SPE 168600

Poroelastic versus Poroplastic Modeling of Hydraulic Fracturing

HanYi Wang, University of Houston, Matteo Marongiu-Porcu, Economides Consultants Inc., Michael
J.Economides, University of Houston

Copyright 2014, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Hydraulic Fracturing Technology Conference held in The Woodlands, Texas, USA, 46 February 2014.
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
With the increasing wide use of hydraulic fractures in the petroleum industry it is essential to accurately predict the behavior
of fractures based on the understanding of fundamental mechanisms governing the process. The prevailing approach for
hydraulic fracture modeling relies on Linear Elastic Fracture Mechanics (LEFM). Generally, LEFM that uses stress intensity
factor at the fracture tip, gives reasonable predictions for hard rock hydraulic fracturing processes, but often fails to give
accurate predictions of fracture geometry and propagation pressure in soft/unconsolidated formations. The reasons are that the
fracture process zone ahead of the crack tip , elasto-plastic material behavior and strong coupling between flow and
stress cannot be neglected in these formations. Recent laboratory testing has revealed that in many cases fracture propagation
conditions cannot be described by traditional LEFM models. Rather, fractures develop in cohesive zones. In this study, we
developed a fully coupled poroelastic and poroplastic hydraulic fracturing model with cohesive zone method which is able to
model crack initiation and growth considering process zone effects. The impact of formation plastic properties on fracture
process is investigated for both short term and long term injection and the results are compared with elastic formation. In
addition, the main factors that affect the poroelastic backstress and the effects of formation plasticity on a hydraulic fracture
are also investigated. The results indicate that both poroelastic backstress and formation plasticity lead to higher net pressure,
but they have an opposite impact on fracture geometry. Poroelastic backstress tends to reduce fracture width while formation
plasticity tends to increase fracture width. It was also observed that plasticity can reduce closure pressure below minimum
horizontal stress. Poroelastic backstress is mostly controlled by the leak-off rate and formation permeability and the effects of
plasticity are mostly influenced by in-situ stress, plastic properties and pore pressure. We also found that the effective
toughness method which is used to quantify fracture process zone effect in soft formation tends to underestimate the fracture
width and overestimate fracture length even if it can match the net pressure. Ignoring plasticity in certain formations can lead
to inaccuracy of fracture geometry, net pressure prediction and fluid efficiency calculations. The deviations of real fracture
geometry from elastic formation model can have great impact on well production, especially in unconventional, lowpermeability reservoirs. So for a more accurate modeling of fracturing in plastic formation, the whole affected region around
the fracture should be considered, especially when the plastic deformation area is large.
Introduction
Hydraulic fracturing was rst applied in the petroleum industry in the 1940s to stimulate productivity in low permeability oilbearing formations. Since then, hydraulic fracturing has been widely used as a common practice to enhance the recovery of
hydrocarbons from low permeability reservoirs and prevent sand production in high permeability reservoirs (Economides and
Nolte, 2000). Recently, a signicant increase in shale gas production has resulted from hydraulic fracturing, which is used to
create extensive articial fractures around (generally cased and cemented) wellbores, for which cement integrity and effective
zonal isolation are crucial for long term productivity, especially if in high-pressure and high-temperature reservoirs (Shadravan
and Amani, 2012) . When combined with horizontal drilling, the hydraulic fracturing may allow formerly unpractical shale
layers to be commercially viable.
Development of theoretical models of hydraulic fracturing started a few decades ago and continues to be of a vivid practical
importance for the oil and gas industry as providing a basis for design and optimization tools (Adachi et al., 2007). As a
stimulation tool, the problem of hydraulic fracturing is in essence one of predicting the shape of the fracture as a function of
time, given the fluid pressure at the wellbore or the flow rate into the fracture. Even in its most basic form, hydraulic fracturing
is not a trivial process to model, as it involves the coupling of at least three physical processes: (i) the rock splitting and

SPE 168600

fracture propagation (ii) the rock deformation and (iii) the non-Newtonian fluid ow in the fracture. An assumption commonly
made is that the rock is linearly elastic and the loss of fluid into the rock is evaluated using a one-dimensional diffusion
equation. The mechanics of fracture opening and fluid loss are considered independent and their interaction is not considered,
such as the KGD model (Khristianovich and Zheltov, 1955; Geertsma and de Klerk, 1969) and PKN model (Perkins and Kern,
1961; Nordgren, 1972). Although classic hydraulic fracturing simulators based on LEFM and analytical or numerical
solutions are convenient to use and can give reasonable predictions for hard rock formation, there are still fundamental
questions on the adequacy of certain physical criteria (assumptions) used in the model and their effect on propagation regimes.
One such assumption is the adequacy of LEFM approach to the description of hydraulic fracture propagation in poorly
consolidated/soft rock. Numerous laboratory studies have shown that deformation of very weak non-cemented sands and
shales can even occur through elastic-visco-plastic constitutive behavior, and the deformations of these types of rocks cannot
be predicted by linear elasticity (Sone and Zoback, 2011). Research involving surveying of net pressures (difference between
the fracturing uid pressure and the far-eld conning stress) indicated that the net pressures encountered in soft formation are
on average 50100% higher than the predicted ones based on LEFM, net pressure is even higher in poorly consolidated
formations. These observations have triggered a series of dedicated studies which looked into the importance of the plastic
deformation in hydraulic fracturing (Papanastasiou, 1997; Germanovich, 1998; Papanastasiou, 1999; Van Dam, 2002). It was
found that plastic yielding near the tip of a propagating fracture provides an effective shielding resulting in an increase in the
effective rock fracture toughness, but these studies have ignored the pressure diffusion and porous behavior of the rock
deformation.
In examining the plentiful literature concerning fracture mechanics, the cohesive zone modeling has been adopted by many
authors to model fracture initiation and propagation. The cohesive zone is a region ahead of the crack tip that is characterized
by microcracking along the crack path. The main fracture is formed by interconnection of these micro cracks. The idea of a
cohesive zone developed from attempts to resolve the tip singularity contradiction predicted by LEFM. The early conceptual
works related to the cohesive zone model, were introduced by Barenblatt (1959; 1962) who proposed the cohesive zone model
to investigate perfectly brittle materials. Dugdale (1960) adopted a fracture process zone to investigate ductile materials
exhibiting small scale plasticity and proposed a critical opening condition as a fracture criterion. Mathematically, the Dugdale
(1960) model is similar to the Barenblatt model with the exception that different conditions were applied to the cohesive zone
parameters. Since then, the cohesive zone method has been extensively used with great success to simulate fracture and
fragmentation and processes in concrete, rock, ceramics, metals, polymers and composites. There are however, few studies
utilizing the cohesive zone method in hydraulic fracturing. Examples of some early studies in this area include the work of
Boone et al. (1986) and Boone and Ingraffea (1990) in which they used the cohesive zone approach to model the fracture
process in impermeable and permeable rocks.
In recent works, Schreer et al. (2006) proposed a model with a tip-velocity provided as part of the solution algorithm and
successfully propagated hydraulic fractures in an unknown path that may enucleate everywhere depending only on the stress
and pressure elds.
Dean and Schmidt (2009) developed a fully coupled geo-mechanical reservoir simulator that combined hydraulic fracture
growth, multiphase/multicomponent Darcy/non-Darcy ow in the porous medium, heat convection and conduction, solids
deposition, and poroelastic/poroplastic deformation in a single application. The program contained two separate criteria that
could be used to model fracture propagation: a critical fracture-opening criterion based on a stress-intensity factor and a
cohesive zone model that used quadrilateral cohesive elements in the fracture. The cohesive zone model includes a cohesive
strength and an energy release rate in the calculations at the tip of a propagating hydraulic fracture. Simulated fracture lengths
without leak-off are compared with analytical solution for both KGD and penny-shaped models, but poroelastic/poroplastic
formation behaviors are not investigated in this study.
Chen et al. (2009) established a model based on the existing pore pressure cohesive nite elements to investigate the
propagation of a penny-shaped hydraulic fracture in an innite elastic medium. The eect of cohesive material parameters and
uid viscosity on the hydraulic fracture behavior are investigated. The nite element results match well with the analytical
solutions where the fracture process is dominated by rock fracture toughness.
Zhang et al. (2010) developed a 3D finite element model of horizontal well fracturing using cohesive zone method. The
simulation results are validated by comparison with data from field measurements. The impact of in-situ stress contrast,
modulus contrast, tensile strength contrast and viscosity of fracturing uid on fracture geometry are investigated in this study.
Mokryakov (2011) proposed an analytical solution for hydraulic fracture with Barenblatts cohesive tip zone based on the
KGD model by assuming impermeable elastic rock without considering leak-off. It is also mentioned that derived (cohesive tip
zone model) solutions can t the experimental data (pressure log) much more accurately for the case of soft rock fracturing
and thus give additional information during interpretation. The relationship between the cohesive parameters and rock fracture
toughness limit is also investigated.
Sarris et al. (2012) developed a poroelastic cohesive zone method and investigated the effects of permeability, injection rate
and formation compressibility on fracture geometry, it is concluded that higher pressures are needed to extend a fracture in a

SPE 168600

poroelastic medium than in an elastic medium, and the created profiles of poroelastic fractures are wider. A constant
permeability of cohesive zone is assumed in this study instead of a rigorous leak off model.
YaoYao (2012) developed a 3D cohesive zone model to predict fracture propagation in brittle and ductile rocks and the
effective fracture toughness method was proposed to consider the fracture process zone effect on the ductile rock fracture. The
results show that with the increase of fracture toughness, the pore pressure cohesive zone model gives predictions that are
more conservative on fracture length as compared with pseudo 3D and PKN models. However, the propagation pressure is not
investigated and the leak-off model used in this study is not consistent between cohesive zone model and analytical models.
In this study, we developed a hydraulic fracture model for both poroelastic and poroplastic formations with the cohesive zone
method. The physical process involves full coupling of the fluid leak-off from fracture surface and diffusion into the porous
media, the rock deformation and the fracture propagation. The model is then solved by finite element method. We investigated
how the plastic deformation can impact the fracture geometry, propagation and closure pressure in an elasto-plastic formation
compared to elastic formation. We also compared the results of an effective toughness method which lumps all plastic
deformation effects into an increased toughness in front of fracture tip, with results from the method that accounts for all
affected areas by plastic deformation.
Mathematical Model
The physical process of the uid driven fracture involves the pumping of a viscous uid that pressurizes the fracture surfaces
which deform. Increasing the pressurization, critical loading conditions will be reached ahead of the tip splitting the rock and
hydraulically driving the fracture. Depending on the formation properties, in-situ stresses and pumping parameters, the fracture
may propagate for more than one hundred meters. In this study, KGD models were developed for plain strain and symmetrical
conditions. This model is appropriate for modeling fractures with fracture height relatively greater than the fracture length.
Furthermore, this model is also appropriate for examining tip effects since the deformation of any arbitrary fracture shape is
approximately planar near the tip.
Fluid Flow
A wide variety of fluids have been used for fracturing including water, aqueous solutions of polymers with or without
crosslinkers, gelled oils, viscoelastic surfactant solutions, foams, and emulsions (Brannon, 2007). Most hydraulic fracturing
fluids exhibit power law rheological behavior and temperature-related properties. In order to avoid additional complexity
added by fluid behavior, incompressible and Newtonian fluid is assumed in this study. The continuity equation which imposes
the conservation of mass in one dimensional flow is (Boone and Ingraffea, 1990)
d d

+ = 0
d d

(1)

where is the local ow rate along the fracture in direction , is the local uid loss in rock formation and is the crack
opening. Equation (1), the local ow rate q can be determined by taking uid leak-off into consideration. Carters Leak-off
model is used in this study to describe leak-off phenomenon (Howard and Fast, 1970):
=

= 2 +

(2)

(3)

where is leak-off velocity, is leak-off coefficient, is the time elapsed since the start of the leakoff process, is the
fluid volume that passes though the surface during the time period from time zero to time , the integration constant is
called spurt loss. The two coefficients and , can be determined from laboratory tests or, preferably, from evaluation of
fracture calibration tests (Valk and Economides, 1999). For a ow between parallel plates, the conservation equation of
momentum balance can be used to relate the pressure gradient to the fracture width for a Newtonian uid of viscosity
(Boone and Ingraffea, 1990):
= =

3 d
12d

(4)

where is the fluid pressure inside the fracture and is the average velocity of fluid over a cross-section of the fracture.
Equation (4) determines the pressure prole along the fracture from the local width and local ow rate. Darcys Law is used to
describe fluid diffusion in the isotropic, porous media:

SPE 168600

where is fluid flux velocity in the porous media, is tensor permeability and is formation pressure gradient.

(5)

Coupled Deformation-Diffusion Phenomena

The basic theory of poroelasticity in which the fully coupled linear elastic rock deformation and pore pressure equations was
initially introduced by the pioneering work of Biot (1941). Since then, many researchers have contributed to its further

through:
development. In fluid filled porous media, the total stresses , are related to the effective stresses ,

, = ,

(6)

The effective stresses govern the deformation and failure of the rock, the poroelastic constant is a rock property that is
independent of the fluid properties and is pore pressure. Ghassemi (1996) demonstrated that variations in the value of
poroelastic constant has no influence on fracture geometry. In this study, the poroelastic constant is assumed to be 1 and
the equilibrium equation in the form of virtual work principle for the volume under its current configuration at time can be
written as (Zienkiewicz and Taylor, 2005):
( ) : = +

(7)

+ = 0

(8)

Where , are effective stress and virtual rate of deformation respectively. and are the surface traction per unit area and
body force per unit volume, I is unit matrix. A porous medium is modeled by attaching the nite element mesh to the solid
phase. Liquid can ow through the mesh. Equating the time rate of change of the total mass of wetting liquid in the control
volume V to the mass of wetting liquid crossing the surface S per unit time gives the wetting liquid mass continuity equation
in the following form:

Where , and are the mass density of the liquid, the porosity of the medium and the average velocity of the liquid
relative to the solid phase. is the outward normal to the surface S.

Fracture Initiation and Propagation

There are a number of fracture propagation criteria provided by the theory of fracture mechanics. The criterion for fracture
propagation is usually given either by the conventional energy release rate approach, which states that a fracture propagates
when the energy release rate reaches a critical value related to fracture toughness; or by the stress intensity approach, which
states that a fracture propagates when the stress intensity factor at the tip exceeds the rock toughness. The energy release rate
and stress intensity approaches are essentially equivalent and uniquely related for linear elastic materials. The most robust
criterion for non-linear mechanics is described by the cohesive zone constitutive model. The cohesive zone model approach
should be clearly contrasted with the conventional fracture mechanics based infinitely sharp fracture models; as such fracture
models have led to a physically meaningless singular stress field near the fracture tip (Sarris et al., 2012). The cohesive
method proposed by Barenblatt (1959; 1962) assumed that the cohesive zone length and the cohesive stress distribution are
material parameters that do not depend on fracture dimensions and loading. The classical Grifths (1921; 1924) brittle elastic
fracture model is a limited case of Barenblatts model: if the cohesive zone length tends to zero (with respective increase of the
cohesive stress) then, in the limit, the stress-strain state corresponding to the classic square root asymptotes will be obtained.
The constitutive behavior of the cohesive zone is defined by the traction-separation relation derived from laboratory tests. The
traction-separation law for the surface is such that with increasing separation, the traction across this cohesive surface reaches
a peak value and then decreases and eventually vanishes, permitting a complete separation. A cohesive potential function is
dened so that the traction is given by
T=

(9)

SPE 168600

where T is traction force and is the displacement across a pair of cohesive surfaces. Various types of traction-separation
relations (potential functions) for cohesive surfaces have been proposed to simulate the fracture process in dierent types of
material systems.
The irreversible bilinear cohesive law (Tomar et al., 2004) is adopted in this study, as shown in Fig.1. This law assumes that
the cohesive surfaces are intact without any relative displacement and exhibit reversible linear elastic behavior until the
traction reaches the cohesive strength (tensile strength) or equivalently the separation exceeds the displacement of
damage initiation, 0 . Beyond 0, the traction reduces linearly to zero up to the displacement of complete failure , and any
unloading takes place irreversibly. For mode-I plane strain, the area under the traction-separation curve equals the fracture
energy GIc , which is the work needed to create a unit area of fully developed crack. For elastic solids this energy is related to
the rock fracture toughness K through (Kanninen and Popelar, 1985):
GIc =

K 2
(1 2 )

(10)

Fig.1. Traction-Separation law and fluid inside cohesive surface

where is Youngs modulus of formation and is Poissons ratio. For bilinear traction-separation law, displacement of
complete failure can be determined by:
2GIc
2K 2 (1 2 )
=
(11)
=

The stress-displacement relation in the initial part of the loading curve corresponds to linear elastic deformation as follows:

= K
(12)
0
Where K is the stiffness of the stress-displacement relation in the loading regime and is assumed to be equivalent to the
stiffness of formation in this study according to the following equation:
=

= K =

(13)

Where d is the initial thickness of cohesive surfaces and the post-peak softening regime the deformation is given by the
following:
( 0 )

(14)
= 1
0
In order to guarantee solution convergence, and to properly capture the details of the deformation eld in the vicinity of the
crack tip and the traction distribution within the cohesive zone, the cohesive element size must be smaller than the cohesive
zone length. The cohesive zone length is an inherent length scale determined by material properties. For Mode-I crack growth
under plane strain conditions, the cohesive zone length is determined by:
=

9K 2

32

32(1 2 )

(15)

Typically, Mode-I-based fracture criteria are employed in conventional hydraulic fracture models; however, it is not sufcient
when a pronounced shear stress component exists. Mode II type fracture can play an important role in rock fracture mechanics
under certain loading conditions. The traction-separation law presented above for Mode-I based fracture criteria can be easily

SPE 168600

extended to mixed-mode fracture criteria. For an isotropic formation, the traction-separation responses in different modes are
assumed to be the same, as shown in Fig. 2.

Fig.2. Traction-Separation law for mix-modes

where t n , t s , t t refer to the nomal, the first, and the second shear stress components; and t 0n , t 0s , t 0t represent the peak values of
the nominal stress when the deformation is either purely normal to the interface or in the rst or the second shear direction;
0n , 0s , 0t corespond to the displacement of initial damage in the normal, the first, and the second shear stress direction and
fn , fs , ft are the displacement of complete failure in these there directions.
Damage is assumed to initiate when a quadratic interaction function involving the nominal stress ratios reaches a value of one.
This criterion can be represented as

+ 0 =1
0
0

16

The symbol used in the above equation represents the Macaulay bracket with the usual interpretation. The Macaulay
brackets are used to signify that a pure compressive deformation or stress state does not initiate damage. The stress
components of the traction-separation model are affected by the damage according to
(1 )
=

damage initated
no damage occurs

17

t are stress components, t are the stress components predicted by the elastic traction-separation behavior for the current strains
without damage. The scalar damage variable, D, represents the overall damage in the material and captures the combined
effects of all the active mechanisms. It initially has a value of 0. If damage evolution is modeled, D monotonically evolves
from 0 to 1 upon further loading after the initiation of damage. For linear softening as shown in Fig.2, the evolution of the
damage variable, D, reduces to (Turon et al., 2006)

where

D=

is effective displacement, defined as

= 2 + 2 + 2

(18)
(19)

The mode mix of the deformation fields in the cohesive zone quantify the relative proportions of normal and shear
deformation. The damage evolution for mixed-mode failure in the current model is dened based on the BenzeggaghKenane
fracture criterion (Benzeggagh and Kenane, 1996), when the critical fracture energies during deformation along the first and
the second shear directions are similar. The energy dissipated due to failure Gc , is defined as
Gc = GIc + (GIIc GIc )

Gshear

Gtotal

(20)

c
c
, Gtotal = Gshear + GIc . And GIc , GIIc , GIII
are the work done by the tractions and their conjugate
where Gshear = GIIc + GIII
relative displacements in the normal, rst, and second shear directions. For isotropic failure in this study where GIIc = GIc , the
cohesive response is insensitive to parameter .

SPE 168600

Simulation Model
A fracture is hydraulically driven with the injection of a uid from the wellbore into the fracture channel. In this model, a predened surface made up of elements that support the cohesive zone traction-opening calculation is embedded in the rock and
the hydraulic fracture grows along this surface, this typical cohesive zone model is illustrated in Fig.3.

Fig.3. Cohesive zone embedded along fracture path (G.M. Zhang et al., 2009)

The fracture process zone (unbroken cohesive zone) is dened within the separating surfaces where the surface tractions are
nonzero. The fracture is fully lled with uid in the broken cohesive zone where no traction from rock fracture exists but uid
pressure is acting on the open fracture surfaces. A two dimensional plane strain uid-driven fracture in porous rock is
simulated, as shown in Fig. 4. The initially unopened fracture is represented by an embedded array of cohesive zone elements
without initial separation along the entire fracture path. An incompressible Newtonian uid is injected at a constant rate. The
cohesive elements at the injection point are dened as initially open to allow entry of the uid at the perforation tunnel(s), so
that the initial ow and fracture growth are possible. Grid size along the fracture path is 0.2 m and all the outer boundaries of
this model are fixed.

.
Fig.4. Initial reservoir and fracture grid around fracture path

The inelastic rock material behavior follows the Mohr-Coulomb flow theory of plasticity for a cohesive frictional dilatant
material. Associative behavior with constant dilatation angle is considered. These assumptions are justified by the presence of
high confining stresses prior to crack propagation and to a decrease in the initial in-situ mean pressure near the crack tip during
propagation. The Mohr-Coulomb criterion assumes that yield occurs when the shear stress on any plane in a material reaches
the same value as shear strength, which is defined as:
= + tan

(21)

where is shear strength, is cohesion strength, is the stress normal to a specific plane, is the friction angle. The
amount of plastic volumetric strain developed during plastic shearing is controlled by shear dilatancy, which is characterized
by a dilation angle , the dilation angle is a measure of the ratio of rate of plastic volumetric strain to the rate of plastic shear
distortion (Vermeer and deBorst 1984), as given by:

SPE 168600

d
|d |

(22)

where is plastic volumetric strain and is the maximum shear strain.

Three different types of formation were investigated, which are poroelastic formation, poroelastic formation with increased
effective toughness and poroplastic formation. For typical quasi-brittle materials, the effective toughness can be 1.141 to 2.236
times higher than original toughness, and for ductile rock, the effective toughness can be determined by investigating softening
behavior experimentally or computationally (YaoYao, 2012). The effective toughness is assumed to be 3 times higher than the
original toughness in this study and the cohesion strength of 3 types of formations are 2 MPa, 4 MPa and 6 MPa respectively,
which represent formations with relative low, moderate and high yield strength in this study. Low cohesion strength can be
explained by weakly cemented formation grains, the presence of microfractures and other planes of weaknesses, or by
reactivation of pre-existing, sealed natural fractures. All the other input parameters are shown in Table.1 unless otherwise
specified.

Table 1-Input Parameters

Elastic Modulus, Gpa
Poissons Ratio
Fluid viscosity, cp
Tensile strength, MPa
Formation Permeability, md
Flow rate Q (3 /s/unit height)
Specific Weight of Fluid, kN/3
Initial Pore Pressure, MPa
Maximum Horizontal Stress, MPa
Minimum Horizontal Stress, MPa
Vertical Stress, MPa
Fracture Toughness , MPa
Carter's Leakoff Coefficient, m/
Spurt Loss Coefficient, m
Porosity
Friction angle
Dilation angle

20
0.25
1
2
10
0.0005
9.8
20
42
37
65
1
2.00E-06
2.5E-04
0.2
27o
8o

The coupled system of equations is solved by use of a Newton-Raphson technique, and the Jacobian matrix is generated for the
entire system of equations, incremental corrections are found by use of a linear solver that includes all the solution variables
and all the variables are updated at the end of each time increment and input as initial values at the start of the next increment.
The program for numerical calculations was developed using FORTRAN and ABAQUS finite element code. Two injection
cases for both short time and long time treatment were simulated. In the first case, a 2 minute injection and fall off was
simulated and the fracture propagated long enough to allow analysis of the results and reach conclusions for long fractures. In
the second case, a 30 minute injection was simulated and the results can provide a direct insight of how the fracture geometry
and fluid efficiency will be different in a more realistic situation if plasticity is considered.
Results and Analysis
In this section we present results of the analysis to demonstrate the importance of the plasticity in modeling hydraulic
fracturing in soft formations. In order to examine the accuracy of this model, the numerical results were compared with the
analytical solution of a 2-D KGD plane-strain tensile fracture model. Because it is not possible to nd the analytical solution
for the coupling process, the comparison to the analytical solution was given only for non-coupling effect. The results are
shown in Fig. 5. The comparison indicates that the simulation results of this model match the analytical solution very well,
except for the net pressure in the early stage of injection; this is because the analytical solution does not account for the period
before breakdown. It also should be mentioned that the numerical model is quite sensitive to the mesh size along the fracture.
A perfect match between the analytical and numerical solution requires the mesh size to be substantially refined.

SPE 168600

12000

0.01

10000

Analytical

8000

Analytical

0.008
Width (m)

Net Pressure (KPa)

Numerical

Numerical

0.006

6000

0.004

4000

0.002

2000
0

0
0

20

40

60
80
Inection Time (s)

100

120

20

40

60
80
Injection Time (s)

100

120

Fig.5. Net pressure and width at wellbore in elastic formation during injection

Case 1: 2-Minute Injection Period

In the first case of 2-minute injection, fracture propagation and fall off were simulated and the results of poroelastic and
poroplastic were investigated. Fig. 6 demonstrates the relationship between net pressure and fracture width at the wellbore as
a function of time. The oscillation of the simulation curve corresponds to halts and sequels in the fracture propagation process
step by step for the series of time increments. The net pressure drops when a new fracture volume is created and a sudden spurt
loss through the newly created surface occurs. The pressure then increases until the cohesive element in front of fracture tip is
again damaged and a new fracture volume created. The results show that after the net pressure reaches a peak value, the
initiated fracture and fracture width start to increase and finally the fracture closes (fracture width at the wellbore drops to
zero) when the net pressure drops to zero.

Fig.6. Net pressure and width at wellbore in elastic formation during injection and falloff

The cumulative leak-off volume per cohesive element surface (0.2 m) based on Carters Leak-off Model was shown in Fig.7,
which indicates a low leak-off rate, so the fracture will propagate relatively longer in a short treatment. When compared to the
poroelastic formation, Fig.8 shows that the net pressure is higher and the fracture is wider in a poroplastic formation. A larger
pressure drop at the end of injection is also observed which can be explained by the fact that the energy absorbed by plastic
rock deformation will not reciprocate as much as that energy in an elastic formation where no energy is absorbed.

Fig.7. Cumulative leak-off volume through per element surface

10

SPE 168600

Fig.8 also indicates that the net-pressure drops to zero while the fracture is still open at the wellbore, this comes from the fact
that some portion of the rock along fracture path will not return to original state in the unloading process because of permanent
plastic deformation. It also suggests that the fracture will close completely at negative net-pressures (fluid pressure less than
the far field stress). Application of classical analysis, which assumes that the fracture closes completely when the fluidpressure drops to the value of the far-field stress, would lead to the underestimation of the minimum in-situ stress.

Fig.8. Net pressure and width at wellbore in plastic formation during injection and falloff

In this study, the effective toughness was assumed to be 3 MPa in a plastic formation and the effective toughness was used
in poroelastic modeling instead of original toughness, to account for the plastic deformation in front of fracture tip. Fig. 9
shows the net pressure of the plastic formation with moderate yield strength and the elastic formation with both original
toughness and effective toughness. As we can see, the net pressure is almost the same in the case of the plastic formation and
the elastic formation with increased effective toughness, and higher than that in elastic formation with original toughness
which does not account for plasticity.

Fig.9. Net pressure at wellbore in elastic and plastic formations

Fig.10. Maximum fracture width in elastic and plastic formations

SPE 168600

11

However, as demonstrated in Fig. 10, the fracture is wider in plastic formation than that in elastic formation with increased
effective toughness, even though they have the same net pressure. So with the same net pressure, the fracture is wider in the
plastic formation when compared to the elastic formation. This discrepancy can be larger in formations with lower yield
strengths. The fracture width profile is shown in Fig.11, and suggests that the method of effective toughness can underestimate
fracture width, and overestimate fracture length under the same net pressure condition because it will only account for the
plastic deformation near the fracture tip. We can also notice a large difference in a relatively small treatment if we do not
consider plasticity at all.

Fracture Width (m)

Poroelastic

Poroelastic, Toughness x3

Poroplastic, Moderate Yield Strength

0.01
0.009
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
0
0

10

12

Distance From perforation tip (m)

Fig.11. Fracture width profile along fracture path after 2 minutes injection

In all cases, the pressure drop along the fracture path is negligible and the pressure inside the fracture is almost equivalent to
the pressure at the wellbore except for the area near the fracture tip. This is because the fracture length is relatively short, fluid
viscosity is small, and according Equation (4), the largest pressure drop occurs at the fracture tip where the width vanishes to
zero. The fluid pressure distribution is shown in Fig. 12. As expected, the fluid pressure drops to pore pressure across the tip
of fracture.

Fig.12. Fluid pressure distribution with 10md permeability

Fig. 13 demonstrates that with higher injection rate, wider fracture profiles can be obtained with faster propagation, which
means that high injection rate can result in higher net pressure and wider plastic deformation. Because the fracture width does
not increase significantly at a higher injection rate, and with low leak-off coefficient in this case, the fracture propagation
speed can be considered as proportional to injection rate. For instance, at the end of the 2-minute injection, the fracture halflength for the case Q x5 injection rate is about 5 times longer than the one created at original injection rate.

12

SPE 168600

30
Fracture half length (m)

Fracture Width (m)

0.01
0.008
0.006
0.004

Poroplastic, Q x1

0.002

Poroplastic, Q x5
2

Poroplastic, Q x1

20
15
10
5
0

0
0

Poroplastic, Q x5

25

4
6
Distance from perforation tip (m)

20

40

60

80

100

120

Injection time (s)

Fig.13. Maximum fracture width and fracture half-length with different injection rate in plastic formation

Fig. 14 shows the net pressure in the plastic formation with moderate cohesion yield strength: with the same far field
horizontal stress, the lower the pore pressure, the lower the net pressure. Fig.15 indicates that with lower pore pressure, plastic
strain decreases and the plastic deformation area becomes smaller, which leads to a longer fracture length.

Fig.14. Net pressure in plastic formation with different pore pressure

Fig.15. Plastic strain with different pore pressure

Theres no doubt that the closer the initial reservoir stress conditions are to the shear failure surface, the more likely shear
failure will be triggered around fracture during propagation. Next, we investigated how stress contrast can have an impact on
plastic deformation by increasing the horizontal stress contrast from 1.1 to 2, the results are shown in Fig.16.

SPE 168600

13

Fig.16. Plastic strain, horizontal stress contrast=2

Compared to Fig. 15, it be can be observed that a higher contrast between the initial minimum and maximum total stresses
can lead to a wider plastic deformation area. When the rock in-situ stress is far away from the shear failure envelop, additional
stress induced by a hydraulic fracture may still not be large enough to cause shear failure at a certain distance from the
fracture. However, when the rock is close to the shear failure envelop under initial condition, stress disturbance induced by
fracture is able to push the rock into the failure envelop more easily, hence can have wider impact around the fracture.
Now we increase the leak-off coefficient 100 times and compare the results with those of original leak-off coefficient. We see
that a higher leak-off rate leads to much shorter fracture length, but this only has a limited impact on net pressure and fracture
width, as shown in Fig.17. Slightly higher net pressure and a smaller fracture width are obtained with higher leak-off rate
because of the poroelastic backstress. Even though the leak-off rate increases 100 times, the permeability is still large enough
to allow the pore fluid to move freely and dissipate energy at the same time scale as the fracture opening, so the pore pressure
around the fracture can be considered nearly constant and the poroelastic backstress induced by leak off volume and plastic
deformation affected by changes of pore pressure is relative small.

Fig.17. Net pressure and maximum width with different leak-off rate in plastic formation, k=10 md

Next, in Fig.18 we reduce the formation permeability from 10 md to 0.001 md and maintain the original leak-off rate. As a
result, we observe a quite large net pressure increase for the lower permeability formation (i.e., the red curve), while only
minor changes occur for the fracture width (i.e., the blue curve).

14

SPE 168600

Fig.18. Net pressure and maximum width with different permeability in plastic formation, = 2E-6 /

Fig.19 provides further explanations on the phenomena depicted in Fig.18. We can observe, in fact, a pore pressure increase
around the fracture (i.e., the red contour indicating a pore pressure value around 25 MPa), because the low formation
permeability impacts adversely the diffusion process, even with the relatively low leak-off rate. We can furthermore see that
the pore pressure around the fracture is almost 5 MPa higher than the original pore pressure. The increased pore pressure
around the fracture leads to larger zones of plastic deformation and poroelastic backstress, which in turn, contributes to higher
net pressure. However, the resulting difference between fracture widths is small due to the large poroelastic backstress, which
offsets some of the effect of higher net pressure and larger plastic deformation.

Fig.19. Fluid pressure distribution in formation matrix, k=0.001 md and = 2E-6 /

In this study, we also found that a non-wetting zone can be developed at the fracture tip in low permeability formation when
the fracture propagates quicker than the fracture uid can ll up new fracture volumes that are created at the tip of a
propagating fracture. This is manifested by negative pressures develop at the tip of a propagating fracture as shown in Fig.20,
the formation permeability was reduced to 0.0001 md and injection rate was increased to 0.05 3 /s/unit height.

Fig.20. Non-wetting zone in front of fracture tip

SPE 168600

15

Three types of plastic formations with different cohesion yield strengths were investigated; the plastic strain and affected area
are shown in Fig.21 and Fig.22.

Fig.21. Plastic strain with high and moderate yield strength

Fig.22. Plastic strain with low yield strength

This shows that the more plastic the formation (with decreasing cohesion strength), the larger is the area affected by plastic
deformation, and the shorter the fracture length. The area that is affected by plastic deformation is within 0.4m and 1 m
around fracture path in the formation with high and moderate yield strength, but it can extend to a few meters in the formation
with relative low yield strength. It was also found that the more plastic the formation is, the higher the needed propagation
pressure is, with resulting wider induced fractures, as shown in Fig.23 and Fig. 24.

Fig.23. Net pressure with different yield strength

16

SPE 168600

Fig.24. Fracture width with different yield strength

Case 2: 30-Minute Injection Period

In the second case, a 30-minute injection was simulated, with Carter's leak-off coefficient of 2 104 m/ and 10md
permeability, in which case the effect the poroelastic backstress is negligible. The fracture length and fracture width are
inherent solutions of each time increment in all models. The total injection fluid can be calculated from injection rate and
pump schedule, the total leak-off volume can be calculated based on equation (2) by integrating the entire fracture surface
domain.
To investigate the impact of deviations of fracture geometry on production, the Unified Fracture Design (UFD) approach
(Economides et al., 2002a) was used, The central idea of the UFD technique is to select the appropriate optimum compromise
between propped fracture length and width, for a given proppant volume and depending on the properties of the reservoir and
the selected proppant. To determine the volume of proppant reaching the target layer in a two-dimensional calculation, the
fracture height must be a known quantity, predicted or presumed. Then

The next step is to calculate the Proppant Number:

( )

=
(1 )

(23)

4 4
2
=
=
2
2

(23)

where is the propped volume inside the pay as calculated from Eq. (23) and is the drainage volume of the reservoir.
From the value of the maximum can be determined, as can the optimum fracture conductivity, , (Economides et al,
2002a). For a square drainage area, the maximum achievable dimensionless pseudosteady-state productivity index as a
function of the Proppant Number is given by:

, =

1
0.990 0.5 In( )

0.423 0.311 0.089 2

6
exp

1 + 0.667 + 0.015 2
6

if 0.1

if 0.1

if ~100

Similarly, the optimal dimensionless fracture conductivity for the entire range of Proppant Numbers is given as

(24)

SPE 168600

17

1.6

if < 0.1

0.583 + 1.48 In( )

, = 1.6 + exp
1 + 0.142In( )

if 0.1 10
if > 10

(25)

Once the optimal dimensionless fracture conductivity is known, the optimal fracture length and width can be readily
determined:

, =
2,
=

(26)

,
2

(27)

Fig.25 shows the relationship and for a range of Proppant Numbers.

Fig.25. Relationship between dimensionless productivity index and dimensionless fracture conductivity

The real fracture geometry that deviated from the optimal design because of formation plasticity can lead to suboptimal
dimensionless fracture conductivity, , which can result in reduction of dimensionless production index, . The
dimensionless fracture conductivity can be calculated by:

=
(28)

Table 2 represents the input data for Unified Fracture Design presented in this work. Based on the input data the optimum
fracture geometry that provides the maximum dimensionless productivity index can be determined. The fracture propagation
model is run independently until its output matches the conclusions of the UFD.
Table 2: Reservoir and fracture input data for the fracture designs

Drainage area, acre

320

Net thickness, h, ft

100

Fracture height, hf, ft

100

Well radius, rw, ft

0.3

Proppant mass, Mp, lbm

200,000

Porosity of proppant pack

0.36

Specific gravity of proppant

2.65

Proppant pack permeability, kf, md

115,000

Reservoir Permeability, k, md

10

18

SPE 168600

The simulation results are shown in Table 3.

Table 3: Simulation results of 30-minute injection

Fracture Length (m)

Fracture Width (mm)
Fluid Efficiency %

Reduction %

Optimum Design

Poroelastic

79.8
11.43
73.87
1.6
0.37
0

80.6
11.52
75.20%
1.6
0.37
0

Poroelastic, x3
0.03
58.8
16.95
80.72%
3.3
0.34
8.1

Poroplastic, Moderate Yield Strength

50.2
20.67
84.06%
4.7
0.33
10.8

From these results, we can see that the plasticity can have a substantial impact on fracture geometry after 30 minutes of
injection. The fracture length can be reduced by nearly 38 % compared to the predicted value in elastic formation in this study.
In addition, the fluid efficiency increases in the plastic formation because the shorter fracture length results in a smaller leakoff surface. These results also indicate that the effective toughness method can underestimate fracture width and overestimate
fracture length even if it can accurately match the net pressure in the plastic formation, because it only accounts for shielding
effects of the process zone in front of the fracture tip and fails to take the entire plastic zone into consideration.
From a production point of view, a 30-minute injection schedule can generate nearly optimal fracture geometry in elastic
formation with maximum dimensionless productivity index (Table 3). However, if the formation exhibits plastic behavior, the
real fracture geometry will be suboptimal and the dimensionless productivity index, , is 10% lower than the optimal value.
The reduction in can be more severe in unconventional reservoirs where formation permeability is extremely low. From
Eq.(23), the lower the formation permeability the higher the Proppant Number. The Proppant Number increases to 3.2 by
reducing the formation permeability to 0.1 md in Table.2. If the dimensionless fracture conductivity, , is 3 times higher
than the optimum value (the same ratio in Table 3) then from Fig.25, it can be seen that the deviation from optimal fracture
geometry can lead to nearly 20% reduction in . In the case of lower cohesion strength formation rock, subjected to higher
in-situ stress and horizontal stress contrast, the fracture half-length can be reduced to around 40% of that predicted by the
elastic model, which can lead to the dimensionless fracture conductivity, , that is 6~7 times higher than the optimal value
leading to a 40~50% reduction in the . Thus reduction in fracture length because of plasticity has much greater impact on the
well production in a low permeability reservoir than that in a high permeability reservoir.
Conclusions
In this study, we developed a numerical model of a hydraulic fracture based on the cohesive zone method to investigate
fracture propagation in elastic and plastic formations. Solid deformation, fluid flow inside the fracture and the fluid diffusion
process are fully coupled in this model. The impact of formation plastic properties on the fracture process is investigated for
both short term and long term injection and the results are compared with an elastic formation. In addition, the main factors
that affect the impacts of poroelastic backstress and formation plasticity on a hydraulic fracture are also investigated. Finally,
we studied how the plasticity can influence well production. From the results of this research, we can conclude that:
1.

2.

3.

4.

In a plastic formation, net-pressure drops to zero while the fracture is still open along a proportion of the original
length; the fracture will close completely at negative net pressure when fluid pressure inside the fracture is less than
far-field minimum horizontal stress.
In comparison to hard rock, plastic and highly deforming formations have exhibited higher breakdown and injection
pressures. The more plastic the formation (lower cohesion strength), the higher is the net pressure required to
propagate the fracture due to increased plastic deformation, which leads to shorter and wider fracture. Under the same
net pressure, a fracture in a plastic formation is wider than that in an elastic formation. This discrepancy is larger in a
formation with lower yield strength where a larger area undergoes plastic deformation.
The effects of formation plasticity on a hydraulic fracture are mostly controlled by in-situ stress, the plastic property
and pore pressure. The closer the initial stress value is to the shear failure value and the greater the total stress
contrast, the more likely shear failure will be triggered around fracture. With the same far-field stress and formation
properties, pore pressure can play an important role on the fracture process in a plastic formation. When pore pressure
increases, net pressure increases due to larger zones of plastic deformation because of strong fluid/solid coupling.
Poroelastic backstress is mostly controlled by leak-off rate and formation permeability. When the leak-off rate is
relatively high and the formation permeability is not large enough to allow pore fluid to move freely and dissipate

SPE 168600

5.

6.
7.

19

energy at the same rate as the fracture opening, the pore pressure around the fracture will increase and induce
compressive stress on the fracture surface.
Both poroelastic backstress and formation plasticity can contribute to a higher net pressure, but they have a different
impact on fracture width. Poroelastic backstress tends to reduce fracture width, while plasticity tends to increase
fracture width.
Non-wetting zone can be developed ahead of the fracture tip in low permeability formations when a fracture
propagates quicker than the fracture uid can ll up newly created micro-cracks.
Ignoring plasticity in plastic formation can lead to inaccuracy of fracture geometry, net pressure prediction and fluid
efficiency calculation. The deviations of real fracture geometry from elastic formation model can have great impact
on well production, especially in unconventional, low permeability reservoirs.

Nomenclature

= Flow rate along fracture, 3 /s

= Leakoff rate, 3 /s

= Fracture width, m

= Carters leakoff coefficient, m/

= Dimensionless fracture conductivity

= Dimensionless production index

= Leakoff rate, m/s

= Cumulative leakoff volume, 3

= Volume of proppant, 3

= Drainage volume of the reservoir, 3

= Volume of proppant, 3

= Leakoff surface, 2

= Spurt loss, m

= Fluid viscosity, cp

= Pressure, Pa

= Stress, Pa
= Effective stress, Pa

= Shear strength, Pa

c
= Cohesion strength, Pa

= Poroelastic constant
= Pore pressure, Pa

= Normal strain

= Shear strain
T
= Traction force, Pa

= Displacement, m
= Effective displacement, m

= Displacement at damage initiation, m

0
= Displacement at failure, m

= Potential function

= Friction angle, degree

= Youngs modulus, Pa

= Poissons ratio

= Rock fracture toughness in Mode I, Pam

I
= Fracture critical energy in Mode I, J
II
= Fracture critical energy in Mode II, J

III
= Fracture critical energy in Mode III, J
= Cohesive tensile strength, Pa

= Cohesive stiffness, Pa/m

= Initial thickness of cohesive surfaces, m

= Cohesive zone length, m

= Fluid mass density, Kg/m3

=Porosity of porous media

= Average velocity of fluid relative to solid phase, m/s

= Fluid flux velocity, m/s

= Tensor permeability in formation, md

= Unit Matrix

20

SPE 168600

= Surface traction per unit area, pa/m2

= Body force per unit volume, N/m3 (non-italic)

References
Adachi, J., Siebrits, E., Peirce, A.P., and Desroches, J. (2007): Computer simulation of hydraulic fractures. International
Journal of Rock Mechanics and Mining Sciences 44(5):739757
Barenblatt, G.I. (1959): The formation of equilibrium cracks during brittle fracture: general ideas and hypothesis, axially
symmetric cracks. Journal of Applied Mathematics and Mechanics, 23:622636
Barenblatt, G.I. (1962): The mathematical theory of equilibrium cracks in brittle fracture. Advanced in Applied Mechanics,
Academic Press, New York, 1962, pp. 55-129
Barree, R.D.: Fracturing Materials, keynote address, SPE Hydraulic Fracturing Conference, The Woodlands, TX, January
19-21, 2007
Benzeggagh, M.L., and Kenane, M. (1996): Measurement of mixed-mode delamination fracture toughness of unidirectional
glass/epoxy composites with mixed-mode bending apparatus. Composites Science and Technology, 56:439449.
Bernhard, A., Schreer, S.S., and Luciano, S. (2006): On adaptive renement techniques in multi-eld problems including
cohesive fracture. Computer Methods in Applied Mechanics and Engineering, 195:444461
Biot, M. A. (1941): General theory of three dimensional consolidation. Journal of Applied Physics, 12(2), 155 164
Boone, T.J., and Ingraffea, A.R. (1990): A numerical procedure for simulation of hydraulically driven fracture propagation in
poroelastic media. International Journal for Numerical and Analytical Methods in Geomechanics,14:2747
Boone, T.J., Ingraffea, A.R., and Wawrzynek, P.A. (1986): Simulation of fracture process in rock with application to
hydrofracturing. International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts, 23:255265
Chen, Z.R, Bunger, A.P, Zhang, and Jeffrey, R.G. (2009) Cohesive zone finite element based modeling of hydraulic
fractures. Acta Mechanica Solida Sinica, 22(5):443-452
Dean, R.H., and Schmidt, J.H. (2009): Hydraulic fracture predictions with a fully coupled geomechanical reservoir
simulator. SPE J. 14(4):707714
Dugdale, D.S. (1960): Yielding of steel sheets containing slits. Journal of the Mechanics and Physics of Solids, 8:100104
Economides, M. and Nolte, K. (2000): Reservoir Stimulation. John Wiley & Sons, Chichester UK, 3rd edition
Economides, M.J., Oligney, R.E., and Valk, P.: Unified Fracture Design. Orsa Press, 2002.
Geertsma, J. and De Klerk, F. (1969): A rapid method of predicting width and extent of hydraulic induced fractures. Journal
of Petroleum Technology, 246:15711581
Germanovich, L.N., Astakhov, D. K., Shlyapobersky, J., Mayerhofer, M.J., Dupont, C., and Ring, L.M. (1998): Modeling
multi-segmented hydraulic fracture in two extreme cases: No leak-off and dominating leak-off. International Journal of Rock
Mechanics and Mining Sciences, 35(4 5), 551 554.
Ghassemi, A. (1996): PhD dissertation Theree Dimensional Poroelastic Hydraulic Fracture Simulation Using The
Displacement Discontinuity Method, University of Oklahoma
Grifth, A.A. (1921): The phenomena of rupture and ow in solids. Philosophical Transactions of the Royal Society,
221:163198
Grifth, A.A. (1924): The theory of rupture. Proceedings of 1st international congress for applied mechanics, Delft, The
Netherlands, pp 5563
Howard, G.C., and Fast, C.R. (1970): Hydraulic Fracturing Monograph. SPE Monograph. Vol. 2. Henry L. Doherty series.
Kanninen, M.F., and Popelar, C.H. (1985): Advanced fracture mechanics. Oxford University Press, UK.
Khristianovich, S., and Zheltov, Y. (1955): Formation of vertical fractures by means of highly viscous uids. In Proc. 4th
World Petroleum Congress, Rome, volume II, pages 579586.
Mokryakov, V. (2011): Analytical solution for propagation of hydraulic fracture with Barenblatts cohesive tip zone.
International Journal of Fracture, 169:159168
Nordgren, R. (1972): Propagation of vertical hydraulic fractures. Journal of Petroleum Technology, 253:306314. (SPE
3009).

SPE 168600

21

Papanastasiou, P. (1997): The influence of plasticity in hydraulic fracturing. International Journal of Fracture, 84(1), 61
79.
Papanastasiou, P. (1999): The effective fracture toughness in hydraulic fracturing. International Journal of Fracture, 96(2),
127 147.
Perkins, T. and Kern, L. (1961): Widths of hydraulic fractures. Journal of Petroleum Technology, Trans. AIME, 222:937
949.
Sarris, E., and Papanastasiou, P. (2012): Modeling of Hydraulic Fracturing in a Poroelastic Cohesive Formation.
International Journal of Geomechanics , Vol. 12, No. 2, April 1, 2012
Sone, H., and Zoback, M.D. (2011): Visco-plastic Properties of Shale Gas Reservoir Rocks. Presented at the 45th U.S. Rock
Mechanics / Geomechanics Symposium, June 26 - 29, 2011, San Francisco, California.
Shadravan, A., and Amani, M. (2012): HPHT 101-What petroleum engineers and geoscientists should know about high
pressure high temperature wells environment. Energy Science and Technology, 4(2), 36-60
Tomar, V., Zhai, J., and Zhou, M. (2004): Bounds for element size in a variable stiness cohesive nite element model.
International Journal for Numerical Methods in Engineering, 61(11): 1894-1920.
Turon, A., Camanho, P.P., Costa, J., and Davila, C.G. (2006): A damage model for the simulation of delamination in
advanced composites under variable-model loading. Mechanics of Materials, 38 (11): 10721089.
Valk, P.P., and Economides, M.J: Fluid-Leakoff Delineation in High-Permeability Fracturing. SPEPF, 14 (2), pp117-130,
May 1999.
Van Dam, D.B., Papanastasiou, P., and De Pater, C.J. (2002): Impact of rock plasticity on hydraulic fracture propagation and
closure. SPE Production & Facilities, 17(3), 149 159.
Vermeer, P. A., and de Borst, R. 1984. Non-Associated Plasticity for Soils, Concrete and Rock. Heron 29(3): 3-64.
YaoYao (2012): Linear Elastic and Cohesive Fracture Analysis to Model Hydraulic Fracture in Brittle and Ductile Rocks.
Rock Mechanics and Rock Engineering, 45:375387
Zhang, G.M., Liu, H., Zhang, J., Wu, H.A., and Wang, X.X. (2010): Three dimensional finite element simulation and
parametric study for horizontal well hydraulic fracture. Journal of Petroleum Science and Engineering, 72:310-317
Zienkiewicz, O.C., and Taylor, R.L. (2005): The Finite Element Method, (5th edition), Vol. 1, the basis. Elsevier Pte Ltd,
London, pp 42-45