Journal of Petroleum Science and Engineering 113 (2014) 17

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: www.elsevier.com/locate/petrol

Sensitivity analysis of hydraulic fracture geometry in shale

gas reservoirs
W. Yu a, Z. Luo a, F. Javadpour b,n, A. Varavei a, K. Sepehrnoori a
a
b

Petroleum and Geosystems Engineering, The University of Texas at Austin, Austin, TX, USA
Bureau of Economic Geology, The University of Texas at Austin, University Station, Box X, Austin, TX 78713, USA

art ic l e i nf o

a b s t r a c t

Article history:
Received 12 October 2012
Accepted 14 December 2013
Available online 22 December 2013

The combination of horizontal drilling and multiple hydraulic fracturing has been widely used to
stimulate shale gas reservoirs for economical gas production. Numerical simulation is a useful tool to
optimize fracture half-length and spacing in a multistage fracturing design. We developed a methodology to use a commercial reservoir simulator to simulate production performance of shale gas reservoirs
after fracturing. We veried our simulation method with the available eld data from the Barnett Shale.
In this work, we performed a sensitivity study of gas production for a shale gas well with different
geometries of multiple transverse hydraulic fractures, in which fractures0 half-lengths vary. Hydraulic
fractures are divided into two outer and inner fracture groups. The simulation results revealed that the
outer fractures contribute more to gas production when fracture spacing is small due to the effect of
fracture interference. Also, we studied the effects of fracture half-length and fracture spacing on gas
production. This work can provide some insights into characterization of hydraulic fracture geometry on
the basis of production data in shale gas reservoirs.
& 2013 Elsevier B.V. All rights reserved.

Keywords:
hydraulic fracturing
shale gas
sensitivity analysis
fracture interference
horizontal well

1. Introduction
Newly developed techniques in the elds of horizontal drilling
and hydraulic fracturing have made possible the current ourishing gas production from shale gas plays in the United States, as
well as the fast-growing investment in shale gas exploration and
development worldwide. Hydraulic fracturing has been used in
the oil and gas industry since the 1940s. Multiple transverse
hydraulic fractures in a horizontal wellbore can create a large
stimulated reservoir volume (SRV), which is the main contributor
to high gas production from shale gas plays with extremely low
permeability (Javadpour et al., 2007; Warpinski et al., 2009;
Javadpour, 2009; Soliman and Kabir, 2012).
Although hydraulic fractures improve gas production from
shale gas wells, hydraulic fracturing is expensive. Long laterals
require greater volume of liquids and proppants, contributing to
higher cost (Kaiser, 2012). Therefore, optimization of hydraulic
fracture parameters, such as fracture spacing and fracture halflength, is important. It is well known that the longer the fracture
half-length and the shorter the fracture spacing, the higher gas
production will be. The greater the number of fractures in the
shale around the wellbore, the faster the gas will be produced

Corresponding author. Tel.: 1 512 232 8068; fax: 1 512 471 0140.
E-mail addresses: farzam.javadpour@beg.utexas.edu,
fgjavad@yahoo.com (F. Javadpour).
0920-4105/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.petrol.2013.12.005

(Kalantari-Dahaghi, 2011). In some cases, more than 20 fracture

stages have been tried in a horizontal well in order to increase
fracture contact with the formation and to produce high initial gas
rates (Castaneda et al., 2010; King, 2010). Meyer et al. (2010)
suggested that initial production increases linearly with number of
fractures, but a high gas ow rate is not sustainable and will
decline sharply once fractures interfere over a given lateral length.
Fazelipour (2011) stated that more hydraulic fractures around the
wellbore increase production rate; fractures are the key component to effective production. Mayerhofer et al. (2010) pointed
out that a large SRV with small fracture spacing could provide
maximum well performance and gas recovery, considering economic optimization in design. Waters et al. (2009) stated that
optimum fracture spacing depends on the incremental cost associated with creating denser fracture systems or productivity
improvement from available fracture networks. All of these studies
suggest the importance of economic factors in optimizing multiple
transverse hydraulic fractures. For example, reduction in fracture
spacing is expensive and may even cause interference and subsequent reduction in gas production. Analytical models do not
incorporate fracture interference and assume equal fracture halflength (Ambrose et al., 2011; Zhao et al., 2012); hence, numerical
simulation is necessary.
In this paper, we performed a sensitivity study of gas production for a shale gas well with uncertain but possible hydraulic
fracture geometry. Hydraulic fractures are divided into outer
fractures and inner fractures (Fig. 1). The performance of the outer

W. Yu et al. / Journal of Petroleum Science and Engineering 113 (2014) 17

fractures is much more benecial for gas production than is the

performance of the inner fractures, owing to the interference
effect of inner fractures. We also studied the effects of both
fracture half-length and fracture spacing on gas production. The
goal of this work is to provide insights into characterization of
hydraulic fracture geometry on the basis of production data in
shale gas reservoirs.

2. Shale gas reservoir modeling

Given the complex nature of hydraulic fracture growth and the
very low permeability of matrix rock in shale gas reservoirs,
coupled with the predominance of horizontal completions, reservoir simulation is a preferred approach to predict and evaluate
well performance. Local grid renement with logarithmic cell
spacing is used in simulation to accurately model gas ow from
the matrix to a fracture, i.e., properly incorporating the transient
ow behavior from the matrix to and within the fracture. In a
block, the hydraulic fracture is explicitly modeled; moreover, in
order to properly simulate the large pressure drop between the
matrix and the fracture, the matrix is described as some sub-cells
whose size increases logarithmically while moving away from the
hydraulic fracture. In addition, a dual-permeability grid is used to
allow simultaneous matrix-to-matrix and fracture-to-fracture
ow. This method can accurately and efciently model transient
gas production from hydraulic fractures of horizontal wells in
shale gas plays (Rubin, 2010; Cipolla et al., 2010). The reservoir is
assumed to be homogeneous and the fractures evenly spaced, with
stress-independent porosity and permeability.
In our simulation, gas is owing into the wellbore only through
fractures, i.e., no matrix-wellbore communication exists. The
gas ow, which is turbulent as a result of the high gas ow
rate in hydraulic fractures, is modeled with non-Darcy ow. The
Outer fractures
Fracture

Fracture
half-length

Horizontal well

Inner fractures
Fig. 1. Sketch of four induced hydraulic fractures in horizontal shale gas production
well.

non-Darcy Beta factor used in the Forchheimer number is determined using a correlation proposed by Evans and Civan (1994):
f 1:485E9=K 1:021

where the unit of K is md and the unit of is ft  1. The (f) correlation was developed using more than 180 data points, including
those for propped fractures, and was found to closely match the
data (Rubin, 2010). Fig. 1 is a diagram of a typical shale gas
completion design, illustrating several important geometric fracture parameters, such as fracture spacing and fracture half-length.

3. Case study
Published average reservoir data for the Barnett Shale in the
Newark East eld were used (Grieser et al., 2009). The Mississippian Barnett Shale, which sits on an angular unconformity
above the Cambrian- to upper-Ordovician-age carbonates of the
Ellenberger Group and Viola Formation and overlying the
Pennsylvanian-age Marble Falls Limestone, is associated with the
late Paleozoic Ouachita orogeny and located in the Fort Worth
Basin area in North-Central Texas (Roy et al., 2013). The least
principle stress is roughly NW to SE, which is the wellbore
orientation, so that multiple transverse hydraulic fractures would
be created (Fisher et al., 2004).
In this case, the well was stimulated by a four-stage fracturing
with a single, perforated interval for each stage. Detailed reservoir
information about this section of the Barnett Shale is listed in
Table 1. For this well, fracture maps were obtained using geophones installed in offset wells, and estimates of fracture halflength at each initiation were provided (Grieser et al., 2009). We
used the simulator CMG-IMEX (CMG, 2012) to model hydraulic
fractures around the wellbore and simulate gas production, as
illustrated in Fig. 2(a). Note the variation in fracture half-length
in this well. History matching of the led data is presented in
Fig. 2(b). It shows a reasonable match between numerical simulation results and actual eld gas ow data.
We set up another shale gas reservoir model with a volume of
990 ft  1980 ft  50 ft, based on average reservoir data from the
Barnett Shale (Table 1). The effect of the number of hydraulic
fractures on cumulative gas production with a base horizontal well
length (660 ft) is presented in Fig. 3. It shows that cumulative gas
production increases linearly at early times of production and then
slows down at later times, nally reaching a plateau. The initial
cumulative production increase is more dramatic in cases with
a higher number of fractures (smaller fracture spacing). Fig. 4
shows the impact of the number of fractures on cumulative gas
production from the base well (660 ft) after 10 years of production.

Table 1
Basic reservoir information.
Parameter

Barnett Shale case

Synthetic case

Unit

Model dimensions (L  W  H)
Initial reservoir pressure
Bottom-hole pressure
Production period
Reservoir temperature
Gas viscosity
Top of reservoir
Initial gas saturation
Compressibility of shale
Fracture height
Fracture conductivity
Matrix permeability
Matrix porosity
Horizontal wellbore length

3500  5000  400 (1066.8  1524  121.9)

3800 (2.62  107)
1500 (1.03  107)
3 (9.47  107)
180 (82)
0.02 (0.00002)
7000 (2133.6)
0.70
3.0  10  6 (4.35  10  10)
400 (121.9)
9 (2.7  10  15)
0.00035 (3.5  10  19)
0.04
2052 (625.4)

990  1980  50 (301.8  603.5  15.2)

3776 (3.60  107)
1000 (6.89  106)
10 (3.16  108)
180 (82)
0.02 (0.00002)
6956 (2120.2)
0.80
10  6 (1.45  10  10)
50 (15.2)
100 (3  10  14)
0.0001(9.9  10  20)
0.08
600 (182.9)

ft (m)
psi (Pa)
psi (Pa)
year (s)
o
F (oC)
cP (Pa s)
ft (m)
fraction
psi  1 (Pa  1)
ft (m)
md-ft (m2-m)
md (m2)
fraction
ft (m)

W. Yu et al. / Journal of Petroleum Science and Engineering 113 (2014) 17

Fig. 2. (a) Four-stage fracture in Barnett Shale and (b) actual versus numerical data of gas ow rate.

Fig. 3. Effect of number of fractures on cumulative gas production in base well.

Fig. 5. Effect of number of fractures on gas ow rate in base well (660 ft).

Fig. 4. Effect of number of fractures on cumulative gas production in base well after
10 years of production.

Fig. 6. Effect of fracture spacing on cumulative gas production in base well with
equal number of fractures.

Below a critical value (  5 fractures), cumulative gas production

increases linearly, and the increasing trend reaches a plateau
beyond the critical value. The change in the cumulative gas
production trend at later times is the result of fracture interference
effects. Fig. 5 shows the effect of the number of fractures on gas
ow rate in the base well (660 ft) for a period of 10 years of
production. Higher number of fractures results in higher production rates. However, as a result of fracture interference effects,
production decline might be more dramatic in cases with a higher
number of fractures, e.g., in a case of 12 fractures in Fig. 5.

Therefore, optimization of the number of fractures is always

necessary.
We also studied the effect of fracture spacing on cumulative gas
production for the case of two fractures in the base well, as shown
in Fig. 6. Cumulative gas production remains constant for fracture spacing larger than 200 ft and decreases with a decrease in
fracture spacing (o200 ft). This observation also implies that,
when the fracture spacing is below a critical distance (which
depends on reservoir characteristics), the interference effect
between fractures becomes important. Therefore, optimal fracture

W. Yu et al. / Journal of Petroleum Science and Engineering 113 (2014) 17

spacing should be selected for practical design of fracture treatment (e.g., the optimal distance of 200 ft in Fig. 6).
Fig. 7 presents the effect of hydraulic fracture half-length on
cumulative gas production in the base well with constant fracture
spacing of 40 ft. The gure shows that cumulative gas production
increases with an increase in fracture half-length, although strong
interference might occur between fractures. Fig. 8 shows that gas
ow rate also increases with increasing fracture half-length. This
correspondence suggests that longer fractures produce morean
inference that is in agreement with the results of Frantz et al.
(2005).

4. Uncertainty in fracture geometry

Fig. 7. Effect of fracture half-length on cumulative gas production in the base

model with 40 ft of fracture spacing.

Because uncertainty exists in hydraulic fracture geometry, we

studied ve different outer-to-inner fracture half-length combinations, as illustrated in Fig. 9. Inner and outer fractures are marked
in this gure.

Fig. 8. Effect of fracture half-length on gas ow rate in base well with 40 ft of

fracture spacing.

Fig. 10. Cumulative gas production of ve outerinner fracture-combination cases

described in Fig. 9.

Case 1: Equal length but short

fractures

Case 2: Inner fractures longer

than outer fractures

Case 4: Outer fractures longer

than inner fractures

Case 5: Equal length but long

fractures

Case 3: Mixed fracture lengths

Fig. 9. Schematic diagrams of ve different innerouter combinations of quadruple transverse fractures in horizontal well.

W. Yu et al. / Journal of Petroleum Science and Engineering 113 (2014) 17

Case 1. Fractures with equal and short fracture half-length of

250 ft.
Case 2. Outer fractures with a half-length of 250 ft, and inner
fractures with a length of 550 ft.

Case 3. Outer fractures with half-lengths of 250 ft and 550 ft, and
inner fractures with lengths of 500 ft and 1100 ft.
Case 4. Outer fractures with a half-length of 550 ft, and inner
fractures with a length of 250 ft.
Case 5. Fractures with equal and long half-length of 550 ft.

Fig. 11. Comparison of gas velocities in outer and inner fractures of Cases 2 and 4.

Fig. 10 compares well performance of the fracture layouts illustrated in Fig. 9. Among these ve cases presented in Fig. 10, Cases
1 and 5 exhibit the lowest and the highest cumulative gas production,
respectively. This outcome is simply because Cases 1 and 5 have the
lowest and the highest total fracture half-length. Interestingly, cumulative gas production gures, particularly at later times, are different in
Cases 24, although the total fracture half-length is the same in these
three scenarios. The cumulative gas production in Cases 3 and 4 is
larger than that of Case 2; in addition, Case 4 performs slightly better
than Case 3. The performance of Case 2 is between that of Case 1 and
Case 5. The reason for the difference between Cases 4 and 2 is the
stronger interference effect between longer inner fractures in Case 2
versus an absence of the interference effect between longer outer
fractures in Case 4. Interestingly, the fracture spacing of the outer

Case 1

Case 2

Case 4

Case 5
Fig. 12. Pressure distribution after 10 years of production for ve cases.

Case 3

W. Yu et al. / Journal of Petroleum Science and Engineering 113 (2014) 17

Fig. 13. Effects of outer-to-inner fracture half-length aspect ratio on cumulative gas
production.

fractures in Case 4 is 240 ft (greater than 200 ft, the critical fracture
spacing). Because the spacing of longer fractures in Case 3 is 160 ft
(less than 200 ft, the critical fracture spacing), a slight interference
between fractures exists, leading to slightly lower cumulative gas
production in Case 3. It should be noted that fracture interference also
exists in the process of fracture propagation because of a stress
shadow effect, resulting in immature fracture growth and fracture
width restriction (Wu and Olson, 2013; Wu, 2013); however, this
factor is not considered in this paper. In addition, Fig. 11 shows that the
difference of gas ow velocity between outer fractures and inner
fractures decreases rapidly with time for Case 2, whereas, in Case 4,
gas ow velocity in the outer and inner fractures continue to be
different for a longer period of time. Fig. 11 also shows that gas ow
velocity of outer fractures in Case 4 is larger than that of inner
fractures and of the other fractures in Case 2. For Case 2, the gas ow
velocities of both outer and inner fractures reach the same value after
6 years of production, and approach the gas ow velocity in the inner
fractures of Case 4. This observation means that contributions of the
longer outer fractures to cumulative gas production and gas ow rate
are higher than the contributions of the inner fractures, owing to
strong fracture interference between inner fractures. This pattern may
suggest longer outer fractures result in improved gas production in
practice.
Fig. 12 shows pressure distribution of the ve cases after 10 years of
production. Clearly, Case 1 has the smallest disturbed volume and Case
5 has the highest disturbed volume. The disturbed volumes in Cases
24, with the same total fracture half-length, are between Cases 1 and
5. The disturbed volume in Case 4 is more than that in Cases 2 and 3.
The disturbed volume in Case 4 is almost as high as that in Case 5. This
result suggests the possibility of equal disturbed volume with less total
fracture half-length (cf. Cases 4 and 5 in Fig. 12).
Fig. 13 shows the effect of the aspect ratio of outer fractures to
inner fractures on cumulative gas production in Case 4. The gure
shows that cumulative gas production increases with the increasing aspect ratio of outer fractures to inner fractures. Note that total
fracture half-length is equal in all these scenarios (1600 ft). The
improvement in cumulative gas production in the scenario of
longer outer fractures is the result of the lack of interference effect
between the outer fractures.

5. Conclusions
Using numerical simulation, we performed a sensitivity study
of gas production for a shale gas well with uncertainty hydraulic
fracture geometry. We rst veried our numerical simulation

method using eld data from a horizontal well with multistage

hydraulic fracturing in the Barnett Shale formation. We then
suggested distinct outer and inner hydraulic fractures and presented a detailed study of ve different combinations of outer and
inner fractures. We also performed sensitivity analyses on the
effects of fracture half-length and fracture spacing on cumulative
gas production and gas ow rate. The following conclusions can be
obtained from the results of this simulation study: (1) reduction of
fracture spacing increases cumulative gas production, until the
point at which fractures start to interfere; (2) cumulative gas
production increases linearly with an increase in fracture halflength; (3) the contributions of outer fractures to cumulative gas
production and gas ow rate are higher than the contributions of
inner fractures; and (4) depending on the characteristics of a shale
reservoir, a scheme of outer and inner fracture scheme can be
potentially designed to optimize gas production rate, cumulative
gas production, and cost associated with the total induced halflength in a multistage fracturing job.

Acknowledgments
The authors thank Computer modeling Group (CMG) for usage of
software, and this work was supported partly by the Reservoir
Simulation Joint Industry Project at the Center for Petroleum and
Geosystems Engineering and the NanoGeosciences Lab at The University of Texas at Austin. Chris Parker edited the manuscript.
Publication authorized by the Director, Bureau of Economic Geology.
References
Ambrose, R.J., Clarkson, C.R., Youngblood, J., Adams, R., Nguyen, P., Nobakht, M.,
Biseda, B., 2011. Life-cycle decline curve estimation for tight/shale gas reservoirs. Paper SPE 140519. In: Proceedings of the SPE Hydraulic Fracturing
Technology Conference and Exhibition, The Woodlands, TX, January 2426.
Castaneda, J.C., Castro, L., Craig, S., Moore, C., Myatt, J., 2010. Coiled tubing
fracturing: an operational review of a 43-stage Barnett shale stimulation. Paper
SPE 130678. In: Proceedings of the SPE/ICoTA Coiled Tubing and Well Intervention Conference and Exhibition, The Woodlands, TX, March 2324.
Cipolla, C.L., Lolon, E.P., Erdle, J.C., Rubin, B., 2010. Reservoir modeling in shale-gas
reservoirs. SPE Res. Eval. Eng. 13 (4), 638653.
CMG: IMEX User0 s Guide, Computer Modeling Group Ltd., 2012.
Evans, R.D., Civan, F., 1994. Characterization of Non-Darcy Multiphase Flow in
Petroleum Bearing Formations. Report, U.S. DOE Contract no. DE-AC2290BC14659, School of Petroleum and Geological Engineering, University of
Oklahoma.
Fazelipour, W., 2011. Development of techniques to integrate hydraulic fracturing
design and reservoir simulation technologies application to forecast production of stimulated wells in unconventional gas reservoirs. Paper SPE 142337. In:
Proceedings of the SPE Middle East Unconventional Gas Conference and
Exhibition, Muscat, Oman, January 31February 2.
Fisher, M.K., Heinze, J.R., Harris, C.D., Davidson, B.M., Wright, C.A., Dunn, K.P., 2004.
Optimizing horizontal completion techniques in the Barnett Shale using
microseismic fracture mapping. Paper SPE 90051. In: Proceedings of the SPE
Annual Technical Conference and Exhibition, Houston, TX, September 2629.
Frantz, J.H., Williamson, J.R.Jr , Sawyer, W.K., Johnston, D., Waters, G., Moore, L.P.,
MacDonald, R.J., Pearcy, M., Ganpule, S.V., March, K.S., 2005. Evaluating Barnett
Shale production performance using an integrated approach. Paper SPE 96917.
In: Proceedings of the SPE Annual Technical Conference and Exhibition, Dallas,
TX, October 912.
Grieser, B., Shelley, B., Soliman, M., 2009. Predicting production outcome from
multi-stage, horizontal Barnett completions. Paper SPE 120271. In: SPE Production and Operations Symposium, Oklahoma, OK, April 48.
Javadpour, F., 2009. Nanopores and apparent permeability of gas ow in mudrocks
(shales and siltstone). J. Can. Pet. Technol. 48, 1621.
Javadpour, F., Fisher, D., Unsworth, M., 2007. Nano-scale gas ow in shale
sediments. J. Can. Pet. Technol. 46, 5561.
Kaiser, M.J., 2012. Haynesville shale play economic analysis. J. Pet. Sci. Eng. 8283,
7589.
Kalantari-Dahaghi, A., 2011. Systematic approach to numerical simulation and
modeling of shale gas reservoirs. Int. J. Oil Gas Coal Technol. 4 (3), 209243.
King, G.E., 2010. Thirty years of gas shale fracturing: what have we learned? Paper
SPE 133456. In: Proceedings of the SPE Annual Technical Conference and
Exhibition, Florence, Italy, September 1922.
Mayerhofer, M.J., Lolon, E.P., Warpinski, N.R., Cipolla, C.L., Walser, D., Rightmire, C.
M., 2010. What is stimulated reservoir volume? SPE Prod. Oper. 25, 8998.

W. Yu et al. / Journal of Petroleum Science and Engineering 113 (2014) 17

Meyer, B.R., Bazan, L.W., Jacot, R.H., Lattibeaudiere, M.G., 2010. Optimization of
multiple transverse hydraulic fractures in horizontal wellbores. Paper SPE
131732. In: Proceedings of the SPE Unconventional Gas Conference, Pittsburgh,
PA, February 2325.
Roy, A., Jayaram, V., Marfurt, K., 2013. Distance metric based multi-attribute seismic
facies classication to identify sweet spots within the Barnett Shale: a case
study from the Fort Worth basin, TX. Paper SPE 168926. In: Proceedings of the
Unconventional Resources Technology Conference, Denver, CO, August 1214.
Rubin, B., 2010. Accurate simulation of non-Darcy ow in stimulated fractured shale
reservoirs. Paper SPE 132093. In: SPE Western Regional Meeting, Anaheim, CA,
May 2729.
Soliman, M.Y., Kabir, C.S., 2012. Testing unconventional formations. J. Pet. Sci. Eng.
9293, 102109.
Warpinski, N.R., Mayerhofer, M.J., Vincent, M.C., Cipolla, C.L., Lolon, E.P., 2009.
Stimulating unconventional reservoirs: maximizing network growth while
optimizing fracture conductivity. J. Can. Pet. Technol. 48 (10), 3951.

Waters, G., Dean, B., Downie, R., Kerrihard, K., Austbo, L., McPherson, B., 2009.
Simultaneous hydraulic fracturing of adjacent horizontal wells in the Woodford
Shale. Paper SPE 119635, SPE Hydraulic Fracturing Technology Conference, The
Woodlands, TX, January 1921.
Wu, K., Olson, J.E., 2013. Investigation of critical in situ and injection factors in
multi-frac treatments: guidelines for controlling fracture complexity. Paper SPE
163821. In: Proceedings of the SPE Hydraulic Fracturing Technology Conference,
The Woodlands, TX, February 46.
Wu, K., 2013. Simultaneous multi-frac treatments: fully coupled uid ow and
fracture mechanics for horizontal wells. Paper SPE 167626. In: Proceedings of
the SPE Annual Technical Conference and Exhibition, New Orleans, LA,
September 30October 2.
Zhao, Y., Zhang, L., Wu, F., 2012. Pressure transient analysis for multi-fractured
horizontal well in shale gas reservoirs. J. Pet. Sci. Eng. 9091, 3138.