Vous êtes sur la page 1sur 10

SPE 138843

Impacts of the Number of Perforation Clusters and Cluster Spacing on


Production Performance of Horizontal Shale Gas Wells
Y. Cheng, SPE, West Virginia University

Copyright 2010, Society of Petroleum Engineers


This paper was prepared for presentation at the SPE Eastern Regional Meeting held in Morgantown, West Virginia, USA, 1214 October 2010.
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 have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is 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 acknowledgment of SPE copyright.

Abstract
Multi-stage hydraulic fracturing has become the key technology to complete horizontal wells in shale gas reservoirs. In each
stage, multiple perforation clusters are used to create multiple transverse fractures. How to place these clusters significantly
affects both the short-term and long-term production performance of horizontal shale gas wells. The authors previous work
has demonstrated that when more than two fractures are created, mechanical interaction among fractures creates strong stress
concentrations around the inner fractures. As a result, the fractures between two edge fractures, i.e., sub-center and center
fractures, are limited to dilate, and their widths are much less than the edge-fractures width.
In this paper, reservoir simulation models were constructed by quantitatively incorporating the findings of the authors
previous work to investigate the impacts of the number of perforation clusters and cluster spacing on production performance
of horizontal shale gas wells. The paper illustrates that with the same cluster spacing, the scenario with more clusters has lower
ultimate gas recovery because of increased number of less-effective inner fractures. Given the same lateral length of a
horizontal well, although reducing cluster spacing increases the total number of fractures, smaller cluster spacing doesnt
necessarily improve well performance. An inadequate small cluster spacing can actually lead to more less-effective and
ineffective fractures, and therefore lower gas rate and ultimate recovery.
Introduction
In gas shale reservoirs, ultra-low permeability matrix is unable to flow gas at a viable rate and to provide an adequate drainage
volume. Horizontal drilling with multi-stage fracture treatment has become the key technology for development of shale gas
reservoirs. The economic viability and recovery enhancement of a production well largely depends on the effectiveness of
hydraulically created fractures. Under current low gas commodity price, it is particularly desirable and critical to create
maximum number of effective fractures along a reachable long horizontal wellbore.
The in-situ stress field has predominant influence on creation of a hydraulic fracture, e.g., impacting the orientation and
geometry of the fracture (Perkins and Kern 1961, Warpinski and Schmidt 1982). On the other hand, creation of a hydraulic
fracture could alter the stress field in the surrounding area of the fracture. The stress change, often referred as stress
concentration or stress shadow, decays with the distance away from the fracture (Solliman et al. 2004, Warpinski and
Branagan 1989). The stress shadow induced from multiple parallel fractures will inhibit the growth of the internal fractures,
i.e., sub-center and center fractures (Fisher et al. 2004, Cheng 2009). Chengs work used boundary element model with the
displacement discontinuity method to calculate the stress distributions around multiple transverse fractures and the geometries
of these fractures. It was illustrated that stress concentration occurs between fractures for the in-situ minimum horizontal
stresses (Fig. 1). The maximum horizontal stress is also altered while stress increase is greater for the minimum horizontal
stress than for the maximum horizontal stress. Also, it was found that the widths of the center and sub-center fractures are
strongly influenced by fracture spacing (perforation clusters spacing) where the width of the edge fractures (two outer
fractures) are relatively insensitive to fracture spacing (Cheng 2009). The closer the fractures are placed, the smaller the
widths of the center and sub-center fractures are. An inadequate small spacing may result in too small fracture width on the
center and sub-center fractures such that proppant transportation may become problematic.
For the multi-stage fracturing treatments, the stress concentration resulted from the previous fracturing stage may have
significant effects on the initiation and propagation of fractures in the subsequent stage when the cluster spacing is too close
(Warpinski and Branagan 1989). Stress concentration reaches the maximum during the fracture treatment, and decays till
fracture closure as pressure in fractures declines over time. However it is not uncommon that this pressure decay process takes
times longer than the duration between two fracturing stages due to limited and slow fluid loss into matrix. To eliminate the

SPE 138843

influence of stress concentration, it has been suggested that an offset distance (away from the fracture face) should be equal to
1.5 times the fracture height when the fracture height is less than fracture length (Fisher et al. 2004).
It is often impractical to place fractures or clusters apart 1.5 times fracture height because this will highly limit the
number of fractures, leading to inefficient drainage of shale gas reservoir. From reservoir engineering standpoint, it is desirable
to create more fractures with shorter spacing (Meyer et al. 2009). Current industry practice typically employs cluster spacing
less than 1.5 times fracture height, for example, cluster spacing of 100 ft vs. fracture height of 250 ft. In this situation, the
effect of stress concentration is not negligible. We need to secure a balance between creating more fractures and mitigating
the impact of stress concentration. To realistically optimize fracture design, stress alteration and variable fracture geometry
due to stress concentration must be taken into account.
In this paper, the fracture geometry and flow capacity are first quantified by incorporating the effects of stress
concentration due to the mechanical interaction in the creation of multiple fractures at a stage. The impact of such interaction
on the opening of the fractures in the subsequent stages is also evaluated through the elevated in-situ minimum horizontal
stress, which tends to restrain the initiation of the parallel fractures nearby. The reservoir simulation models are then
constructed to forecast the short-term and long-term production performance of horizontal shale gas wells where the
perforation clusters are closely located and the effect of the stress concentration is not negligible. Based on the forecasted
production performance, the economics of the well is evaluated. Four cases are presented to illustrate how the number of
perforation clusters and cluster spacing affect wells gas recovery and economics.
Case Study to Quantify Variation in Fracture Width
In this study, four cases are considered. All cases have the same lateral length (3000 ft), but vary with the number of
stages and cluster spacing.
Case 1 has three transverse fractures per stage and the cluster spacing is 100 ft. It has a total of ten stages. It is referred to
as the base case in this study.
Case 2 has five transverse fractures per stage with a total of six stages. The cluster spacing is 100 ft, same as case1.
Case 3, which is similar to case 1, has three transverse fractures per stage, but the cluster spacing is reduced to 50 ft.
Therefore it has a total of twenty stages
Case 4, which is similar to case 2, has five transverse fractures per stage, but the cluster spacing is reduced to 50 ft.
Therefore it has a total of twelve stages
Boundary element model with the displacement discontinuity method is used to calculate the geometries of fractures for
these cases. Other basic geomechanical properties of reservoir and fracture dimensions are listed in Table 1.
Table 1 Reservoir geomechanical and fracture properties
Poissons ratio
= 0.25
Youngs modulus
= 4.5*106 psi
Overburden stress
= 7,000 psi
Initial maximum horizontal stress (xx)
= 6,000 psi
Initial minimum horizontal stress (yy)
= 4,000 psi
Initial shear stress (xy)
= 0 psi
Net pressure at the center of fractures
= 500 psi
Fracture height
= 250 ft
Fracture half length
= 500 ft
For Cases 3 and 4 with a cluster spacing of 50 ft, the effect of the stress concentration resulted from the previous stage is
considered for all stages except the first one by elevating the minimum horizontal stress by 250 psi. This incremental in in-situ
stress is quite conservative based on the findings from the authors previous work (Cheng 2009). For example, for the case of
three fractures simultaneously created with a spacing of 100 ft, the minimum horizontal stress can be increased by 400 psi at
the location 200 ft away from the nearest edge fracture (Fig. 1). Although the effect of the stress concentration would
gradually decay with time due to pressure reduction in fractures, it could be a rather slow process and require a long time for
fractures to get closed due to the low permeable nature of shale formations. It is quite likely that the created fractures have not
closed yet when the subsequent stage treatment begins. On the other hand, the stress concentration is expected to be stronger
at a distance of 50 ft from the nearest edge fracture than at a distance of 200 ft. Therefore, 250 psi is used as a conservative
approximation to the increase in the minimum horizontal stress around new fractures to be initiated in the subsequent stage.
For Cases 1 and 2, the stress concentration caused by the previous stage is assumed to be negligible on the fractures formed in
the subsequent stage treatments. Note that it is not the intent of this study to precisely quantify the stress incremental for each
case. For the purpose of case comparison, it is adequate to focus on the relative magnitude of stress concentration among
cases.
The resultant fracture geometries at one stage for Cases 1 to 4 are shown in Figs. 2 to 5. Clearly, when more than two
fractures are created, the set of fractures at different positions display different geometric shapes. The edge fractures and subcenter fractures are asymmetrical and no longer elliptic-like. Note that the fracture width has been magnified by 1000 times for

SPE 138843

clarity of visualizing these fractures. The fractures between two edge fractures, i.e., sub-center and center fractures, are limited
to dilate due to inter-fracture interaction. Therefore, they have much less fracture width than that of the edge fractures. It can
be further understood that the widths of the center and sub-center fractures are strongly influenced by fracture spacing
(perforation clusters spacing) because they are suppressed from two sides, while the edge fracture width is relatively
insensitive to fracture spacing as they are not suppressed from one side by other fractures. From the previous study, it can be
seen that the closer the fractures are placed, the smaller the widths of the center and sub-center fractures will be.
The profiles of fracture width along the fracture half length from the fracture center to the tip are given in Figs. 6 to 9. As
can be observed, for all cases, the width of the center or sub-center fracture is much less than the width of the edge fractures.
For Cases 1 and 2 with the cluster spacing of 100 ft, the width of the center or sub-center fracture is less than half of the width
of the edge fractures. For Cases 3 and 4 with the cluster spacing of 50 ft, the width of the center or sub-center fracture is more
than four times less than the width of the edge fractures. Moreover, as the cluster spacing is reduced, the width of all the
fractures, either center or edge one, correspondingly shrinks (Figs. 8 and 9). For example, comparing Fig. 6 with Fig. 8, the
width of edge fractures is reduced by more than one time, and the width of center fracture is reduced by more than three times,
as the cluster spacing is reduced from 100 ft to 50 ft.
Reservoir Simulation Models and Forecasted Well Performance
Reservoir simulation models are constructed to simulate gas production from shale reservoirs for each individual case
defined in the previous section. The desorption mechanism is included into the simulation. The common properties of reservoir
and fluid used in the simulation models are given as follows in Table 2.
Table 2 Reservoir and fluid properties
Initial reservoir pressure
= 2,886 psia
Reservoir temperature
= 125 F
TOC
= 4.5%
Isotherm Langmuir volume
= 74.4 scf/ton
Isotherm Langmuir pressure
= 473 psia
Net pay thickness
= 120 ft
Matrix porosity
= 0.05
Matrix permeability
= 250 nd
Fracture half-length
= 500 ft
Gas gravity (air=1)
= 0.6
Gas initial viscosity
= 0.019 cp
In construction of simulation models, the variation of fracture widths along the fracture length determined in the previous
section is translated into variable fracture conductivity based on a relationship of cubic power between fracture width and
conductivity.
Effect of number of transverse fractures per stage on gas production
The gas production performance for Case 1 is demonstrated in Fig. 10 for a time period of 60 years, in terms of daily gas
production rate and cumulative gas production. The initial gas production rate is about 7.8 mmscf/d. The cumulatively
recovered gas volume is 4.54 bscf in 30 years and 5.17 bscf in 60 years. Fig. 10 also provides the gas production performance
for Case 2. The initial gas production rate is about 5.3 mmscf/d. The cumulatively recovered gas volume is 4.22 bscf in 30
years and 4.94 bscf in 60 years.
The difference between Cases 1 and 2 is the number of perforation clusters per stage. As shown in the previous section,
the width of fractures is different, depending on their relative positions. So are the flow capacities (conductivities) of the
resultant fractures. Compared with Case 1, Case 2 has a lower initial production rate, i.e., 2.5 mmscf/d less than Case 1s (a
relative reduction of 32.3%). The cumulative production of Case 2 is 0.32 bscf less than Case 1s in 30 years (a relative
reduction of 7.1%) and 0.24 bscf less than Case 1s in 60 years (a relative reduction of 4.6%). These results indicate that
increase of the number of clusters can result in less efficient production performance and gas recovery because the number of
less-effective fractures increases as the number of perforation clusters increases in one stage.
By comparison of Cases 3 and 4 with a closer cluster spacing (50 ft), we can observe similar characteristics. The gas
production performances of Case 3 and 4 are displayed in Fig. 11. The initial production rate of Case 4 is 1.4 mmscf/d less
than that of Case 3, corresponding to a relative reduction of 32.1%. The cumulative production of Case 4 is 0.40 bscf less than
Case 3s in 30 years (a relative reduction of 10.4%) and 0.31 bscf less than Case 3s in 60 years (a relative reduction of 6.8%).
Effect of perforation cluster spacing on gas production
Case 1 is taken as a base case to compare with Cases 3 and 4. Fig. 12 shows the production performances in terms of daily gas
production rate and cumulative gas production for Cases 3 and 4 along with Case 1. As we can see, the early-time production
rate for the case with larger spacing (100 ft of Case 1) is higher than those with smaller spacing (50 ft of Cases 3 and 4),
regardless of the number of fractures per stage. The initial gas rate for Case 3 is 4.4 mmscf/d, 43.1% less than that of Case 1.

SPE 138843

For Case 4, the initial gas rate is 3.0 mmscf/d, 61.4% less than that of Case 1.
Although Cases 3 and 4 both have a total of sixty fractures, which doubles the number of fractures that Case 1 has, they
produce less cumulative gas volumes. In 30 years, the cumulative gas production of Case 3 is 3.86 bscf compared with 4.54
bscf of Case 1, representing a reduction of 14.9%. The cumulative gas production of Case 4 is 3.46 bscf, representing a
reduction of 23.7%. In 60 years, the cumulative gas productions are 4.55 bscf for Case 3 and 4.24 for Case 4, corresponding to
a reduction of 12.0% and 18.0% respectively relative to Case 1.
Without question, the cluster spacing is an important factor influencing the production behavior of shale gas wells. Against
intuition, increasing the number of clusters doesnt necessarily improve well production performance. Particular attention
needs to be focused on this parameter in hydraulic fracturing treatment design.
Economic Evaluation
Considering a typical drilling & completion (D&C) cost of 4.5MM$/well and a flat strip gas price of $5/Mscf, the
economics of these cases are evaluated, with results demonstrated in Fig. 13. Clearly, the impacts of cluster numbers and
spacing on wells economics are significant. All four cases have same lateral length, but their economics are quite different
simply because of different completion designs in terms of numbers of clusters per stage and cluster spacing.
The best case is Case 1, which provides the highest NPV, 4.95 MM$ at a discount rate of 10 %. The rate of return is 68%,
representing an excellent economics. Recall that Case 1 has three perforation clusters/stage with a spacing of 100 ft. Case 2s
economics becomes less attractive compared with Case 1. Its NPV is decreased by 28%, to 3.58MM$ with the rate of return
dropping to 45%. Note that Case 1 and 2 have the same cluster spacing (100ft), but differ in the number of clusters. Case 2 has
five clusters per stage.
Compared with Case 1, Case 3s economics becomes much less attractive. Its NPV is decreased by 41%, to 2.9 MM$ with
the rate of return dropping to 37%. Note that Cases 1 and 3 have same numbers of cluster per stage but differ in cluster
spacing. Case 3s cluster spacing is 50 ft while Case 1s is 100ft. This means that Case 3 doubles the numbers of stages, which
would increase the completion cost. In this straightforward comparison, the difference in D&C cost is not included in the
evaluation. If the increase in completion cost was considered, the economics of Case 3 would be much worse.
The worst case is Case 4. Its NPV is only 1.47 MM$. Compared with Case 1, this represents 70% reduction. Accordingly,
the rate of return is only 20%. Recall that Case 4 has five perforation clusters/stage with a spacing of 50 ft. Note that reducing
cluster spacing may increase the completion cost. Again, in this straightforward comparison, the difference in D&C cost is not
included in the evaluation. Therefore, the estimate of Case 4s economics is at the optimistic end. Even though, its economics
is unattractive.
Although the evaluation results above are related to the specific reservoir petrophysical and geomechanical properties used
in this study, the principle findings hold in general. That is increasing numbers of clusters and/or reducing cluster spacing
doesnt necessarily improve wells production performance and economics. The reason is that wells performance is dictated
by the numbers of effective fractures that are actually created in the reservoir instead of the designed number of fractures.
Therefore, completion design should be optimized by focusing on maximizing the numbers of effective fractures.
Conclusions
This study investigates the impacts of the number of clusters and cluster spacing on production performance and economics
for a horizontal well with multi-stage multiple transverse fractures. The variations of fracture geometry due to the variations of
the number of clusters and cluster spacing are first estimated using the boundary element model. The resultant variable flow
capacities of fractures are incorporated into reservoir simulation for prediction of the production performance. Economics of
the studied cases are evaluated based on the predicted production performance and generic economic parameters.
The main conclusions from this study include:
1. The magnitude of mechanical interactions of fractures is closely related to the number of fractures simultaneously
created and fracture spacing. As a results, the width of fractures differ significantly with different number of fractures
and fracture spacing
2. Increasing the number of perforation clusters in one stage can lower the initial gas rate and the cumulative gas
production because the width growth of center and sub-center fractures is inhibited under certain fracture spacing and
geomechanical properties of the reservoir.
3. Decreasing the cluster spacing as to increase the total number of fractures can significantly reduce gas production
when the cluster spacing is reduced to an inadequately small size, with which the width growth of fracture is strongly
inhibited due to the mechanical interaction.
4. Wells production performance and economics are dictated by numbers of effective fractures that are actually created
in the reservoir instead of the designed number of fractures. Fracture treatment design should be optimized by taking
into consideration of the variations of fracture geometry and resultant flow capacity due to stress concentration effect.

SPE 138843

References
Cheng, Y. 2009. Boundary Element Analysis of the Stress Distribution around Multiple Fractures: Implications for the
Spacing of Perforation Clusters of Hydraulically Fractured Horizontal Wells. SPE 125769, SPE Eastern Regional Meeting,
Charleston, WV, 23-25 September .
Fisher, M.K. et al. 2004. Optimizing Horizontal Completion Techniques in the Barnett Shale Using Microseismic Fracture
Mapping. SPE 90051, SPE Annual Technical Conference and Exhibition, Houston, Texas, 26-29 September.
Perkins, T.K. and Kern, L.R. 1961. Widths of hydraulic fractures. JPT, 13 (9): 937-949.
Meyer, B. R. et al. 2010. Optimization of Multiple Transverse Hydraulic Fractures in Horizontal Wellbores. SPE 131732,
SPE Unconventional Gas Conference, Pittsburgh, PA, 2325 February.
Soliman, M. Y. et al. 2004. Geomechanic Aspects of Multiple Fracturing of Horizontal and Vertical Wells. SPE 86992, SPE
International Thermal Operations and Heavy Oil Symposium and Western Regional Meeting, Bakersfield, CA, 16-18
March.
Warpinski, N.R. et al. 1982. In-situ Stresses: the Predominant Influence on Hydraulic Fracture Containment. JPT, 34 (3): 653664.
Warpinski, N.R. and Branagan, P.T. 1989. Altered-stress Fracturing. JPT, 41 (9): 990-997.

SPE 138843

600

h,min , psi

-600
-600

600

Fig. 1 Altered minimum horizontal stress field for three parallel fractures with a spacing of 100 ft (after Cheng 2009)

200
150
100
50
0
-600

-400

-200

200

400

600

-50
-100
-150
-200
Fracture opening magnified by 1000 times

Fig. 2 - Fracture geometry of three parallel fractures with a spacing of 100 ft for Case 1
300
200
100
0
-600

-400

-200

200

400

600

-100
-200
-300

Fracture opening magnified by 1000 times

Fig. 3 - Fracture geometry of five parallel fractures with a spacing of 100 ft for Case 2

SPE 138843

200
150
100
50
0
-600

-400

-200

200

400

600

-50
-100
-150
-200
Fracture opening magnified by 1000 times

Fig. 4 - Fracture geometry of three parallel fractures with a spacing of 50 ft at the subsequent stage for Case 3
200
150
100
50
0
-600

-400

-200

200

400

600

-50
-100
-150
-200
Fracture opening magnified by 1000 times

Fig. 5 - Fracture geometry of five parallel fractures with a spacing of 50 ft at the subsequent stage for Case 4

0.5
Edge fractures
Center fracture
Fracture width, inch

0.4

0.3

0.2

0.1

0
0

100

200

300

400

500

600

Distance from wellbore, ft

Fig. 6 - Fracture width profiles of three parallel fractures with a spacing of 100 ft for Case 1

SPE 138843

0.5
Edge fractures
Sub-center fractures
Center fracture

Fracture width, inch

0.4

0.3

0.2

0.1

0
0

100

200

300

400

500

600

Distance from wellbore, ft

Fig. 7 - Fracture width profiles of five parallel fractures with a spacing of 100 ft for Case 2

0.5
Center fracture
Edge fractures
Fracture width, inch

0.4

0.3

0.2

0.1

0
0

100

200

300

400

500

600

Distance from wellbore, ft

Fig. 8 - Fracture width profiles of three parallel fractures with a spacing of 50 ft at the subsequent stage for Case 3

0.5
Edge fractures
Sub-center fractures

Fracture width, inch

0.4

Center fracture

0.3

0.2

0.1

0
0

100

200

300

400

500

600

Distance from wellbore, ft

Fig. 9 - Fracture width profiles of five parallel fractures with a spacing of 50 ft at the subsequent stage for Case 4

SPE 138843

Gas rate for Case 2


Cum gas for Case 2

100,000

6.0

10,000

4.5

1,000

3.0

100

1.5

10
1

10

100

1,000

10,000

Cum. gas production, bscf

Gas production rate, mscf/d

Gas rate for Case 1


Cum gas for Case 1

0.0
100,000

Time, days

Fig. 10 Production performance for Cases 1 and 2.


Gas rate for Case 4
Cum gas for Case 4

100,000

6.0

10,000

4.5

1,000

3.0

100

1.5

10
1

10

100

1,000

10,000

Cum. gas production, bscf

Gas production rate, mscf/d

Gas rate for Case 3


Cum gas for Case 3

0.0
100,000

Time, days

Fig. 11 Production performance for Cases 3 and 4.


Gas rate for Case 1

Gas rate f or Case 3

Gas rate for Case 4

Cum gas f or Case 1

Cum gas for Case 3

Cum gas f or Case 4

10,000

4.5

1,000

100

1.5

10

10

100

1000

10000

100000

Time, days

Fig. 12 Production performance for Cases 1, 3 and 4.

Cum. gas production, bscf

Gas production rate, mscf/d

100,000

10

SPE 138843

NPV (10), MM$

IRR: 78%

IRR: 45%
IRR: 37%

3
4.95

2
3.58

IRR: 20%
2.9

1
1.47

0
Case 1

Case 2

Case 3

Case 4

Fig. 13 Economics of Case 1 to Case 4 with D$C cost of 4.5 MM$ and gas price of $5/Mscf

Vous aimerez peut-être aussi