lfi>(LnCo+2S), is line. Sageev6 [1986] shows that the correlation parameter presented
applicable for the case when the effects of the wellbore storage and by Ramey et af [1975] is applicable for certain ranges, and that the
skin in the observation well are negligible. The second correlation log-log early time presentation of the slug test is needed to address the
group, (CofL5)[LnCo + 0.8(Sl + Sz)], is applicable when wellbore effects of wellbore skin.
storage values in both wells are identical, and skin values are Several combinations of reservoir and well properties and their
arbitrary. Interference slug testing in various composite reservoirs is effects on the slug test are discussed in the literature. Double-porosity
discussed, including oil-gas well interference, and interference in the effects are considered by Dougherty and Babu7 [1984] and Grader
presence of a linear boundary. and Ramel [1988]. The effects of a finite thickness wellbore skin on
the slug test are considered by Moench and Hsieh9 [1985]. Some
INTRODUCTION researchers consider the effects of wellbore hydraulics on the slug test
In a slug test, we introduce a sudden pressure change at the response, such as Shinohara and RameylO [1979], Kippll [1985], and
active well and measure the response of the reservoir by recording the SaldtJna 12 [1983]. High energy variable storage slug tests such as the
pressure as a function of time at the active well or at observation closed chamber test or air chamber test are discussed by Simmons and
wells. Several properties, such as formation transmissivity and Grader13 [1988], Salas and Sageev l4 [1987], and Mfonfu and
wellbore skin can be estimated by interpreting the pressure-time Grader lS [1989]. Other reservoir configurations, such as linear and
response. In a double-porosity reservoir, other parameters, such as spherical flow reservoirs and their effects on the slug test are
interporosity-flow coefficient and fracture storage may be obtained. discussed by Karasaki et al16 [1988].
The major stages of the development of slug test theory As the advances in high-sensitivity pressure measuring devices
correspond to the complexity of the paramenters involved in the become a reality, interference slug testing should become a jraetical
formulation. Initially, the slug test is introduced as an instantaneous and useful diagnostic tool in reservoir engineering. Sageev [1986]
line-source by Ferris and Knowles l [1954]. They use the solution describes the characteristic pressure responses for the interference slug
presented by Theis 2 [1935] and observe that the residual head is test, as well as the effects of wellbore skin. The author shows that
inversely proportional to time. Hence, they propose to graph the wellbore skin lowers the maximum dimensionless crressure at the
residual head as a function of inverse time and use the resulting observation well, but not significantly. Karasaki et at! [1988] discuss
straight line to determine transmissibility. The late-time data that the factors which affect the duration and volume of the rock perturbed
by the slug test. They also point out several advantages of interference
slug tests. Past works on interference slug testing are all based on the
References and illustrations at end of paper. model that assumes that the observation well is infinitesimally small
and has no storage and skin, and that the reservoir is homogeneous.

287
2 TRANSIENT PRESSURE ANALYSIS FOR AN INTERFERENCE SLUG TEST SPE23444

This paper presents a general solution for an interference slug These Laplace solutions are numerically inverted by an algorithm
test conducted in a two-composite system. The active well and the developed by Steh!estt8 [1970].
observation well may have storage and skin. The solutions for an
interference slug test in a homogeneous reservoir are obtained by VALIDATION OF THE SOLUTIONS
simplifying the composite model. The effects of wellbore skins and
storages, as well as property ratios between the two reservoir domains The solutions derived in the previous section are checked by two
. are examined. Two new correlation groups for interference slug methods. First, a direct verification shows that the solutions satisfy
testing are advanced. the associated boundary conditions and governing equations. Second,
we prove that the available solutions for interference slug testing in
homogeneous reservoirs with a vanishing radius of the observation
MATHEMATICAL FORMULATION well are subsets of the newly presented model.
In this section, a solution for the interference slug test in a two- Using an opposite procedures as shown in Appendix A, it is
composite system is presented. The solution for the homogeneous readily shown that the solutions (shown as Eqs. (A.ll) and (A-12»
system is readily obtained from general fonn by letting the properties satisfy the inner boundary conditions of the wells and interface
in both regions be identical. The following assumptions are made: (I) boundary conditions. For lack of space, we prove that the solutions
the system consists of two finite radius wells with a single phase fluid also satisfy the governing equations (Eqs. (A-I) and (A-2». By
flowing in a horizontal and infinite reservoir (see Fig. 1), (2) each invoking the addition theorem for Bessel functions, Eqs. (A·ll) and
region is a homogeneous, unifonn and isotropic porous medium, (3) (A-12) are written as:
the fluid is slightly compressible with a constant viscosity, (4) rock
properties do not depend on pressure, and (5) gravitational forces are
negligible. By invoking Laplace transfonnation and addition theorem
POI = Dil i: e.. cos( nOt) [ Kn(ro{S)ln«(
n=O
01 {S)+AnK,.(rov'S)]

of Bessel function 17, the following Laplace solutions for the two-well
system are obtained (see Appendix A): + Di2 i: en cos( n02)BnK,.(ro{S)
n=O
(18)

PWOl = ~l{ (W2+d2Mv[ Ko({S) + SI{SK1({S)] P02 = Di2~ e.. cos( nov[ Kn«D2VSlltvln(roVSllt:z)+Cnln(roVSlltv]

+ (d3W2+d2d3M:rdtd~V[ Io({S)-Stv'Slt({S)] } (1)


-
+ DiI:E e.. cos( nOt)Dnln(roVSlltv
n..o
(19)

where 0t = 10 - epOt l and O2 = 10 - ep021.


Pwm = ; t dt""S11t2[ Ko(VSlllvlt(VSlltV + K t(VSlltvIo(VSlllV] (2)
g EQuations (18) and (19) are a combination of the tenns
K,.(ro"S) cos[ n(O + P») and In(roVSlltv cos[ n(O + P»), respectively,
where the tenns d t , d2, d3, d4, Wt, W2, M t ' M2 and.1g are derived in where B is an arbitrary constant. As indicated in Appendix B,
Appendix A, and given by Eqs. (A-21) - (A-24), (A-44) - (A-47), and K,.(roVs) cos[ n(O + P») and In(ro{S) cos[ n(O + Ii») are the solutions
(A-27) respectively. The variables are defined in the Nomenclature. of the two-dimension diffusion equation. Hence. by in"Jking the
For a homogeneous system, 11t2 = 1 and A12 = 1, then Eqs. (A-32) - superposition principle. and noting that Ko(ro{S) and Io(ro Slltv are
(A-35) are simplified as: specific fonns of the tenns K,.(ro"S) cos[ n(O + Ii)] and
An = 0 (3) In(roVSlltv cos[ n(O + Ii»), it is readily proved that Eqs. (18) and (19),
hence, Eqs. (1) and (2) are solutions of the governing equations (Eqs.
Bn = In «(D2{S) (4) (A-I) and (A-2».
Co = 0 (5) For the homogeneous system, substituting Eqs. (3), (7), (8), (9),
Dn = K n «(01 {S), (6) and (10) into Eq. (A-27) yields:
Ag = W t W2 - K5(Lov'S) M tM2 (20)
Accordingly, Eqs. (A-21) - (A-24) are reduced to:
If we assume that the observation well is infinitesimally small and has
dt =Ko(Lo{S) (7) no wellbore storage, and we refer to Eqs. (A-38) - (A-4I), it follows
d2 =0 (8) that:
d3 =0 (9) W t = CD\SKo({S) + {SKo(v'S)(CD\StS + 1) (21)
d4 = Ko<Lo{S) (10) W2 = 1 (22)
M t = CD\slo({S) - {SIt({S)(CDiSts + 1) (23)
Substituting the above coefficients into Eqs. (1) and (2) yields the
interference homogeneous solutions: M2 = 0 (24)

Pwot = ~t { W2[ Ko({S) + Stv'SKt({S)] Substituting Eqs. (21) - (24) into Eq. (20) gives:
Ag = CotsKo({S) + {SKo({S)(CD\StS + 1) (25)
- K6(Lo{S)M2[ Io({S) - St{Slt({S)] } (11)
Substituting the above equations into Eqs. (A-25) and (A-26) yields:
Pwo2 = ~t Ko(Lov'S)v'S[ Ko({S)lt({S) + Kt({S)lo(v'S)] (12)

where (26)
W t = COtsKo(v'S) + {SKt(v'S)(CD\SSt + 1) (13)
W2 = CD2sKo(v'S) + {SKt (v'S)(CD2sS2 + 1) (14)
(27)
M 1 = CD\slo(v'S) - {S It({S)(CDlsS t + 1) (15)
M2 = CD2slo(v'S) - {S 11(v'S)(CD2sS t + 1) (16) Substituting Eqs. (3), (26) and (27) into Eq. (A-ll) gives:
Ag = W2W t - K5(Lov'S)M tM2 (17)

288
SPE23444 L.CFnJ AJij) A.S.G~DER 3

K (R fS)
o D1 The uppermost curve represents the active well response. It is clear
pD!=---;:::---;::------=---=..:------ (28) that the data fall on the 45 degrees decline. From the time match of
- - - + s [ Ko(fS)+s!fSK!(fS)~
fSK!(fS) the data to the model, the transmissibility is determined. However, it
CD! is not possible to calculate the storativity of the system. Ferris and
Knowles! do not provide other aquifer information. If we assume
Equation (28) is identical to the solution presented by Sageel' [1986]. that the formation thickness is 30 ft, that the porosity is 0.20 and that
Substituting Eq. (28) into Eq. (A-6) generates the dimensionless the total compressibility is 5xlO--{j l/psi, we can then calculate
pressure at the wellbore, which is same as the solution presented by interference responses for this test. Four such calculations are
Ramey et al!9 [1972]. demonstrated in Fig. 3 for an observation well located either 100 ft
(curves A and B) or 500 ft (curves C and D) away from the active
INTERFERENCE IN A HOMOGENEOUS SYSTEM well. Curves A and C represent storage only in the active well.
In this section we discuss interference slug testing in general, Curves Band D represent identical storage in both wells. Figure 3
and then examine interpretation methods for the following demonstrates that even at an observation well located 500 ft away
homogeneous cases: 1. interference testing with storage and skin only from the active well, with wellbore storage (assuming the wells have a
in the active well, 2. interference testing with identical storage in both similar configuration), the pressure trace is significant and analyzable.
wells and different wellbore skins in the wells, and 3. interference If data were collected at such a well starting at about 20 seconds after
testing with arbitrary storage and skin values in both wells. the start of the test, the interference pressure record would exhibit a
maximum. Hence, the data would yield the transmissivity as well as
General Discussion the storativity of the aquifer.
Interference slug test analysis requires a detectable pressure The effects of wellbore storage in the observation well are
record in the observation well in response to an active slug test. The demonstrated in Fig. 4. As observation well storage increases, the
mathematical formulation presented here (as well as in most other magnitude of the pressure response decreases, the maximum pressure
formulations in the literature) is a diffusive one using parabolic forms response is delayed and the shape of the pressure response flattens. If,
of partial differential equations. Hence, in such a formulation there is for example, the detectable dimensionless pressure is 0.001
no support for shocks, such as in a hyperbolic formulation. We note (represented by the dashed line on Fig. 4), the case when the
that in this discussion we ignore pressure shock waves that travel at dimensionless storage Coo = 106 is almost undetectable. However, the
the speed of sound in the reservoir. These waves are beyond the case when Coo = 105 is clearly detectable.
scope of discussion in most interference testing methods, and are not The effect of wellbore skin in the active well on interference
accounted for in the present formulation. Since shocks are not pressure responses is demonstrated in Figure 5. In this case, the
allowed, every portion of the connected reservoir domain contributes active wellbore storage is fixed at CD! = 10,000. There are two
the pressure response of the reservoir. Hence, the concept of families of curve in Fig. 5 corresponding to two values of observation
detectable pressure change is advanced. The detectable pressure well storage, CO2 = 50 (the light curves), and Coo = 10,000 (the dark
change in an observation well is determined by the pressure measuring curves). Each family of curves contains four curves representing
equipment. With the advent of electronic pressure measuring devices, different values of skin at the active well. Active well skin has an
the resolution of the detectable pressure change is improving. Also, effect on the observation well response regardless of the value of the
the temporal resolution is greatly improved. As the pressure response storage at the observation well. Also, we note that the larger the active
at an observation well in response to a slug test exhibits a rise and a wellbore skin, the lower the response at the observation well.
decline (in dimensionless units), the pressure response curve is
The effect of wellbore skin in the observation well on
intercepted twice by the detectable pressure level. The time period
interference pressure responses is demonstrated in Figure 6. The
between these two intersection points is defined as the detectable time
active wellbore dimensionless storage is fixed at 10,000 and active
period.
wellbore skin is set to zero. There are two families of curves in Fig.
The two-composite model presented in this paper is simplified to 6 representing two values of dimensionless storage in the observation
simulate a homogeneous reservoir by setting the properties of region 2 well, CO2 = 50 (the thin lines) CO2 = 10,000 (the thick lines). Figure
to be identical to the properties of region 1. However, we keep the 6 shows that the effects of observation well skin are not significant
option of having wellbore storage and skin effects in the active well when observation wellbore storage is small. However, the effects of
and in the observation well. Typical interference responses are shown wellbore skin become quite significant when observation well storage
in Fig. 2. There are two families of three curves in this figure. The is large. This is consistent with the fact that more fluid must be
thick lines represent interference responses without wellbore storage in transmitted out of the observation well during the initial phase of a
the observation well, CO2 = 0 and CD! = 10,000. The time axis is drawdown slug test when wellbore storage is large. This enhances the
tn/CD! in all the diagrams unless defined differently. The thin group signature of wellbore skin. Wellbore skin has no effect on the Ferris
of curves represents interference responses when both wells have and Knowles 45 degrees decline period. We note that when wellbore
storage, CD! = 10,000 and Coo = 10,000. In all cases, wellbore skin storage values at both wells are identical, the families of curves in
is negligible. Each family of curves represents three dimensionless Figs. 5 and 6 are identical. Hence, the case when S! > 0 and S2 = 0
interwell distances, Lo = 100, 200, 500. It is demonstrated that if, for is identical to the case when S! = 0 and S2 > O. However, when
example, the detectable dimensionless pressure is 0.001, all these slug observation wellbore storage is not identical to the storage in the
test cases yield a significant pressure response in the observation well, active well, interchanging wellbore skin values leads to different
with a detectable time period greater than two log cycles. We note responses. The family of curves for CO2 = 50 in Fig. 5 is not
here that a dimensionless detectable level of 0.001 is not unusual. identical to the same family of curves in Fig. 6.
Ferris and Knowles! [1954] measured active well dimensionless
pressure values between 0.01 and 0.001. Storage and Skin Only in the Active Well
The interference pressure responses of Fig. 2 show that at late Figure 7 shows a modified presentation of interference slug
time, all the responses join the 45 degrees Ferris and Knowles line. testing. There are three curves in this figure, each curve is composed
Hence, the combined effects of wellbore storage of both wells of three interference responses denoted by squares, circles and a solid
diminish during the late time flow period. Physically, this represents line. The various combinations of dimensionless storage,
the time that all reservoir volume that contains the two wells is at a dimensionless distance, and wellbore skin are shown in the inset table
very similar pressure, returning to the initial pressure at the same rate. on the figure. The correlating parameter for this case is
Any pressure data collected during the 45 degrees period yield only (CnfL5)(LnC o + 2S). This correlation parameter is similar to the
the value of transmissivity, as shown by Ferris and Knowles!, if the parameter presented by Ogbe and Brigham20 [1989] and by Ehlig-
actual wellbore storage can be calculated (which is usually the case). Economides and Ayoub2 ! [1986]. The time axis is modified to
Figure 3 shows a match of the data presented by Ferris and Knowles! tOL = tn/L5, and the pressure axis is modified to p; = Pwm/CD!/L5.
to a homogeneous slug test model with storage only in the active well. The various combinations of properties with the same correlating

289
4 TRANSIENT PRESSURE ANALYSIS FOR AN INTERFERENCE SLUG TEST SPE23444
parameter yield identical responses as presented in Fig. 7. Given Figure 10 demonstrates the effectiveness of this correlating parameter.
these three correlating parameters, a type curve for this slug test Each one of the curves in Fig. 10 denoted as A, B, and C contains
configuration is presented in Fig. 8. The modified dimensionless time, data from three different configurations that match well to each other
tOL,peak' when the peak pressure change occurs is calculated by the (squares, circles, and solid line). The details of these various cases
correlation parameter, denoted here as X, with the following equation: are summarized in the inset table in Fig. 10. Figure 11 presents a
type curve for interference slug test analysis when both the active well
TOL,peak = - 0.051889 + 0.48967X + 0.06838X2 - 0.0043554X3 (29) and the observation well have the same storage but may have arbitrary
The peak modified dimensionless pressure drop is calculated using the values of wellbore skin. This type curve is used in the same fashion
correlation parameter, X, by: as the type curve presented in Fig. 8, but will yield only the sum of
wellbore skins and not the individual values. The independent
P;,peak = - 0.40523 - 0.38301X - 0.1281X2 + 0.012704X3 (30) determination of each wellbore skin requires another set of pressure
Equations (29) and (30) are used in the design phase of an data, possibly from the active well.
interference slug test. Using estimations of the dimensionless storage, The modified dimensionless time, tOL,peak at the peak pressure
the dimensionless distance and wellbore skin, the correlating parameter change can be calculated using the correlation parameter, denoted here
X is evaluated, yielding estimates of the maximum pressure change as F, by the following equation:
expected in the wellbore, and when this pressure change will occur.
In general, one log cycle of data on both sides of the peak provide a ToL,peak = 0.23288 + 0.69113F + 0.09155F2 - 0.011201F3 (31)
significant and analyzable pressure record. where F = (CofL5)[Ln Co + 0.8(SI + Sz)]. The peak modified
A synthetic example is presented to demonstrate the application dimensionless pressure drop is calculated using the correlation
of the new type curve for interference slug testing. The pressure-time parameter, F, by:
data are given in Table I, where the actUal pressure drop data in the
P;,peak = - 0.51157 - 0.47069F - 0.13268F2 + 0.16538F3 (32)
well are converted to dimensionless values by dividing the pressure
drop values by the initial pressure drop value. The relevant Equations (31) and (32) are used in the design phase of an
welVreservoir data for this test are: h = 35 ft, rw = 0.25 ft, ~ = 0.90 interference slug test. Using estimations of the dimensionless storage,
cp, P = 62.0 Ib/cu. ft, and L = 26.25 fl The match of the data to the the dimensionless distance and wellbore skins, the correlating
type curve is demonstrated in Figure 9. In actual testing, we do not parameter F is evaluated, yielding estimates of the maximum pressure
expect to have a test period that spans 5 log cycles. The large time change expected in the wellbore, and when this pressure change will
span is used only for a demonstration. The curves in the type curve occur. In general, one log cycle of data on both sides of the peak
are unique, and would require shorter time spans to be identified. For provide a significant and analyzable pressure record.
example, in the case we are considering, data from 0.01 to 1 hour
would identify the curve uniquely. The parameter match is Arbitrary Storage and Skin in Both Wells
(CIfL5)(Ln Co + 2S)match = 2~. The pn:ssure axis match is There is no simple way to correlate this case. The interference
[(Po)/PO]match = 1/1.77. The nme match IS [toL/t]mateh = 35.0. response of a well without wellbore storage, as shown in Fig. 8,
Estimates of transmissivity, storativity, and wellbore skin are obtained exhibits a longer flat response as the value of the correlating parameter
in the following. The wellbore storage constant, C., for a moving increases. However, this is not the case when the observation well
liquid column in the wellbore is determined by: has storage, as shown in Fig. 11. The newly developed model can
generate interference responses for any combination of storage and
C. = 7tr;(I44) = 0.0807 bbllpsi skin values. Hence, the case when the storage in the observation well
Pwf is different from the storage in the active well must be interpreted by
From the pressure match, it follows: generating a specific type curve for the test and by use of an
automated parameter matching technique.
-p;] 2
= LD/C 27t4>h<;L 4><; 1
o = _.:........::...-=_-.:...-..:.-.--:-=:-::......
[ Po match 5.615 C. 3.009xlO--6 1.77 INTERFERENCE IN A COMPOSITE SYSTEM
Hence, the storativity is determined as: The parameters that control interference slug test responses in a
composite system are: two values of wellbore storage, two values of
4>ct = 1.715xlO--6 lIpsi wellbore skin, transmissivity ratio between the two regions of the
From the time match we obtain: reservoir, storativity ratio between the two regions of the reservoir,
and three geometrical parameters. In total, there are nine
= 0.000;64: = 35.0 well/reservoir parameters. The goal of this section is to demonstrate
[todt ]
match 4>~trwLO the composite reservoir effects on interference slug test responses, and
show that in certain cases, these effects are significant and must be
Substituting the determined value of the storativity, viscosity, wellbore taken into considerations. Figure 12 shows some of the simplified
radius, and dimensionless distance into the above equation and solving cases that we consider in this section.
for the permeability yields: Figure 12a corresponds to the general mathematical formulation
k= 141 md of an active well near a circular subregion containing an arbitrarily
placed observation well with storage and skin. Figure 12b represents
From the matched correlation parameter, it follows: a composite case where the observation well is located in the middle
of a circular subregion. This case may represent interference testing
C~ (LnCo + 2S)]
[ Lo
= 25 near a gas cap, or near a gas injection'well in a gas storage aquifer.
match Such a case may occur in EOR projects, where the observation well
may represent an injector. Figure 12c represents a similar
Substinning the matched values for Co and Lo yields the value of
configuration as Fig. 12b, except that the observation well is located in
wellbore skin: the same reservoir unit as the active well. For this case (Fig. 12c), the
S = 2.19 formulation presented in this paper does not allow us to assigJl'
The interpretation results agree with the input data: 4><; = 1.7 X 10--6, k wellbore storage and skin to the observation well.
= 141 md, and S = 2.12. Figures 12d, 12e, and 12f represent a two-composite reservoir
where the two portion of the reservoir are separated by a permeable
Identical Storage in Both Wells With Wellbore Skin linear boundary. This configuration is obtained by using a large radius
The interpretation technique for this case is similar to the case for the circular subregion with respect to the interwell distances.
when there is only wellbore storage and skin in the active well. The Figure 12d is the simple case where the active well and the
correlating parameter for this case is: (CofL5)[LnCo + 0.8(SI + Sz)]. observation well are placed symmetrically on both sides of the

290
SPE23444 L. CHU AND A. S. GRADER 5

boundary. In Fig. 12e, the observation well is placed arbitrarily in the in the simple geometrical case shown in Fig. 12d. In this case, the
reservoir region not containing the active well. In Fig. 12f, the two wells are located symmetrically on both sides of the boundary.
observation well is placed in the same reservoir region that contains The interwell distance is 100 wellbore radii. When the mobility ratio
the active well. Again, in this case (12t), we cannot assign wellbore is greater than one (A12 > 1) the effect of the boundary is not very
storage and skin to the observation well. significant (thick curves in Fig. 16). However, when the mobility
ratio is less than one (A12 < 1) the effect of the boundary is
Circular Boundary Composite System significant, as shown in the thin curves in Fig. 16. The curves for
In this section we consider two interference slug test cases. In A12 = 0.01 and 0.001 exhibit an upward bend following the peak
the first case, the observation well is at the center of the circular pressure change. The upward bend is indicative of the boundary
subregion (Fig. 12b). In the second case, the observation well is between the two wells. Following this bend, the pressure response
located in the same region as the active well (Fig. 12c). Figure 13 resumes the decline that approaches the Ferris and Knowles 45 line.
demonstrates the effects of the mobility ratio (A12 = AlAi> on The effects of the storativity ratio are shown in Fig. 17. In this
interference responses of an observation well. This case simulates case, as the storativity ratio decreases, the magnitude of the pressure
oil-gas or gas-water interference testing. Both wells have wellbore change decreases, and the peak pressure change is delayed. In this
storage (CD! = Co2 = 104) and zero skin. The wells are 300 wellbore case, where the mObility ratio is of unit value, it is not possible to
radii apart, and the subregion radius is 100 wellbore radii. Also, the identify a significant diagnostic effect of the linear boundary, unlike
storativity ratio has a value of I, 0)12 = 1. The thick curves represent the case where mobility ratio was lower than one. However, changes
the cases where the mobility ratio is greater than 1. As AI2 increases, in the mObility ratio have a significant effect on interference pressure
the interference response exhibits a sharp rise prior to the peak response for this well-reservoir configuration.
pressure change. This rise is sharper than the rise in the homogeneous We now consider the case described by Fig. 12e, where the
case. Also, as AI2 increases above one, the maximum pressure change observation well is located not on the symmetry line. The effects of
at the observation well decreases. This represents the physical the location of the observation well are presented in Fig. 18. We
characteristic of the system allowing most of the response to be consider five locations along an arc with a radius of 100 wellbore radii
controlled by the mobile outer region. The thin curves in Fig. 13 around the active well, with 10 degrees between each location. This
represent cases when AI2 ~ 1. As the mobility ratio decreases below a configuration is shown in detail in the inset diagram in Fig. 18. In
value of one, the magnitude of the response decreases and the pressure this case, '1112 = I, AI2 = 100, CD! = Co2 = 104, and SI = S2 = O. In
rise and decline becomes wider. the cases described in Fig. 18, as the angle of rotation increases, the
The effects of the storativity ratio on interference slug test observation well is closer to the boundary. Since the active well
responses are shown in Fig. 14 for cases when the observation well region is 100 times more mobile than the observation well region, the
has no storage (thick curves), and cases when the observation well has pressure change in the observation well increases as the observation
storage (thin curves). As the storativity ratio decreases, the maximum well is located closer to the boundary. The location of the observation
pressure response at the observation well decreases, and the peak well has a major effect on the pressure response, similar to the effects
pressure change occurs at a later time. This reflects the fact that at in the case of a constant rate active well. This indicates that multiple
low storage ratios, the subregion provides more pressure support to interference slug tests would yield the location of the interface in the
the active well than the high storage ratio cases. Hence, the overall reservoir. However, as there are several parameters involved, the
pressure change in the subregion decreases when storage ratio development of a simple analysis technique, as suggested by Vel?
decreases. As shown in Fig. 14, as observation well storage increases, [1977] for the linear impermeable boundary case, is not possible. The
the magnitude of the pressure change in the observation well analysis would involve a trial and error procedure and include some
decreases. However, for small storativity ratio values, the value of automated analysis techniques. The advancement of such an
observation well storage has no effect on the pressure response. For interpretation method is beyond the scope of this paper.
example, the two curves for AI2 = 0.001 are practically identical. We now consider the effect of interference testing near an
Figure 15 shows the effects of a constant pressure circular impermeable linear boundary (Fig. 12t). In this case, the subregion is
subregion on interference slug test responses. Five observation wells set to be impermeable with a large radius. The observation well in
are positioned on a circular arc around the subregion as shown in the this case has zero wellbore storage and skin. We consider five
inset diagram in Fig. 15. The location of the observation well has a observation wells placed on a circular arc around the active well 45
significant effect on the pressure response. The peak pressure change degrees apart. The distance between the observation wells and the
at the well closest to the active well (well A, a = 0) is 2 orders of active well is 80 wellbore radii (Ro = 80), and the distance between
magnitude larger than the maximum pressure change recorded at the the active well and the boundary is 100 wellbore radii (Lo = 1(0), as
well on the far side of the subregion (well E, a = 180). If the shown in the inset diagram in Fig. 19. The log-log response in this
detectable dimensionless pressure is 0.001 for example, we will not figure shows the effects of the boundary. The upper most interference
record a significant pressure change at observation wells beyond 135 curve (curve A) represents the well closest to the boundary (well A).
degrees. However, the other wells have significant pressure records. Three more curves are provided in Fig. 19 for reference. These
curves are the responses of the active well. The uppermost curve
The constant pressure subregion represents an extreme case of (curve F) is the response of the active well in the presence of the
the true-composite configuration. The newly developed model allows linear boundary. The second curve from the top (curve H) represents
us to consider any combinations of properties for the two reservoir the active well case in a homogeneous system. The third curve from
regions. Two reference curves are also provided in Fig. 15. The the top (curve G) represents the active well response in the presence
uppermost curve in the decline portion, denoted by AAA, represents of a constant pressure boundary. The lowermost interference curve in
the response of the closest well (well A, a = 0) when the subregion is Fig. 19 (curve AA) represents the interference response of an
impermeable. The second thick curve from the top during the decline observation well located between the active well and a constant
portion, denoted as AA, represents the response of the closest well in pressure linear boundary (well A). In this case, the pressure change in
a homogeneous system. For this observation well, all the
combinations of a composite subregion must fall between the the observation well does not exhibit the 45 degrees slope within the
impermeable subregion and the constant pressure subregion responses dimensionless presure range shown in Fig. 19, and the pressure
(between curves A and AAA). returns to the initial pressure faster than the case without a constant
pressure source.
Linear Boundary Composite System Figure 20 is a semi-log presentation of the interference data
In this section we discuss interference slug testing in a two- shown in Fig. 19. The location of the observation well has a
composite system where the two regions are separated by a linear significant effect on the magnitude of the pressure change in the
boundary, as shown in Figs. 12d, 12e, and 12f. Figures 16 and 17 observation well. In this case, where the observation wells are located
present the effects of mobility ratio and storativity ratio, respectively, on a circle around the active well, the semi-log response is close to
being symmetric, the maximum pressure change occurs at the same
291
6 TRANSIENT PRESSURE ANALYSIS FOR AN INTERFERENCE SLUG TEST SPE23444

time, and the magnitude of the pressure change decreases as the COj dimensionless wellbore storage of the well j
obsetvation well is placed at a large distance from the linear boundary. C. wellbore storage
We also note that the time axis in this case is normalized by the Ct total system compressibility
dimensionless distance between the active well and the linear On = coefficients
boundary, tofLfi. Dij coefficients
dl = coefficients
CONCLUSIONS dz coefficients
The notion that the slug test probes only the near wellbore d3 = coefficients
region adjacent to the screen, as suggested by Ferris and Knowles l <4 coefficients
[1954], is not correct. This has been demonstrated in all the Dij = coefficients set by inner boundary condition of well j
interference slug test cases presented in this paper. Moreover, F = correlating parameter
resetvoir properties such as the presence of a linear boundary, or a fZl = coefficient
composite system have significant signatures on interference slug test In = modified Bessel function, first kind, n'th order
responses. These responses are always lower in magnitude than the I'n = spatial derivative, modified Bessel function, first kind
response at the active well, but are clearly detectable even using the h formation thickness
technOl0f! presented in 1954 (the acoustic "popper", Ferris and kj = permeability for region i
Knowles). With the advent of electronic real-time pressure recording K,. = modified Bessel function, third kind, n'th order
devices, the resolution of the pressure measurements and the temporal K'n = spatial derivative, modified Bessel function, third kind
frequency of measurements have increased dramatically. Today, we L = distance between active well and boundary
can measure pressure changes in wells caused by the relative motion Lo = dimensionless distance between active well and boundary
of the earth with respect to the sun and the moon, known as earth Ml = coefficients
tides. These pressure changes are on the order of 1/100 - 1/10 of a Mz = coefficients
psi. We can also measure pressures at high frequencies, providing a m coefficient
solid platform for derivative analysis. Hence, the detectable Po = dimensionless pressure of active well at time zero
dimensionless pressure value can be reduced to 0.0001 and even lower Pi = pressure at region i
than that This level of detectability makes interference slug testing
feasible, and dispels the notion that only the near wellbore domain Poi = dimensionless pressure at region i, po;=Pn-P/PufO
p = modified dimensionless pressure, P =PwD2/CDl lLo
contributes to the slug test response. In view of the material presented
Pii initial pressure at region i
in this paper we make the following conclusions:
Pwoj = dimensionless pressure at the wellbore for well j
1. A new mathematicd model for interference slug testing in a Laplace transform of POi
two-composite system is advanced. This two-well system POi =
Pwoj = Laplace transform of Pwoj
considers finite radius wells with wellbore storage and wellbore R Oj = dimensionless form of Rj. R Oj =Rjrb
skin in both wells. . dimensionless radius of well j
Rwoj =
2. Based on the general solutions, the new solution for a
homogeneous resetvoir is derived.
R.i = distance from well j to pressure point
distance from system geometric center to pressure point
r
3. Based on the general solutions, the new solutions for an rl radius of subregion
interference slug test near a linear boundary is derived. rb arbitrarily chosen base radius
4. The new solution simulates interference slug testing with ro dimensionless form of r, ro = r/rb
wellbore storage and skin in both wells in a composite resetvoir rDl dimensionless fonn of rl' rDl = rl/rb
separated by a linear boundary. rOWj = dimensionless fonn of rwj. rOWj=Tw/rb
rp = radius of flowing pipe
5. A correlation parameter, (CofLo)(LnCo + S), is advanced for the
rwj = radius of the well j
homogeneous case when the wellbore storage in the obsetvation
r'j = distance from system geometric center to well j
well is negligible. A corresponding interpretation technique is
r'Oj dimensionless fonn of r'j' r'OJ = r'/rb
presented, which allows the estimation of the transmissibility,
Sj wellbore skin of the well j
storativity and skin factor. A correlation parameter,
s = Laplace variable
(CofLo)[LnCo + 0.8(Sl + S~], is advanced for the homogeneous
t time
case when wellbore storage in the obsetvation well is the same
to = dimensionless time, to = k l t/( C\> ~ Ctrwz)~
as the wellbore storage in the active well. A corresponding
interpretation technique is presented, which allows the estimation tOL = modified dimensionless time, tOL = tofLo
of the transmissibility, storativity and the sum of the skin factors. Wl coefficient
Wz = coefficient
6. Correlating equations for determining the time of the peak X correlating parameter
pressure change and the magnitude of the peak pressure change
for the homogeneous case are presented for test design purposes. Greek
7. The effects of mobility ratio, storativity ratio, and location of the (X = angie around the well
obsetvation well in a composite system on interference slug test ~ arbitrary angle
responses are demonstrated. The new model can simulate slug E" = constant
test interference between oil and gas wells, and between gas and coefficient
L\
water wells. T1ij = diffusivity ratio, T1ij = OJCO)ij
9. The newly presented model should allow us to develop new p density
simple interpretation techniques for limiting cases (as was shown 0 = rotation angle of the pressure point
for the homogeneous resetvoir). This model should also allow OJ = angie between the well j and theJ!i.ressure point
us to consider analysis of complicated interference tests in true A;j = apparent mobility ratio, ~ = (')..1 T1)ij
composite systems using automated analysis techniques. A.;j = mobility ratio, A.;j = (kI~)/(kI~)j
J.l.i fluid viscosity in region i
NOMENCLATURE C\> = porosity
C\>Oj initial angle of the well j
COij storativity ratio, CO;j = (C\><;)/(C\><;)j
coefficients
coefficients
coefficients

292
SPE23444 L. CHU AND A. S. GRADER 7

Subscripts 18. Stehfest, H.: "Algorithm 368, Numerical Inversion of Laplace


D dimensionless fonn Transfonns," Communications of the ACM, D-5 (Jan. 1970) 13,
= region of the composite system No. 1,47-49.
j = 1 denotes active well. 2 denotes observation well 19. Ramey, H. J., Jr., Agarwal, R. G., and Martin, I.: "Annulus
n = order of Bessel function Unloading Rates as Influenced by Wellbore Storage and Skin
w = wellbore Effect," Soc. Pet. Eng. J., 453-463. 1972.
20. Ogbe, D.O., and Brigham, W. E.: "A correlation for
ACKNOWLEDGEMENTS Interference Testing with Wellbore-Storage and Skin Effects,"
The authors wish to thank the Petroleum and Natural Gas SPEFE (Sep. 1989) 391-396.
Engineering Section and the Pennsylvania State University for making 21. Ehlig-Economides, C. A.. and Ayoub, J. A.: "Vertical
this research possible. Interference Testing Across a Low-Penneability Zone," SPEFE
(Oct. 1986) 497-510.
REFERENCES 22. Vela, S.: "Effect of a Linear Boundary on Interference and Pulse
1. Ferris, J. G., and KnOWles, D. B. : "The Slug Test for tests - The Elliptical Inference Area," J. Pet. Tech. (Aug. 1977)
Estimating Transmissibility," U.S. Geol. Surv. Ground Water 947-950.
Note, 26, 1-7, 1954. 23. Abramowitz, M., and Stegun I. A.: Handbook of Mathematical
2. Theis, C.V. "Relation Between the Lowering of the Piewmetric Functions, Dover Publications. INC., New York, 1964.
Surface and the Rate and Duration of Discharge of a Well Using
Ground-Water Storage," Eos Trans. AGU, 16(2) (1935), pp APPENDIX A
519-524. The general solution for interference slug testing in a two-
3. Carslaw, H.S. and Jaeger, J.C.: Conduction of Heat in Solids, composite reservoir is derived here. The wells, both active and
2nd ed. Oxford University Press, 1960. observation, have wellbore storage and skin. The coordinate system
and geometric parameters are shown in Fig. 1. Based on the
4. Cooper. H. S., Jr.. Bredehoeft, J. D.. and Papadopulos, I.S. : assumptions described in the Fonnulation section, the flow of the fluid
"Response of a Finite Diameter Well to an Instantaneous Charge in the system is idealized by the following two dimensional partial
of Water," Water Resources research, 3(1). 263-269,1967. differential equations in Laplace space (with respect to time) as:
5. Ramey, H. 1., Jr., Agarwal, R. G., and Martin, I.: "Analysis of
"Slug Test" or DST Flow Period Data," J. Can. Pet. Technol., iP POl +..!... dpOl + _1_ a2 POl = SPol (A-I)
37-47, 1975. aro2 ro aro ro2 ae 2
6. Sageev. A. : "Slug Test Analysis," Water Resources Research,
(Aug. 1986) 22, No 8, 1323-1333. 2 2
a PD2 1 dPo2 1 a Po2
7. Dougherty, D. E., and Babu, D. K.: "Flow to a Partially - -2- + - - - + -2- -2- = 1112 sPD2 (A-2)
dro ro dro ro de
Penetrating Well in a Double-Porosity Reservoir," Water
Resources Research, (1984) 20, No.8, 1116-22. The associated inner boundary conditions and interface conditions are
8. Grader, A. S., and Ramey, H. J., Jr.: "Slug Test Analysis in as follows:
Double-Porosity Reservoirs," SPEFE (June 1988) 329-339. (A-3)
9. Moench A. F., and Hsieh, P. A.: "Analysis of Slug Test Data in
a Well with Finite Thickness Skin," Paper presented at the (A-4)
I.A.H. 17th International Congress on Hydrology of Rocks of
Low Penneability, Int. Assoc. of Hydrogeol., Tucson, Ariz.,
Jan. 7-12, 1985. (A-5)
to. Shinohara, K., and Ramey, H. J.• Jr.: "Slug Test Data Analysis,
Including Inertial Effect of the Fluid in the Wellbore,"
Proceedings of the SPE Annual Conference and Exhibition. SPE
of AIME, Las Vegas, Nev. Sept. 23-26, 1979.
11. Kipp, K. L., Jr.: "Type Curve Analysis of Inertial Effects in the
Response of a Well to a Slug Test," Water Resources Research,
Vol. 21, No.9, 1397-1408, Sept. 1985.
12. Saldana, M.: "Flow Phenomenon of Drill Stem Test With Inertial
and Frictional Wellbore Effects," PhD Dissertation, Stanford
University, Stanford, CA. (Oct. 1983).
13. Simmons, J. F., and Grader, A. S.: "Application of Closed
Chamber Theory to Backsurge Completion Testing," SPEFE
(November 1988) 527-535.
14. Salas 8., and Sageev, A.: "Closed Chamber Testing: The
Effects of Wellbore Friction and Fluids Compression," SPE
16800, Proceedings of the SPE Annual Technical Conference
and Exhibition, SPE of AIME, Dallas, Texas (September 1987).
From Eqs. (A-7) and (A-8):
15. Mfonfu, G. B. S. and Grader, A. S.: "Closed Chamber Tests as a
_ 1 ap02
Moving Boundary Problem," SPE 19832, Proceedings of the
SPE Annual Conference and Exhibition, SPE of AIME, San 21t b
r
Co2SP02(RW02.S) - (C02SS2 + 1)- 21<[ Ro - : > -]
oRo RD = RDW2
da = 0
Antonio, Texas (8-11 October 1989).
(A-tO)
16. Karasaki, K., Long, C. S., and Witherspoon, P. A.: "Analytical
Models of Slug Tests," Water Resources Research, (Jan. 1988) According to the geometry of the system and features of Bessel
24, No.1, 115-126. functions, we construct the general solutions for each region as
17. Watson, G. N.: A Treatise on the Theory of Bessel Functions, follows:
2nd Ed. Cambridge at the University Press, 1962.

293
8 TRANSIENT PRESSURE ANALYSIS FOR AN INTERFERENCE SLUG TEST SPE23444

AIZ [ I'n(rDl v'S)K,.(r'Dlv'S)+AnK'n(rDl v'S)] =Dnl'n<rDlvsTlI:z) (A-30)

..
+ DiZL En cos( nB:z)BnK,,(rov'S) (A-ll)
~(r' ozVSTh:z) K'n(rodsTlI:z)+Cnl'n(rDlvsllI:z)=AIZBnK'n(rDlv'S) (A-3I)
Combining Eqs. (A-28) - (A-29) yields:
n-=O
~ K,.(r' 01 v'S) [ AIZI'n(rDl v'S)In<rDl~:z)-~(rDl v'S)1'n(rDl~]
Pm = DiZ [ Ko(RmVSTlI:z) + ~ En cos( nB:z)c"In(roVSTlI:z)] K,,(rol v'S)I'n(rDlVST) I:z)-A12K'n(rDl v'S)~(rDl VST)l:z)

+ Dil
..
L En cos( nBI)DnIn(rovsTh:z) (A-I2) (A-32)
n-=O
~(r' mVsT)I:z> [ AIZKn(rDlVSTJ I:z)K'n<rDl v'S)-K'n(rDlVsT)I:z)K,.(rOlv'S)]
By invoking the addition theorem of Bessel function, Eqs. (A-II) and
(A-I2) are rewritten as: C
n
K,,(rOlv'S)I'n(rOl~:z)-AlZK' n(rOl v'S)~(rOl~
.. (A-33)
pDl=Dil Ko(RDl v'S)+ L En(DilA n + DZiBnfzl )
.. n-=O
In(r' m~[K,.(rOl~I' n<rOl~:z)-K' n(rOl~~(rOl~l
L (-I)me." cos( mal)Kn-m(r'OlVs)Im(ROlv'S) n
(A-l3) B K,.(rOlv'S)I'n(rDlVST) l:z)-AlZK'n(rOlv'S)In(rOlVSTJl:z)
m=-oo

Pm = DiZKo(RozvsT)l:z) +
..
L En (DiZCO + DilDnfl:z)
(A-34)

AIZK,.(r'01 v'S) [ I'n(rmv'S)K,.(rOlv'S)-In(rOlv'S)K'n(rOl v'S)1


.. n-=O
n (A-35)
L (-I)me." cos( ma:z)In-m(r'mVST) l:z)Im(RmVST) I:z) (A-I4) D K,.(rOlv'S)I'n(rodSTJ l:z)-AlZK'n<rOlv'S)In<rDlVSTJ l:z)
Substituting Eqs. (A-13), (A-25), and (A-26) into Eq. (A-6) yields:
where fij = cosB;lcosBj. Substituting Eqs. (A-13) and (A-I4) into Eqs.
(A-9) and (A-lO) gives: PWOl = ~l{ (Wz+dzM:z)[ Ko<RWOlv'S) + RWOlSlv'SKl(RWOlv'S)]
Dil(Wl + d3M l) + DiZd4M l = COl (A-I5)
D;z(Wz + dzM:z) + DildlMz = 0 (A-I6) (d3WZ+dzd3Mz-dld4M:z)[ IO(RWOlv'S)-SlRWOlv'SII(Rwolv'S)] }
where
(A-36)
WI = COlsKo(RWOlv'S) + RWOlv'SKl(RwOlv'S)(COlSSl + 1) (A-I7)
Substituting Eqs. (A-I4), (A-25), and (A-26) into Eq. (A-8) yields:
Wz = CozsKo(RwmVsT)l:z) + RwmVSTJlzKl(RwozVSTJl:z)(CmSSz + 1)
(A-I 8) Pwoz= ~I {-dIMz[Ko(RWOZVSTJl:z) + RwozSZVSTJ12Kl(RWOZVSTJI:z)]
M I = COls IO(RwOlv'S) - RWOlv'S Il(RWOlv'S)(COlSS l + 1) (A-I9)

dl =
..
Mz=Cozs Io(RwozVST)l:z)-RWD2VSTJlzIl(RWD2Vsnl:z)(CmsSz+I) (A-20)

L Encos(naa)DnIn<r'ozVSTJl:z) (A-2I)
+ d l Wz [ Io(RwoZVST)l:z)-SzRWOZVSTJlzIl(RWD2VSTJl:z)] } (A-37)

If the active well and the observation well have identical radii, then
n-=O
.. RWOl = RWD2 = 1, and Eqs. (A-36) and (A-37) can be simplified as:

= ~l { (Wz+dzM:z)[ Ko(v'S) + Slv'SKl(v'S)]


dz = L Enc"In(r'mVSTJl:z) (A-22)
Pwol
..
n-=O

L
(d3WZ+dzd~z-dld4M:z)[ Io(v'S)-Slv'SIl(v'S)] }
d3 = EnAnK,.(r'01 v'S) (A-23)
.. n-=O + (A-38)
d4 = L
n-=O
EnBncos(naa)K,,(r'Olv'S)

Where no = laz-<Xli and ~. Bn. Cn. and Dn are undetermining


(A-24)
PwoZ== ;1 g
dlVSTJlZ[ Ko(VSTJI:z)Il(VST)I:z)+KI(VSTJI:z)Io(VSTJI:Z>] (A-39)

coefficients. Combining Eqs (A-I5) and (A-I6) yields: where


COl(WZ + dzM:z) WI = CDlsKo(v'S) + v'SKI(v'S)(COlsS I + 1) (A-40)
Dil = (A-25)
~g Wz = CDzsKo(VSTJI:z> + VSTJ12KI(VSTJI:z)(COZSSz + 1) (A-4I)
-eOldlMz
DiZ = (A-26) M l = COlsIo(v'S) - v'S II(v'S)(COlSS l + 1) (A-42)
~g
Mz = CozsIo(vsT)l:z) - VSTJlzIl(VST)I:z)(CmsSz + 1) (A-43)
where
APPENDIX B
~g = (WI + d~l)(WZ + dzM:z) - dld~lMz (A-27) We hereby prove that the term Kn(rDv'S)cos[ n(B + 13)] and the
Substituting Eqs. (A-ll) and (A-I2) into Eq. (A-3) and rearranging term ~(rov'S)cos[ n(B + 13)] are the solution of the two-dimensional
gives: diffusion equation (Eq. (B-1». We start with the Kn terms.
In<rOlv'S K,,(r'Olv'S» + ~K,,(rOlv'S) = Dn~(rOlvsT)l:z) (A-28) az Po 1 aPo 1 POI az
-- + -- + ---
arDz z = sPo
rD aro roz aB
(B-1)
~(r' ozVSTJ dK,.(rOl VSTJ l:z)+Cn~(rOl VST) lz}=BnK,.(rOlv'S) (A-29)
We assume that:
Substituting Eqs. (A-ll) and (A-12) into Eq. (4) and rearranging
yields: Po = K,.(rov'S)cos[ n(B + /3)] (B-2)

294
SPE23444 L. CHU AND A. S. GRADER 9
2
Taking the derivatives with respect to fD and e, respectively, yields:
_-.!!...-COS[ n(8+~)] K.,(fD..JS) = s cos[ n(~)]K.,(fD..JS) (B-6)
~D
apD = _l...JScos[ n(8+~)] [ Kn-l(fD..Js)+K..+l(fD..JS)1 (B-3)
afD 2 :J According to the Recurrence fonnulae of Bessel functions 23 , it

-~cos[ n(8+~)][ n~~K.,_l(fD..JS)


follows:
azpD =
a~D 2 fD'IS Kn-l(fD..JS) - K..+l(fD'fs) =- 2~K.,(fD..JS) (B-7)
fD'IS
-2KnCrD..JS)- n+~Kn+l(fD..Js)J
fD'IS
(B-4)
Substituting Eq. (B-7) into Eq. (B-6) shows that both sides of the
equation are identical. It proves that the tenn cos[ n(8+~)]Kn(fD..Js~ is
a2- a solution of Eq. (B-1). Following the same procedure, and recalling
~2
= -n2cos[ n(8+~)] K.,(fD..JS) (B-5)
the space derivatives and Recurrence fonnulae fOf In shown in Eqs.
ae D
(B-8) - (B-9), the tenn cos[ n(8+~)]In(fD..Js) is a solution to Eq. (B-1).
Substituting Eqs. (B-2) - (B-5) into Eq. (B-1) and rearranging gives:
l'nCfD..Js) = ~ [ In--l(fD..Js)+ln+1(fD..JS)] (B-8)
s cos[ n(e+~)]K.,(fD..JS)- ns rcos [ n(8+~)][ K.,-l(fD..Js)-K.,+1(fD..JS)]
2fD'IS
In-l(fD..Js) - In+l(fD..Js) = 2nrVfD ..Js) (B-9)
fD'IS

TABLE 1
DATA FOR EXAMPLE PROBLEM

.1t(hrs) PWD2 .1t(hrs) PwD2

0.00107 0.‫סס‬OO207 0.23533 0.1195304


0.00171 0.0003395 0.42787 0.0982708
0.00235 0.0012938 1.06967 0.0451845
0.00428 0.0072714 1.71147 0.0241296
0.D1070 0.0323376 2.35327 0.0152936
0.01711 0.0513869 4.27868 0.0069951
0.02353 0.0653722 10.6967 0.0025134
0.04279 0.0911967 17.1147 0.0015350
0.10697 0.1195965 23.5327 0.0011048
0.17115 0.1231838 42.7868 0.0005998

295
SPE 2344\4

lOt I' "lilt!' iii 1111


I._11'1'
,
'''I ..' ; , i i ii~

--
active
well E
'a
m
(1)
.c
co
~ 0.1
'a
"Cii
...
(1)

region 1
0.01 ' '" I ! II!! ,«!!, It!' I !, "! ! ! I ! ! ! " ! 1 "'W
0.001 0.01 0.1 10
time (min)
Figure 3: Match of the Ferris and Knowles 1 [1954] data.

Figure 1: Schematic of the composite reservoir configuration

..
Q)
::;,
,=
- - Co 10'C02 0
= ,=
- - CO2 Co 10'
= ~
Q)
0.1
CO2 0=
=
10'
=
10 2
=
:g 10'

..
III
= = =
~
5, 52 0 10'
Q) 0.1 =
10'
C. C. 0.Q1 =
10'
III III
:n :n
'E 'E
o o 0.001
.~ 0.01 'iii
C
Q) Q)
E
:s .50.0001
'C

0.00"'0.1 1 10 100 1000


dimensionless time I dimensionless storage
Figure 2: Effect of interwell distance on interference responses of a Figure 4: Effects of observation wellbore storage on the interference
homogeneous system. pressure responses in a homogeneous reservoir.
SPE 2 J 44 It

!:::J
III
!
:::J
:g 0.1
S2= 0.0
L D= 100.0
- C D,=10oo0
Coo= 50
- - C Ol = 10000
Coo: 10000
e
c.
III
! S,=
0.0
III
C. CD
III 3.0 "2
m 0.01 6.0 o
9.0 'iii
"2 I:
o CD
"0 E
I:
~ 0.001 :c
:c "CD
:;::

h1' ""H'l"I"I1~ " ,,"~III I I 'I"~*n I II"~I\MI I),,~~~O


:co
E
dimensionless time/dimensionless storage
Figure 5: Effetcs of wellbore skin in the active well on interference pres-
sure responses.
modified dimensionless time
Figure 7: Verification of the correlating parameter for the homogeneous
case with storage and skin in the active well,
(CD1L6)(/nCD + 28).

!
:::J
III
- C D,=10ooo
! S,=O.O
Coo= 50
III
:::J L D= 100.0
= ! 7. 5.0 13.100

:g 0.1 - - C Ol 10000
Coo: 10000
C.
III
0.1 8. 7.5 14.250
9. 10 15.500
! S2=
0.0
. III
CD
10.25 16.750
C. 11.50 17.1000
III 3.0 "2 12.75

m 0.01 6.0 o
9.0 'iii 0.01
"2 I:
o CD
'iii E
I:
CD 0.001 :c
E " 0.001
:c CD
:;::

b1 I I I "'g'l' I I ,,"I~ I I I "I'~'b I I I ,,~ltb I I I ,,~Woo I ')"i~txJo


:co
E O.OOOb.lff
dimensionless time/dimensionless storage 1 10 100 1
modified dimensionless time
Figure 6: Effetcs of wellbore skin in the observation well on interference
Figure 8: Type curve for interference slug testing in a homogeneous
pressure responses. reservoir with wellbore storage and skin in the active well.
seE 2., 44 4

...;:,
Q) ...;:,
Q)

Co (Ln Co+ 0.8(S, + S2))/~o


1/1 1/1
1/1
1/1
...
Q)
Q.
...
Q)
Q.
1.0.10
2.0.25
7. 5.0
8. 7.5
13.100
14.250
3.0.50 9. 10 15.500
1/1 1/1 4.0.75 10.25 16. 750
1/1 1/1 5.1.00 11.50 17.1000
Q)
Q)
0.01 C 2.50 12.75
C
0 ..
0
0
'iii 0.01
'iii c:
c: Q)
~ 0.001 E
:e :e'C 0.001

-:e
'C
.2!0.OO01

0
E
0.1
tim. (hr8)
1
Q)
:;::
:e0
Eo.OOOb.
1e-0 6.01
modified dimensionless time
Figure 11: Type curve for interference slug testing in a homogeneous reservoir
Figure 9: Type curve match of the example. Interference slug testing in
with identical wellbore storage values in both wells and arbitrary skin
a homogeneous reservoir.
values in both wells.

•.
N
:g

A.... . ...."."..
1.0' s,
~
0
955.8 200 1.5
14874.4 1000 7.5
3518.0 500 10
53601.7 460 7.5
active
"

T~
l!! 8 s 176955.4 800 5.0
:> I 595336.3 1500 4.0
'"
'"
[Q1
c 108716.0 200 5.0

~
l!!
Q,
c s 9580717.6 2000 2.5

..'"'"
I 1640159 800 4.0
0.01
C
0

E
..
'iii
l:

0.001


'6
'tI
.!!!
:s
E
0
0.0001
... • +J ·
D
1e-()50!1 l!! [ II!I~ ! I f II ~b ! II t;'&o ' t ! !IiYoo I ! t ~1Moo ! !);! ~:15

Figure 10: Verification of the correlating parameter, (CDIL6)(lnCD+0.8(S,+S2))' Figure 12: Composite reservoir configurations derived from the general case.
SPE 2.3 44 4

A. 2 : ..
CI)

~A~A e
E+ut
10 ::s
0.1 ~ -, --, 100 ~ 0.1
1000
~
I!!::> C.

~
~
III
III
I!! 0.01
a. 0.01 CI)

III "2
III
Gl
o
C 'iii 200
0 c
.~ 0.001 ~ 0.001
Gl
E
'6 :0

0.0001 10000 1e+05


dimensionless time I distance square
1e-0!b!1 I ! !",,,~r , , ,,, !lJ'x ! ,!, 1 'J~ft ' , , "Jut,,"' !, '!WA Figure 15: Effetcs of the location of the observation well around a constant
pressure circular subregion on interference responses. Curve AA
represents the interference response in a homogeneous system of
well A. Curve AAA represents the interference response of well A
Figure 13: Effetcs of mobility ratio on interference pressure near an impermeable circular boundary. Curves A-E represent
responses of an observation well located in the interference responses near a constant pressure circular boundary
of the corresponding wells.
center of a subregion.

'"
:g

..
CI)
0.1

..
CI)
::s
fII
0.1
= 0.01
= 0.001
-
CD,= 10' CD2 = 0
CD2= CD,= 10"
5, =5. = 0
::s

..
fII
fII
CI)

fII C. 0.01
A12 = 1
~ fII
fII
C.
fII ~
fII C
0.01 .2
~ 0.001
C fII
0 C
CI)
'iii
cCI) E
:00.0001
E 0.001
:0

1e-D!b.1 1 10 100 1000 10000


0.00010:1 '" ,I! '''~ ,! I ",~
o dimensionless time I dimensionless storage
dimensionless time I dimensionless storage Figure 16: The effects of mobility ratio on interference
responses of an observation well. The active well
Figure 14: Effetcs of storativity ratio on interference pressure responses of an and the observation well are located symmetrically
observation well located in the center of a subregion. on both sides of the boundary. The interwell dis-
tance is 100 'w'
seE 2344 4
O.1~ 'iiiili'! i i~' "Iilli! i """1

~ ~
:::J :::J

~C.
III
III
0.01 CD
Q. 0.1
III
:g III
III
CD
'2 '2
o o
.~ 0.001 'iii 0.01
CD c
CD
E E
:c :c
0.000'10.1 1 10 100 1000
dimensionless time / dimensionless storage dimensionless time/distance square
Figure 17: The effects of storativity ratio on interference Figure 19: The ellects of an impermeable linear boundary and the location
responses of an observation well. The active well of the observation well on interference responses. The wells are
and the observation well are located symmetrically on located on an arc with a radius of 80 wellbore radii, with a
both sides of the boundary. The interwell distance is separating angle of 45 degrees. The a~tive well is 100 wellbore
100,w' radii from the linear boundary. CD = 10 , and S = O.

0.3

..
CD
Angle:
40
30
1]12=

A12 =
1
100
C02 = CD,= 1ll"
..
CD
:::J
III
I A

:::J 20 III
5, = 52 = 0
..
III 10 CD
III 0 Q.0.2
CD 0.1 Distance=100 rw
C. III
III III
III CD
CD '2
'2 0
0 'iii 0.1
~ 0.01 C
CD
CD E
E
:c :c

dimensionless time / distance square


dimensionless time/dimensionless storage
Figure 20: The ellects of an impermeable linear boundary and the location
Figure 18: The effects of the location of the observation well on of the observation well on semi-log interference responses. The
interference responses in a composite reservoir. The wells are located on an arc with a radius of 80 wellbore radii,
wells are located on an arc with a radius of 100 with a separating angle of 45 degrees. The active well is 100
wellbore radii, with a separating angle of 10 degrees. wellbore radii from the linear boundary. CD = 104 • and S = O.

Potrebbero piacerti anche