Documente Academic
Documente Profesional
Documente Cultură
Well Interference
J. E. WARREN
MEMBER AIME GULF RESEARCH & DEVELOPMENT CO.
J. H. HARTSOCK PITTSBURGH, PA.
JUNIOR MEMBER AIME
p, - p(r,t)
70.7 qfL f3 . ( - 948 'cP fL C r) 70.7 q fL f3
kh El kt 948 cP fL CD'
a=
(1) k
In the derivation of Eq. 1, it is assumed that a well of and
infinitesimal radius is producing at a constant rate from x = rAID; (1 - x) = rBID.
an infinite reservoir. Because of the assumption of a Let u be a quantity that satisfies the following equa-
vanishing well radius, the solution is valid only for large tion.
values of time. Horner5 discusses the limitations on the !:::.<I> = - Ei ( - u') . (3)
use of this equation; Elkins· presents field data which
Then,
support Horner's conclusions. If two or more wells are
producing from the same reservoir, the resulting pres- , ax'
UA = -1-'
sure distribution is obtained by the superposition of
the solutions for the individual wells. and
The object of this paper is to develop approximate a (l - X)2
UB' =
techniques for determining the time and the place at 1
which pressure disturbances originating from two ad- Solving for t and x,
jacent wells begin to interact, or interfere, significantly.
Two distinct approaches will be followed. First, the ef-
fect due to each well will be considered separately;
then, the cumulative effect of the two wells will be ex- x = ---,--- (4)
amined. U + U
A B
1.0
~ The superposition of two solutions having the form
of Eq. 2 leads to the following expression.
o. 1 1\ where
\
~
:>
(ApA + AP/J) k h
~ 70.7 qA JL f3
W and
A = qB/qA'
.0 I
Differentiating Eq. 5 with respect to x,
-ax' -a(1-x)'
O<P AH 2e-t- 2Ae t
---ax =-x-
-
T (1 - x) (6)
For finite production rates from both wells and
.00 I o~ X ~ 1, the derivative is equal to zero only at the
point of maximum pressure. Equating the derivative to
zero and re-arranging,
a 1 AX
- =--ln~-. (7)
t 1-2x I-x
.000 I
.ol 0.1 1.0 10.0
u The values of aft that satisfy Eq. 7 for various com-
binations of x and A are indicated graphically in Fig. 2;
FIG. 1- u VS - Ei (- u'). for convenience, it is assumed that A :2: 1 and x ~ .5.
With the relationship between aft and x in hand, A<PAB
EXAMPLE may be expressed as a function of aft and A. via Eq. 5;
The following data were used in the example given the results are shown in Fig. 3.
by Stevens and Thodos.' For specified values of <P AB , a and A., the interference
k = 270 md, fJ. = 45 cp, characteristics can be found quite simply. First, aft is
cf> = .22, C = 5.15 X 10-6 pst', read directly from Fig. 3; then, x is obtained from Fig.
h = 57 ft, qA = 550 STB/D, 2 for that aft and the appropriate A..
D = 1,000 ft, qB = 1,100 STB/D,
EXAMPLE
f3 = 1.10, APA = ApB = 25 psi.
Using the same data as in the first example, required
Required calculations are the following. calculations are as follow.
(25) (270) (57)
(50.) (270) (57)
,M'A = (70.7) (550) (45) (1.10) = .200;
(70.7)" (550) (45) (1.10) = .400;
(25) (270) (57)
(.948) (.22.) (45) (5.15 x 10-') (1000)'
A<PB = (70.7) (1100) (45) (1.10) = .100; a=
(270)"
(948) (.22) (45) (5.15 X 10-') (1000)'
a=
(270) "A finite pressure drop must be specified because the form of
the equation is such that an infinitesimal disturbance is propagated
179 hours. at a velocity that approaches infinity.
From Fig. 1, u" is found to be 1.02 and UB is 1.20.
Using the equations of Eq. 4.
o. 6
179 hours
t = (1.02 + 1.20) , = 36.3 hours;
)..:.1.0
--
o. 5
1.02
Therefore,
x = (1.02 + 1.20) = .459 .
fA is 459 ft and fB is 541 ft.
0.4
)."1.5
A'" 2
,.- -...---: --
~~~
f--
..- -t::
--b ;::
~ ~~
~
/"'
This calculated time of interference does not agree
with that obtained in the referenced example because
X 0.3
, ., --
1-,- '--
~ ~v
~
the latter contains a computational error in the argu-
-- ----------
)..' 4
0.2 '--
5
ment of the exponential integral. )..<
).< 6 V
).=8
o. I '" 10
SECOND METHOD
Since it seems difficult to attach any particular phys- o
0.1 1.0 10
ical significance to the definition of interference pre- alt
sented in the preceding section, we propose an alterna- FIG. 2-VALDES OF alt THAT SATISFY EQ. 7 FOR VARIOUS
tive definition: interference occurs when the maximum COMBINATIONS OF X AND y.
NOMENCLATURE
6e C = system compressibility, psi-'
D = well spacing, ft
I
4r-~--, ~~~~~~~~ __ 00
r _,
2r-~+--t-t-+---i-t-R~~~~-i----'--,---;--,
~t--
- Ei ( - y) = exponential integral, J _e_ dt,
~
I
dimensionless y
S = saturation, dimensionless
Z = gas deviation factor, dimensionless
h = formation thickness, ft
= 179 hours; ko = effective permeability to oil, md
( 1100) p = pressure, psi
,\ = (550) = 2.0 . q = oil production rate, STB/D
r = radial co-ordinate, ft
From Fig. 3, aft is found to be 5.0. Therefore, t = time, hours
u = positive root of Ei (- u ) + ..0.<1.> = 0,
2
(179)
t = (5J)T = 35.8 hours. dimensionless
x = distance from Well A along the line
From Fig. 2 with aft = 5.0 and ,\ = 2.0, x is .450. centers, dimensionless
Consequently, rA is 450 ft and rn is 550 ft. .0.<1.> = potential change, dimensionless
These results do not differ greatly from those obtained a = time characteristic of system, hours
by the first method; however, for larg<>r pressure drops f3 = oil formation-volume factor, reservoir
and/or larger values of '\, the answers do diverge sig- bbI/STB
nificantly. It can be shown that the procedure of Stevens ,\ = ratio of production rates, dimensionless
and Thodos actually leads to upper bounds on both t f-t = viscosity of oil, cp
and x for a given set of conditions. ¢ = porosity, dimensionless
SUBSCRIPTS
CONCLUSION A = Well A
B = Well B
Two direct methods for determining the presence of
g = gas
interference between two adjacent wells have been pre-
i = initial conditions
sented; both are based on the continuous line-source so-
0 = oil
lution that applies to an ideal physical system. Simple
p = pore volume
examples have been used to illustrate the techniques de-
w = water
scribed.
we = connate water
The validity of the results obtained is limited by the
assumptions made in the derivation of Eq. 1. In gen- REFERENCES
eral, there is not sufficient information available to jus-
tify the use of anything other than average values for 1. Stevens, W. F. and Thodos, G.: "Prediction of Approximate
the formation parameters, i.e., ko, ¢ and h. The com- Timc of Interference Between Adjacent Wells", Trans.,
AIME (1959) 216, 77.
pressibility and the viscosity, however, will be pressure-
2. Muskat, M.: Flow of Homogeneous Fluids Through Porous
dependent although only one mobile fluid is present. Media. McGraw-Hill Book Co., Inc., N. Y. (1937) Sec-
Parsons' correctly indicated that the system compres- tion 10.15.
sibility may be defined by the following expression. 3. Bruce, W. A.: "Pressure Prediction for Oil Reservoirs",
[+ - ~ G!)
Trans., AIME (1943) 151, 73.
4. Kelvin, Lord: Mathematical and Physical Papers, II, 41.
C = t ] Sy + C w Sw + c" So + C p
5. Horner: Proc., Third World Petroleum Congress, Sec. II,
(8) 503, E. J. Brill, Leiden (1951).
The use of this definition leads to a nonlinear form 6. Elkins, L. F.: Drill. and Prod. Prac., API (1946) 109.
7. Parsons, R. L.: "Discussion on Prediction of Approximate
which has no formal solution. Therefore, a necessary Time of Interference Between Adjacent Wells", Trans.,
additional restriction must be imposed; i.e., there is no AIME (1959) 216, 436. ***
!SEPTEMBER, 1960 91