Professional Documents
Culture Documents
SPE Paper - Pressure For Bfield PDF
SPE Paper - Pressure For Bfield PDF
formation below a shale break initially was believed Well 244 was pulsed first; well 245 was pulsed
to be isolated from the upper Tar Spriugs. Later 17 hours later after perturbations from the first
pressure tests indicate at least some coimaunication test had attenuated. Pressure responses to these
across the shale. pulses are shown in Figures 4 and 5. Background
drifts in pressures were caused partly by nearby
Well completions and air permeabilities are waterflood operations. For example, the initial
given in Figures 2 and 3. Average porosity is downward trend resulted from startup of a produc-
estimated to be @ = 0.178. Average compress- ion well, BFU23, the previous day. The overall
ibility of this liquid filled reservoir at resid- upward trend in pressures beginning after the
ual oil saturaton after waterflood (Sorw~ 0.28) first pulse in well 244 was caused by the higher
is calculated to be 6,8 x 10-6 v/v/psi. Reservoir post-pulse injection rate. Changes in rates in
temperature at the start of the pilot was 95°F; waterflood injectors outside the pilot also could
preflood water viscosity was 0.86 cp. During have affected pattern pressures. These effects
polymer drive the estimated viscosity was 2.0 cp. are undesirable during pulse tests since interpre-
Air permeabilities were approximately equal to tation depends upon deviation in pressure from a
brine permeabilities for a wide ra,,&ti of samples. known trend.
The ratio of brine permeability at Sorw to air
permeability averages about 0.17. See Table 1 and The initial pulse responses were interpreted
Figure 3. by the simplified procedure derived in Appendix A.
This procedure was developed to analyse these
The history of injection operations pertinent tests and the reverse pulse tests described in
to the pressure measurements is given in Table 2. section 3 following. (See Appendix B.)
Additional details of the flooding history during
pilot operations are discussed by French et al.l 2. Simultaneous Pressure Buildup-Falloff
Tests In Five-Spot Pattern Wells
TRANSIENT TEST METNODS
Wring initial operation of the preflood,
Four types of transient teats were perfartaecf water uas imjecte4 i*ecE the four eertter wells 241,
6iuring Ehe pilot: standad pttlse, pressure buildup- 242, 244, and 2&5. end water md td% were produced
fakloff, reverse puEse, WA produetiee drswdown. froai the eent=al wdl 243. Se=on sfter etsrwp,
(1) Shoptiy sfter begitmittg iajection ittto two of howEver, it beeessa evideztk khak tselk 242wmkd -t
-. the Pikot imjeceors ak oppoeike eormsre 04 @te tekts Sufffcht fk&4= St&la enuae=ceee~kkl
one-acre pattern, stendsrd pulses~ ware ebeerved wrEmvert 242 wes pgeced en wi~ psedeekfm W
i~ edjsces~ shut-itl isjeccors, (21 After eevarak ele=n sp kite wall. &fte= Z EE2 mi=~bs ef opere-
sorttha of waterflood operation, peessure falloffa tiw the entire pilot wee &hut-is an Febru=ry itl,
and bttildups were =aaure& follwin~ siteulka=ous l%ft i= arder tosAify surfsee facilities. The
shut-in of injectors and producers. status of each waEI just befcre shut-in is given
in Table 4. Aaerada type bottots+ole pressure
Later in the project two additional tests bocsbs were run down tubing in each injector while
were performed in the vicinity of injector 241. still injecting, and in producer 243 while still
(3) Responses from single production pulses from producing. (Well 243 had an extra string of tub-
shut-in observers were observed at the nearby ing to permit logging without interrupting produc-
injector 241 during injection. (4) In-situ mobil- tion.) The bomb in well 242 was run down the
ity measurements were made at the observers during tubing-casing annulus 30 minutes after shut-in.
flowing production that resulted from injection Injectors were equipped with 2 3/8-inch OD tubing
into nearby well 241. with bottom hole packers in the tubing-casing
annulus, The two producers had 2 3/8-inch OD
1. Standard Pulse Tests production tubing without packers in the 7-inch OD
(6.538-inch ID) casing.
After 20 years of waterflooding, new wells
were drilled in the watered-out reservoir in the Simultaneous shut-in was accomplished by
pilot area. Fairly standard pulse tests were shutting down the production pumping units and
performed by injecting into two of the future closing valves at a header located near well 244.
injection wells (244 and 245) and observing pres- Surrice pressures were monitored from the instant
sures in the other two shut-in injectors (241 and of shut-in until the wellhead pressure fell to
244) , All wells were equipped with tubing-casing zero. In a prior falloff test in 245, the well
annulus packers. In response wells 241 and 242, initislly was shut in at the header, but as soon
bottom hole tubing plugs were run above the pres- as the surface pressure fell to zero the wellhead
sure bombs to provide complete bottom hole shut-in, was bull-plugged to prevent surface injection line
The Sperry-Sun recording pressure gauges permitted drainage into the well, In subsequent falloffa,
105 hours of measurement with a precision greater this precaution was not taken and little difference
than ~ 0,005 psi and $ 0,5 minutes. Absolute was observed, For the first 72 hours pressures
accuracy was not critical since all measurements were recorded continuously with the Amerada bombs.
depended upon values relative to initial conditions, After 36 days shut-in, bombs were rerun for the
final me4surementa. Surface pressure measurements,
Injection and production rates prior to and converted to bottom-hole conditions, are combined
during the pulse tests are shown in Table 3, An with the bottom hole measurement to display over-
injection sequence in each injector consisted of all pressure behavior. See Figure 6.
three, 40-minute shut-in pulses 40 minutes apart.
!mm \lnRti GIMIRGEL. STSGEMHSR 3
3. Reverse Pulse Test DecesAer 21, 1968, about three months after start
of the final waterflood. (See Table 5.) Aa during
The praxiraity o: cnree observation wells, the reverse pulse tests, the observation wells
2M), 246 and 247, to the injector well 241 pro- were produced while injecting at a constant rate
vided a unique opportunity to perform precise in well 241. The wellhead pressures at the pro-
interwell pulse tests. Favorable conditions were: ducers were almost instantaneously reduced from
(1) a “hard” system, that is, completely liquid- the shut-in values and maintained at O psig for
filled injectors, reservoir, and producers, ten minutes. Again, produced fluids were collected
(2) single phase flow, and (3) enough pressure to in an 18-inch diameter calibrated vessel, Cumula-
make the producing observers flow. At the begin- tive volume was typically less than one-half barrel,
ning of August, 1968, after 2 1/2 months of con- (See Figure 9.)
tinuous injection into well 241, observation wells
began to flow without pumping. By September, the RESULTS
chemical flood fronts had passed the region between
Standard Pulse Tests
injector and observer, so that the formation was
at residual oil saturation after chemical flood,
The standard pulses shown in Figures 4 and 5,
Sore, with only aqueous polymer phase flowing. At and the reverse pulses shown in Figure 7 were
well 241, surface pressure while injecting was
analyzed by the method given in Appendix B. Results
about 630 psi. At well 246 shut-in surface pres-
are summarized in Table 6. Because of the uncer-
sure about 160 psi. The usual pulse direction was
tainty in extrapolating the original pressure
reversed, that is, pressure changes resulting from
trenda in the standard pulse tests, we decided to
a production pulse in the “observation” wells were
interpret only the first pulse in a sequence. In
sensed at the injection well, Production was
general, this is the preferred design since a
obtained by simply opening a valve at the observer
single pulse yields a maximum pressure for a mini-
wellhead and the reverse pulse response in injec-
mum extrapolation in time of the “base line” pres-
tion pressure was measured at the injector well-
sures. The advantage of multiple pulses is the
head.
assurance that the pulse response is real and
reproducible. If Ehe validity of response taust be
es~atslished, the second pulse ahcdd be delsged
un=il the ~imk szaximaais elesrlg esEe&liehed and
a new trcmd ha* seabiliaed. When iejeceion or
prchctiottkisteey is eoeqlicated, teatching the
rape by msm -E- s~~ks- caEct&*-
kions is neeessar~; howwer, the rmdks still m’e
Sajece m the eafeeEislita&ion 4 me aeemaq? ef
the exkrapelatiort ef the original tremls over the
entire period ef a~lysis.
If we assume that the open interval, h, ia Applying the same analyaes to these “reverse”
25 feet, and use the other values given in Table 6, pulses as to the previous standard pulses revealed
the calculated geometrical mean permeabilities for similarities and some interesting differences.
the pulse well pairs 244 + 241 and 245 + 241 are (See Table 6.) Because a polymer drive after
67 md and 44 md, respectively. Alternatively, the chemical flood is compared to initial waterflood,
directional permeabilities calculated from Equation both processes and pulse methods must be compared, ,
B,l do not require measurement of Apm since,
First, the calculated nobilities (~/IJ),
Kt$f@Cr2 which are determined independently of viscosities,
k~= are generally higher during the polymer drive. We
At
would expect this condition if residual oil had
The comparable directional permeabilities, using been removed by the chemical slug but if polymer
c = 6.8 x 10-6 vlvlpsi, are 48 md and 46 md. The viscosity were inadequate. Careful sampling of
near equivalence of these two directional perme- injected and produced polymer, and Brookfield
abilities, taken at 90° appeara to justify our measurements, established the viscosity to be near
later assumption of isotropic permeability, These 2 Cp, For that value, the calculated permeabilities
values also are in good agreement with the water are about three times those prior to chemical
permeabilities at Sorw in the “B” sand in well flood.
241. See Table 1. However, the geometrical perme-
abilities calculated above appear to be too high Second, calculated thickneasea of formation
for the mean of the whole 25-foot interval. The measured by the nearby observation well pulses
higher value of geometric permeability associated were considerably smaller than the previous values
with well 244 compared to well 245 suggests that calculated for longer distances across the pattern.
the formation may be thicker near well 244. An These valuea corresponded to the high permeability
estimate of thickness can be obtained directly “B” zone. (See Figure 10.)
from the maximum pressure if k@ is assumed equal
to kc. A combination of the oilfield unit Equa- Third, when a visectaity of 2 cp ia used, the
tions Ii,l and B.2 gives a form eitailar to Equation edculated penaaabilitfee ta pdyner sre shout
A.IT: hakf the ai~ Permasbkities d the cmes. Fer che
tkwea ebaerve= pu%ees dwing P&pa? &Eive, EMS
r&eie GE pokymer pemadtiliey eo &iE ~sbilitg
.- ‘“[-l[*l”?; is ~bkme. iE ri8Aduei 41 kad beeE4Eee-
For smt putees, f{k~>h my be eakeri *S miqr; eivd~ ~ved frcaa =he sake near MM &sjeetes,
however, * carreceie~, Fft@, can be 0be8i~ khe I-E ddeeed ~maabiEities Would indis*Ee
from Figure A.3 ao theE: a pdyaar resistance facta in addition te the
1 ~&hra pdytaer vieceeisyc
h.~. .—
36.5 @r2~
[ 1[ [i + F(tLB)] 1 Fourth, when a series of varyin& pulse length
For the 244 + 241 pulse, the correction ia only tests are cmpared for a given well pair, we find
about 2.5 percent. The thickness appears to be that, with two exceptions, the tsobility changee
34.6 feet compared to 24,4 feet for the 245+ 241 only elightly; however, there ia a trend of increaa-
pulse. Thus, it appears that the reservoir near ing thickness as the time to maximum response
well 244 likely is connected with the underlying increases, if the one-minute pulses are excluded.
zones. Also, the calculated kh is about half that
measured from air permeabilities of cores from the Fifth, effect of pulse length on the size of
25-foot interval. For this reservoir at residual maximum pressure response follows the prediction
oil from waterflood, the implied relative perme- of linearity in Equation A.18 since f(tLD)/me is
ability is too high, Measured core permeabilitiea nearly unity for the short pulses. Predicted
from the “B” sand in well 241 give a relative deviations for the one- and two-minute pulses are
permeability of 0.17. The effect of layering of less than four percent. For the five- and ten-
greatly dissimilar permeability zones accounts for minute pulses, corrections up to 40 percent are
at least some of this result, obtained from Figure A.3 but these still give
values less than observed, Contributions from
Reverse Pulse Tests layera with increasingly lower permesbilities
diminish the pressure response for these longer
The reverse pulse teats, in which observations pulses,
were made in an injector during injection, appeared
to behave very much like the standard pulses, Simultaneous Pressure Buildup-Falloff Tests
Care must be taken, however, to avoid feedback in
the system if a group of injectors are connected The pressure history of the pilot ehut-in
on a single header. A necesasry condition for from three seconds to 42 days is given in Figure 6.
this simple analysia is that the rate of injection Surface resdings at very early times are not accu-
be constant, If a change in rate of one of the rate as these measurement were taken viouslly
other injectors or producers affects the rate in from rapidly changing pressure Sauges at the header.
the observing-injector, the small pressure pulse Surface readings were continued until boteom-hole
will be distorted. preaaures had declined to 920 psi (zero surface
pressure). However, 30 seconds afeer shut-in,
bottom-hole readings also became available for
comparison with surface readinga. Because trends
ore comparable, the surfsce pressure readings were
WE 1Eueo Urtunufi k . mL Gw&rTe Len >
2 - displacing phase
APPENDIX A
ACKNOWLEDGEMENTS
I. DIMENSIONLESSPERMEABILITYFUNCTIONSFOR PULSE
TESTS
Over the years I have had a number of informal
communications with Shell co-workers which provided
The pressure response of a $ingle change in
a background for the ideas developed in this report.
injection or production from a line source or sink
I particularly wish to acknowledge assistance from
is given3 by the well-known relationship:*
informal reports of C. Bremer and J. A. Dillabough, ● .
&he computational work of N. A. Hyhill, the example
caaes of Appendix B provided by M. Prata, the
conduek of some of the field tests by J. O. Farmer
-d M. S. French, and the review of this treport by where
E. 1+. Mach.
mm
For raservoi~s tdtlt &rti*otropie persst&bHitg,
1. Frenek, H. S., et al: Weld Test af an Aquetws Coliiea4 has shown ehet the tw Perreeabflities
!$urfaetant System for Oil Recovery, Beaten occurring in Equation A.1 are net idemtieal. %Ms
Field, Illinois,” J$T, February k9731 resuit was ge=~eti=ad *Y Fspadoptikas$ and diaettsea4
pP. i95-2(34. by others.6S7*~ k@ is a directional permeability
in the direction of tha pulse response. k~ is a
2. Johnson, C. R., Greenkorn, R. A., and Woods, geometrical rsean permeability of the principal
E. G.: “Pulse-Testing: A New Method for Describ- directional parmeabilities.
ing Reservoir Flow Properties Between Wells,”
JPT, December 1966, pp. 1599-1604. ke = ~kmax “ %nin
3. Matthews, C. S., and Russell, D, G,: “Pressure For a single pulse of duration At, Figure
Buildup and Flow Tests in Wells,” SPE Monograph Al, the pressure response is the superposition of
Volume 1 (1967), the effects of the two changes in rate at the
beginning and end of the pulse:2
4. Collins, R, E,: Flow of Fluids Through Porous
Materials, Reinhold Publishing Corporation, Ap = * [Ei(+) - E&%)] (A.3)
New York (1961), p. 115.
where
Substituting Equation A.7 into Equation A.3 and Values of k@D and kGD with consistent units, are
rearranging gives a dimensionless group, kGD, shown as functions of tLD in Figure A.2.
expressed as a function of pulse length, At, and
time to maximum response, tm. Examples of the application of these charts
for design and interpretation using oil field
units are given in Appendix B. See Equations 9.1
and 9,2 and Figures 9.1 and 9.2.
k$At
ki$/D = ‘- = -
1 1
(A.1O f(tLD) = ~ =
()
~pcr2 4 tLD+ 1 ‘* ~lg, [-tLnln~~’v~i~D+,)ln (~)]1
tLD(tLD+l)ln 1
()—tLD
(A.15)
Frtxa the further relationships of the dimensionless
groups in Equations A.i@ end Ail, this funccion
[ t~(t@ll In (* )1
lnA-lnB-A+B=O
and
dt
(A,19)
SPE 11086 GEORGEL. EGE14EIER 9
*1t
For equal pulse lengtha:
(tLDl+l)
() L)
= Lie tLDl+l
cLW- in —
tLf)l tLDl
M. —, (A.23)
LD*k t LD tLD2 tLD2+l-)
!l(C~}(tLD*l)2hl —
~LD In —
tLD2
()
t()
tfJ)+l IV. ANALYSIS OF EARLYTIME RESPONSEFROMINTER-
‘LDln ~
FERENCETESTS
ne
= Lim
t LD +m LD+ 1
H
Accuracy of reservoir properties derived from
‘LDln ~ interference tests are maximized if the entire
pressure response is matched to the exponential
integral function; however, when only short time
= Lim
t LD+m
()
111+~
t LD
‘Li)
observations
is theoretically
associated
are practical,
possible
with successive
early characterization
by observation
inflection
of events
points of
the Ei function and its derivatives, These are
w=’ obtained from Equation A,l as follows:
and
k#D Ap= G[-Ei(-K/t)] (A.24)
Lim = Re (A.20)
t LD+= ~
() Where
G = -6&qW4wk~h
This simple relationship, which also may be
shown ftom the instantaneous 1 ne source solution,2 K E constant defined in Equatien A.6
p~evidea inai~ht Em many siaip ificationa in Dukaa *=+J”+
at K/t
U*2S}
M4ci
(A131)
,
la
.- -..
I?fPERkSKL PiMMtllM
.s----- — .. —----—
TESTIKG METiiOBSPoll PIELB PILOTS
-------- -— .----— - --- - -——- - -— ---
aPE Ileiki
--------
where n = O, 1, end 2. Angles are in degreeg. distance is wall-known; but reservoir trmtideei-
bility, kb/B, and hydraulic diffusitivity, kl+~c,
Tha ahapao of these funetiona are shown in which are the quantities that we wish to determine
Figure A.4. Compared to the inflection point of from the pulse test, must be estimated in order to
the Ei function, the maxima of the successively design the test. Of the variables comprising
higher derivatives occur at earlier times. these quantities, porosity, thickness and viscoeity
are known better than permeability and compressi-
For pulse tests, the earliest
lag time to bility,
maximum response occurs length becomes
as the pulse
small. This can be shown by rearranging Equation To use Figures B.1 and B.2, for design,
A.1O, and taking the limit as tLD approaches assume ke D kc = k, and:
infinity:
1) Calculate the value ofKGD= tkh/p) “
tLD+l ‘LD (Apm/9Aq) from estimated transmissibility, flow
= Lim (tLD+l)ln ~ rate and the maximum lwessure disturbance that is
+ tL@= () desired.
9
*PE
--- } iQs6
----- -------- L.
f$E&M3ff -. STWZEMRZER
- . --- ... . . .. 11
Exea!ple 2: Deaim for Pul&e Remonee ta SCcp# 3 and 4 may be eoetbined to give &
5)
Produe~icm from a Gaa Wall value for reservoir atorege, $cht U.udly thick-
ness is known batter thm kot&l eoatpresoibiliey;
Estimated Conditions: however, in the Banton ease diseuaaed in the body
of the report, compressibility maamremente mada
p = 3000 gsi on this well-consolidated sandstone provided an
T m 20130F accurate value for this liquid-filled reservoir,
In that case the thieknees, h, can be calculated
k ❑ 300 md from
h = 50 ft.
hm~~;
p = 0.02 Cp
Aq = -10 MMSCF/day Anisotropic Permeability
TABLE 1
[1
kh A% (300)(50)(1) 5.615
- 71.4
rw=l 0.02}(0.00$!4$ ~~ AWRAm OP O03USARAM8EB
Prom Fi&we R.}, WPERR4R *&zmmmwEE
ktik
t~=o,li and - 332t3
hVEftAO# AM
There Fese
+W2
AT= ~um
M)
& = 33~fj ( G.2)(Q.Q21(lQ-31{wQ)2 * ~hr~or
A * 25
300 12.9 deye
md B s 200
c 16 40
t~ = C~(At) ● (0.1E)(M38) ● %.3 hre. or
1.4 daya Average of eeven ecmples froa
ttBII sand in wlt 241*
INTERPRETATION
‘w@Sorw ~ 64 md,
= 0.17
Once the actual maximum pressure response ●nd kair 378 met.
the time at which it occurs are determined from a
field test, the FiSures B.1 and B,2 can be used in kair * kbrine @ SW . },o
reverse to obtain values of reservoir properties,
Isotropic Permeability
..
2&l Prodm i ng
—
240, ?.$fr ,2*7 Shnt-in 1, 2. $, b 10 min. production 1
P!llws. 1
?51 .3+%*M%
.%$
25kt, z5%,247
Il?js%tlnkt
Pt@lw tlt
Prnd4reinR
S&t-in Wrfntre prcrctUFed VQttrm+z dur-
I
ilq Iu mitt. mats in *II*
2#,%45, end 247,
I
Thtkt.S3
m_in’oNv!i-l/W$W-.I tult r
(mu t
TIIP
(pIii)(Btm) (1’:!
)
Nit,,
tVtttOIW 1ftN
‘TAME 5
PRODUCTION
TESTS - WELL246
Test 1 Test 2
Well 241
Injection Rate 65 0/D 73 B/D
Surface Injection Pressure 668 psig 450 psig
Well 246
I
Stabilized Rate at Surface 41.5 B/D 30 B/D
I Pressure = O psig
I Shut-in Surface Pressure 161 PSig 60 psig
TABLE 6
*Assumin& k~ = kG.
TABLE 7
TABLE B.1
it-
-i-t
\
*&
INJECTION WELL
PRODUCTION WELL
OBSERVATION WELL
561
w“
5s9”
558”
557”
556-
554.
553”
2s0”
244
200
1
,00 245
o~
IS 16 17 18 19 20 21 22 23 24
TIME (HRS)
200
000
800
600
400
200
*
.8 0.01 0.1 1.0 10 100 1( ‘o
TIME AFTER SHUT-IN (HOURS)
Ffg. ~ sw~tiw.
\240
‘2 MIN 5
At 240
‘1
0 1=1 I I I I I I I I 1 I 1 I
Id
.,. LL
T
2105
5 10 1s
:;.’
TIME FROM STFIRT OF PULSE tm [mini
. . I 1%
,1.
5
u 1“
i3 J :“2
1-
IA. ..
,,. :.. ..,.
.3
.2
.$
+
2115
II
1(
I
o 200 400 0
1 ‘r
VRLUES
200
OF h
EXTRRPOLRTED
TO ~t=O
o 2 4 6 8 !0
TIHE IHIN)
TIME c
10
.. . .
...a’.
.
..
., . . ..-
. ..... ..-_mm
.....- -
----- .: .::: . ..
.. ....
.
. .... ....=.- .. . . .
,,, ,, ....
...+
.. .. I
:;: ,.
.. ..+
..
. . . ...... . .,. - . . ... .....
. . . ... ., ..’ . .
‘,
. . . .
. ...7
., 0.1
...
..-.
::.
::.
.,. . .,, ::. . 001
0001
~@
~b
I
.oo~,,
1 10 A’
’10
$
*“n
o
-’
0 L &s 1.0
DIMENSIONLESS TIME (!.]
,,, .
,.. i ..
,,
,..
,. .,
.!
,..
‘< .,
,,, ,.. .
:,
.i
1: II
1000
I :
100