IPTC 13338

Sensitivity Studies of Horizontal Wells with Hydraulic Fractures in Shale

Gas Reservoirs
X. Zhang, SPE, C. Du, SPE, F. Deimbacher, SPE, M. Crick, SPE, and A. Harikesavanallur, SPE, Schlumberger

Copyright 2009, International Petroleum Technology Conference

This paper was prepared for presentation at the International Petroleum Technology Conference held in Doha, Qatar, 7–9 December 2009.

This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily
reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society
Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum Technology
Conference 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 where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435.

Abstract

Producing natural gas from shale gas reservoirs has gained momentum over the past few years in North America and will
become an increasingly important component of the world’s energy supply. A shale gas reservoir is characterized as an
organic-rich deposition with extremely low matrix permeability and clusters of mineral-filled “natural” fractures. Shale gas
storage capacity is defined by the adsorbed gas on the organic material within the shale matrix and free gas in the limited
pore space of the shale rocks. Horizontal drilling and hydraulic fracturing are the primary enabling technologies to obtain
economical production from the shale gas reservoir.

This paper presents a comprehensive reservoir simulation model to study the impact of reservoir and hydraulic fracturing
parameters on production performance of a shale gas reservoir. The simulation model was constructed as a multi porosity
system with matrix sub-grids to account for transient gas flow from the matrix to the fracture. The extended Langmuir
isotherm was used to model the desorption process of multiple components during the production. Primary hydraulic
fractures perpendicular to the horizontal wellbore were modeled explicitly with thin grid cells that preserved the finite
conductivity. The hydraulically-induced fracture network around the horizontal well was characterized by the matrix-fracture
coupling factor (sigma) and permeability of the fractures.

The study was aimed to quantify the influence of the reservoir and hydraulic fracture parameters using experimental design,
including porosity and permeability of the reservoir matrix and fracture, matrix-fracture sigma factor, matrix subdivisions
and, primary hydraulic fracture half-length, height, spacing and conductivity, rock compaction, non-Darcy flow coefficient,
as well as gas content. Sensitivity tests were performed to identify the most influential reservoir and hydraulic fracture
parameters and provided important insights into the impact of uncertainties on shale gas production forecasts, which can be
critical for fracture treatment design and production scheme optimization.

Introduction

Fundamentals of unconventional gas reservoirs have been reviewed1 and a multidisciplinary workflow was established for
geological characterization of shale gas reservoirs.2 Studies have been performed to detect natural fractures using 3D-seimic
data and estimate natural fracture patterns in gas shales using production data.3,4 To enable production from a shale gas
reservoir, the key is to stimulate the existing natural fractures or rock fabric by hydraulic-fracture treatments that pump large-
volume, high-rate water with different size of proppents.5 The application of microseismic mapping has revealed
complexities of the created fracture networks and improved our understanding of hydraulic-fracture growth in the gas shale
formation. A semi-analytical pseudo 3-D geomechanical model has been developed to characterize the induced fracture
network.6

A shale gas reservoir with shale rock and natural fractures can be typically modeled as a dual-porosity system, proposed by
Warren and Root.7 The key aspect of this approach is to separate the flow through the fractures from the flow inside the
matrix. The reservoir model is represented by two overlapping continua – fracture networks acting as main flow channels and
matrix blocks serving as a major storage source. The interaction between the two continua is controlled through a transfer
2 IPTC 13338

function, also called shape or coupling factor that can be evaluated with typical dimensions of the matrix blocks or distances
between fractures.8

Fracture

Matrix

Actual Reservoir Model Reservoir

Fig. 1: Dual-porosity idealization (Warren and Root, 1963)

Traditional dual porosity models assume that the matrix to fracture flow is in pseudo steady state. However, shale gas
reservoirs often have a matrix porosity system in which the transient behavior in the matrix becomes important. A discretized
matrix model is available to sub-divide each matrix cell into a series of nested sub-cells, allowing the simulator to predict the
transient behavior.9

The geological complexity and connectivity of naturally fractured reservoirs can be characterized using a discrete fracture
network (DFN) model.10 The stochastic approach of the DFN model enables multiple realizations for a quantitative measure
of uncertainty and validation. Effective techniques have been developed to upscale fracture permeability, porosity, and
matrix-fracture sigma factor from the DFN model to the dual-porosity simulation model, taking into account the complex
fracture system geometry and connectivity.

DP Continuum
DFN
Matrix System

Fracture System

Porosity
Dir. Perm.
σ-factor

Fig. 2: Upscaling of properties from the DFN model to dual porosity system (Golder Associates)

Construction of the discrete fracture network (DFN) models can be constrained with microseismic events generated in the
hydraulic-fracture treatment.11 Fracture locations were interpreted by using event amplitude as an intensity measure, and
fracture sizes were estimated by the relationship between the amplitude and the seismic moment. The resulting geocellular
model would provide a quantitative framework to improve the history-matching process.

The performance of Barnett Shale production from vertical and horizontal wells was evaluated with an integrated approach
involving reservoir characterization (G&G, petrophysics, rock mechanics, etc.), completion/stimulation design, and reservoir
simulation.12,13 An integration methodology and workflow was developed for the Barnett Shale gas reservoir characterization
and simulation.13 The workflow incorporates seismic interpretations and attributes, borehole image and log interpretations,
core analysis, hydraulic fracturing treatments, and microseimic data to construct reservoir models and discrete fracture
networks that are then upscaled to dual-porosity reservoir models for numerical simulation.

Sensitivity studies

A base simulation model was constructed to evaluate sensitivities of reservoir and hydraulic fracture parameters for a
horizontal well with multiple hydraulic fractures in a shale gas reservoir. The model dimensions are 4000 ft in length, 2000 ft
in width, and 300 ft in height. Ten transverse hydraulic fractures with 300 ft spacing are placed along a 3000 ft horizontal
IPTC 13338 3

wellbore. The hydraulic fractures are explicitly represented by grid cells with a width of 0.01 ft. The top of the reservoir is
assumed to be 7000ft.

Matrix subdivision
The model is a dual porosity system with shale matrix and natural fracture. To account for transient flow of free gas from the
matrix to the fracture, the Multi-Porosity option in ECLIPSE14 was applied to divide each matrix cell into a series of nested
sub-cells that are logarithmic away from the fracture face. Matrix sub-grids forms a series of separate one-dimensional flow
systems connected to corresponding fracture cells. The gas in the matrix pore space migrates into fracture channels, driven by
an unsteady-state pressure gradient. Fig. 3 shows the pressure profile vs. time in nested matrix sub-cells divided spherically,
called “Russian doll.” From Fig. 4, it can be seen that the transient flow facilitated by matrix subdivision has a significant
impact on early gas production.

3000

Pressure (psi)
d
2000
T = 0 YR
T = 1 YR
Matrix T = 2 YRS
1000
T = 5 YRS
T = 10 YRS
T = 20 YRS
0
0 10 20 30 40 50
d (ft)

Matrix Cell Center Fracture Face

Fig. 3: Pressure distribution in matrix sub-cells

6000

Matrix Subdivision = 1
Gas Production Rate

Matrix Subdivision = 4
4000
(Mscf/day)

2000

0
0 5 10 15 20
Year

Fig. 4: Impact of transient flow due to matrix subdivision on production rate

Reservoir parameters
The ultra-low permeability of the shale matrix typically ranges from 10 to 100 nano-Darcies (or 0.00001 to 0.0001 md). The
shale matrix rock has porosity of 2-8%. The hydraulic fracturing process “opens” the calcite-filled natural fractures and
“breaks” the matrix rock. As a result, a vast, complex fracture network, known as the ESV (estimated stimulated volume), is
4 IPTC 13338

formed around the wellbore to effectively drain the gas. Dominate flow paths within the ESV can be approximated as
primary hydraulic fractures and hydraulically induced fractures.

A sensitivity study was performed on stimulated fracture network permeability ranging from 0.0001 to 0.001 md, and matrix-
fracture sigma factor ranging from 12 to 1.2E-5, corresponding to matrix block sizes (or distance between fractures) from 1
to 1000 ft. The variation in the porosity of the stimulated fracture network is assumed to be 0.2-0.8%. To account for the
transient behavior of the flow from the matrix to the fracture, each matrix grid block was divided into 1 to 5 sub-cells. Fig. 5
shows the impact of the reservoir parameters on the production performance. The hydraulic fracture parameters such as
spacing (300 ft), conductivity (5 md-ft), height (200 ft), and half-length (300 ft) were fixed in this sensitivity study.

Stimulated Fracture Network

Matrix-Fracture Sigma
Permeability
8000 8000

Kf = 0.0001 md
Production (MMSCF)

Sigma = 1.2E-5

Production (MMSCF)
Kf = 0.0005 md
Cumulative Gas

6000 Sigma = 1.33E-4

Cumulative Gas
6000
Kf = 0.001 md Sigma = 12

4000 4000

2000 2000

0 0
0 5 10 15 20 0 5 10 15 20
Year Year

(a) (b)

Stimulated fracture
Matrix Subdivision 36%
network permeability
8000
Matrix-fracture sigma 33%
Production (MMSCF)

Subdivision = 1
Cumulative Gas

6000 Subdivision = 3 Matrix porosity 10%

Subdivision = 5

4000 Matrix permeability 7%

Fracture porosity 7%
2000
Matrix sub-division 6%

0
0% 20% 40%
0 5 10 15 20
Year Influence of Reservoir Parameters

(c) (d)

Fig. 5: Impact of reservoir parameters on cumulative gas production. (a) Impact of stimulated fracture network permeability;
(b) Impact of matrix-fracture sigma; (c) Impact of matrix subdivision; (d) Influence of all reservoir parameters

Fig. 5 demonstrates that the most influential reservoir parameters were the stimulated fracture network permeability and
matrix-fracture sigma factor. From a standpoint of hydraulic-fracture treatments in the shale gas reservoirs, it is
understandable that the stimulation processes can enhance fracture flow capability by activating existing natural fractures and
increase contact areas by cracking shale rocks. It is expected there will be an increase in stimulated fracture network
permeability and matrix-fracture sigma factor within the ESV constrained by microseismic activities. This implies that
reservoir characterization is very important in the evaluation of shale gas production. With integration of the DFN model,
hydraulic-fracture treatment data, and microseismic events, reservoir properties within the ESV can be better characterized
and upscaled for reservoir simulation. Fig. 6 shows that an enhancement in stimulated fracture network permeability can
result in a larger drainage area around the well.
IPTC 13338 5

Stimulated Fracture Network Stimulated Fracture Network

Permeability = 1.0E-4 md Permeability = 1.0E-3 md

Fracture Pressure (Psia) After 20 Years

Fig. 6: Impact of stimulated fracture network permeability on the drainage area

Hydraulic fracture parameters

In this study, it was assumed that a planar, primary fracture with a width of 0.01 ft is hydraulically initiated from the
perforation cluster in each stage. The conductivity of the primary fracture ranges from 1 to 50 md-ft. The fracture spacing can
vary from 600 to 200 ft, corresponding to 5 to 15 fractures. The fracture height is limited from 100 to 300 ft. The fracture
half-length can range from 100 to 500 ft.

A sensitivity study of hydraulic fracture parameters has been performed with fixed reservoir properties such as stimulated
fracture network permeability (5E-4 md) and porosity (0.6%), matrix permeability (5E-5 md) and porosity (6%), matrix-
fracture sigma factor (1.33E-4), and matrix subdivision (3). The impact of hydraulic fracture parameters on the cumulative
gas production is illustrated in Fig. 7. The impact of hydraulic fracture half-length, spacing, and conductivity on the drainage
area around the horizontal well is shown Fig. 8.

Hydraulic Fracture Half-Length Hydraulic Fracture Spacing Hydraulic Fracture Height

8000 8000 8000
Production (MMSCF)

Xf = 100 ft
Production (MMSCF)
Production (MMSCF)

Spacing = 600 ft Height = 100 ft

Cumulative Gas

6000 Xf = 300 ft 6000

Cumulative Gas

6000 Height = 200 ft

Cumulative Gas

Spacing = 300 ft
Xf = 500 ft Spacing = 200 ft Height = 300 ft

4000 4000 4000

2000 2000 2000

0 0 0
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20
Year Year Year

(a) (b) (c)

Hydraulic Fracture Conductivity Fracture half-

33%
8000 length
Production (MMSCF)

Conductivity = 1 md-ft 29%

Fracture spacing
Cumulative Gas

6000 Conductivity = 5 md-ft

Conductivity = 50 md-ft

4000 Fracture height 23%

2000 Fracture
16%
conductivity

0
0% 20% 40%
0 5 10 15 20
Year Influence of Hydraulic Fracture Parameters

(d) (e)

Fig. 7: Impact of hydraulic fracture parameters on cumulative gas production. (a) Impact of hydraulic fracture half-length; (b)
Impact of hydraulic fracture spacing; (c) Impact of hydraulic fracture height; (d) Impact of hydraulic fracture conductivity;
(e) Influence of all hydraulic fracture parameters.
6 IPTC 13338

Hydraulic Fracture Half-Length = 100 ft Hydraulic Fracture Half-Length = 500 ft

(a)
Hydraulic Fracture Spacing = 600 ft Hydraulic Fracture Spacing = 200 ft

(b)
Hydraulic Fracture Conductivity = 1 md-ft Hydraulic Fracture Conductivity = 50 md-ft

(c)
Fracture Pressure (Psia) After 20 Years

Fig. 8: Impact of hydraulic fracture parameters on the drainage area. (a) Impact of hydraulic fracture half-length; (b) Impact
of hydraulic fracture spacing; (d) Impact of hydraulic fracture conductivity

Other parameters
In this section, the impact of rock compaction, non-Darcy flow, and gas content on the shale gas production is investigated.
Considerable compaction effects are usually seen in propped hydraulic fractures. Effects of rock compaction are tested using
rock compaction tables with pore volume and transmissibility multipliers. The pore volume multipliers are computed from
the rock compressibility (Cr) ranging from 1E-6 to 1E-4 1/psi. The transmissibility multipliers are assumed to be a cubic
power of the pore volume multipliers. The turbulent flow due to high gas flow rates in and around fractures is modeled with
the Non-Darcy flow coefficient varying from 0 to 20. The gas adsorption is measured by gas content that is a volume
constant in the Langmuir isotherm. The gas contents range from 0 to 350 SCF/ton in the model. All reservoir and hydraulic
fracture parameters are fixed to investigate sensitivities of rock compaction, non-Darcy flow, and gas content. The results are
shown in Fig. 9. It is expected that significant rock compaction and non-darcy flow behavior occur in and around hydraulic
fractures, illustrated in Fig. 10.
IPTC 13338 7

Rock Compaction Non-Darcy Flow

6000 6000
Cr = 1E-4 1/psi Non-Darcy Coeff = 20

Production (MMSCF)
Cr = 1E-5 1/psi Non-Darcy Coeff = 5

Production (MMSCF)
Cumulative Gas
Cr = 1E-6 1/psi Non-Darcy Coeff = 0

Cumulative Gas
4000 4000

2000 2000

0 0
0 5 10 15 20 0 5 10 15 20
Year Year

(a) (b)

Gas Content
6000 Rock
53%
Gas Content = 0 scf/ton compaction
Gas Content = 100 scf/ton
Production (MMSCF)

Gas Content = 350 scf/ton

Cumulative Gas

4000
Non-Darcy flow 32%

2000
Gas content 15%

0
0% 20% 40% 60%
0 5 10 15 20
Year Influence of Other Parameters

(c) (d)

Fig. 9: Impact of other parameters on cumulative gas production. (a) Impact of rock compaction; (b) Impact of non-Darcy
flow; (c) Impact of gas content; (d) Influence of other parameters

Transmissibility Reduction after 2 Years Forchheimer Non-Darcy Flow Factor after 2 Years

(a) (b)
Fig. 10: Impact of rock compaction and non-Darcy flow. (a) Significant transmissibility reduction around fractures; (b)
Deviation of the flow behavior from Darcy’s law in fractures.

All parameters
With variation of all parameters within their individual data ranges, linear experiments of Plackett-Burman type have been
designed to screen influential parameters including reservoir, hydraulic fracture, rock compaction, non-darcy flow, as well as
gas content. The results are demonstrated in Fig. 11. The matrix-fracture sigma factor and stimulated fracture network
permeability are identified as the most influential parameters. Primary hydraulic fracture spacing, conductivity, and half-
length are also important. The rock compaction around the propped hydraulic fractures has a significant, negative impact on
the gas production.
8 IPTC 13338

Matrix-fracture sigma 15.28%

Stimulated fracture network permeability 14.68%

Hydraulic fracture spacing 11.48%

Hydraulic fracture conductivity 11.25%

Rock compaction 10.73%

Hydraulic fracture half-length 8.18%

Natural fracture porosity 6.60%

Matrix sub-division 6.55%

Matrix permeability 4.97%

Matrix porosity 3.41%

Hydraulic fracture height 2.44%

Non-Darcy flow 2.33%

Gas content 2.10%

0% 10% 20%

Influence of All Parameters

Fig. 11: Impact of all parameters on cumulative gas production

Summary

The matrix-fracture sigma factor and stimulated fracture network permeability are the most influential parameters. The
discrete fracture network (DFN) model conditioned with hydraulic-fracture treatment data and microseismic events should be
able to characterize fracture intensity, width, and permeability, and/or fracture conductivity within the ESV. The updated
DFN model can be upscaled to dual-porosity system properties such as the matrix-fracture sigma factor and stimulated
fracture network permeability. The resulting reservoir model can improve the history matching process and provide reliable
forecasts of the shale gas production performance.

Hydraulic fracture conductivity, spacing, half-length are all important factors. The primary hydraulic fractures initiated from
the perforation clusters in multiple stages likely receive most of the proppants and possess high conductivity. The highly
conductive primary fractures have significant impacts on the gas production. The optimal hydraulic fracture spacing is
essential for achieving high recovery factor and minimizing costs. The hydraulic fracture length and stimulated region size
determine the well drainage area, which is a crucial factor to estimate well spacing.

Rock compaction in propped hydraulic fractures, especially in geopressured environment, is an important factor. As the
pressure declines, the fractures start to “close” due to compaction, proppant embedment, proppant crushing, and fine
migration, etc. The rock compaction has a negative impact on the gas production.

The matrix subdivision can describe the transient behavior of gas flow from the matrix pore space to the fracture channels.
The transient behavior dominates during early decline of the gas production.
IPTC 13338 9

Acknowledgement

The authors wish to thank Schlumberger for permission to publish this paper. Thanks are extended to Walt Sawyer and Li
Fan for reviewing the manuscript and valuable suggestions.

References

1. Jenkins, C.D. and Boyer, C.M.: "Coalbed- and Shale-Gas Reservoirs," paper SPE 103514, Distinguished Author
Series, JPT, February 2008.

2. Slatt, R.M., Singh, P., Philp, R.P., Marfurt, K.J., Abousleiman, Y., and O’Brien, N.R.: ” Workflow for Stratigraphic
Characterization of Unconventional Gas Shales,” paper SPE 119891, presented at the 2008 SPE Shale Gas
Production Conference, Dallas, Texas, USA, 16-18 November 2008.

3. Gray, F.D.: “Fracture Detection in Unconventional Gas Plays Using 3D-Seismic Data,” paper SPE 114902,
presented at the CIPC/SPE Gas Technology Symposium, Calgary, Alberta, Canada, 16-19 June 2008.

4. Gaskari, R. and Mohaghegh, S.D.: “Estimating major and Minor Natural Fracture Pattern in Gas Shales Using
Production Data,” paper SPE 104554, presented at the 2006 SPE Eastern Regional Meeting, Canton, Ohio, USA,
11-13 October, 2006.

5. Cipolla, C.L.: “Modeling Production and Evaluating Fracture Performance in Unconventional Gas Reservoir.” Paper
SPE 118536, Distinguished Author Series, JPT, September 2009.

6. Xu, W., Calvez, J.L., and Thiercelin, M.: “Characterization of Hydraulically-Induced Fracture Network Using
Treatment and Microseismic Data in a Tight-Gas Formation: A Geomechanical Approach,” paper SPE 125237,
presented at the 2009 SPE Tight Gas Completions Conference, San Antonio, Texas, USA, 15–17 June 2009.

7. Warren, J.E. and Root, P.J.: “The Behavior of Naturally Fractured Reservoirs,” paper SPE 426, SPEJ, September
1963.

8. Kazemi, H., Merril, L.S. Jr., Porterfield, K.L., and Zeman, P.R.: “Numerical Simulation of Water-Oil Flow in
Naturally Fractured Reservoirs,” paper SPE 5719, SPEJ, December 1976.

9. Shale Gas, Chapter 59, ECLIPSE Technical Description manual 2009.1.

10. Dershowitz, B., Lapointe, P., Eiben, T., and Wei, L.: “Integration of Discrete Feature Network Methods with
Conventional Simulator Approaches,” paper SPE 62498, SPE Reservoir Eval. & Eng. 3 (2), April 2000.

11. Williams-Stroud, S.: "Using Microseismic Events to Constrain Fracture Network Models and Implications for
Generating Fracture Flow Properties for Reservoir Simulation," paper SPE 119895, presented at the 2008 SPE Shale
Gas Production Conference, Dallas, Texas, USA, 16-18 November 2008.

12. Frantz, J.H. Jr., Williamson, J.R., Sawyer, W.K., Johnson, D., Waters, G., Moore, L.P., MacDonald, R.J., Peracy,
M., Ganpule, S.V., and March, K.S.: "Evaluating Barnnet Shale Production Performance Using an Integrated
Approach," paper SPE 96917, presented at the 2005 SPE Annual Technical Conference and Exhibition, Dallas,
Texas, USA, 9-12 October 2005.

13. Du, C., Zhang, X., Melton, B., Fullilove, D., Suliman, B., Grant, D. and Calvez, J.: "A Workflow for Integrated
Barnett Shale Gas Reservoir Modeling and Simulation," paper SPE 122934, presented at the 2009 SPE Latin
American & Caribbean Petroleum Engineering Conference, Cartagena, Colombia, 31 May -3 Jun 2009.

14. ECLIPSE Reservoir Simulator, Schlumberger Information Solutions.