Sie sind auf Seite 1von 18

Journal of Petroleum Science and Engineering 57 (2007) 303 – 320

www.elsevier.com/locate/petrol

Well test analysis of finite-conductivity fractured wells producing at


constant bottomhole pressure
Ibrahim Sami Nashawi ⁎, Adel H. Malallah
Department of Petroleum Engineering, College of Engineering and Petroleum, Kuwait University, P.O. Box 5969, Safat 13060, Kuwait
Received 20 May 2005; received in revised form 28 October 2006; accepted 30 October 2006

Abstract

Conventional pressure buildup and drawdown tests are generally influenced by wellbore storage effects. These effects may dominate
the early well test data prohibiting good formation characterization of the area surrounding the wellbore. One of the advantages of
constant bottomhole pressure tests is that they are immune to these adverse effects. Constant pressure test data can be used with
confidence to provide good description of the formation around the wellbore in addition to full-scale reservoir interpretation.
This paper presents an analysis technique for finite conductivity fractured wells producing at constant bottomhole pressure from
closed reservoirs. The reciprocal rate and reciprocal rate derivative data are directly used to determine the fracture and reservoir
parameters without recoursing to type curve matching. All the dominant flow regimes such as early time bilinear, pseudo-radial,
and boundary-dominated flow are analyzed using log–log plots of the reciprocal rate and reciprocal rate derivative data. The slopes
of the straight lines of the different flow regimes are very distinct and are used to determine various reservoir and fracture
parameters such as fracture conductivity, reservoir permeability, skin factor, drainage area, and shape factor.
Furthermore, a 0.65 slope straight line equation describing the transition between the infinite acting pseudo-radial and the
boundary-dominated flow period in rectangular systems is presented. It is shown in the paper that this straight line can be used to
either determine the formation permeability in the absence of the pseudo-radial flow, or calculate the drainage area. It is also
illustrated that the intersection points of the various straight lines can be used to verify the accuracy of the results obtained from the
different flow regimes.
A systematic step-by-step procedure illustrating the methodology of the proposed technique for the analysis of bilinear, pseudo-
radial, and boundary-dominated flow regimes is described. The applicability of the method is illustrated using two simulated cases.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Fractured well; Finite-conductivity fracture; Constant-pressure analysis; Closed systems; Bilinear flow; Pseudo-radial flow; Boundary-
dominated flow

1. Introduction mercial rates. Massive hydraulic fracturing (MHF)


stimulation treatments are extensively used in tight
Many reservoirs around the world have low perme- reservoirs to boost reservoir performance and meet
ability, which restricts hydrocarbon production at com- increasing demand for oil and gas. A good fractured
well surveillance is essential for optimal reservoir
⁎ Corresponding author. Fax: +965 4849558. exploitation and long-term strategic plan development.
E-mail addresses: nashawi@kuc01.kuniv.edu.kw (I.S. Nashawi), Accurate well characterization dictates a good knowledge
malallah@kuniv.edu (A.H. Malallah). of the type of fracture being created in the reservoir. Three
0920-4105/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.petrol.2006.10.009
304 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

major types of fractures are commonly considered: (i) due to technical and/or economical restrictions, the
uniform flux fracture, (ii) infinite conductivity fracture, well test analyst may be forced to terminate the test
and (iii) finite conductivity fracture. before observing the pseudo-radial flow and/or the
Several flow regimes are observed in fractured wells. boundary-dominated flow regimes; thus, valuable
One of the responsibilities of the well test analyst is to use information may be lost due to these difficulties. In
the appropriate tools to predict the type of flow regime these circumstances, the only alternative for the well
that may develop in the fracture around the wellbore. The test analyst is to use pre-fracture data and rely on his
flow regimes that are generally considered in well test experience to fit all the pieces together to obtain a good
analysis of fractured wells are (a) fracture linear flow, (b) description of the reservoir.
bilinear flow, (c) formation linear flow, (d) pseudo-radial Over the years, numerous papers describing the
flow, and if the test is run long enough boundary- behavior of fluid flow in the vicinity of vertically
dominated flow may be also observed. Fig. 1 displays the fractured wells were published. Several methods were
flow regimes that are commonly observed in fractured proposed for well test analysis of fractured wells. The
wells. presented techniques were based on a variety of
Appropriate well test analysis of each of these flow numerical (Russell and Truitt, 1964a,b; Raghavan et
regimes yields valuable information about one or more al., 1972a,b; Locke and Sawyer, 1975; Agarwal et al.,
of the fracture and reservoir parameters. Fracture half- 1979a,b; Narasimhan and Palen, 1979; Bennet et al.,
length, xf, is determined from the analysis of the 1982) and semi-analytical solutions for both finite and
fracture linear or formation linear flow data. Fracture infinite fracture conductivity (Prats, 1961; Prats et al.,
conductivity, kfwf, is calculated from the bilinear flow 1962a,b; Gringarten and Ramey, 1973a,b; Gringarten
period. Reservoir permeability, k, and skin factor, s, et al., 1975a,b; Cinco-Ley and Samaniego-V, 1977,
are obtained from the pseudo-radial flow period. Well 1981; Lee and Brockenbrough, 1986).
drainage area, A, and reservoir shape factor, CA, are Prats (1961) and Prats et al. (1962a,b) investigated the
determined from the boundary-dominated flow regime. effects of infinite capacity vertical fractures on well
Combining all the information together should yield a performance. Prats examined the flow of incompressible
clear picture about the created fracture and help the fluids, whereas, Prats et al. inspected the flow of
reservoir engineer better understand and characterize compressible fluids and presented solutions for wells
the reservoir being investigated. producing at either constant rate or constant bottomhole
Unfortunately, however, some of the flow regimes, pressure. Russell and Truitt (1964a,b) pioneered applica-
especially those pertaining to early well test time such tions of methods based on the assumption of pseudo-radial
as fracture linear or bilinear flow, may often be masked flow in infinite-conductivity fractured wells. They
by wellbore storage effects. Furthermore, sometimes tabulated their numerical results in terms of dimensionless

Fig. 1. Sequence of flow regimes in a vertically fractured well.


I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 305

pressure drop as a function of time and fracture half- diligence should be exercised in its application and cross
length. Gringarten et al. (1975a,b) claimed that the study validation of the results with conventional plots should
of Russell and Truitt is unsuitable for short-term well tests. be made whenever possible.
Thus, they presented analytical solution for infinite- Tiab (1989, 1994, 1995), Tiab et al. (1999), and
conductivity and uniform-flux fractures that alleviates Mongi and Tiab (2000) introduced a new technique for
this shortcoming. Their solution is applicable to both the analysis of pressure derivative data of fractured
closed square and infinite reservoirs. Cinco-Ley et al. wells without resorting to conventional type-curve
(1978) modified the work of Russell and Truitt and matching. The technique was proven to be very
presented a semi-analytical model and type curves for the successful in determining the various fracture and
analysis of pressure data of wells intersected by finite- reservoir parameters. The power of the method
conductivity vertical fractures in infinite slab reservoirs. originates from the fact that it uses slopes and
Cinco-Ley and Samaniego-V (1977) investigated the intersection points of exact analytical solutions. In
effects of wellbore storage and fracture damage on the 2003, Nunez et al. successfully extended the applica-
well test data. They concluded that wellbore storage and tion of this technique to fractured gas wells (Nunez
damage have detrimental consequences on the well test et al., 2003). However, one of the problems facing this
and thus, should be seriously considered if effective and technique as well as any other conventional technique
accurate results are to be obtained. Hanley and is the presence of wellbore storage effects which span a
Bandyopadhyay (1979) presented a simple semi-analyt- long period of testing time especially in finite-
ical model for a well in a square reservoir with a uniform- conductivity fractured wells. Multirate test and con-
flux fracture fully penetrating the formation in the stant-pressure test have proven to be two of the
horizontal direction (xe /xf = 1). In addition to wellbore remedial well tests that minimize the adverse effects
storage, they also examined the effect of fracture length of wellbore storage (Earlougher, 1977; Agarwal et al.,
and fracture conductivity. Their results confirmed the 1979a,b). The theory of constant-pressure test has long
conclusions of Cinco-Ley and Samaniego-V regarding been well-established in the oil industry. Many papers
the adverse impact of wellbore storage on well test data. published in the literature addressed the applications of
Furthermore, they have shown that the pressure behavior constant-pressure tests to oil and gas reservoirs either
of a finite-conductivity fracture at intermediate time homogeneous or fractured (Samaniego-V and Cinco-
values is similar to the linear flow behavior of infinite- Ley, 1980; 1991; Guppy et al., 1981; 1988; Thompson,
conductivity fracture plus an extra pressure drop that is 1981; Thompson and Reynolds, 1986; Lio and Lee,
function of the fracture conductivity. Cinco-Ley et al. 1994; Berumen et al., 1997; Nashawi, 2003). Further-
(1989) observed this pressure behavior for dimensionless more, constant-pressure tests have an advantage over
fracture conductivity greater than 5π (FCD N 5π) and for constant-rate tests because they are immune of well-
limited time range which they defined as tDxf (FCD)2 N 1. bore storage effects which distort early time pressure
They called this flow regime “pseudolinear” to distin- data and may lead to wrong results. Constant-pressure
guish it from the real linear flow characterizing the well tests enable the well test analyst to analyze early
infinite-conductivity fracture. In 1982, Ahmed presented time data and obtain better description of the area
a detailed study on all the available specialized plots used around the wellbore in addition to reducing the well
to analyze various types of fractures and formation flow test duration. Since the initiation of the first constant-
regimes (Ahmed, 1982). pressure well test, the oil industry has witnessed a huge
Type-curve matching techniques are also being technological advancement in terms of new state-of-
extensively used for well test analysis of fractured the-art tools that are capable of measuring small rate
wells (Gringarten et al., 1974a,b; Cinco-Ley and variations with good accuracy (Piers et al., 1987; Eyl et
Samaniego-V, 1977; Baker and Ramey, 1978; Cinco- al., 1994; Chandran et al., 2005; Poulisse et al., 2006).
Ley et al., 1978; Agarwal et al., 1979a,b; Wong et al., This paper uses log–log reciprocal rate and
1986). Type-curve matching methods have been con- reciprocal rate derivative plots for the analysis of
sidered as excellent diagnostic tools because they are finite-conductivity fractured wells producing at con-
more general than the specialized analytical plots in the stant bottomhole pressure from closed reservoirs. The
sense that they span the entire range of flow regimes analysis covers the bilinear, pseudo-radial, and bound-
and even include the intervening transition zones. ary-dominated flow regimes. A step-by-step procedure
Consequently, the beginning and the end of each flow is presented to illustrate the applicability of the
period can be recognized. However, due to the non- technique. Two simulated examples are used to
uniqueness of the type-curve results, extreme care and demonstrate the validity of the method in minimizing
306 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

the wellbore storage effects and providing accurate test 3. Mathematical development
results.
The various flow regimes that commonly prevail
2. Numerical model during a well test of a massive hydraulically fractured well
with finite-conductivity vertical fracture are illustrated in
Gassim3B simulator (Lee and Wattenbarger, 1996) Fig. 1. However, if the fracture conductivity is low and the
was used to generate the data required for the analysis. well test is performed long enough until the effects of
This is a single-phase numerical simulator that production reach the outer reservoir boundary, three main
simulates liquid and real gas flow in the reservoir. flow regimes can be observed: (i) bilinear flow, (ii)
Gassim3B is a two-dimensional model that would be pseudo-radial flow, and (iii) boundary-dominated flow.
used with either x–y or r–z geometries. In all simulated The development of the equations corresponding to
examples, the reservoir model was considered to be each flow regime and their applications to calculate the
composed of a horizontal porous medium that has various fracture and reservoir parameters are discussed in
isotropic rock properties and uniform pay zone the subsequent sections.
thickness. The well, which is intersected by a vertical
finite-conductivity fracture, fully penetrates the pro- 3.1. Bilinear flow regime
ducing formation and produces oil at constant bottom-
hole pressure. The fracture has symmetrical properties Cinco-Ley and Samaniego-V (1981) introduced this
on both sides of the well and penetrates the entire flow regime to the well test literature. They claimed that
vertical extent of the pay zone. The grids perpendicular this regime prevails when the oil entering the wellbore is a
to the fracture are very fine in the vicinity of the combination of two linear flow regimes, (i) fracture linear
wellbore, increase until the middle of the fracture, then flow and (ii) formation linear flow (Fig. 1), hence, the
decrease to the tip of the fracture, and then increase name bilinear flow.
again to the reservoir boundary. The fine grids around The equation describing the flow of oil at constant
the wellbore and the tip of the fracture allow accurate bottomhole pressure during the bilinear flow regime is
modeling of the flow behavior where large pressure given in dimensionless form as (Guppy et al., 1981):
drops occur in the reservoir near the fracture. The grid
1 2:7222 0:25
spacing in the direction parallel to the fracture face is ¼ pffiffiffiffiffiffiffiffi tDxf ð1Þ
also fine near the fracture and increases toward the qD FCD
reservoir boundary (Nashawi, 2006). Fig. 2 is a
All the equations presented in this work are expressed
schematic illustration of a vertically fractured well
in the customary field units illustrated in Table 1.
located in the center of a closed square reservoir.
The dimensionless time based on fracture half-length,
tDxf, dimensionless fracture conductivity, FCD, and
dimensionless rate, qD, are respectively defined as:
0:0002637kt
tDxf ¼ ð2Þ
/lct x2f

kf wf
FCD ¼ ð3Þ
k xf
141:2 qBl
qD ¼ ð4Þ
khDp
where the pressure drop Δp is defined as:
Dp ¼ pi −pwf ð5Þ

Substituting Eqs. (2), (3), and (4) into Eq. (1) and
solving for 1 /q yields:
 0:25
1 48:98Bl 1
¼ pffiffiffiffiffiffiffiffiffi t 0:25 ð6Þ
Fig. 2. Schematic illustration of a vertical fracture in a closed system. q hDp kf wf /lct k
I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 307

Table 1
SI units and customary field units
Parameter Customary field units SI units
A 2
ft x 9.290 E − 02= m2
ct psi− 1 x 1.450377 E − 01= kPa− 1
h ft x 3.048 E − 01= m
k md x 9.869233 E + 04= m2
p psi x 6.894757 E + 00= kPa
q STB/D x 1.589873 E − 01= STD m3/D
rw ft x 3.048 E − 01= m
μ cp x 1.0 E + 00= mPa.s
kf md x 9.869233 E + 04= m2
wf ft x 3.048 E − 01= m
xf ft x 3.048 E − 01= m

let: The fracture conductivity can also be calculated from


 0:25 the reciprocal rate derivative plot. This is achieved by
48:98Bl 1
mB ¼ pffiffiffiffiffiffiffiffiffi ð7Þ differentiating Eq. (8) with respect to time as follows:
hDp kf wf /lc tk
1 dq
then Eq. (6) can be written as: − 2 ¼ 0:25mB t −0:75 ð14Þ
q dt
1
¼ mB t 0:25 ð8Þ taking the logarithm of both sides of Eq. (14) yields:
q  
1 dq
the subscript B indicates bilinear flow. log − 2 ¼ −0:75logðtÞ þ logð0:25mB Þ ð15Þ
q dt
In conventional well test analysis, a Cartesian plot of
1 /q versus t0.25 yields a straight line passing through the 1 dq
A log–log plot of − versus t should yield a
origin of the axes and having a slope equals to mB. The q2 dt
fracture conductivity, kfwf, can be obtained from the slope straight line with a negative slope of 0.75. For t= 1 h, Eq.
of the straight line as: (15) becomes:
   0:5  
48:98Bl 2 1 log − 2
1 dq
¼ logð0:25mB Þ ð16Þ
kf wf ¼ ð9Þ
hDpmB /lct k q dt B1h
Taking the logarithm of both sides of Eq. (8) yields: Eq. (16) can be written as:
   
1 1 dq
log ¼ 0:25logðtÞ þ logðmB Þ ð10Þ − 2 ¼ 0:25mB ð17Þ
q q dt B1h
If the bilinear flow behavior prevails in the fracture, a substituting mB from Eq. (17) into Eq. (7) and solving for
log–log plot of 1 /q versus t should yield a straight line
k f wf yields:
with a slope of 0.25. For t = 1 h, Eq. (10) becomes:

½ 
  2
1
log ¼ logðmB Þ ð11Þ  0:5
q B1h 12:245Bl 1
kf wf ¼   ð18Þ
1 dq /lct k
Eq. (11) can be written as: hDp − 2
  q dt B1h
1
¼ mB ð12Þ
q B1h 1 dq
Even though a plot of − 2 versus t provides a direct
q dt
substituting Eq. (12) into Eq. (7) and solving for kfwf means to calculate the fracture conductivity from Eq. (18),
yields: it is always desirable in well test analysis to observe the

½ 
2
quarter slope line during the bilinear flow regime. This can
 0:5
48:98Bl 1 be accomplished by multiplying Eq. (14) by t:
kf wf ¼   ð13Þ
1 /lct k 1 dq
hDp −t ¼ 0:25mB t 0:25 ð19Þ
q B1h q2 dt
308 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

Taking the logarithm of both sides of Eq. (19) yields: well test analyst with the opportunity of drawing the
  quarter slope line across the reciprocal rate derivative data
1 dq
log −t 2 ¼ 0:25logðtÞ þ logð0:25mB Þ ð20Þ with more confidence keeping in mind that at t= 1 h the
q dt reciprocal rate is four times the reciprocal rate derivative
1 dq (Eq. (24)).
The log–log plot of −t 2 versus t resulting from
q dt
Eq. (20) should have the familiar slope of 0.25 character-
3.1.1. Well in a closed reservoir
istics of bilinear flow of finite-conductivity fracture.
The dimensionless reciprocal rate during bilinear flow
Setting t = 1 h, Eq. (20) becomes:
regime for a well located in a closed reservoir can be
 
1 dq written as:
log −t 2 ¼ logð0:25mB Þ ð21Þ  
q dt B1h 1 2:7222 A 0:25 0:25
¼ pffiffiffiffiffiffiffiffi 2 tDA ð26Þ
Eq. (21) can be written as: qD FCD xf
  where the dimensionless time based on drainage area tDA
1 dq
−t 2 ¼ 0:25mB ð22Þ is defined as:
q dt B1h
0:0002637kt
Comparing Eqs. (17) and (22) demonstrates that the tDA ¼ ð27Þ
/lct A
1 dq 1 dq
two plots of − 2 and −t 2 versus t have the same
q dt q dt A log–log plot of 1/qD versus tDA should result in a
value at time t = 1 h. straight line with a slope of 0.25 during the bilinear flow
    period.
1 dq 1 dq
− 2 ¼ −t 2 ð23Þ Differentiating Eq. (26) with respect to tDA yields:
q dt B1h q dt B1h
 
Furthermore, comparing Eqs. (12) and (22) illustrates that 1 dqD 0:6805 A 0:25 −0:75
− 2 ¼ pffiffiffiffiffiffiffiffi 2 tDA ð28Þ
at time t = 1 h: qD dtDA FCD xf

    multiplying Eq. (28) by tDA yields:


1 1 dq
¼ 4 −t 2 ð24Þ  
q B1h q dt B1h 1 dqD 0:6805 A 0:25 0:25
−tDA 2 ¼ pffiffiffiffiffiffiffiffi 2 tDA ð29Þ
qD dtDA FCD xf
The fracture conductivity can be obtained from the plot 1 dqD
1 dq Eq. (29) implies that a log–log plot of −tDA versus
of −t 2 versus t by either substituting Eq. (23) into q2D dtDA
q dt tDA should result in a straight line with a slope of 0.25.
Eq. (18) or Eq. (24) into Eq. (13) as follows: For the special case of a square reservoir, Eq. (29) can
2 32
 0:5 be written as:
6 12:245Bl 7 1  
kf w f ¼ 4   5 ð25Þ 1 dqD 0:9624 xe 0:5 0:25
hDp −t q12 ddqt /lct k −tDA 2 ¼ pffiffiffiffiffiffiffiffi tDA ð30Þ
B1h qD dtDA FCD xf
1 dq
The plot of −t versus t may be more desirable
q2 dt
1 dq 3.2. Pseudo-radial flow regime
than the plot of − 2 versus t for two reasons: (1) the
q dt
slope of the former has a value of 0.25 which is a well For pseudo-radial flow regime, the reciprocal of the
recognized indicator of the presence of bilinear flow dimensionless flow rate of an oil well producing at cons-
characteristic of finite-conductivity fracture and (2) since tant bottomhole pressure is given as (Earlougher, 1977):
1 dq 1
the two plots, −t 2 versus t and 1 /q versus t, have ¼ 0:5½lnðtD Þ þ 0:80907 þ 2s ð31Þ
q dt qD
positive slopes, they can be drawn on the same graph
without any crossing between the data points. where the dimensionless time, tD, is defined as:
Usually the reciprocal rate derivative data are more
0:0002637kt
sensitive to rate variations than the reciprocal rate data. tD ¼ ð32Þ
Thus, having both plots on the same graph provides the /lct rw2
I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 309

substituting Eqs. (4) and (32) into Eq. (31) and solving for where tR is any convenient time taken from the horizontal
1 /q yields: line corresponding to the pseudo-radial flow regime and
    (1 /q)R is the matching value on the reciprocal rate curve.
1 70:6Bl 0:0002637kt
¼ ln þ 0:80907 þ 2s
q khDp /lct rw2 3.2.1. Well in a closed reservoir
ð33Þ The dimensionless reciprocal rate during pseudo-radial
Eq. (33) can be written as: flow regime for a well located in a closed reservoir is
    given as:
1 70:6Bl k
¼ lnðtÞ þ ln −7:43163 þ 2s    
q khDp /lct rw2 1 A
¼ 0:5 lnðtDA Þ þ ln 2 þ 0:80907 þ 2s ð40Þ
ð34Þ qD rw
Expressing Eq. (34) in base 10 logarithm yields:
    Differentiating Eq. (40) with respect to tDA yields:
1 162:6Bl k
¼ logðtÞ þ log −3:23 þ 0:869s 1 dqD 1
q khDp /lct rw2 − 2 ¼ 0:5 ð41Þ
qD dtDA tDA
ð35Þ
multiplying Eq. (41) by tDA yields:
Eq. (35) is used in conventional well test analysis to
determine the formation permeability and mechanical skin 1 dqD
−tDA 2 ¼ 0:5 ð42Þ
factor using a semilog plot of 1 /q versus t. qD dtDA
Differentiating either Eq. (34) or Eq. (35) with respect 1 dqD
A log–log plot of −tDA 2 versus tDA should result in
to t yields: qD tDA
    a horizontal line if the test is conducted long enough to
1 dq 70:6Bl 1 reach the pseudo-radial flow period.
− 2 ¼ ð36Þ
q dt R khDp t R
multiplying Eq. (36) by t yields:
3.2.2. Intersection between bilinear and pseudo-radial
 
1 dq 70:6Bl straight lines
−t 2 ¼ ð37Þ The bilinear and the pseudo-radial flow straight lines
q dt R khDp
intersect at a distinctive point. The coordinates of this
the subscript R in Eqs. (36) and (37) designates pseudo- point can be determined by equating Eqs. (29) and (42),
radial flow regime. and solving for tDA as follows:
1 dq
A log–log plot of −t 2 versus t should yield a  
q dt 1 kf wf 2
horizontal line if pseudo-radial flow prevails during the tDA ¼ ð43Þ
3:431A k
well test.
The formation permeability can be calculated from Eq. Eq. (43) can be written in real time as:
(37) as follows:
70:6Bl ðkf wf Þ2
k¼   ð38Þ tBR ¼ 1105:272ð/lct Þ ð44Þ
1 dq k3
hDp −t 2
q dt where tBR is the intersection time between the bilinear and
  R
1 dq pseudo-radial straight lines.
where −t 2 is determined by extrapolating the Eq. (44) can be used to verify the accuracy of the
q dt R
horizontal line drawn through the pseudo-radial data to the fracture conductivity and permeability calculated from
vertical axis. Eqs. (25) and (38), respectively.
The pseudo-radial flow period can also be used to
calculate the mechanical skin factor, s. Taking the ratio of 3.3. Boundary-dominated flow regime
Eqs. (35) and (37) and solving for s yields:
2   3
1 For long producing time, when the effect of
 
6 q R ktR 7 production reaches the outer reservoir boundary, the
s ¼ 1:15136 40:4343  1 dq −log /lct r2 þ3:235
7
reciprocal rate derivative starts to deviate from the
−t 2 w
pseudo-radial horizontal line given by Eq. (37). During
q dt R
this period, for constant bottomhole pressure produc-
ð39Þ tion, the reciprocal rate changes exponentially with
310 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

time. The equation describing the dimensionless flow where tRBD is the time at which the reciprocal rate
rate during the boundary-dominated flow period is derivative plots of the pseudo-radial and the boundary-
given as (Raghavan, 1993): dominated flow regimes intersect.
  Furthermore, El-Banbi and Wattenbarger (1998) and
1 2ktDA
qD ¼ exp − ð45Þ Wattenbarger et al. (1998) have shown that the exponen-
a a
tial behavior of the dimensionless reciprocal rate starts at
where a is defined as: tDA = 0.25. Thus, substituting 0.25 for tDA into Eq. (52)
 
1 4A yields:
a ¼ ln g þs ð46Þ
2 e CA rw2 ktbBD
A¼ ð54Þ
Taking the inverse of the dimensionless flow rate, Eq. 948/lct
(45) becomes:
  where tbBD is the starting time of the boundary-dominated
1 2ktDA flow regime on the reciprocal rate curve.
¼ a exp ð47Þ
qD a It has been already shown in this work that the
Differentiating Eq. (47) with respect to the dimension- boundary-dominated flow regime starts at tDA = 0.0625 on
less time tDA yields: the reciprocal rate derivative curve. This value is four times
  less than the one given by El-Banbi and Wattenbarger
1 dqD 2ktDA
− 2 ¼ 2k exp ð48Þ (1998) and Wattenbarger et al. (1998) for the reciprocal
qD dtDA a rate curve (tDA =0.25). This fact provides the equation
Multiplying Eq. (48) by tDA gives: presented in this work for drainage area calculation (Eq.
  (53)) with favorable time advantage over Eq. (54).
1 dqD 2ktDA
−tDA 2 ¼ 2ktDA exp ð49Þ Moreover, the information embedded in the boundary-
qD dtDA a dominated data can be used to calculate the shape factor,
CA, of the reservoir under investigation. Solving Eq. (46)
Equating the boundary-dominated flow equation (Eq.
for CA gives:
(49)) and the pseudo-radial flow equation (Eq. (42)) to
determine the intersection point of the two curves yields:
  4A
2ktDA CA ¼ expð2s−2aÞ ð55Þ
2ktDA exp ¼ 0:5 ð50Þ eg rw2
a
Eq. (50) can be written as: The reciprocal rate derivative during the boundary-
 
2ktDA 1 dominated flow period (Eq. (49)) can be written as:
tDA exp ¼ ð51Þ
a 4k      
1 dq 141:2Bl 2ktDA
Eq. (51) cannot be easily solved analytically because it −t 2 ¼ 2ktDA exp ð56Þ
involves, in addition to time, many unknown parameters q dt BD khDp a
such as reservoir shape factor, drainage area, and skin
factor. Thus, meticulous numerical simulation runs were It has been shown that during the pseudo-radial period, the
performed in this work for various reservoir geometries, reciprocal rate derivative is constant (Eq. (37)). Substitut-
formation properties, and fracture lengths to determine a ing Eq. (37) into Eq. (56) yields:
value for tDA that provides accurate solution for Eq. (51).      
It was concluded that a tDA value of 0.0625 highly satisfies 1 dq 1 dq 2ktDA
−t 2 ¼ 4ktDA −t 2 exp
Eq. (51) under a wide range of reservoir parameters. q dt BD q dt R a
During the boundary-dominated flow regime Eq. (27) ð57Þ
can be used to calculate the drainage area, A, of the well as
follows: Eq. (57) can be written as:
0:0002637kt
A¼ ð52Þ 2  3
/lct tDA 1 dq
  −t 2
Substituting 0.0625 for tDA into Eq. (52) yields: 2ktDA 1 6 7
6  q dt BD 7
exp ¼ 4 5 ð58Þ
ktRBD a 4ktDA 1 dq
A¼ ð53Þ −t 2
237/lct q dt R
I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 311

solving Eq. (58) for a yields: It is worthy to mention that the results obtained from
2ktDA Eqs. (61) and (62) should not be compared to Dietz's
a¼ 2   3 ð59Þ (1965a,b) shape factors, which are determined from the
1 dq
6 −t 2 7 constant-rate solution. Helmy and Wattenbarger (1998)
ln6 q dt BD 7 have shown that using the constant-rate shape factors for
4 1 dq 5
4ktDA −t 2 wells producing at constant pressure introduces an error as
q dt R
high as 10% in the calculation of production forecast and
substituting Eq. (59) into Eq. (55) yields: ultimate recovery.
 
A
CA ¼ 2:2459 2 4. Rectangular systems
rw
8 9
>
> >
>
>
> >
> Tiab and Crichlow (1979) and Tiab and Kumar (1980)
>
> >
>
>
> >
> have shown that the time period required for the flow
>
> >
>
>
< >
= regime to change from infinite-acting radial flow into fully
4ktDABD
 exp 2s− 2  3 developed pseudosteady state flow is much longer for
>
> 1 dq >
>
> −t 2   >
>
>
rectangular systems than for square reservoirs. Tiab (1994)
>
> 6 7>
>
> ln 6  q dt BD 1 7>
>
>
illustrated that for a four-to-one rectangle the transition
>
> 4 4ktDABD >
5>
>
:
1 dq
−t 2 >
; period for uniform-flux and infinite-conductivity fractured
q dt R wells yields a straight line having a slope of 0.5 on the
ð60Þ log–log pressure derivative plot of constant-rate test. This
straight line corresponds to the influence of two closest
Eq. (60) can be written as:
  parallel boundaries (Tiab and Crichlow, 1979; Tiab and
A Kumar, 1980; Tiab, 1994).
CA ¼ 2:2459 2
r An exhaustive computer simulation study conducted in
8 w 9
> > this work for various reservoir and fracture properties has
>
> >
>
>
> >
> shown that the straight line transition between the infinite-
>
> >
>
>
> >
> acting pseudo-radial period and the boundary-dominated
>
> >
>
< 4ktDABD = flow period also exists on the log–log reciprocal rate
 exp 2s− 2   3 derivative plot of constant-pressure test. However, the
>
> 1 dq >
>
>
> −t 2 >
> slope of the line was found to be 0.65. Fig. 5 displays the
>
> 6 7 >
>
> ln 6 q dt
 BD 7−lnð4ktDABD Þ> >
> straight line behavior for three different rectangular
>
> 4 5 >
>
>
:
1 dq
−t 2 >
; systems.
q dt R The general equation describing the straight line
ð61Þ performance during the transition period of finite-con-
where tDABD is the dimensionless time based on drainage ductivity fractured wells producing at constant pressure
area calculated at any time t greater than or equal to tBD was found to be:
(t ≥tBD) during the boundary-dominated flow period.
If the pseudo-radial horizontal line is not observed, the 1 dqD
shape factor can be determined from both the reciprocal −tDA ¼ CtDA
0:65
ð63Þ
q2D dtDA
rate and reciprocal rate derivative plots during the
boundary-dominated flow period as: For three-to-one, four-to-one, and five-to-one rectangular
2   3
1 systems, the constant C is equal to 6, 7, and 8, respectively.
 
A 6 q BD 7 For a four-to-one system, expressing Eq. (63) in real
CA ¼ 2:2459 2 exp6 4 2s−ð4ktDABD Þ   7 5 time and rate gives:
rw 1 dq
−t 2
   
B l0:35
q dt BD
1 dq 1 0:65 0:65
ð62Þ −t ¼ 4:663 t ð64Þ
      q2 dt hDp k /ct A
1 1 dq 1
where and −t 2 are values of and let:
q BD q dt BD q
 
   
B l0:35
1 dq corresponding to tBD, respectively. The
−t 2 1 0:65
q dt mCB ¼ 4:663 ð65Þ
derivation of Eq. (62) is given in Appendix A. hDp k /ct A
312 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

then, Eq. (64) can be written as: Eq. (72) can be written in real time as:
" #1=4
1 dq k
−t ¼ mCB t 0:65 ð66Þ tBCB ¼ 11:174ð/lct A 1:625
Þ ð73Þ
q2 dt ðkf wf Þ5

where the subscript CB designates closest parallel where tBCB is the intersection time between the bilinear
boundary. and the closest parallel boundaries lines.
Taking the logarithm of both sides of Eq. (66) yields: Eq. (73) can be used to verify the accuracy of the
fracture conductivity and permeability obtained from Eqs.
  (25) and (69), respectively.
1 dq
log −t 2 ¼ 0:65logðtÞ þ logðmCB Þ ð67Þ
q dt
4.2. Intersection between pseudo-radial and closest
For t = 1 h, one can write: parallel boundaries straight lines

  The intersection point between the pseudo-radial


1 dq
−t 2 ¼ mCB ð68Þ horizontal line and the closest parallel boundaries straight
q dt CB1h
line can be determined by equating Eqs. (42) and (63) as
follows:
If the pseudo-radial straight line is not defined on the
reciprocal rate derivative plot, then Eqs. (65) and (68) can 0:65
7tDA ¼ 0:5 ð74Þ
be solved for the formation permeability k as:
Solving Eq. (74) for the dimensionless time tDA yields:


 

½
20 1 20
  7 tDA ¼ 13 ð75Þ
1 137 B 14
k ¼ 81:366l   ð69Þ
/ct A 1 dq
hDp −t 2 Eq. (75) can be written in real time as:
q dt CB1h  
/lct A
tRCB ¼ 65:406 ð76Þ
k
On the other hand, if the pseudo-radial horizontal line
is observed, Eqs. (65) and (68) can be used to determine where tRCB is time at which the pseudo-radial line and the
the drainage area A as: closest parallel boundaries line intersect.

½ 
Eq. (76) can serve many purposes. It can be used to
20
either (i) calculate the formation permeability or (ii)


13

1 l137 B drainage area and (iii) validate the accuracy of both values.
A ¼ 10:683   ð70Þ For completeness sake, the constant values associated
/ct k 1 dq
hDp −t 2 with Eqs. (63), (69), (70), (73), and (76) for three-to-one,
q dt CB1h
four-to-one, and five-to-one rectangular systems are
presented in Table 2.
4.1. Intersection between bilinear and closest parallel
boundaries straight lines 5. Features of finite conductivity vertical fracture

The bilinear straight line and the closest parallel Figs. 3 and 4 are log–log plots of dimensionless
boundaries straight line intersect at a unique point. This reciprocal rate and dimensionless reciprocal rate deriva-
point is determined by equating Eqs. (29) and (63) for a tive, respectively, versus dimensionless time based on
four-to-one rectangular system as follows: drainage area for a finite-conductivity fractured well inside
a square reservoir. The figures illustrate the behavior of
 
0:6805 A 0:25 0:25 reciprocal rate and reciprocal rate derivative for several
0:65
7tDA ¼ pffiffiffiffiffiffiffiffi 2 tDA ð71Þ values of xe /xf.
FCD xf
The plotted data span the three flow regimes, bilinear,
Solving Eq. (71) for the dimensionless time tDA yields: pseudo-radial, and boundary-dominated flow. As has been
  discussed in the previous section and shown graphically in
1 k 1:25
tDA ¼ A0:625 ð72Þ Figs. 3 and 4, each flow regime has a unique feature that can be
339:371 kf w f used to determine a specific fracture and/or reservoir parameter.
I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 313

Table 2
Constants associated with rectangular systems equations
Rectangular system Eq. (63) Eq. (69) Eq. (70) Eq. (73) Eq. (76)
Three-to-one 6 52.380 8.428 16.428 82.911
Four-to-one 7 81.366 10.683 11.174 65.406
Five-to-one 8 119.161 13.120 8.003 53.260

The bilinear flow period exhibits a quarter slope on both the fracture conductivity determined from the former
figures (Figs. 3 and 4). This is a distinctive feature of this period and formation permeability determined from the
flow regime. The slope of this line, on real time and rate latter period, respectively, by virtue of Eq. (44).
plot, can be used to determine the fracture conductivity For the case of rectangular systems, the reciprocal rate
using either Eq. (13) and Fig. 3 or Eq. (25) and Fig. 4. derivative plot exhibits a straight line with a slope of 0.65
The pseudo-radial flow period displays a horizontal during the transition time from the infinite-acting pseudo-
line as was demonstrated analytically in Eq. (42) and radial flow regime to the boundary-dominated flow
illustrated graphically in Fig. 4. This line is an exclusive regime as illustrated in Fig. 5. The straight line can be
characteristic of the pseudo-radial derivative plot. It can be used to either determine the formation permeability from
used to calculate the reservoir permeability and skin factor Eq. (69) provided that the drainage area is known or to
using Eqs. (38) and (39), respectively. calculate the drainage area using Eq. (70) provided that the
The boundary-dominated flow regime has exponential permeability is calculated from the pseudo-radial line or is
behavior on both reciprocal and reciprocal rate derivative known from a pre-fracture well test.
plots as shown in Figs. 3 and 4 and illustrated by Eqs. (47) Furthermore, the transition straight line, i.e., 0.65 slope
and (49), respectively. The data recorded during this flow line, intersects the quarter slope line of the bilinear period
regime can be used to calculate the drainage area of the and the horizontal line of the pseudo-radial period at two
well using either Eq. (53) and Fig. 4 or Eq. (54) and Fig. 3. exclusive points. The first intersection point can be used to
Furthermore, the reciprocal rate derivative plot can be verify the accuracy of the results calculated from the
used alone or in conjunction with the reciprocal rate plot to quarter slope line and the 0.65 slope line using Eq. (73).
determine the shape factor of the reservoir as illustrated in Whereas, the second intersection point serves many
Eqs. (61) and (62). purposes. It can be used to either calculate the reservoir
Moreover, the intersection point between the quarter permeability, the drainage area, or validate the accuracy of
slope line of the bilinear period and the horizontal line of these values using Eq. (76).
the pseudo-radial period serves a vital purpose in the Thus, the reciprocal rate and reciprocal rate derivative
analysis procedure. It can be used to verify the accuracy of allow the well test analyst to take full advantage of the data

Fig. 3. Reciprocal dimensionless rate variation as a function of Fig. 4. Reciprocal dimensionless rate derivative variation as a function
dimensionless time. of dimensionless time.
314 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

of the reciprocal rate plot and calculate kfwf


from Eq. (13).
Step 3: Determine the value of [− t(1 / q2 )dq / dt]R
corresponding to the pseudo-radial horizontal
line and calculate the formation permeability
k using Eq. (38).
Step 4: Determine the intersection time tBR of the
bilinear and pseudo-radial straight lines from
the log–log plot of the reciprocal rate
derivative curve. Also calculate tBR from Eq.
(44) using the values of kfwf and k determined
in Steps 2 and 3, respectively. If the two tBR
values are reasonably equal, then kfwf and k
are correct; however, if they are widely
different, shift one or both straight lines and
repeat Steps 2 through 4 until they become
Fig. 5. Reciprocal dimensionless rate derivative variation as a function equal. Generally, if the radial flow line is well
of dimensionless time for various closed rectangular systems. defined, then the bilinear line is the one that
should be moved. An upward shift will
included in each flow regime to determine the reservoir decrease the value of kfwf whereas a down-
and fracture properties. ward shift will increase it.
Step 5: Determine the values of (1 / q)R and [− t(1 / q2)
6. Analysis methodology dq / dt]R corresponding to any convenient time
tR during the pseudo-radial period and calcu-
A properly designed constant pressure test of a well late the skin factors using Eq. (39).
intercepted by a finite-conductivity vertical fracture in a Step 6: Determine the approximate time tRBD at which
closed system should provide all the required information the boundary-dominated flow starts on the
that enables the well test analyst to calculate the fracture reciprocal rate derivative plot and calculate the
conductivity k f wf, formation permeability k, skin factor s, drainage area A using Eq. (53); or determine
drainage area of the well A, and reservoir shape factor CA. the time tbBD at which the boundary-dominated
The objective of this work is to present a simple and yet flow begins on the reciprocal rate curve and
accurate method based on analytical solutions that fulfill calculate A using Eq. (54).
the purpose of such well test. Step 7: Determine the values of [− t(1 / q2)dq / dt]R and
The following step-by-step methodology illustrates the [− t(1 / q2)dq / dt]BD corresponding to any con-
application of the proposed technique for the case where venient time tBD during the boundary-domi-
all the flow regimes, i.e., bilinear, pseudo-radial, and nated period and calculate the shape factor CA
boundary-dominated flow, are observed. using Eq. (61); or determine the values of (1 /
q)BD and [− t(1 / q2)dq / dt]BD corresponding to
Step 1: Plot 1 / q and − t(1 / q2)dq / dt versus t on log– tBD and calculate CA from Eq. (62).
log paper. The appearance of a quarter slope Steps 1 through 7 apply for any closed system.
line indicates that the well is intercepted by a If the reciprocal rate derivative plot exhibits a
finite-conductivity fracture. The presence of 0.65 slope line feature of rectangular systems,
horizontal line signifies pseudo-radial flow. A proceed the calculations as follows:
straight line having a slope of 0.65 means that Step 8: Determine the value of [− t(1 / q2)dq / dt]CB1h
the well is located in a closed rectangular from the 0.65 slope transition line or its
system. extrapolation and calculate either the forma-
Step 2: Determine the value of [− t(1 / q2)dq / dt]B1h tion permeability k or the drainage area A
from the quarter slope line and calculate the using either Eqs. (69) or (70), respectively.
fracture conductivity kfwf using Eq. (25). If Step 9: Determine the intersection time tRCB between the
for any reason, the bilinear reciprocal rate pseudo-radial line and the 0.65 slope line from
derivative line is not well defined, obtain the the log–log plot of the reciprocal rate derivative
value of (1 / q)B1h from the quarter slope line curve. Calculate tRCB by substituting the values
I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 315

of k and A obtained from Steps 3 and 8,


respectively, into Eq. (76). If the two tRCB values
are approximately equal, then k and A are correct;
however, if they are extremely different, shift one
or both straight lines and repeat Steps 3 and
8 until they become equal.
Step 10: Determine the intersection time tBCB between
the bilinear line and the 0.65 slope line from
the log–log plot. Also calculate tBCB by
substituting the values of kfwf and k obtained
from Steps 2 and 8, respectively, into Eq. (73).
If the two tBCB values are approximately
equal, then kfwf and k are correct; however, if
they are different, shift one or both straight
lines and repeat Steps 2 and 8 until they
become equal. Fig. 6. Reciprocal rate and reciprocal rate derivative variation as a
function time, case no. 1.

7. Applications
7.1. Simulated case no. 1
Two synthetic wells are used to illustrate the
applicability of the proposed technique. The first case The synthetic well of this example produces at a
presents a fractured well located in the center of a constant bottomhole pressure of 2900 psi from a low
square reservoir whereas the second case is for a well permeability formation (k = 2 md). The well is intercepted
located in the center of a four-to-one rectangular by a vertical fracture having a conductivity of 1250 md-ft.
system. Pertinent oil, reservoir, and fracture properties After 350 h of well test the reservoir boundary started to
used to simulate the synthetic cases are reported in affect the data.
Table 3. Both cases have a dimensionless fracture Fig. 6 displays the reciprocal rate and reciprocal rate
conductivity FCD value of 5. derivative data for the entire testing time. The reciprocal
rate derivative curve reflects the presence of bilinear,
pseudo-radial, and boundary-dominated flow regimes. A
Table 3 straight line of slope 0.25 is drawn through the bilinear
Oil and reservoir properties — simulated cases flow data. A horizontal line is drawn through the infinite
Oil properties acting pseudo-radial flow data. The boundary-dominated
Oil viscosity, μ (cp) = 0.72 flow regime is identified by the peculiar exponential
Oil formation volume factor, B (RB/STB) = 1.475 behavior of the plotted points. Furthermore, a quarter
slope straight line is drawn through the bilinear data of the
Well/reservoir properties
reciprocal rate curve.
Productive thickness, h (ft) = 100
Wellbore radius, rw (ft) = 0.25 The values of the reciprocal rate and reciprocal rate
Hydrocarbon porosity, ϕ (%) = 0.23 derivative determined from the bilinear straight lines at
Reservoir pressure, pi (psi) = 3000 t = 1 h are 0.00307 (STB/D)− 1 and 0.000768 (STB/D)− 1,
Bottomhole pressure, pwf (psi) = 2900 respectively. Using these values, the fracture conductivity
Total compressibility, ct (psi− 1) = 3 × 10− 6
is calculated from Eqs. (13) and (25) to be 1252.17 md-ft
Formation permeability, k (md)
Case no. 1 = 2 and 1250.54 md-ft, respectively, and the dimensionless
Case no. 2 = 3 fracture conductivity is calculated from Eq. (3) to be 5.01
Drainage area, A (ft2) and 5.0, respectively. The pseudo-radial horizontal line
Case no. 1 = 6,250,000 intersects the vertical axis at [−t(1 /q2)dq / dt]R = 0.00239
Case no. 2 = 9,000,000
(STB/D)− 1. This value is substituted into Eq. (38) and the
Fracture properties formation permeability is determined to be 2.09 md.
Fracture half-length, xf (ft) = 125 To calculate the skin factor, s, two values are selected
Fracture conductivity, k f wf (md-ft): from Fig. 6 at t= 33 h, one from the reciprocal rate curve,
Case no. 1 = 1250 (1 / q)R = 0.0102 (STB/D)− 1, and the other from the
Case no. 2 = 1875
reciprocal rate derivative curve, [− t(1 / q2)dq / dt]R =
316 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

Table 4 of 3 md whereas the fracture intercepting the wellbore has


Results of simulated case no. 1 a conductivity of 1875 md-ft.
Parameter k, md k f wf, md-ft A, ft2 CA Fig. 7 illustrates the behavior of the reciprocal rate and
Simulator 2 1250 6,250,000 29.34 reciprocal rate derivative data for the entire testing time. In
values addition to the bilinear, pseudo-radial, and boundary-
Calculated 2.09 Eq. (13): Eq. (53): Eq. (61): dominated flow regimes, the reciprocal rate derivative
values 1252.17 6,212,757 29.23
curve reveals the existence of no-flow boundary parallel to
Eq. (25): Eq. (54): Eq. (62):
1250.54 6,212,757 31.21 the fracture orientation. The presence of this boundary is
confirmed by drawing a 0.65 slope line through the tran-
sition data from the pseudo-radial to the boundary-
dominated flow regimes.
0.00239 (STB/D)− 1, then s is determined from Eq. (39) to Selecting (1 / q)B1h = 0.002288 (STB/D)− 1 and [− t
be −4.9. (1 / q2 )dq / dt]B1h = 0.000572 (STB/D) −1 from the bilin-
Fig. 6 shows that the bilinear and the pseudo-radial ear straight lines of the reciprocal rate and reciprocal
straight lines intersect at tBR = 94 h. If this value is correct, rate derivative, respectively, the fracture conductivity
it should be approximately equal to the one evaluated from is calculated from both Eqs. (13) and (25) to be
Eq. (44) using the permeability and conductivity values 1872.32 md-ft. The dimensionless fracture conductiv-
calculated from the quarter slope line and the horizontal ity is calculated from Eq. (3) to be 4.993. The
line. Thus, using Eq. (44), tBR is determined to be 94.1 h, calculated kf wf is within 0.14% of the one used in the
which highly agrees with the one obtained from the simulator input file. The horizontal straight line crosses
figure. This good match also indicates that the fracture the vertical axis at [− t(1 / q2 )dq / dt]R = 0.00165 (STB/
conductivity and formation permeability calculated from D)− 1 ; substituting this value into Eq. (38), the forma-
Eqs. (25) and (38), respectively, are correct. tion permeability is evaluated to be 3.03 md. The skin
Fig. 6 illustrates that the boundary-dominated flow factor is calculated from Eq. (39) to be − 4.92, using
regime starts at tRBD = 350 h and at tbBD = 1400 h on the (1 / q)R = 0.00562 (STB/D)− 1 selected from the recip-
reciprocal rate derivative and the reciprocal rate curves, rocal rate plot at t = 10 h.
respectively. Substituting tRBD into Eq. (53) and tbBD into A tBR value of 69.19 h is determined from Eq. (44)
Eq. (54), and calculating the drainage area A, it was using k f wf and k calculated from Eqs. (25) and (38),
found out that both equations yield the same result, respectively. This value matches tBR =69 h obtained from
A = 6,212,757 ft2. the intersection of the bilinear and the pseudo-radial
Eqs. (61) and (62) can be used to evaluate the shape straight lines. This very good match verifies that the
factor CA. Eq. (61) requires reading two values from the calculated kfwf and k are accurate.
reciprocal rate derivative curve at an arbitrary time tBD For closed four-to-one rectangular systems, as the case
during the boundary-dominated period. [−t(1/q2)dq / of this example, Eq. (70), developed in this work, can be
dt]R = 0.00239 (STB/D)− 1 is selected from the horizontal
line whereas [−t(1 /q2)dq /dt]BD = 0.014676 (STB/D)− 1 is
obtained from the exponential curve, at tBD = 1543.54 h.
Using these values, CA is calculated to be 29.23. To use
Eq. (62), (1/q)BD =0.025351(STB/D)− 1 is selected from
the reciprocal rate curve at tBD = 1543.54 h. This value is
substituted with [−t(1 /q2)dq /dt]BD = 0.014676 (STB/D)− 1
into Eq. (62), then CA is determined to be 31.21.
Table 4 is a summary of the entire results obtained from
this example. It is clearly illustrated in this table that all the
calculated parameters are in excellent agreement with the
input values.

7.2. Simulated case no. 2

This case demonstrates the applicability of the pro-


posed technique to a well producing from a closed four-to- Fig. 7. Reciprocal rate and reciprocal rate derivative variation as a
one rectangular system. The formation has a permeability function time, case no. 2.
I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 317

Table 5 2. Log–log plots of reciprocal rate and reciprocal rate


Results of simulated case no. 2 derivative are used to calculate the fracture conduc-
Parameter k, md k f wf, md-ft A, ft2 CA tivity, formation permeability, skin factor, well
Simulator 3 1875 9,000,000 3.95 drainage area, and reservoir shape factor.
values 3. Straight line equations corresponding to various flow
Calculated 3.03 Eq. (13): Eq. (70): Eq. (61): regimes encountered during testing of finite-conduc-
values 1872.32 9,150,380 3.99
tivity fractured wells are developed and used to
Eq. (25):
1872.32 calculate the reservoir and fracture parameters.
4. The presented equations are based on constant-
pressure solution; thus, wellbore storage effects are
minimized which allows the analysis of early test
used to calculate the drainage area. This requires data and provides better description of the area
knowledge of [−t(1 /q2)dq / dt]CB1h, which is obtained around the wellbore.
from the 0.65 slope straight line to be 0.0000837 5. The points of intersection of the straight lines provide
(STB/D)− 1. Thus, A is calculated to be 9,150,380 ft2. a unique opportunity to validate the accuracy of the
Moreover, the time of intersection tBCB between the obtained results as well as to determine some of the
bilinear and the 0.65 slope straight lines can be used to unknown parameters in certain circumstances where
validate the accuracy of k f wf, k, and A calculated from one of the flow regimes is not observed.
Eqs. (25), (38), and (70), respectively. A tBCB value of 6. A straight line of 0.65 slope was developed for
122 h is determined from Fig. 7. This value highly agrees rectangular closed systems. The straight line indi-
with tBCB = 122.07 h calculated from Eq. (73). This cates the presence of nearby parallel boundaries and
excellent match between the two tBCB values confirm the can be used to determine the formation permeability
accuracy of the calculated k f wf, k, and A. in the absence of pseudo-radial flow.
Furthermore, the accuracy of k and A evaluated from
Eqs. (38) and (70), respectively, can be established by Nomenclature
comparing the tRCB value calculated from Eq. (76) and the A drainage area of well, ft2
one obtained from the intersection between the horizontal B oil formation volume factor, RB/STB
and the 0.65 slope straight lines. Eq. (76) yields tRCB = CA reservoir shape factor
98.13 h, whereas tRCB =98 h is determine from Fig. 7. It is ct total compressibility, psi− 1
very obvious that there is an excellent agreement between dq / dt rate derivative, STB/D/h
the two tRCB values which confirms the precision of the FCD dimensionless fracture conductivity, Eq. (3)
calculated k and A. h net pay zone thickness, ft
Finally, to calculate the shape factor, [−t(1/q2)dq/ k formation permeability, md
dt]BD =0.57977 (STB/D)−1 and [−t(1/q2)dq/dt]R =0.00165 kf fracture permeability, md
(STB/D)−1 are selected from the reciprocal rate derivative mB slope of the bilinear flow plot, Eq. (7)
exponential curve and the pseudo-radial horizontal line, pi initial reservoir pressure, psi
respectively, at tBD =10,369 h. Substituting these values into pwf wellbore flowing pressure, psi
Eq. (61), a shape factor value of 3.99 is determined. q well flow rate, STB/D
Table 5 reports the results of this example. As it was the qD dimensionless flow rate for the constant-pressure
case of simulated example no. 1, it is clearly shown in the test, Eq. (4)
table that the calculated parameters are in excellent rw wellbore radius, ft
agreement with the input values. s skin factor
t production time, h
8. Conclusions tBCB intersection time between bilinear and closest
parallel boundaries lines, h
Several important conclusions summarize the work tbBD time corresponding to the beginning of the
presented in this paper: boundary-dominated flow regime, h
tBR intersection time between bilinear and pseudo-
1. The presented technique is simple and straightfor- radial straight lines, h
ward. No type-curve matching of any sort is required tD dimensionless time, Eq. (32)
which provides confidence in the uniqueness of the tDA dimensionless time based on drainage area, Eq.
results. (27)
318 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

tDABD dimensionless time based on drainage area cor- Appendix A


responding to boundary-dominated flow period
tDxf dimensionless time based on fracture half- Taking the ratio of Eqs. (47) and (49) yields:
length, Eq. (2)  
tBD production time corresponding to boundary- 1
 
dominated flow regime, h qD a 1
 ¼ ðA  1Þ
tR production time corresponding to the pseudo- 1 dqD 2k tDA
radial flow regime, h −tDA 2
qD dtDA
tRCB intersection time between pseudo-radial and
closest parallel boundaries lines, h solving Eq. (A-1) for a yields:
tRBD intersection time between the reciprocal rate
derivative of the pseudo-radial and the bound- 
1
ary-dominated flow regimes, h qD
wf fracture width, ft a ¼ 2ktDA   ðA  2Þ
1 dqD
xf fracture half-length, ft −tDA 2
qD dtDA
Greek symbols
γ constant = 0.5772 It is imperative to state that Eq. (A-2) is valid for tDA
Δp pressure drop, psi, Eq. (5) values greater than 0.25, i.e., this is when the boundary-
ϕ formation porosity, fraction dominated flow regime prevails on both the reciprocal rate
μ oil viscosity, cp and reciprocal rate derivative plots.
Eq. (A-2) can be written in terms of actual time and rate
Subscripts as:
A area  
1
B bilinear  
BCB bilinear and closest parallel boundaries 0:0002637kt q
a ¼ 2k   ðA  3Þ
intersection /lct A 1 dq
−t 2
BR bilinear and pseudo-radial intersection q dt
CB closest parallel boundaries
CD dimensionless conductivity substituting Eq. (A-3) into Eq. (55) and simplifying yields:
D dimensionless  
f fracture A
CA ¼ 2:2459
i initial rw2
2   3
R pseudo-radial 1
 
RCB pseudo-radial and closest parallel boundaries 6 ktBD q BD 7
intersection exp6
42s− 301:773/lct A  1 dq 5
7
BD boundary-dominated −t 2
q dt BD
RBD pseudo-radial and boundary-dominated flow
intersection ðA  4Þ
t total
Eq. (A-4) can be written as:
w well
wf well flowing 2   3
1
1h one hour  
A 6 q BD 7
CA ¼ 2:2459 2 exp6
42s−ð4ktDABD Þ   7
5
rw 1 dq
−t 2
Acknowledgments q dt BD
ðA  5Þ
The authors express their sincere appreciation to Prof.      
1 1 dq 1
Robert A. Wattenbarger of Texas A&M University for where and −t 2 are values of and
providing the computer codes for Gassim3B simulator.  q BD q dt BD q
Special thanks are also due to Prof. Djebbar Tiab for 1 dq
−t 2 corresponding to tBD, respectively.
inspiring the idea of the paper. q dt
I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320 319

References Gringarten, A.C., Ramey Jr., H.J., Raghavan, R., 1975a. Applied
pressure analysis for fractured wells. J. Pet. Technol. 887–892
Agarwal, R.G., Carter, R.D., Pollock, C.B., 1979a. Evaluation and (July).
performance prediction of low-permeability gas wells stimulated Gringarten, A.C., Ramey Jr., H.J., Raghavan, R., 1975b. Applied
by massive hydraulic fracturing. J. Pet. Technol. 362–372 pressure analysis for fractured wells. Trans. AIME 259.
(March). Guppy, K.H., Cinco, H., Ramey Jr., M.J., 1981. Transient flow
Agarwal, R.G., Carter, R.D., Pollock, C.B., 1979b. Evaluation and behavior of a vertically fractured well producing at constant
performance prediction of low-permeability gas wells stimulated pressure. SPE 9963 available at SPE headquarters, Richardson, TX.
by massive hydraulic fracturing. Trans. AIME 267. Guppy, K.H., Kumar, S., Kagawan, V.D., 1988. Pressure-transient
Ahmed, U., 1982. Transient pressure analysis of hydraulically analysis for fractured wells producing at constant pressure. SPE
fractured wells in the Western tight sands. SPE 10878. Rocky Form. Eval. 169–178 (March).
Mountain Reg. Meet., Billing, MT. Hanley, E.J., Bandyopadhyay, P., 1979. Pressure transient behavior of
Baker, B.J., Ramey Jr., H.J., 1978. Transient flow to finite the uniform influx finite conductivity fracture. SPE 8278. Annu.
conductivity vertical fractures. SPE 7489. Annu. Tech. Conf. Tech. Conf. Exhib., Las Vegas, Nevada.
Exhib., Houston, TX. Helmy, M.W., Wattenbarger, R.A., 1998. New shape factors for wells
Bennet, C.O., Rosato, N.D., Reynolds, A.C., Raghavan, R., 1982. produced at constant pressure. SPE 39970. Gas Technol. Symp.,
Influence of fracture heterogeneity and wing length on the response Calgary, Canada.
of vertically fractured wells. Soc. Pet. Eng. J. 681–698 (Aug.). Lee, S.T., Brockenbrough, J., 1986. A new analytic solution for finite
Berumen, S., Samaniego, F., Cinco, H., 1997. An investigation of conductivity vertical fractures. SPE Form. Eval. 75–88 (Feb.).
constant-pressure gas well testing influenced by high-velocity Lee, J., Wattenbarger, R.A., 1996. Gas reservoir engineering. SPE
flow. J. Pet. Sci. Eng. 18, 215–231. Textbook Series, vol. 5. Society of Petroleum Engineers, Inc.,
Chandran, T., Talabani, S., Jehad, A., Al-Anzi, E., Clark Jr., R., Wells, Richardson, TX, U.S.A.
J.C., 2005. Solutions to challenges in production logging of Lio, Y., Lee, W.J., 1994. New solutions for wells with finite-conductivity
horizontal wells using new tool. IPTC 10248. Int. Pet. Technol. fractures including fracture-face skin: constant well pressure cases.
Conf., Doha, Qatar. SPE 28605. Annu. Tech. Conf. Exhib., New Orleans, LA.
Cinco-Ley, H., Samaniego-V, F., 1977. Effect of wellbore storage and Locke, C.K., Sawyer, W.K., 1975. Constant pressure injection test in a
damage on the transient behavior of vertically fractured wells. SPE fractured reservoir — history match using numerical simulation
6752. Annu. Tech. Conf. Exhib., Denver, CO. and type-curve analysis. SPE 5594. Annu. Tech. Conf. Exhib.,
Cinco-Ley, H., Samaniego-V, F., 1981. Transient pressure analysis for Dallas, TX.
fractured wells. J. Pet. Technol. 1749–1766 (Sept.). Mongi, A., Tiab, D., 2000. Application of Tiab's direct synthesis
Cinco-Ley, H., Samaniego-V, F., Dominguez-A, N., 1978. Transient technique to multi-rate tests. SPE 62607. Western Reg. Meet.,
pressure behavior for a well with a finite-conductivity vertical Long Beach, CA.
fracture. Soc. Pet. Eng. J. 253–264 (Aug.). Narasimhan, T.N., Palen, W.A., 1979. A purely numerical approach
Cinco-Ley, H., Samaniego-V, F., Rodriguez, F., 1989. Application of for analyzing fluid flow to a well intercepting a vertical fracture.
the pseudolinear-flow model to the pressure-transient analysis of SPE 7983. California Reg. Meet., Ventura, CA.
fractured wells. SPE Form. Eval. 438–444 (Sept.). Nashawi, I.S., 2003. Pressure transient analysis for wells with variable
Dietz, D.N., 1965a. Determination of average reservoir pressure from sandface flow rate. J. Can. Pet. Technol. 42, 44–53 (July).
build-up surveys. J. Pet. Technol. 955–959 (Aug.). Nashawi, I.S., 2006. Constant-pressure well test analysis of finite-
Dietz, D.N., 1965b. Determination of average reservoir pressure from conductivity hydraulically fractured gas wells influenced by non-
build-up surveys. Trans. AIME 234. Darcy flow effects. J. Pet. Sci. Eng. 53, 225–238.
Earlougher Jr., R.C., 1977. Advances in well test analysis. SPE Nunez, W., Tiab, D., Escobar, F.H., 2003. Transient pressure analysis
Monogr. Series, vol. 5. Richardson, TX, U.S.A., pp. 38–40. for a vertical gas well intersected by a finite-conductivity fracture.
El-Banbi, A.H., Wattenbarger, R.A., 1998. Analysis of linear flow in SPE 80915. Prod. Operations Symp., Oklahoma City, OK.
gas well production. SPE 39972. Gas Technol. Symp., Calgary, Piers, G.E., Perkins, J., Escott, D., 1987. A new flowmeter for
Alberta, Canada. production logging and well testing. SPE 16819. Annu. Tech.
Eyl, K.A., Chapellat, H., Philippe, F., Flaum, C., Whittaker, S.J., Conf. Exhib., Dallas, TX.
Becker, A.J., Groves, J., 1994. High-resolution density logging Poulisse, H., van Overschee, P., Briers, J., Moncur, C., Goh, K.C.,
using a three detector device. SPE 28407. Annu. Tech. Conf. 2006. Continuous well production flow monitoring and surveil-
Exhib., New Orleans, LA. lance. SPE 99963. Intelligent Energy Conf. Exhib., Amsterdam,
Gringarten, A.C., Ramey Jr., H.J., 1973a. The use of source and The Netherlands.
Green's functions in solving unsteady flow problems in reservoirs. Prats, M., 1961. Effect of vertical fractures on reservoir behavior —
Soc. Pet. Eng. J. 285–296 (Oct.). incompressible fluid case. Soc. Pet. Eng. J. 105–118 (June).
Gringarten, A.C., Ramey Jr., H.J., 1973b. The use of source and Prats, M., Hazebrock, P., Strickler, W.R., 1962a. Effect of vertical
Green's functions in solving unsteady flow problems in reservoirs. fractures on reservoir behavior — compressible fluid case. Soc.
Trans. AIME 255. Pet. Eng. J. 87–94 (June).
Gringarten, A.C., Ramey Jr., H.J., Raghavan, R., 1974a. Unsteady- Prats, M., Hazebrock, P., Strickler, W.R., 1962b. Effect of vertical
state pressure distribution created by a well with a single fractures on reservoir behavior — compressible fluid case. Trans.
infinite conductivity vertical fracture. Soc. Pet. Eng. J. 347–360 AIME 225.
(Aug.). Raghavan, R., 1993. Well Test Analysis. Prentice Hall, Inc., Engle-
Gringarten, A.C., Ramey Jr., H.J., Raghavan, R., 1974b. Unsteady- wood Cliffs, New Jersey, U.S.A., p. 115.
state pressure distribution created by a well with a single infinite Raghavan, R., Cady, G.C., Ramey Jr., H.J., 1972a. Well test analysis
conductivity vertical fracture. Trans. AIME 257. for vertically fractured wells. J. Pet. Technol. 1014–1020 (Aug.).
320 I.S. Nashawi, A.H. Malallah / Journal of Petroleum Science and Engineering 57 (2007) 303–320

Raghavan, R., Cady, G.C., Ramey Jr., H.J., 1972b. Well test analysis Tiab, D., 1995. Analysis of pressure and pressure derivative without
for vertically fractured wells. Trans. AIME 253. using type-curve matching — skin and wellbore storage. J. Pet.
Russell, D.G., Truitt, N.E., 1964a. Transient pressure behavior in Sci. Eng. 12, 171–181.
vertically fractured reservoirs. J. Pet. Technol. 1159–1170 (Oct.). Tiab, D., Crichlow, H.B., 1979. Analysis of multiple-sealing fault
Russell, D.G., Truitt, N.E., 1964b. Transient pressure behavior in systems and bounded reservoirs by type curve matching. Soc. Pet.
vertically fractured reservoirs. Trans. AIME 231. Eng. J. 378–392 (Dec.).
Samaniego-V, F., Cinco-Ley, H., 1980. Production rate decline in pres- Tiab, D., Kumar, A., 1980. Detection of location of two parallel sealing
sure-sensitive reservoirs. J. Can. Pet. Technol. 75–86 (July–Sept.). faults around a well. J. Pet. Technol. 1701–1708 (Oct.).
Samaniego-V, F., Cinco-Ley, H., 1991. Transient pressure analy- Tiab, D., Azzougen, A., Escobar, F.H., Berumen, S., 1999. Analysis of
sis for variable rate testing of gas wells. SPE 21831. Rocky Mountain pressure derivative data of finite-conductivity fractures by the
Reg. Meet. Low-Permeability Reservoirs Symp., Denver, CO. “direct synthesis” technique. SPE 52201. Mid-Continent Opera-
Thompson, J.K., 1981. Use of constant pressure, finite capacity type tions Symp., Oklahoma City, OK.
curves for performance prediction of fractured wells in low Wattenbarger, R.A., El-Banbi, A.H., Villegas, M.E., 1998. Production
permeability reservoirs. SPE/DOE 9839. Low Permeability analysis of linear flow into fractured tight gas wells. SPE 39931.
Reservoirs Symp., Denver, CO. SPE Rocky Mountain Reg./Low Permeability Reservoirs Symp.
Thompson, L.G., Reynolds, A.C., 1986. Analysis of variable-rate Exhib., Denver, CO.
well-test pressure data using Duhamel's principle. SPE Form. Wong, D.W., Harrington, A.G., Cinco-Ley, H., 1986. Application of
Eval. 1, 453–469 (Oct.). pressure derivative function in the pressure-transient testing of
Tiab, D., 1989. Direct type-curve synthesis of pressure transient tests. fractured wells. SPE Form. Eval. 470–480 (Oct.).
SPE 18992. SPE Joint Rocky Mountain Reg./Low Permeability
Reservoirs Symp. Exhib., Denver, CO.
Tiab, D., 1994. Analysis of pressure and pressure derivative without
type-curve matching: vertically fractured wells in closed system.
J. Pet. Sci. Eng. 11, 323–333.

Das könnte Ihnen auch gefallen