Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

SPE 109876

Real-Time Performance Analysis of Water-Injection Wells


B. Izgec, SPE, Chevron ETC/Texas A&M University and C.S. Kabir, SPE, Chevron ETC

Copyright 2007, Society of Petroleum Engineers


permeability, skin, and drainage-area pressure. Because
This paper was prepared for presentation at the 2007 SPE Annual Technical Conference and formation parting is quite common, van den Hoek (2005)
Exhibition held in Anaheim, California, U.S.A., 11–14 November 2007.
presented a method for discerning shrinkage of fracture height
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
or length from falloff tests.
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to While great strides have been made in interpreting falloff
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at tests, Hall (1963) plot appears to be one of the few tools
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
available for ongoing performance monitoring. Others have
for commercial purposes without the written consent of the Society of Petroleum Engineers is attempted modifications of the Hall plot by introducing the use
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous of bottomhole flowing and reservoir pressures (Buell et al.
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.
1990) or evaluating the reservoir pressure from a slope-
analysis method (Silin et al. 2005a, 2005b). Ideally, the Hall
Abstract method is suitable for either early injection period or during
This work presents a complete reformulation of the Hall the post-breakthrough period, because the notion of single
method involving both pre- and post-breakthrough situations. reservoir pressure is entertained. Both reciprocal injectivity
Two approaches involving both transient and material-balance index or RII (Hearn 1983, Abou-Sayed et al. 2007) and
methods produced very similar solutions, which were verified evolving skin (Zhu and Hill 1998) are other alternatives to
with the results of coupled geomechanical/fluid-flow monitoring real-time well performance.
simulations. The new formulations allow tracking the This study presents a new formulation of the Hall analysis.
expanding water-bank radius from inception to breakthrough. To that end, the development of an analytic derivative
Pressure of this bank at the water/oil interface is evaluated at expression turns out to be much more discriminating for
every timestep, thereby allowing continuous update of the yielding the desired diagnostic clues. Ascertaining variable
‘external pressure’ in Hall’s formulation. We show that Hall’s radial distance of the injection bank and pressure at the
formulation is a particular case of the proposed approach. water/oil interface (pe) makes the new formulation robust and
Several simulated and field examples demonstrate the value of suitable for prebreakthrough situation. Our study shows that pe
reformulated Hall analysis. practically becomes time invariant in postbreakthrough
Because Hall formulation involves an integral, the situation, suggesting applicability of the original Hall
resultant signature, by nature, is insensitive in revealing clues formulation.
about subtle changes that may occur during formation
fracturing or plugging. We observed that the derivative of Problem Statement & Solution Approach
modified-Hall integral, obtained analytically, provides This study presents reformulation of the Hall (1963) approach
definitive signatures about fracturing or plugging. for monitoring performance of water-injection wells. The
The new interpretation approach is particularly suitable for intrinsic idea was to update the outer-bank pressure or pe at
projects at the inception of flooding. Mature projects can every timestep for proper application of the original Hall
benefit equally from the new tool. Perhaps the biggest appeal formulation. Two approaches are presented for updating pe,
of the proposed tool lies in the green fields, where real-time transient and pseudosteady-state, as shown in Appendices A
data are readily available. and B, respectively. We also obtained an analytic expression
for the derivative of the Hall integral, as shown in Appendix
Introduction C.
Significant advances have been made in recording, Using coupled geomechanical/fluid-flow simulations, we
transmitting, filtering, and interpreting real-time production present synthetic cases showing fracturing, nonfracturing, and
data. However, data interpretation from injection wells has not plugging of the formation. In particular, comparing and
gained as much attention. Traditional water-injection well contrasting the derivative curve with the Hall integral provides
evaluation involves pressure-transient analysis, which the definitive clues. To summarize, the two curves trace the
predictably improved over the years, as testified by the same path in matrix-dominated flow when neither fracturing
contributions from Hazebtoek et al., 1958; Kazemi et al., nor plugging occurs. The derivative curve goes below the Hall
1972; Marrill et al. 1974; Sosa et al., 1981; Abbaszadeh and curve in fracturing situation and rides above it when formation
Kamal, 1989; Yeh and Agarwal, 1989; and Bratvold and plugging occurs. Thereafter, field examples corroborate these
Horne, 1990. Falloff analysis allows estimation of findings.
2 SPE 109876

Diagnosis with Simulated Examples pseudoskin will increase DHI, whereas negative skin will
Using a commercial simulator Reveal (2006), which is capable decrease it.
of capturing fluid flow and geomechanics of injection-induced Separation of the two curves is indicative of fracturing of
fracturing, we generated several cases to develop the the formation, as shown in Fig. 4. Note that unlike its
diagnostic signatures in typical waterflooding situations. pressure-transient counterpart, where time-derivative of skin is
Finite-element 3D thermal fracture formulation is used, where zero, skin is left intact when the cumulative-injection-
shape of the fracture evolves with its propagation. The model derivative of the pseudosteady-state equation is sought, as
contains 1,875 cells (25×25×3) and has a producer and an shown in Eq. 1.
injector. Ordinarly, in transient-pressure or transient-rate analysis,
time-variant entities (normalized pressure or rate) are graphed
Formation Parting. Fig. 1 shows that high-rate, cold-water against a time function on a log-log grid to maintain the aspect
injection was initiated to induce formation parting. Fig. 2 ratio for quantitative analysis. However, because we are
presents the fracture growth with time and the attendant seeking just diagnostic clues, the Cartesian graph, which
flowing-bottomhole pressure. As shown, the fracture growth avoids the scale compression, turns out to be more revealing.
starts after 40 days of injection. As with any geomechanical Fig. 5 illustrates this point.
simulator, the initial fracture length was assigned at 0.1 ft. 5,150 400

30,000 300
5,100

pe , psia
F

re , ft
0

iw 150 200
Rate, STB/D

Temperature,

20,000 5,050
100 PSS 100
Transient
re
10,000 5,000 0
T wf 50
0 20 40 60 80 100 120
0 0 Time, days
0 50 100 150 Fig. 3 − Estimating water-bank pressure by transient and PSS
Time, days methods.
Fig. 1 − High-rate, cold-water injection to induce formation
parting. 1.E+05
Hall Integral/ DHI , psi-D

Start of Fracing
300 5,800
Hall Plot
p wf Numeric Derivative
5,600 Analytic Derivative
200
pwf , psi
Xf , ft

5,400
100 xf
5,200
1.E+03
0 5,000
1.E+04 1.E+06
0 20 40 60 80 100 120 Cumulative Injection, STB
Time, days Fig. 4 − Downward separation of the derivative trace implies
Fig. 2 − Fracture propagation with injection time. formation parting on log-log graph.

Pressure at the outer edge of the waterbank is calculated 4.E+04


Hall Integral/ DHI , psi-D

with both the transient and pseudosteady-state formulations, as Start of Fracing


discussed in Appendices A and B. Overall; we note good
agreement between the two approaches. When the derivative
of the Hall integral is taken, both numerically and analytically, 2.E+04
downward separation of this signature occurs from the Hall
trace because negative skin occurs in the numerator.
Hall Plot
Inspection of Eq. 1 or Eq. C−12 makes this point clear: Numeric Derivative
Analytic Derivative
0.E+00
DHI = α1Wi ⎧⎨ln (re / rw ) + s* ⎫⎬ (1)
⎩ ⎭ 0.E+00 7.E+05 1.E+06
Cumulative Injection, STB
where DHI is the derivative of Hall integral, Wi is the Fig. 5 − Downward separation of the derivative prominent on
cumulative water injection and the constant α1 and pseudoskin Cartesian graph.
s* are defined in Appendix C. Eq. 1 shows that positive
SPE 109876 3

Another way of viewing the same data is through evolution Fig. 11 showing the RII graph (Hearn 1983; Abou-Sayed
of skin in a waterflood. This approach is used in acidizing et al. 2007) also makes the same point. Note that the reciprocal
(Zhu and Hill 1998) and may be adopted in the context of injectivity index is simply a ratio of pressure drop (pi – pwf)
water injection. Fig. 6 shows the evolving skin and significant and injection rate, iw. Just as in Hall formulation, the notion of
lowering after 40 days suggest induced fracturing. pi or average-reservoir pressure becomes hard to define in
ever-changing injection environment. That is why the use of pe
80 Start of Fracing 5,900 is a prudent approach when the RII method is used.
5,400 2,500
60

pwf , psi
p wf 5,600 iw
Skin

2,000

Rate, STB/D
40 s* 5,200 Start of Plugging

pwf , psi
5,300 1,500
20 p wf
1,000
0 5,000 5,000
500
0 20 40 60 80 100 120
Time, days 4,800 0
0 20 40 60
Fig. 6 − Skin evolution provides diagnostic clues.
Time, days
Matrix Injection. When clean water dominates matrix Fig. 8 − Increasing BHP mimics formation plugging.
injection, both the Hall integral and its derivative should
overlay on top of each other. Fig. 7 makes this point amply 5,060 100
clear. The slight bending of the curves before reaching
105 STB cumulative water injection is owing to decreased re 75
injectivity before breakthrough. As the steady-state between pe , psia 5,050

re , ft
the producer and the injector is attained, the curvature is 50
replaced by a straight line. One important lesson learned here 5,040
25
is that the separation of the two curves should signify PSS
Transient
fracturing, not bending of the curves, per se. 5,030 0
0 20 40 60
1.E+07 Time, days
Hall Integral/ DHI , psi-D

Fig. 9 − Efficient voidage replacement results minimal pe changes.

1.E+03 2.E+04
Hall Integral/ DHI , psi-D

Start of Plugging
D HI
Hall Plot Analytic Derivative

1.E-01 1.E+04
1.E+03 1.E+05 1.E+07
Cumulative Injection, STB
Fig. 7 − No separation of the curves implies matrix injection. 1.E-01
1.E-01 5.E+04 1.E+05
Formation Plugging. Formation inevitably plugs up when
Cumulative Injection, STB
unfiltered water is injected. We simulated a case where skin
increases exponentially with time according to the following Fig. 10 − Increasing Hall-derivative trend suggests plugging.
relation:
0.10
s = 1.008 ∗ ⎛⎜⎜ e 0.996 ∗ t ⎞
∆p/q , psi/STB/D

⎟⎟ (2)
⎝ ⎠
Start of Plugging
0.05
Fig. 8 depicts increasing BHP trend at a constant-injection
rate of 2,000 STB/D. Efficient voidage replacement in the
model resulted in minimal changes in pe, as indicated by both
the transient and PSS methods. Fig. 9 demonstrates this point. 0.00
As shown in Fig. 10, the Hall derivative curve rides above the 0.E+00 5.E+04 1.E+05
Hall curve, suggesting plugging. This outcome is expected
Cumulative Injection, STB
because the skin term occurs in the numerator of the Hall-
derivative expression, as shown in Eq. 1. Fig. 11 − Reciprocal injectivity graph also suggests plugging.
4 SPE 109876

Field Applications 600


The use of Hall graph and its derivative is not restricted to
monitoring water-injection wells in a waterflood, but is 20
applicable equally to other situations, such as during

pe, psia
re

re, ft
produced-water disposal, matrix acidizing, and steam 500
injection. We shall discuss several field cases of water- 10
Transient
injection wells and also present an example where real-time PSS
matrix acidizing occurs. re
400 0
Case 1−Injection in a diatomite reservoir. Fig. 12 shows 0 100 200 300 400 500
that constant-rate injection occurred for about 60 days, Time, days
followed by constant-pressure injection in this thick (~ 900 ft), Fig. 13 − Slow growth of water bank in thick formation.
low-permeability reservoir. The precipitous decline of
injection rate and ultimate shut down for a period is a
100,000
reflection of obvious injectivity issue. When injection was

Hall Integral/ DHI , psi-D


resumed and much of the injectivity restored thereafter, albeit 80,000
Acid
at a higher pressure, progression of the water bank was never job
too far from the well in this thick reservoir, as Fig. 13 60,000
D HI
HI
suggests. Acid
Not surprisingly, evaluation of pe suggested time-invariant job
40,000
behavior because of minimal movement of the water bank. As
suspected, the late-time decline of injection rate is associated 20,000
with plugging. Fig. 14 showing the Hall and its derivative Shut-in period
illustrate this point. The Hall integral (HI) by itself cannot 0
reveal the evolving story in this complex situation. The benefit 0 20,000 40,000 60,000
of first acid job is not demonstrated. In fact the well was shut Cumulative Injection, STB
in soon after. When injection commenced, low injectivity
Fig. 14 − Injectivity issues prominent in early life.
becomes readily apparent from the DHI signature. But injection
above the fracture gradient restored injectivity, and finally the
beneficial effects of the second acid job became transparent. 20
Although the RII graph captures the major events, as
Δ p/q, psi-D/STB

shown in Fig. 15, discerning subtle changes, such as late-time 15


plugging and the beneficial effects of the second acid job is
difficult. Note that we used variable pe to construct the RII 10
graph.
5
Case 2−Injection in a Carbonate reservoir. Fig. 16 shows
the injection rate and the corresponding bottomhole pressure 0
in this high-transmissivity reservoir where peripheral water
0 20,000 40,000 60,000
injection occurs. Comparing the two profiles over the six-year
period is quite complex in that many events may be inferred. Cumulative Injection, STB
For simplicity, we have identified three major time domains,
Fig. 15 − RII graph shows formation plugging at late times.
designated by phases I, II, and III.

1,800
400 1,000 II III
16,000 I
Rate, STB/D

800 1,400
pwf , psi

300
Rate, STB/D

12,000
pwf, psig

600
200 8,000 1,000
400 Injection Rate
Bottomhole Pressure
100 4,000 600
200
0 500 1,000 1,500 2,000 2,500
0 0
Time, days
0 150 300 450 600
Time, days Fig. 16 − Three time domains identified in injection history.

Fig. 12 − Constant-injection rate followed by constant-pressure Continuous decline in pe, as shown in Fig. 17, suggests
injection.
inadequate voidage replacement in the pattern; both PSS and
transient methods produce consistent results. Fig. 18 depicts
SPE 109876 5

early-time matrix injection (Phase I), followed by plugging in


Phase II, and finally fracturing in the remaining period, Start of Fracing
designated by Phase III. In contrast to Fig. 18, the RII method

∆p/q , psi/STB/D
0.10
produces inconclusive diagnosis, as shown in Fig. 19.
o Variable p e
Δ Constant p e
0.06
PSS
1,100 Transient re 2,000
pe , psia

900

re , ft
0.02
1,000
700
0.E+00 1.E+07 3.E+07
Cumulative Injection, STB
500 0 Fig. 19 − RII graph shows inconclusive diagnosis.
0 500 1,000 1,500 2,000
Time, days 12,000 3,400
Fig. 17 − Inadequate voidage replacement identified.

Rate, STB/D
2,800
8,000

pp , psi
2,200
Hall Integral/ DHI , psi-D

Start of Fracing
III
1.E+06 4,000
D HI 1,600
II 0 1,000
5.E+05 0 100 200 300 400 500
I
Tim e, days

Fig. 20 − Injection history of Well A10.


0.E+00
0.E+00 1.E+07 2.E+07 8.E+05
Hall Integral/ DHI , psi-D

Cumulative Injection, STB


Fig. 18 − Normal injection, plugging, and fracturing identified.
Filter Change

Case 3−Injection in a sandstone reservoir. Sharma et al. 4.E+05 Acidizing


(2000) presented Well A10 from the Gulf of Mexico, where
D HI
injection occurs in an unconsolidated formation. Fig. 20
presents the injection history. The authors used the Acidizing
conventional Hall analysis to discern the state of injection. 0.E+00
Given distinct events that transpired after acid treatment and 0.E+00 1.E+06 2.E+06 3.E+06
filter change (finer to coarser), identifying issues with the
Cumulative Injection, STB
conventional Hall analysis became feasible.
When the same analysis was done with the derivative, our Fig. 21 − Effect of acid treatment or lack thereof easily identified.
results essentially corroborated the previous findings.
However, the ease of diagnosis becomes transparent when the
two curves are used in tandem; Fig. 21 makes this point. Note 150 1,000
that the use of coarser filter manifests in terms of positive
pwf , psi

skin. The late-time data suggest rapid decline of the beneficial 100 p wf
Skin

800
effects of acidization and is tied to the water-quality issue. s*
50 600
Case 4−Matrix acidizing in a Carbonate reservoir.
Application of the proposed methodology for matrix acidizing
0 400
is important because very small region around the wellbore is
affected by this process. Fig. 22 presents the conventional 8 28 48
approach of skin evolution with time, presented originally by Time, min.
Zhu and Hill (1998) as Example 3. The Hall derivative plot is Fig. 22 − Skin evolution demonstrates effectiveness of acidizing.
able to identify even the small permeability changes around
the wellbore, as Fig. 23 demonstrates.
6 SPE 109876

100
method the pe values are generated by either transient or an
adjusted form of PSS formulation. For the derivative
Hall Integral/ DHI , psi-STB

perm eability alteration


calculations we use only the adjusted PSS formulation, which
is equivalent to the transient solution.
As pointed out by Yeh and Agarwal (1989), a reduction in
10 injectivity may develop because of changes in relative
permeability, fluid viscosities, compressibilities, and initial
water saturation. The combined effect of all these elements
D HI between the injected and reservoir fluid and mechanical
damage can be calculated by using a transient equation, as
1 pointed out by Zhu and Hill (1998).
10 100 In the following we offer some insights into our approach,
Cumulative Injection, STB although we employ a more rigorous treatment using the
Fig. 23 − Hall diagnosis shows effectiveness of acidization. superposition principle. The PSS (Eq. A-9) and transient
(Mathews and Russell 1967) equations for any radius are
Discussion given, respectively, as
The original Hall formulation was predicated upon the use of
tubinghead pressure and also ignoring the oil/water interface ⎡ ⎛ 2 2⎞ ⎤
i Bμ ⎢ re2 ⎛ r ⎞ 1 ⎜⎝ r − rw ⎟⎠ ⎥
pressure (pe) at the floodfront. Although this method has pr = pwf − w ⎢ ln ⎜ ⎟ − + s* ⎥ (3)
served the industry for over four decades, the very nature of 2πkh ⎢ ⎛ r 2 − r 2 ⎞ ⎜⎝ rw ⎟⎠ 2 ⎛ r 2 − r 2 ⎞ ⎥
⎜ e⎢⎣ ⎝ w⎟ ⎜ e w⎟ ⎥⎦
integral has a smoothing effect on the signature produced. ⎠ ⎝ ⎠
Therefore, the Hall integral’s ability to diagnose subtle
changes remained unsatisfactory. In fact, unless large event ⎡ ⎤
qμ ⎢ ⎛⎜ 4kt ⎞⎟
occurred, such as significant plugging or fracturing, the p(rw , t ) = pi − ⎢ln + 2s* ⎥⎥ (4)
4πkh ⎢ ⎜ φμc r 2 ⎟ ⎥⎦
effectiveness of this tool was often called into question. ⎣ ⎝ tw⎠
Although improvements have been offered, the
fundamental issue concerning pe remained. In this respect, the Incorporating the pseudoskin calculated by transient
variable-pe formulation presented here is an improvement over equation into the PSS equation and assuming a radius value
the traditional approach. To augment understanding of subtle very close to the wellbore, one can obtain the original transient
changes that may otherwise remain undetected, the notion of expression. This excercise shows that the PSS formulation can
derivative of the Hall-integral was introduced. The main be used to approximate the transient behavior as long as the
advantage is that both the curves are used in tandem to skin is calculated by transient formulation. Of course, the
diagnose the status of injection. In this context, derivative of changes in injection rate can be accounted for by using the
the pressure-normalized rate integral (Blasingame et al. 1989) superposition principle. Our numeric and field examples
for rate-transient analysis may be construed as an analog. demonstrate this point.
The variable-pe formulation for the Hearn plot or RII is
also an improvement over the existing practice. However, Conclusions
because two curves are used in tandem in the modified-Hall 1. This study shows that the widely used Hall method
formulation, it offers distinct advantage over the modified- works best in cases where the oil/water interface
Hearn approach. pressure of the moving bank (pe) chages minimally.
As discussed, the original Hall formulation presupposes Post-breakthrough situation is a case in point. The
steady-state flow, thereby assumes waterbreakthrough at the new analytic-pe formulation is general and works at
time of application. Nonetheless, the Hall method also applies all times for both Hall and Hearn plots, thereby
in situations where steady-state flow is experienced inside the proving its suitability in monitoring real-time
assumed radius re in which minimal changes in pe occurs. situations.
Mechanical damage around the wellbore or the contrast 2. When used in tandem with the Hall integral, the
between the reservoir and injected fluid may act as skin and derivative curve, obtained analytically, provides
distort the steady-state behavior within the assumed radius re. unambiguous diagnosis of a well’s performance
The changes in slope of the Hall plot are unreliable under status. For example, no separation between the two
these circumstances. curves occurs for matrix injection. However,
For prebreakthrough cases, a transient formulation is fracturing is indicated by downward separation of
required that can handle both mechanical damage and derivative from the original trace and plugging is
deviations from steady-state flow behavior caused by indicated by upward separation. In this regard, the
differences in reservoir and injected fluid properties. However, reformulated Hall approach is superior to the Hearn
transient equations do not provide the particular form we seek method.
to generate the Hall integral. The new PSS formulation allows 3. Both simulated and field examples demonstrate that
us to conserve the original form of Hall integral and track the Hall-integral derivative is applicable to wells
changes in pe before the onset of steady state, within a given injecting water in various formations and is equally
radius or between an injector/producer pair. In the new effective in monitoring matrix acidization.
SPE 109876 7

Acknowledgment Silin, D.B., Holtzman, R., Patzek, T.W., and Brink, J.L. 2005a.
We are indebted to Chevron management for permission to Paper SPE 93879 presented at the SPE Western Regional
publish this work. Meeting, Irvine, CA, 30 March−April 1.
Silin, D.B., Holtzman, R., Patzek, T.W., Brink, J.L., and Minner,
M.L. 2005b. Waterflood Surveillance and Control:
Nomenclature Incorporating Hall Plot and Slope Analysis. Paper SPE 95685
B = formation volume factor, RB/STB presented at the SPE Annual Technical Conference and
ct = total system compressibility, 1/psi Exhibition. Dallas, TX, 9−12 October.
h = formation thickness, ft Sharma, M.M., Pang, S., Wennberg, K.E., and Morgaenthaler,
DHI = derivative of Hall integral, psi-D L.N. 2000. Injectivity Decline in Water-Injection Wells: An
iw= water-injection rate, STB/D Offshore Gulf of Mexico Case Study. SPEPF 15 (1): 6−13.
k = permeability, md Sosa, A., Raghavan, R., and Limon, T.J. 1981. Effect of Relative
m = semilog slope (=162.6 Bμ/kh), psi/log-cycle Permeability and Mobility Ratio on Pressure Falloff Behavior.
JPT 33 (6): 1125−1135.
p = average-pressure in waterbank, psia
van den Hoek, P.J. 2005. Dimensions and Degree of Containment
pe = oil/water interface pressure, psia of Waterfloos-Induced Fractures From Pressure-Transient
pi = initial pressure, psia Analysis. SPEREE 8 (5): 377−387.
pwf = flowing bottomhole pressure, psig Yeh, N-S. and Agarwal, R.G. 1989. Pressure Transient Analysis of
pwh = flowing wellhead pressure, psig Injection Wells in Reservoirs With Multiple Fluid Banks. Paper
r = radial distance, ft SPE 19775 presented at the SPE Annual Technical conference
re = reservoir-boundary radius, ft and Exhibition, 8−11 October, San Antonio, TX.
rw = wellbore radius, ft Zhu, D. and Hill, A.D. 1998. Field Results Demonstrate Enhanced
Matrix Acidizing Through Real-Time Monitoring. SPEPF 13
s = mechanical skin, dimensionless
(4): 279−284.
s* = pseudoskin, dimensionless
Sor = residual-oil saturation, fraction Appendix A – Pseudosteady-State Approximation for
t = injection time, hr Pressure Inside the Waterbank
tD = dimensionless injection time
Wi = cumulative water injection, STB Starting with the radial diffusivity equation
μ = fluid viscosity, cp
φ = porosity, fraction 1 ∂ ⎛⎜ ∂p ⎞⎟ φμct ∂p
r = (A-1)
∆tsup = summation term in superposition r ∂r ⎜⎝ ∂r ⎟⎠ k ∂t

References Assuming constant-injection rate for a small time increment,


Abbaszadeh, M. and Kamal, M. 1989. Pressure-Transient Testing
change in pressure with respect to time inside the waterbank
of Water-Injection Wells. SPERE 4 (1): 115−124.
Abou-Sayed, A.S., Zaki, K.S., Wang, G., Sarfare, M.D., Harris, can be represented by the following relation:
M.H. 2007. Produced Water Management Strategy and Water
Injection Best Practices: Design, Performance, and Monitoring. i B
dp w
( )
SPEPO 22 (1): 59−68. =− (A-2)
Blasingame, T.A., Johnston, J.L., and Lee, W.J. 1989. Type-Curve dt
φhπ r 2 − r 2 c
Analysis Using the Pressure Integral Method. paper SPE 18799 e w t
prepared for presentation at the SPE California Regional
Meeting, Bakersfield, CA, 5-7 April. Incorporating Eq. A-2 into Eq. A-1, we get
Bratvold, R.B. and Horne, R.N. 1990. Analysis of Pressure-Falloff
Tests Following Cold-Water Injection. SPEFE 41 (1): 293−302.
⎛ ⎞
1 ∂ ⎛ ∂p ⎞ φμct ⎜ ⎟
Buell, R.S., Kazemi, H., and Poettmann. F.H. 1990. Analyzing iw B
Injectivity of Polymer Solutions With Hall Plot. SPERE 5 (1): ⎜r ⎟ = ⎜− ⎟ (A-3)
41−46. r ∂r ⎝ ∂ ⎠
r k ⎜ φhπ ⎜ r 2 − r 2 ⎞⎟c
⎛ ⎟
⎜ ⎟
Hall, H.N. 1963. How to Analyze Waterflood Injection Well ⎝ ⎝e w⎠ t ⎠
Performance. World Oil (October): 128−130.
Hazebroek, P., Rainbow, H., and Mathews, C.S. . 1958. Pressure Defining parameter x and integrating
Fall-Off in Water Injection Wells. Trans., AIME 213: 250−260.
Hearn, C.L. 1983. Method Analyzes Injection Well Pressure and
iw Bμ
Rate Data. Oil and Gas J. (April): 117-120. x= (A-4)
Kazemi, H., Merrill, L.S., and Jargon, J.R. 1972. Problems in πkh ⎛⎜ re2 − rw2 ⎞⎟
Interpretion of Pressure Fall-Off Tests in Reservoirs With and ⎝ ⎠
Without Fluid Banks. JPT 24 (9): 1147−1156.
Mathews, C.S. and Russell, D.G. 1967. Pressure Buildup and Flow
Tests in Wells. Monograph Series. SPE of AIME, vol. 1. dp x 1
=− r+x (A-5)
Merrill, L.S., Kazemi, H., and Jargon, J.R. 1974. Pressure Buildup dr 2 1 r
and Falloff Analysis of Water Injection Tests. JPT 26 (7):
809−818; Trans., AIME, 257. We define the no-flow, outer-boundary condition for a given
Reveal Reservoir Simulator, v. 2.3. 2006. Petroleum Experts Ltd., time increment x1 as
Edinburgh, UK.
8 SPE 109876

x 2 ⎡ ⎛ ⎤
x1 = re (A-6) k ⎞⎟
2 b = m ⎢⎢log⎜⎜ − 3.23 + 0.868 s* ⎥⎥ (B-3)
⎜ φμc r 2 ⎟⎟
⎣⎢ ⎝ t w⎠ ⎦⎥
Combining Eqs. A-5 and A-6, we have ⎛⎜ i − i ⎞⎟
( )
n
Δtsup = ∑ ⎝ wj wj − 1 ⎠
log tn − t j −1 (B-4)
j =1 in
⎡ 2 ⎤
dp x ⎢ re
= − r⎥ (A-7)
dr 2 ⎢ r ⎥ The final form of the equation is
⎣ ⎦
pi − pwf
Seperating and integrating again across the waterbank = mΔtsup + b (B-5)
qn

⎡ ⎤
pr r ⎢ r2 ⎥ Parameter b can be updated at every timestep during injection
x ⎢ e ⎥ as follows:
p wf
∫dp =
2 ⎢

rw r
∫− r ⎥dr

(A-8)

⎢⎣ ⎥⎦ pi − pwf
b= − mΔtsup (B-6)
qn
The final form of the equation for pressure inside the bank is Assuming constant-reservoir and -fluid properties, parameter
m remains constant over the injection period. Then pseudoskin
⎡ ⎛ 2 2⎞ ⎤ can be updated continuously with the following equation
iw Bμ ⎢ re2 ⎛ r ⎞ 1 ⎜⎝ r − rw ⎟⎠ ⎥
pr = pwf − ⎢ ln⎜⎜ ⎟⎟ − + s* ⎥ (A-9)
2πkh ⎢ ⎛ r 2 − r 2 ⎞ ⎝ rw ⎠ 2 ⎛ r 2 − r 2 ⎞ ⎥
⎜ w ⎟⎠ ⎜ e w ⎟⎠ ⎡ ⎛ ⎤
⎣⎢ ⎝ e ⎝ ⎦⎥ 1 ⎢b k ⎞⎟
s* = − log⎜⎜ + 3.23⎥⎥ (B-7)

0.868 m ⎜ φμc r 2 ⎟⎟
⎣⎢ t w ⎝ ⎠ ⎦⎥
*
The variable s is a pseudoskin factor representing the
resistance to injection. Pseudoskin is a dynamic parameter and In our method, the initial-reservoir pressure corresponds to
changes as a function of injection rate, bottomhole pressure the current reservoir pressure. Also, the calculated skin factor
and average reservoir pressure, as pointed out by Yeh and represents the combined effects of all types of resistances to
Agarwal (1989). injection.
The assumption of no-flow boundary condition in this
derivation introduces the limitation that every point inside the Appendix C – Analytic Derivative of Hall Integral
waterbank has the same rate of pressure increase during
injection. This assumption does not hold if the mobility ratio is Starting with pseudosteady-state equation
greater than one. Under these circumstances the regions within
the highly water-saturated areas close to the wellbore exhibit a 141.2iw Bμ ⎡ ⎛ re ⎞ *⎤ (C-1)
higher pressure increase, whereas the regions farther away will pwf − pe = ⎢ln⎜⎜ ⎟⎟ − 0.5 + s ⎥
be affected less. Nonetheless, this condition can be addressed kh ⎢⎣ ⎝ rw ⎠ ⎥⎦
by using the pseudoskin factor.
Integrating both sides with respect to time
Appendix B – Evaluation of Resistance to Injection
by Means of Skin 141.2Wi Bμ ⎡ ⎛ re ⎞ ⎤
∫⎛⎜⎝ pwf − pe ⎞⎟dt =
⎠ kh
⎢ln⎜⎜
r
⎟ − 0.5 + s* ⎥

(C-2)
⎣⎢ ⎝ w ⎠ ⎦⎥
Following the methodology outlined by Zhu and Hill (1998)
for real-time monitoring of matrix acidizing, the line-source
The Hall plot is generated by plotting the integral term
solution for transient flow during injection is
against cumulative injection, Wi. The derivative term can be
obtained by differentiating the integral with respect to natural
pi − pwf n ⎛⎜ i − i ⎞⎟
logarithm of cumulative injection. Designating the derivative
= m∑ ⎝ j j −1⎠
log⎛⎜ tn − t j − 1 ⎞⎟ +
in i ⎝ ⎠ as DHI, we have
n
j =1 (B-1)
d ∫⎛⎜ pwf − pe ⎞⎟dt
⎡ ⎛ k ⎞ ⎤
m ⎢log⎜ ⎟ − 3.23 + 0.868 s* ⎥ ⎝ ⎠
⎜ 2 ⎟ DHI = (C-3)
⎣⎢ ⎝ φμct rw ⎠ ⎦⎥ d ln(Wi )

Defining Replacing the term in the parenthesis with the right-hand side
and defining parameters
162.6 Bμ
m= (B-2)
kh 141.2Bμ (C-4)
α1 =
kh
SPE 109876 9

1/ 2
⎛ 5.615Wi B ⎞⎟
re = ⎜⎜ (C-5)
⎜ πhφ (1 − S ) ⎟⎟
⎝ or ⎠

1/ 2
⎛ 5.615B ⎞⎟ 1
α 2 = ⎜⎜ (C-6)
⎜ πhφ (1 − S ) ⎟⎟ rw
⎝ or ⎠

Combining Eqs. C-3 through C-6, one obtains

⎡ ⎛ ⎞⎤
d ⎢Wiα1⎜ ln⎛⎜Wi1 / 2α 2 ⎞⎟ − 0.5 + s* ⎟⎥
⎝ ⎝ ⎠ ⎠⎦
DHI = ⎣ (C-7)
d ln(Wi )

Expanding the terms, Eq. C-7 becomes

( ) ( )
d ⎡⎢0.5Wiα1 ln Wi + Wiα1 ln α 2 − 0.5Wiα1 + s*Wiα1 ⎤⎥
DHI = ⎣ ⎦
d ln(Wi )
(C-8)

Using the following relation

d(x )
= eln x , (C-9)
d(ln x )

we have the final form of the analytic derivative as

DHI =
Wiα1
+e
ln W
i
( )⎛⎜ α1 ln(Wi ) + α ln(α ) − α1 + s*α ⎞⎟
2 ⎜ 2 1 2 2 1⎟
⎝ ⎠
(C-10)

Upon manipulation, one obtains

Wiα1 ⎧ ln (Wi ) 1 ⎫
DHI = + α1Wi ⎪⎨ + ln(α 2 ) − + s* ⎪⎬ (C-11)
2 ⎪⎩ 2 2 ⎪⎭

Finally, we can write

DHI = α1Wi ⎧⎨ln (re / rw ) + s* ⎫⎬ (C-12)


⎩ ⎭

You might also like