Beruflich Dokumente
Kultur Dokumente
AUGUST, 1962
SPE 177 899
comparatively low and a high pressure differential exists .6.t. For the linear case it is assumed that the production
across the formation during the filling of the tool. Since is from an area of cross-section A, for the cylindrical case
there is no provision in the present formation tester for from the face of a cylinder of height h, and for the
measuring the instantaneous flow rate, it is necessary to spherical case from the surface of a sphere of radius Q.
make certain assumptions concerning this variable. Two The build-up equation appropriate for linear flow, included
idealized cases will be considered: (l) where the flow for completeness only, may be derived from the work
rate is constant during drawdown and zero after the tool of van Everdingen and Hurst' as
is shut in and during build-up, and (2) where the pressure 2fL V[ _ _]
is constant at the perforation during drawdown, and the pet) Pi - - - = - - yT + .6.t - Y.6.t
= (1)
Y7rakA T
flow rate is zero during the build-up.
The equation for cylindrical flow, based on Horner's de-
FLOW GEOMETRY- velopment' is
ITS EFFECT ON PRESSURE BUILD-UP pet) = p, - 4:kh ~ [In(T + .6.t) - In(.6.t)] , (2)
Fig. 2 shows a schematic comparison between the flow and the equation for spherical flow as developed in the
geometry for the wireline formation tester and the spheri- present paper is
cal flow geometry assumed in the mathematical develop-
= Pi - 4:k ~: ;-[).6.t - ~ .6.J
ment. The arrows indicate the direction of fluid flow
during the drawdown period. As will be shown, the pres- pet) yT (3)
sure disturbance produced in the formation during the where fL = fluid viscosity (poises),*
drawdown period will extend for some distance away k = formation permeability (sq cm),
from the perforation. The volume of formation which a = fLCcp/k (seconds/sq cm),
is affected by the pressure disturbance (and which con- c = compressibility (sq cmj dyne) ,
tributes to the subsequent pressure response during build- V = total volume of fluid produced (cc),
up) is large enough so that for practical purposes the T = total time the tool was open (seconds),
influence on the build-up of the impervious borehole and .6.t = time after the tool is closed (seconds) (time
the nonspherical perforation becomes unimportant. for build-up),
The choice of the size of the spherical sink (or "per- Pi = initial reservoir pressure (dynesjsq cm),
foration") to be used in the mathematical model, which pet) = pressure at any time t (dynesjsq cm),
during drawdown will have a "flow conductance" equiva- A = cross-sectional area (sq cm), and
lent to that of the actual perforation, will be discussed h = bed thickness of interval tested (cm).
in detail in the section, "Effect of Perforation Shape". An inspection of these three equations reveals that: (1)
It is in the essentially spherical flow geometry in thick each has a different time dependence; (2) the factor
beds that the formation tester differs most strikingly a (which is dependent upon the fluid compressibility)
from conventional drillstem tests. For most production enters into the equations for linear and spherical flow, but
and drillstem tests, it is reasonable to assume that the not into the one for cylindrical flow; and (3) a geometri-
flow has cylindrical symmetry about the borehole axis. cal dimension (A, or h) enters into the equations for
This is also true for the formation tester in thin beds. linear and cylindrical flow, but not into the one for spheri-
Different flow geometries lead to completely different cal flow. This last point suggests that for spherical geom-
equations for the analysis of the pressure build-up re- etry the exact size and shape of the perforation do not
sponse. To illustrate this, consider the three idealized flow control the pressure build-up curve and that the pressure
geometries that can be handled exactly-(l) linear, (2)
"Note that during the theoretical development egs units are used. In
cylindrical and (3) spherical flow. For each case the the examples. practical units are used as indicated.
formation is assumed homogeneous, at an initial pressure
Pi, and produced for a period of time T at a constant
rate. The total volume of fluid produced in time T is
denoted by V in each case, and the time after shut-in by
SHUT-IN STATIC
\ I
/
i OPEN
~ t
RESERVOIR
PRESSURE:
PERFORATION
~\/
'"a:=> Pi
V)
V)
'"a: (0 )
Q.
~~
k LARGE
TIME --+
t
STATIC
RESERVOIR
PRESSURE:
/' /1 '.T.PAD
\
BOREHOLE SPHERICAL
....EL.
'"a:=> (IMPERVIOUS) PERFORATION
V)
V)
OPEN
'"a: (b)
Q.
~ k SMALL (0 ) (b)
FIG. 2-COMPARISON OF FOR~IATION TESTER FLOW GEOMETRY (A)
FIG. I-TYPICAL PRESSURE CURVES RECORDED BY WITH THE SPHERICAL FLOW GEOMETRY (B) ASSUMED IN THE
WIRELINE FORMATION TESTERS (A) IN A HH;lI· MATHEMATICAL DEVELOPMENT. THIN ARROWS INDICATE DIRECTION
PERMEABILITY BED, AND (B) IN A LOW-PERME- OF FLUID FLOW DURING DRAWDOWN. SAMPLE CHAMBER OF TESTER
ABILITY BED. AND CONNECTING TUBES NOT SHOWN.
- e
--;;;;z J T + At] t> T
erfc1~' (9)
of time the flow in the vicinity of the perforation will have
- - .fLQ { f ja(r - a)' _ [J-'-(r---a~)' FOR THE CONDITIONS SHOWN. NOTE THAT THE PRESSURE ACTUALLY
p(r,t) - Pi 4-rrkr er c 1 4t exp 1 a' OBSERVED WITH THE TOOL DURING DRAWDOWN MAY DIFFER FROM
CURVE SHOWN HERE.
P ( r,
t) -
- Pi
_ (Pi - Pd)a{ f
r er c
J a(r - a)'
1 4 (T + At)
2200 t=20
At-IO
~-----;:-
4:k[ 2In(~)1
center of a bed. For this problem the solution for draw-
c- (13) down can be put into either of the two forms shown in
Eq. 14 (see Appendix B).
Eq. 13 has the same form as the C found from Eq. 5a CJ) 2 [ nhya
if a is replaced by a e !! where p(t) = p(t) ,pher;e.' - 2:: 11 erf 7. I A
ut
L n = 1 - v
ael! = --(""'""2--=-L~)'
(13a) _ erf nhy--:X ][~/L ~]
21n If" 2yT + At 47rkh T '
If the dimensions L
aet!
= 6 in. and R = VB in. are used,
turns out to be 0.66 in. or approximately 2 cm. The
p(t) = p(t)eYlind,;e., - n~l 2{ Ei [ -(27l'n)'
corresponding correction of the spherical flow solution
would be of importance only for drawdown. The shape (T :2Al)] _Ei [ -(27rn)' (:~2)]}
of the perforation does not affect the build-up analysis.
This model accounts for the effect of perforation shape.
However, it must be emphasized that, since the perfora-
[4~h ~] . (14)
tion is formed in a rather violent manner, one cannot be In these formulas, pet) ,phr,"ea' and p(t) eylindr,ea' represent
certain that the formation next to the perforation has not the solutions for spherical and cylindrical flow given by
been damaged in some way. For example, if the rock Eqs. 3 and 2, respectively. The summation represents cor-
surrounding the perforation were fractured, it would have rections applied to these simple cases. From the form of
the effect of making R somewhat larger and, hence, bring the corrections, it is clear that for t < <ah' the first form
a'l! nearer to L. On the other hand, any action that would of the solution is preferable, while for t > > ah' the sec-
reduce the effectiveness of the perforation would tend to ond form is preferable.
decrease R. In any case, the uncertainties inherent in the Eq. 14 may be used to study the effect of bed thickness
evaluation of the conductance of the perforation tend to on the pressure build-up curves for a small perforation.
make permeability based on drawdown data somewhat This equation has been plotted in Fig. 5 for beds 3.3-,
questionable. Even if these limitations can be overcome, 10.4- and 32.8-ft thick, where aa' = 1 second, p, = 2,550
2000:
-,(1'ICi''- ~I=-
~v ------
1
'--. --. -.A.-----+---+--f.---?./
I ! I
2200 sl'~
400:
I i 600 1---- f----.. -.-- - - --------
ilt/ (Hill) -----'--0> -1 I 1--- _._.
o - ___ ------- -- j - - - . _ ! --- -.-1. ---
.001 .01 .1 1.0 200
IIt/(Ttllt}-~ 1
FIG. 5-THEORETICAL PRESSt'RE BUILD-UP PLOTS. CURVES A ARE o
,001 ,01 I.
COMPUTED FROM EQ. 14 FOR THIN BEDS, 3.3-, 10.4- AND 32.8·FT
THICK. CURVES B ARE THE CORRESPONDING PLOTS FOR FIG. 6--COMPARISON OF COMPUTED PRESSURE BUILD-UP CURVE
CYLINDRICAL FLOW. FOR 3.3-FT BED WITH THAT FOR A THICK BED (SPHERICAL FLOW).
f-------+----+--
DEPTH·64881t.
Q::3.58cc/sec, water
T. 510 sec.
-J ,
-~1401~+
MV
o- - - ----
W'
MICRONORMAL 10 2" 20
-- --- ---- - _ .. - - _·t --- ------. - -- ----.-- ---
m'475psi/cycle !
:s --. -
i
2525 psi. .......... :
SHALE
. ~ r- STREAKS
V>
C.
.~
•.J."
-
TEST DEPTH
1500 ~------~~~--+-------~--- .. 6488
::> ~ <T>
(]I -.
V> 0
V> .":,~
1000 ~ ------t---t----t----.-.--+----i D 0
Q.
\ .'
500 ,0 -- [l: F
III/(T+III) ~
O~~====~-~-~+_~----~-~
.01 .05.1 .5 I.
FIG. 8-LOG EXAMPLE ILLUSTRATING A CASE WHERE TESTED INTER-
FIG. 7-PRESSURE BUILD-UP PLOT OF DATA FROM FORMATION TEST VAL Is EFFECTIVELY A THIN BED BECAUSE OF NEARBY IMPERVIOUS
OF A WELL IN NORTH TEXAS IN SECTION SHOWN IN FIG. 8. STREAKS ABOVE AND BELOW.
1. Fischer, J. S., Havard, E. L. and Van Ness, A. E.: Paper Taking the Laplace transform of Eq. A-5 and using
1244-G presented at SPE Drill. and Prod. Prac. Meeting in Eq. A-4,
Lafayette, La. (April 24, 1959).
2. Petty, L., Jr. and Phillips, H. c.: Drill. and Prod. Prac., - 471'kA ( - )
API (1958) 48. Q(r,s) = -fL- I + vas r e
-y--;;:,
(A-6)
3. Milburn, J. D. and Howell, J. c.: lour. Pet. Tech. (Oct.,
1961) 987. Now the transform of the total flow rate Q(t) [=
4. Finklea, E. E.: Drill. and Prod. Prac., API (1959) 131.
5. Horner, D. R.: Proc., Third World Pet. Congo (1951) Sec. II, Q(a,t)] into the spherical perforation is Q(s). Thus,
503.
6. Lebourg, M., Fields, R. Q. and Doh, C. A.: Trans., AIME Q(s) Vasa
(1957) 210, 260. A=~ e
7. van Everdingen, A. F. and Hurst, W.: Trans., AIME (1949) 471'k (1 + vasa)
186,305. so that Eq. A-4 now becomes
8. Horner, D. R.: Proc., Third World Pet. Congo (1951) Sec. II,
503.
9. Stratton, J. A.: Electromagnetic Theory, McGraw·Hill Book per,s) = Pi _~ Q(s) e'Y"*U) (A-7)
Co., Inc., N. Y. (1941) 207. s 471'kr (l + vas a)
10. Ruhl, W. and Schmid, c.: Geolog. lahrb. (1957) LXXIV, and if we let r = a,
447.
11. Kunz, K. S. and Moran, J. H.: Geophysics (1958) 23, No.4 Pi fL Q(s)
770. ' pes) = pea,s) = - - -- . (A-8)
12. Campbell, G. A. and Foster, R. M.: Fourier Integrals for s 471'ka (1 + vas a)
Practical Applications, D. Van Nostrand Co. (1948).
The inverse transform of this equation may now be
found by applying the convolution theorem"'* to the last
APPENDIX A
term on the right to find Eq. 4 of the text.
HOMOGENEOUS BEDS, SPHERICAL SYMMETRY
CASE OF CONSTANT PRESSURE p. IN
DEVELOPMENT OF EXPRESSION FOR pet) PERFORATION DURING DRAWDOWN
IN TERMS OF FLOW RATE
The differential equation for the pressure in isotropic If the pressure in the spherical perforation during draw-
media is down is held at a constant value Pd, then pea,s) = P./s;
2op and, for r = a, Eq. A-4 becomes
\1p=a- (A-I)
at -v'~a
where a = fL'c/lc/k and where \1' is the Laplacian operator. P. = Pi +A _e_ _ , (A-9)
If P has spherical symmetry, so that it depends only on s s a
the radius r and the time t, Eq. A-I becomes from which
-Vas a
I a( ,op) op e
? a,:- r a;: = a at' A = - (Pi - Pd) a - -
S
(A-lO)
which may be written in the form and Eq. A-4 can now be written
o'(rp) o(rp)
---aT = a ---at . (A-2)
- a e
- yas(r-a)
Pi fLQ e
-V~r-a) I /z
p(rs) = - - - - - - - - = , - -
, s 47rkr s(1 + Vasa)
(A-14)
Iz~ SECOND ORDER I MAGE
The inverse transform of Eq. A-14 yields Eq. 10 of the FIG. B-1
text.
For the pressure at the tester, set r = a to get the pressure at the spherical perforation then is given by
Eq. B-3 because, there, rn z nh.
pes) = pea,s) = ~ __ fLQ [ 1 ]
s 47rka s(l
(A-15 )
The inverse transform of Eq. A-15 yields Eq. 8 in the
+ Vas a)
- -
pes) = pea,s) = P(S)'Phc,'cn'
-
+
fLQ(S) [
47rk 2
j J
n:: e:~a,
text. (B-3)
To obtain p(t), first consider the term
APPENDIX B
-
,- [
7r 00 -n"va,]
FINITE BEDS
G,(s) = 2 ... /_ ~ _e_ _ .
1 a n=l nh
The method of images may be used to extend solutions Then, take its inverse transform
for infinitely thick beds to the case of a finite bed of
a]
~ "l 00 [
thickness h. Assume that the tester is at the center of the G,(t) = exp ~ -(nh)--4 . 0
(B-4)
t n = 1 t
bed. The distance r n from the nth image to a point P in
the bed having spherical co-ordinates (r, 8) is With this result for G 1 (t), the convolution theorem
can be used to obtain the inverse transform of the whole
rn = Vr2 + (nh)' - 2nhr(cos 8), (B-1) last term of Eq. B-3, yielding the first Eq. 14 of the text.
as illustrated in Fig. B-1 for the second-order image below This assumes that the flow rate is constant during draw-
the bed. down.
Adding all terms of the form of Eq. A-7, Using Poisson's sum rule, an alternate series for G,(t)
,..",. -Va'" r can be written (Eq. B-5) which is a more convenient form
- Pi fLQ(S) n= +00 e " for use when t is large.
per,s) = - + -- ~ -- (B-2)
s 47rk n=
where the approximation has been made that
-00 rn
G,(t) = ~-;;ht
J-; 2[+ 2"I - (14~-:;;:t
J ah')
+ n:: exp (4 -- :h
'tn')]
Vasa
__e"-----==-_ z 1.
00
1 2
(B-5)
(1 + vasa)
Eq. B-5 can be used to obtain the second Eq. 14 of the
The term corresponding to n = 0 can be separated out text since the middle term of Eq. B-5 will just cancel
for convenience and called P(S)'PhH,ca,. The transform of P(t)'Ph."ca, for large values of t.t. ***