Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

SPE

SOS3BQ.Jof Petroleum ErigIrwws

SPE 21724

Maximizing Injection Rates in Wells Recently


Converted to Injection Using Hearn and Hall Plots
P.M. Jarrelland M,H, Stein,Amoco Production Co.
SPE Members

:Opyrigh! 1991, Socialy of Petroleum Engineers, Inc.

h18 papar waa prapared for presentation at tha Production Operations Symposium held in Oklahoma city, Oklahoma, April 7-9, 1991.

“his paper waa !Ielacted for presentation by an SPE Program Committee following raview of information contained in an abstracl submitted by the author(s), Conlenls 01 Iho paper,
9 preaenled, have not been reviawed by the Soclaty of I%trolaum Engineers and are aubjec: 10 correction by the author(s). The material, aa presented, does not necessarily rellr,’1
nY Poaitlon of the SOO[etY of Pa!rol@um En9inaers, ita officors, or mefnbera. papers presented at SfT meetbw are aubpcl 10 publication review by Editorial Committees of the Soc[eiy
f Pelrolaum Engineers, Permlas!on to copy is restricted 10 an abotract of not more Ihon 302 worde. Illustiallona may not be copied. The abstract ehou[d contain conspicuous acknowledgment
f whera and by whom the paper la presented, Write Publications Managar, SPE, P.O. Box 833836, Richardson, TX 750S3-3836 U.S.A. Talax, 730989 SPEDAL.

ABSTRACT INTRODUCTION

A new technique is prese~ted wh~ch aids ~n maxi- An upcoming 50-well conversion program (from pro-
mlzirtgwater injection rates In wells recently ducing wells to water Injection wells) in a mature
converted to injection, thereby accelerating patterned-flood presented the sltuatton where an
offset otl production response. This method fs eng+neer would have to devote cons~derable time
applicable to waterflood operations where injec- monitorlng lnject~on rates, pressures and volumes
tion wells are surface pressure cntrolled and to dec~de timing for periodic step-rate tests.
where bottom-hole Injection just below formation Applying accepted guidelines for scheduling
parting pressure (FPP) is desired. step-rate tests during conversion start-up, as
introduced by Robertson and Kelm’, suggests an
Two plotttng methods are shown to be Instrumental average of three tests per well per year to keep
In monttorl~g the acceleration of fill-up and bottom-hole fnjection pressures as close to FPP as
average reservoir pressure growth in an actual possible, Since FPP increases as conttnued water
fteld case. One is the Hall methodi: plotting a Injectton Increases avarage reservoir pressure”,
function of bottom-hole in.jectlonpressure versus it was recognized that, ideally, injection pres-
cumulatlve water Injection, and the other was sures could be increased on a more frequent basis
introduced by Hearn*: plotting the tnverse lnjec- (i.e., monthly) instead of three times per year
ttvity Index versus cumulative water injection, during conversion start-up.

After tnitiatlng inject{on Into several converted The challenge was finding a valid, cost effective
wells at pressure limits well below the average method to Increase injection pressures frequently
offset parting pressures, perlodfc surface pres- and at small enough increments to remain below FPP
sure Increases were made at each well over a throughout the first year after conversion, The
per~od of several months. Monitoring the Hall and benefits would be: 1) accelerated injection rate
Hearn plots as the pressure and rates increased and average reservoir pressure build-up to accelr?-
rendered qualitative interpretations of whether rate offset otl response, and 2) lower operatfnq
the rates were being maintained below FPP. Accel- costs from reducing the numbr?rof step-rate tests
erated reservofr pressure growth was achieved from three to one during th@ conversion start-up
which resulted in earlier than expected offset oil period,
response,
Butlding and monitoring Ilalland Hearn plots from
Application of these plots also reduce, and somv- dallY surface injection rates and pressures
tfmes ellmtnate, the need to perform perfodfc allowed these small monthly pressure increases to
step-rate tests designed to determine FPP during be made for each converted well. Only one step-
the injection start-up period, thereby signifi- rate test was run on each well when bottom-hole
cantly reducing operating costs, injection pressures approached the average of the
existing offset injection wells, approximately
nine to twelve months after fnttlal Injection,
fieferencesand figures at end of paper

821
2“ MAXIMIZING INJECTION RATES IN WELLS RECENTLY CONVERTiD.- TO
.- INJECTION USING.- NEARN AND HALL P1OTS
. ..- -. -..2)?74. .
SPF

TJLEQRY 141,2 Bwpw ln(re/rwa)


jApdt= – Wf ..,,,,(4)
!-le~e~al kwh

The qualitative use of both plotting methods where Ap = pw-pe.


offers similar detections in reservoir condition
changes. Both are founded on the radial, steady Hall plotted the integral of the pressure drop
state form of the Darcy flow relationship: with respect to time versus cumulative injection.
This Is referred to as the “liallplot”. Observa-
0.00708 kwh (pw-pe) tion of this pressure function plot reveals that
Qw = ._.— if an injection well “ stimulated, there should
Bw IJw ln(re/rwa) .,..,.,.....(1) be a decrease in slop~, and if a well is damaged,
the slope should increase (Fig, 1). Regardless of
-s the constant value assumed for p , the slope
where rwa = rwe . changes mentioned above would stilleoccur. Only
the intercepts would change.

Assumptions applied in both plotting methods are: When ( Ap dt is plotted versus cumulative injec-
piston displacement; steady state, radial single- ton, the slope of the Hall plot. is equal to
phase, single-layer flow; pe is constaot; and no Eq. 5:
residual gas saturation in the water and oil
banks. Factors that are not required are a unity 141.2 Bw ILwln(re/rwa)
mobility ratio, or constant injection rate and m=
pressure. The only data required are injection kwh ,,......,.(5)
rate and surface injection pressure, and an
estimate of pet
This slope assumes single-phase flow with only one
The primary difference between each method is when fluid bank, It is also possible to analyze the
they should be applled, relative to fillup or off”- Hall plot using the concept of multiple fluid
set well pressure interference. A straight-line banks. Based on the slope of the Hall plot, if
relationship describes the reservoir transmissi- the skin is known, the transmissibility can b[!
bility from both plots. The Hearn plot develops a calculated, or vice versa. For single-phase flow,
constant slope prior to offset well pressuro the transmissibility will not change significantly
interference (e.g. during fill-up), A second, with time; therefore, any change in the slope of
horizontal straight-line should result after fil’i- the Yall plot will be due to skin effects. Buell,
up. The Hall plot develops a straight-line slope et.al.s showed that, in general, the fluid bank
after pressure interference (?.g. after fill-up). near wellbore will dominate the Hall plot slope.
Prior to fill-up the Pe increases and causes In the case of this paper, the dominating fluid is
upward curvature. water.

Hall Plot The Hall plot is a steady-state analysis method,


which provides a continuous injection monitoring
}ialloriginally proposed this plot method to quan- tool. Integrating the pressure data with the Hall
titatively analyze the performance of waterflood method has a smoothing effect on the data,
injection wells, Hall devised the following Smoothing the injection data over an extended
approach to eliminate the complications of both period of time results In more representative
pressure and injection rate variations. Ibis reservoir condition interpretations.
method was based on Darcy’s law for single-phase,
steady-state, Newtonian flow of a well centered in Acquisition of the data for the Hall plot is
a circular reservoir, which is given in Eq. 1. Inexpensive, since all that is required is the
Hall integrated both sides with respect to time recording of cumulative injection and surface
(Eq. ?), and rearranged the resulting relationship pressures, The surface pressures are then con-
to form Eq. 3: verted to bottom-hole pressures, correcting for
hydrostatic head and friction losses. The Hall
kwh j (pw-pe) dt plot can be used to qualitatively interpret what.
jqwdt=_ is happening in the reservoir when changes occur
141,2 Bwpw ln(re/rwa) .,,,.,,(2) in the slope of the curve,

]IearnPlot
kwh
Wi = ( ( Pwdt- ( pe dt) Applicability of this method is during a well’s
141,2 Bwpw ln(re/rwa) initial injection period, Hearn introduced this
,,,,,.......(3) plot technique as a means to quantitative}
estimate permeability to water and well skin
where Wi = cumulative injection, barrels, factor from the same plot. This metilodassumes a
simplified saturation distribution (F:g, 2) and no
By consolidating the pressure terms to the left residual gas saturation in the \.’aterand oil
side of Eq, 3, the final relationship plotted by banks. Hearn modified Muskat’s’ technique for
Hall developed as Eq. 4: constant pressure well analysis to allow for

.,..
.

*“e m,v~” D M ,lADQFI


v.. I M M Crrru ?
arc LSIt C.I I .!!. !.,....-, !!.!!. “0..0.

. .

varying rate or injection pressure. As ihown by APPI.ICATIONOF PLOTTING METHODS


Muskat, the In.jectfonrate can be present+d as:
General

0.00708 \h (pw-pe) Over a period of two years, a 50-well conversion


% = program was Implemented throughout the Central
Bwpw (In ro/rwa + M In re/ro) ,......(6) Mallet Unft.,Slaughter (San Andres) Ffeld, Hockley
County, Texas. By converting existing producing
wells to water fnjection wells, the flood pattern
where, M = (kw po/ko I@, or mobility ratio, realignment was accomplished, The previous
“Chfckenwfre” patterns were transformed into
Hearn set M to unttya, but we have chosen to leave skewed line-drives (Fig. 4), which are expected to
M arbitrary in the following derivation. Our end result In significantly htgher recoveries.
result Is fdentfcal to Hearn’s derivation when M
Is unfty. Oaily surface rate and pressure data was recorded
from the automation system and stored in a data-
0.00708 kwh (pw-pe) base. Each dafly set of data was then manipulated
qw = into the pressure and volume functions required by
Bwpw in (re/rwJ the Hall and Hearn plots. Through a program
developed on the mafnframe computer system, the
data acqufsitfon, manipulation, plotting and
0,01416 kwh (pw-pe) storage was achfeved automatically on a daily
=
— basis. Four plots were maintained on a sfngle
Bw IJwIln (ro2/rwa2) + M in (rez/ro’)) page display for each well; 1) rate and pressure
versus tfme; 2) Hall plot; 3) Hearn plot; and 4)
.,.,$.,,,,.,,,!.,(7) cumulative water injectfon and injectivity index
(AP/q ) versus time. This automated plotting
It can be shown by material balance that r 2 in routi~e provided for a very efficfent means by
Eq. 71s directly proportional to the cumul’%tlve which the performance of the fffty wells could be
water fnjection volume, or: evaluated,

Each well commenced water injectfon at a surface


rea =Cwf pressure equal to the average of the exi tfng
,,, ,,.,,, ... ,,,., (8)
offset injectfon pressures less a 200 to 300 psfg
margfn. As cumulative fnjectfon increased,
safeti.1
where C = constant = 5,615/mh # S the average reservofr pressure and FPP fncreased.
9’ The normal, accepted approach for increasing the
operatfng pressure fs to allow for stabilized
Also by material balance: fnjectlon rate at the fnftfal start-up control
pressure, then perform a step-rate test to deter-
ro2=Fw .,,,,,,,,,,.(9) mfne part.fngpressure. Robertson and Kelm showed
f that this inftfal perfod should allow the water
where F = constant = 5.615/mh # (iwBT- Swc), bank radius to extend to 150 to 300 ft, whfch
occurs when the fnjectfon rate drops to roughly
Combfnfng Eqs. 7, 8 and 9 gfves the relatfonshtp one-half the fnftfal fnjectfon rate for West Texas
in Eq, 10: San Andres floodsa. Subsequent operatfng pressure
Increases would be achieved in the same manner
through the start-up llfe until each wells’
Ap bottom-hole operating pressure approached the
=alnW1+a(lnL +Mln ;)
~ r= ..!, (lO) offset operating pressures. Thfs start-up life
wa typically lasts from twelve to efghteen months fn
West Texas San Andres floods, wfth two to three
where, a = BwIJw/(0.01416 kwh). step-rate tests befng performed durfng that tfme
to justify operating pressure increases (Ffg, 5).

Plottlng Ap/q (which is the inverse injectivity It was the authors’ fntentfon to honor the forma-
Index) versuswln W results in a strafght-lfno tfon parting pressure limitation while maintaining
erIed to fn this paper as the the most frequent operat~ng pressure fncreases, as
~~$~~”~~~~!~ Yl!e slope of the 1ine 1s a, and the is the fntention of the guidelines proposed by
intercept Is the right-hand term of Eq, 10, The Robertson and Kelm’. The only difference is that
mobility ratio and saturations wfll appear only fn the method detafled in thfs paper attempts to
the Intercept term; therefore, permeability from accelerate the process (Fig, 6). Application of
the slope Is not Influenced by the mobiltty ratln the sam~ technology offered by step-rate tests
or the simplified fluid dl$trfbutfon, through the Hall and Hearn plots can provide thfs
acceleration,
As shown in Fig, 3, several Interpretations of ths?
slope changes and curve shifts describe reservoir
conditions result~ng from pluggfng, fracturing,
pressure Interference, or pressure transient
effects.

I
,.#..
4 MAXIMIZING INJECTION RATES IN HELLS RECENTLY CO :RTEO TO INJECTION USING HEARN AND HALL PLOTS SPE Z1724

Field Data time. If the slope of the downward shifted data


in Fig. 12 had shown to be shallower than the
An example of how the Hall and Hearn plots were first slope, then the fracture could be presumed
qualitatively applled is illustrated by the to be extended out of pay.
start-up life of Central Mallet Unit (CMU) Well
tie. 258 (Fig. 7). The plan was to make 50 to 80 A step-rate test was performed at the end L? the
psig pressure increases about every thirty to last pressure increment for CMU Well No. 258,
forty days until the operating pressure was within This test revealed the FPP was 25 psig higher than
about 50 psig of the average offset operating the current pressure, which confirmed that the
pressures. Then run a step-rate test to confirm well was not operating above FPP. AISCI,an injec-
FPP and an injection profile to reveal any out-of- tion profile performed after the step-rate test
pay injection, showed no out-of-pay fnjection.

After monitoring the rate and pressure data for CONCLUSIONS


stabilization, an initial pressure increase of
50 psig was affected. The Hearn plot developed a The application of Hall and Hearn plots to
straight-line slope through this initial stabili- accelerate waterflood response was demonstrated.
zation period, as expected (Fig. 8), It was These plots can be used instead of step-rate tests
observed that the second stabilization period, to justify increases in bottomhole injection
after the first pressure increase, continued to pressure. As a result, operating costs a:sociated
follow the straight-line fit of the first period with step-rate testing can be reduced and oil
on the Hearn plot. This was interpreted as below response accelerated.
Fpp injection based on the guidelines noted in
Fig. 3. A second pressure increase was made after NOMENCLATURE
enough data was collected to estimate that the
first increase maintained rate below FPP. The B = formation volume factor, dimension’ess
Hearn plot shifted downward at this second pres-
sure increase, Indicating possible near wellbore c = constant given in Eq, 8
fracturing, but recovered to the initial injection dt = time increment, days
period fit. This “shift and recover” occurrence =
is interpreted as merely a pressure transient D true vertical hole depth, feet
effect, Since the equations that describe this 9 = acceleration of gravity, 32.2 ftf$=
plot assume steady- state conditions, sudden rate =
or pressure changes would cause the minor distur-
h formation thickness, feet
bances in the Hearn plot. This same “shift and F = constant in Eq. 9
recover” performance was also exhibited after the
third pressure increase of 80 psig, Observing the k. = permeabl’ity to ojl, md
Hall plot during these pressure increase periods kw = permeabi ity to water, md
show how the inherently smoothed data reveal no
significant slope deviation (Fig. 9), which m= slope of the Ha 1 plot, (ps g-dtys)/barrel
corroborates the judgment that fracture propaga- M = mobility ratio, dimensionle s
tion was not occurring after each pressur~
P = pressure psig
increase.
P = pressure at the external drainage radius,
If significant fracture propagation had occurred e
psig
after one of the sequential pressure increases, it
would have been evident on the Hall and Hearn P. = pressure at outer radius water bank, psig
plots. This can be illustrated by the arbitrary surface ubing pressure, psig
Ps =
adjustment of the rate data for CMU Well No. 259
at the the third pressure Increase (Fig. 10). The Pw = bottomho e injection presslre, psig
Hall plot shown in Fig. 11 reveals a sharp down- q = rate, barrels/day
ward departure at.the time of th{s rate increase,
r = external drainage radius, eet
which indicates increased transmissibility, or in e
this case a fracture, Likewise, the Hearn plot in r = outer radius of water bank feet
Fig, 12 illustrates the same fracture propagation o
r = wellbore radius, feet
from a marked downward sl,iftof the data. From w
Eq, 10, it is recognized that the change in v = apparent wellbore radius, eet
magnitude of the inverse injectivity index value wa
s = skin, dimensionless
(y-axIs) is inversely proportional to the apparent
wellbore radius (r ), Earlougher’ showed that Sg = initial gas saturation, dimensionless
for a uniform flux Y$acture, r = xf/e, or r =
0,367 xf. Therefore, fracture ~ength can als~abe sWc = connate water saturation, dimensionless
estimated from Hearn plot shift magnitudes, giving average water saturation behind the water
an idea of how long the fracture may be propagat- iwBT =
front, dimensionless
ing. If the fracture growth is de~;;:ined to be
within an acceptable range, pressure t = injection time, days
increases could be continued until the target
pressure is re~ched, realizing that the average Wi = cumulative water injected, barrels
reservoir pressure will continue to Increase and
)(f = fracture radius, feet
thereby allow closure of a minimal fracture with

824
.

SPE 21724 P.M. JARRELL, M.H. STEIN 5

. .
Greek Symbols

AP = pressure difference, psi


Af)f = pressure loss due to frictfon, psl
At = time increment, days
P = fluid density, lbm/ft’
+ = porosity, dimensionless
P = Viscosity, Cp
IJo = oil Viscosity, Cp
~w = water viscosity, cp

~KNOWLEDGEMENTS

The authors wish to thank Amoco Production Company


for permission to publish this paper.

REFERENCES

1. Hall, H. N., “How to Analyze Waterflood


Injection Well Performance”, World Oil,
October 1963, pp. 128-130.

2. Hearn, C. L,, “Method Analyzes Injection Well


Pressures and Rate Data”, Oil and Gas
Journal, April 1983, pp. 117-120.

3. Robertson, D. C. and Kelm, C. H. “Injection


Well -Testing to Optimize Waterflood
Performance,” paper SPE 5130 presented at the
SPE-AIME 49TH Annual Fall Meeting, Houston,
Texas, October 6-9, 1974.

4. Hubbert, M. K., and Willis, D. G., “Mechanics


of Hydraulic Fracturing”, AIME Trans., V,
210: 153-166, 1957,
5. Buell, R. S,, Kazemi, H., and Poettmann,
F. H.,” Analyzing Injectivity of Polymer
Solutions with the Hall Plot”, paper SPE
16963 presented at the 62nd Annual Technical .
Conference and Exhibition of SPE in Dallas,
Texas, September 27-30, 1987,

6. Muskat, Morris, “Physical Principals of Oil


Production”, McGraw-Hill Book Co, Inc,, New
York, 1949, pp.682-686.

7. Earlougher, R. C,, llAdvances in Well Test.


Analysis”, Millet the Printer, Inc., Dallas,
Texas, 1977, Volume Five of the Henry L,
Doherty Series of Monographs, p, 154,

S1 METRIC CONVERSION FACTORS

bbl X 1.589 873 E-Ol=m’


Cp x 100* E-03 = Pas
ft X 3.048* E - 01 =m
lbm/ft’ x 1.601 846 E+r)l= kg/m*
md x 9,869 233 E-04=pmz
PSi X 6.894 757 E+(lo=k Pa

*conversion factor is exact

---
.
~(;E 2172*

qw = Water injection fate


WI= Cumulative water Injected

WellBore Plugging
/
it
v ‘ Fracture
Extension
a
Fracturing
neartheWell
E
4
$ 141.26~ PW In(r./rW) Connete water
E
-5
E
8
Cum. Inj,
Water,(BBL)

Fig. 1 Hall Plot Simplified


fluid
distribution
~f~~n Method(fromreference)

. .

,.
I

ol~
1000
Cumulative InjectedWater, 1000 BBLs.

6. E811vplueQlq me Ifsctuflw nur MM, O. Ftwlurlw nar WI (chin down)


Ud Pfwaufs trmwlcnl Cilwto
Il. Msourmwl al marvatr MI & W8ctvm oxtmdcn (lkfm Chmpe)

c. w911
* pIUOOIn.J
mm UP) F. Prmvae Imuiwbna pwnt
—.
Fig, 3 Hearn plot Fig.4. ConveralonPattern
Reallnment
f2hlckenwire
toSkewed LYne.Drive

020

.
-. ●
SPE 21724

Average Offeet injection Pressure Average offset InJection preseure


l,5co~
------------------
---................... 1,50!) . . . . . . . . . . . . . . . - —. _ - _ _ _ _
1,400- 1,400 - ->,(+,0 ,,,,,
A
rJ
~ 1,200- — ..—.— ~ 1,200 — ‘+5CPslQIncremente
- ‘< - - - >T(+150 Psl~) . 1
< SRT (+75 Pslg)
SRT (+75 Palg)
~- l,ooo-
f?
~
i soc-
g
I ~ 6W -
2Z

! 4w-

a “4
o~’ o 2 4
Cums 110 MBW

6 8 10 12
Months sftar Injection Skwt.Up
Cum = 230 MBW

14 16
‘ 20:k_+ o 2 4 6 Q 10 12
Months otter Inlection Start-Up
Cum,26CMSW

14 1$

FI,6
NormalA[~?o?ch-SRTS Desiredi pproach
More Fre uentPressurelncreasee
Periodic
Injection
PressureIncreases
after
Conversion
@ ~mallerlncremente

(3) +80 PSI


(2)+80 PSI ,
1,200-

1,000- 7,., ,. .,, ,,!!,~’”’.’ ”’”-’


,,

s ‘
400- ;
r~;y:cy~ty”-.’.’”-.’”
wlo pressure 0.501 . .... . . . .. . .... ;,:, :...,.,.,,. ;
200- Increcaes
0.00 I I I 1~
10 20 30 40 50 6070 200
O!lrcl 1 I I I CumulatlveWaterlnjected
(bblxlO1)
Jsnusry‘February’
March April May June
FI .8
Fig.7 CMUWell No.258 Y
CMUWelNo.258
Rateand Pressure Data tleern Plot

*Z7
300 1,2001

1 1,000 -,i, . . . . . . .
,., ..,..
Wrrmx
In].
presoum
~ 250- ,,, -.,,,,,,,
‘8
~200-

4
●a 150- inl:;~
4
.g ,fjQ-

E
3 eo-

0 I I I I I I I I I O!llrl, !, r,, I I I I
0 10 20 30 40 6(I 61) 70 OD 90 January February’ Merch April May
Cumulative Water InjectIon (Iablax 103 )
Fig.10 CMUWell No,258
Simulated Frac8ureCase
FI .9
CMU We ? I No, 25S
Hall Plot

300
.-
1 .’
,,’

,’
*250- ,’
? ,’
/“
~ 2W ..
4

$150- t
Fracture
g /“” Opening
3 IN.

5
50-

0 I I I I I I [ I 1 I
o 10 20 30 40 50 60 70 eo 90 100
CumulativeWaterInjeoted(bblax 103 ) CumulativeWaterInjactad(bblx 103 )

CMU %ilvo. 258 CMU %il _!i20. 258


Simulated Fracture Case !Nmulated Fracture Caee
tlall plot Hearn Plot

You might also like