SPE 168622

Limited Height Growth and Reduced Opening of Hydraulic Fractures due to Fracture
Offsets: An XFEM Application
Safdar Abbas, Schlumberger, Elizaveta Gordeliy and Anthony Peirce, University of British Columbia, Brice Lecampion, Dimitry Chuprakov
and Romain Prioul, Schlumberger

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, 4–6 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

Hydraulic fracture propagation through a layered medium often exhibits a complex fracture path due to lateral shifting of the fracture
path after passing through an interface. The existence of these offsets in the fracture path has been confirmed by mapping of mined
fractures and by laboratory experiments. In addition to the stress contrast and material contrast, these offsets act as another mechanism
for fracture-height containment. In order to investigate fracture height growth problem, we have considered a case of plane-strain
fractures propagating from an injection point and propagating in a predetermined path. Fracture offsets of a given length and angle are
prescribed within the predetermined path. In this way, we parametrize the problem of fracture offsets in order to quantify their effects
on the fracture height growth and fracture opening reduction at the fracture offset (pinching effect). This is done while keeping a
hydrostatic state of confining stresses in order to observe only the geometric effect of the fracture offset. We use a recently developed
hydraulic fracturing code based on the eXtended Finite Element Method (XFEM). This code solves the fluid flow in the fracture and
the elastic response of the fracture in a fully coupled manner solving for the fracture velocity using the complete hydraulic fracture
tip asymptotics.
For different combinations of independent parameters (i.e., formation moduli, far-field stresses, fluid injection rate, ratio of offset
length to the length of the straight fracture and the offset angles) we investigate the effect on the fracture height growth and the
reduction in fracture opening at the fracture offset. A detailed parametric study shows that while each parameter affects the fracture
height containment, the fracture opening reduction is dominated by the fracture offset angle which is a geometric effect of the fracture
offsets. This has a profound effect on the proppant bridging at the fracture offset.

Introduction

Hydraulic fracture treatments are designed in order to restrict the fracture within the pay zone. If the fracture crosses into the adjacent
rock formations, an excessive amount of material and effort is wasted in fracturing unproductive zone. In contrast, if the designed
height of the hydraulic fractures is not achieved then a large area of the productive zone is not stimulated which affects the production
rates. For this reason, numerical prediction of the fracture height growth has been an important field of research in hydraulic fracture
modeling. Traditionally, fracture height growth has been associated with stress (Palmer and Carroll Jr. (1983); Jeffrey and Bunger
(2009); Adachi et al. (2010)) and modulus contrast (Van Eekelen (1982); Gu and Siebrits (2008)) between adjacent layers. During a
number of field studies it was found that the fracture height growth is less than that predicted by various numerical models considering
modulus and stress contrast only. This led to the research in the fracture height containment due to weak mechanical interfaces
(Van Eekelen (1982); Teufel and Clark (1984); Barree and Winterfeld (1998); Miskimins and Barree (2003); Athavale and Miskimins
(2008)). When a hydraulic fracture interacts with a weak interface, either it crosses the interface, it is blunted and stopped at the
interface or the hydraulic fracture develops a fracture offset (Beugelsdijk et al. (2000); Wu et al. (2004)).
There has been a number of studies which involve hydraulic fracturing before mining operations and then mining through the
formation and observing the hydraulic fractures (Boyer II et al. (1986); Fisher and Warpinski (2012); Jeffrey et al. (2009)). One
of the common features of hydraulic fractures found in all of these studies is the presence of fracture offsets. Many laboratory
experiments have also verified these findings (Chudnovsky et al. (2001); Warpinski (1991); Hanson et al. (1981); Zhang and Jeffrey
(2008); Zhang et al. (2007); Thiercelin et al. (1987)). There has been a number of studies which focused on the mechanism that leads
to the development of these fracture offsets. These offsets develop as a result of the interaction between a propagating hydraulic
fracture and the bedding planes which act as the planes of weakness. Different models were developed as the result of these studies
which explained the factors affecting fracture crossing, arresting or offsetting. Stress contrast, angle and interface friction have been
identified as the parameters affecting the result of this interaction. In a recent development, fluid injection rate/viscosity has been
2 SPE 168622

identified, experimentally by Beugelsdijk et al. (2000), semi analytically by Chuprakov et al. (2013a,b); Kresse et al. (2013) and in a
hydraulic fracture model by Kresse et al. (2013), as one of the important factors affecting the fracture outcome of such an interaction.
In this study, we are interested in finding out the effect of a fracture offset after it has been created. Once the fracture offset has
been created, it affects the fracture height growth by slowing down fracture propagation in vertical direction. The fracture offset
also acts as a restriction in the fluid flow path which causes an increase in the treating pressure. This reduced fracture opening also
results in proppant bridging in the fracture offset. This may result in the loss of the stimulated area ahead of the fracture offset
due to fracture closure after the fluid pumping is stopped at the end of the fracturing treatment. This investigation will quantify the
effect of geometric parameters of a fracture offset namely its length and angle. These parameters affect the fracture height growth
and reduce the fracture opening in the fracture offset as discussed by Jeffrey et al. (2009, a) and Jeffrey et al. (2009, b). In both of
these studies, the opening at the fracture offsets is affected by the difference in the far field stresses which are higher in the direction
of fracture propagation as compared to the direction of fracture opening. In this contribution, we have assumed that there is no
difference in the vertical confining stress and the minimum horizontal principal stress thus eliminating any effect associated with the
difference in confining stresses in the two directions. This is done in order to study the solely geometric effects of the fracture offsets.
Using a recently developed numerical code based on the eXtended Finite Element Method (XFEM) (Gordeliy and Peirce (2013a,b))
a hydraulic fracture is propagated symmetrically through a prescribed path which includes fracture offsets at a given distance. This
allows for the simulation of hydraulic fracture propagation taking into account the full fluid-solid coupling using the correct fracture
tip velocity. In addition to the geometry of the fracture offsets we have also studied the effects of injection rate, material properties
and the confining stresses on the fracture height growth and the reduction in the fracture opening at the fracture offset.

Numerical Scheme

The hydraulic fracture propagation is simulated using a recently developed numerical algorithm based on the eXtended Finite Element
Method (XFEM). This algorithm relies on the XFEM to solve the elastic component of the coupled elasto-hydrodynamic equations.
The XFEM has an advantage over the traditional finite element method in simulating fracture mechanics (Belytschko and Black
(1999); Moës et al. (1999); Fries and Belytschko (2010)). The finite element mesh is not needed to conform to the fracture geometry
which avoids any kind of mesh manipulation during the fracture propagation. In addition to that, the singular fields at the fracture
tip are resolved by enriching the approximation space by using appropriate enrichment functions (Fries and Belytschko (2010)) thus
avoiding mesh refinement at the fracture tip. For the hydrodynamic part of the coupled problem, the conservation of the viscous
fluid flowing within the fracture is described by the Reynolds lubrication equation which represents a degenerate PDE. This equation
is solved in a fully coupled manner with the equations of elastic equilibrium for a solid body in the state of plane strain. In order
to avoid solving a degenerate PDE for the singular pressure at the fracture tip and in order to incorporate a tip asymptotic solution,
applicable at the computational length scale, a mixed scheme is used. In this mixed scheme, the fracture is divided into two regions,
i.e. a channel region and a tip region. The tip asymptotic solution is prescribed in the tip region as the displacement jump using the
localized mixed hybrid formulation of Zilian and Fries (2009). For the channel region, the applied fluid pressure, obtained from the
lubrication equation, serves as the boundary condition at the crack faces within the XFEM formulation. The resulting fracture opening
in the channel is computed from the XFEM solution for the fracture with these mixed boundary conditions. . The dynamic fracture
propagation involves locating the fracture tip for each time step which represents a free boundary in space. This free boundary is
located by using the Implicit Level Set Algorithm (ILSA) (Peirce and Detournay (2008)). This algorithm is used in the context of the
XFEM by Gordeliy and Peirce (2013b).

C C
C
A A
B B B
A

h0

O O O

A′ B′ A′ A′
B′ B′

C′ C′ C′
Fig. 1—Fracture propagation along a predetermined path.

In the current application we consider a hydraulic fracture growing in an impermeable elastic rock in a state of plane strain. The
rock stiffness is characterized by the Young’s modulus E and Poisson’s ratio ν. The rock fracture toughness KIc characterizes the
breaking strength of the rock. The fracture is assumed to be driven by a Newtonian fluid with a viscosity µ. The fluid is injected at a
SPE 168622 3

Table 1—Material parameters for simulations.

E ν K
√ σv = σh E0 K√0
(psi) (-) (psi in) (psi) (psi) (psi in)
Limestone 3.75 × 106 0.24 480 6474 3.99 × 106 1531.93
Shale 2.176 × 106 0.17 1500 6063 2.24 × 106 4787.3

constant volumetric injection rate Q0 per unit length in the out-of-plane direction from a point source. The initial fracture geometry
corresponds to a longitudinal fracture propagating from a horizontal wellbore with initial height h0 . The initial fracture opening and
pressure corresponds to the analytical solution (Adachi and Detournay (2002)) of a symmetric plane strain hydraulic fracture for a
given rock material parameters, fluid viscosity, constant injection rate and fracture half-length (initial height h0 , see Figure 1). As the
fluid is injected with a constant rate, the pressure inside the fracture increases with time and the fracture starts to grow symmetrically
along the path described by the dashed lines OABC and OA0 B0C0 in Figure 1.

Results and Discussion

hj

wst
α
hst
hof wj
hf

Injection point Injection point

wj

Fig. 2—Introduction of the opening and the length terms in the case of a straight fracture and a fracture with an offset.

Table 2—Parameter space for simulations.

hj
Q hf hj h0 α
hf
bpm (-) (ft) (ft) (ft) (-)
[1, 40] [0.01, 0.04] [95, 25] [1, 1] [31.5, 22.5] [60, 70, 80, 90]

In order to compare the test cases with different fluid injection rates in different rock materials the dimensionless toughness K
is computed for each case. The dimensionless toughness K for plane strain hydraulic fracture geometry propagating in an elastic
impermeable medium is expressed as (Detournay (2004)):
K0
K= , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1)
(µ0 E 03 Q0 )1/4
where µ0 is dynamic viscosity given as µ0 = 12µ. The dimensionless toughness less than 3 indicates that the fracture is propagating
in the viscosity dominated regime which means that there is a large fluid pressure drop inside the fracture and more energy is used
to push the fluid in the fracture than it is needed to fracture the rock. We have run a number of simulations using the XFEM code
for the hydraulic fracture propagation through a prescribed path. The fracture is assumed to have a propagation path with offsets of a
given length and angle. It is assumed that the vertical stress is equal to the minimum principal stress and that there is no material or
stress contrast across the fracture offset. In this way only the geometric effect of the fracture offsets is considered. All the test cases
considered are symmetrical where two longitudinal fractures are initiated from the top and the bottom of a horizontal wellbore. These
fractures propagate and encounter fracture offsets of a given size h j and angle α at a given fracture length h f as shown in Figure 2. In
the test cases, limestone is considered as an example of a harder rock and shale is considered as an example of a softer rock. Material
properties and the confining stresses for each rock type are based on the real data from Eagleford shale (1). Water is considered as
the fracturing fluid in all the test cases with the viscosity of 1 cP. The fluid is pumped at two different rates in order to see the effect
4 SPE 168622

130 6500
α = 90° α = 90°
120 α = 80° α = 80°
°
α = 70 α = 70°
°
110 α = 60 6450 α = 60°
Straight Fracture Straight Fracture

100
Fracture Height (ft)

Inlet Pressure (psi)

90 6400

80

70 6350

60

50 6300

40

30 6250
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Time (sec) Time (sec)

hj
Fig. 3—The effect of fracture offset angle on the fracture height and inlet pressure for Shale with Q = 4 bpm, = 0.01 ft, h f = 95 ft, h0 =
hf
31.5 ft.

of high and low injection rates on the fracture height growth. Two cases of fracture offset condition are considered. One with the
short offset as compared to the fracture length, i.e. smaller ratio h j /h f , and the other with a longer offset as compared to the fracture
length, i.e. larger ratio h j /h f , see 2. In all the test cases, the time delay and the reduction in fracture opening at the middle of the
fracture offset (i.e., at h j /2) are evaluated for several values of fracture offset angle α given in 2.

0.25
α = 90° α = 90°
10 α = 80° α = 80°
Fracture Opening in Fracture Offset (in)

α = 70 °
α = 70°
9 α = 60 °
0.2 α = 60°
Straight Fracture
8
Height Containment (%)

7
0.15
6

0.1
4

0.05
2

0 0
0.02 0.04 0.06 0.08 0.1 0.12 0.14 30 40 50 60 70 80 90 100 110 120 130
Time (sec) Length (ft)
85
α = 80°
80 α = 70°
α = 60°

75
Reduction in Opening (%)

70

65

60

55

50

45

40

35
95 100 105 110 115 120 125
Fracture Length (ft)

Fig.
 4—Height containment and
 opening reduction for Shale with Q = 40 bpm, K = 0.2538 small-offset
hj
= 0.01 ft, h f = 95 ft, h0 = 31.5 ft .
hf

The effect of the angle of the fracture offset on the fracture height growth can be seen from Figure 3 (left). This is the case of
SPE 168622 5

5 100
α = 90° α = 90°
°
4.5 α = 80 α = 80°
α = 70° 90 α = 70°
°
α = 60 α = 60°
4

Reduction in Opening (%)

Height Containment (%)

80
3.5

3
70

2.5

60
2

1.5 50

1
40
0.5

0 30
0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11 96 98 100 102 104 106 108 110
Time (sec) Fracture Length (ft)

Fig.
 5—Height containment and  opening reduction for Limestone with Q = 40 bpm, K = 0.0527 small-offset
hj
= 0.01 ft, h f = 95 ft, h0 = 31.5 ft .
hf

80
10 α = 80° α = 80°
α = 70° α = 70°
9 α = 60° α = 60°
70

8
Reduction in Opening (%)
Height Containment (%)

7 60

50
5

4
40
3

2
30

0 20
0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 95 100 105 110 115 120
Time (sec) Fracture Length (ft)

Fig.
 6—Height containment and
 opening reduction for Shale with Q = 1 bpm, K = 0.6384 small-offset
hj
= 0.01 ft, h f = 95 ft, h0 = 31.5 ft .
hf

a fracture propagating in the shale formation (lower confining stress) with the high fluid injection rate Q = 40 bpm. Length of the
fracture before the fracture offset is h f = 95. In the case of 90◦ fracture offset, the fracture slows down as it reaches the fracture offset
and stops just after crossing the offset. In the case of 80◦ and 70◦ offset angles, fracture propagation velocity is considerably reduced
but the fracture is not stopped. The fracture inlet pressure at the wellbore is plotted against time for this case in Figure 3 (right).
This plot reveals that a significant increase in the wellbore pressure is observed in the case of 90◦ , 80◦ and 70◦ fracture angles. This
pressure increase at the fracture inlet can be considered as one of the diagnostic features of the situation when a fracture encounters
a fracture offset in the propagation path.
Next we look at the effect of fracture offsets on fracture height growth. The fracture height containment is computed using the
fracture height versus time plot in Figure 3 (left). The height containment is defined as the percentage reduction in fracture height at
a given time in fracture propagation (see Figure 2).
hst − ho f
% Height Containment = × 100, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (2)
hst
where hst is the height of the fracture with no fracture offset at the same time when the height of the offset fracture is ho f . The
reduction in fracture height growth with time for the previous case is plotted for each fracture offset angle in Figure 4 (top-left). It
is obvious from this plot that the fracture offsets affect the fracture height growth depending upon the offset angle. Higher the offset
angle, higher is the fracture height containment in this case. As soon as the fracture reaches the fracture offset, it starts to lose fracture
height. The fracture height containment becomes nearly constant after achieving a maximum value. The same characteristic behavior
is observed for all the test cases considered, see Figures 5 to 11 (left figures). In most of the test cases, the fracture with the 90◦
offset was not able to propagate through the fracture offset. Only in the case of the limestone, with high rate of flow and smaller ratio
h j /h f , the fracture with 90◦ offset propagated through the offset, see Figure 5 (left). In the test case for shale with Q = 1 bpm and
large ratio h j /h f , the fracture with 80◦ offset was also not able to propagate through the fracture offset, see Figure 10 (left).
Another effect of the fracture offsets is that the energy used to open the main fracture is consumed in shearing the fracture offset.
6 SPE 168622

90
α = 90° α = 80°
°
α = 80 α = 70°
12
α = 70° α = 60°
° 80
α = 60

Reduction in Opening (%)

10
Height Containment (%)

70

60

50
4

40
2

0 30
0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 95 100 105 110 115 120 125 130 135
Time (sec) Fracture Length (ft)

Fig.
 7—Height containment and  opening reduction for Limestone with Q = 1 bpm, K = 0.1325 small-offset
hj
= 0.01 ft, h f = 95 ft, h0 = 31.5 ft .
hf

66
20
α = 90° α = 80°
α = 80° 64 α = 70°
18
α = 70° α = 60°
°
α = 60 62
16
Reduction in Opening (%)
Height Containment (%)

60
14

12 58

10 56

8 54

6 52

4 50

2 48

0 46
0.015 0.016 0.017 0.018 0.019 0.02 0.021 0.022 26 27 28 29 30 31 32
Time (sec) Fracture Length (ft)

Fig.
 8—Height containment and
 opening reduction for Shale with Q = 40 bpm, K = 0.2538 Large-offset
hj
= 0.04 ft, h f = 25 ft, h0 = 22.5 ft .
hf

This causes a fracture opening reduction (or pinching) at the fracture offset. This is defined as the percentage difference in the opening
at the fracture offset as compared to the straight fracture (see Figure 2).
wst − w j
% Width Reduction = × 100, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (3)
wst
where w j is measured at the middle of the fracture offset i.e. at offset distance h j /2 and wst is measured at a distance of h f + h j /2
from the injection point. For the case of shale with Q = 40 bpm and small ratio h j /h f , the percentage reduction in fracture opening
at the fracture offset is plotted in Figure 4 (right). The fracture opening is considerably reduced at the fracture offset. As soon as the
fracture enters the fracture offset, the fracture opening in the offset decreases very quickly, see Figure 4 (right). After the fracture has
propagated ahead of the fracture offset, the fracture opening keeps increasing in the fracture offset with nearly constant rate. Thus
the percentage opening reduction in the fracture offset remains constant. It is worthwhile to remind the reader that in all the test cases
we have considered a hydrostatic state of confining stresses, i.e. σv = σh . If we consider a deviatoric state of confining stresses with
σv > σh , the rate of increase in fracture opening at the offset will be even lower. It can be observed in all the test cases that there is a
huge influence of fracture offset angle on the fracture opening reduction at the fracture offset. Generally, there is a 10 to 15 percent
reduction in fracture opening for each 10◦ increase in fracture offset angle α. There is no plot of fracture opening reduction for the
larger offset angles in some cases. This is due to the fact that the fracture was not able to propagate through the fracture offset in
those cases.
Next we investigate the parameter space given in Tables 1 and 2 to find out the effect of each parameter on fracture height growth
and the fracture opening reduction. Each investigation covers all the fracture offset angles. In Figures 4 to 11 the fracture height
containment is given as a function of time (figures on the left side) and the opening reduction as a function of fracture length (figures
on the right side). For the sake of comparison the fracture height containment is considered at the time where the difference between
the values is maximum for different offset angles. For example, the fracture height containment for the case in Figure 4 is computed
at the time 0.1267 sec. Similarly, the opening reduction is considered at the fracture length where the difference between the values
SPE 168622 7

70
α = 90° α = 80°
16 °
α = 80 α = 70°
α = 70° 60 α = 60°
°
14 α = 60

Reduction in Opening (%)

Height Containment (%)

50
12

10 40

8
30

20
4

10
2

0 0
0.011 0.012 0.013 0.014 0.015 0.016 0.017 0.018 25 26 27 28 29 30 31 32
Time (sec) Fracture Length (ft)

Fig.
 9—Height containment and opening reduction for Limestone with Q = 40 bpm, K = 0.0527 Large-offset
hj
= 0.04 ft, h f = 25 ft, h0 = 22.5 ft .
hf

14 60
α = 90° α = 70°
α = 80° α = 60°
°
12 α = 70 58
°
α = 60
Reduction in Opening (%)
Height Containment (%)

10 56

8 54

6 52

4 50

2 48

0 46
0.2 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29 0.3 26.5 27 27.5 28 28.5 29 29.5
Time (sec) Fracture Length (ft)

Fig.
 10—Height containment and
 opening reduction for Shale with Q = 1 bpm, K = 0.6384 Large-offset
hj
= 0.04 ft, h f = 25 ft, h0 = 22.5 ft .
hf

is maximum for different offset angles. For example, fracture offset width reduction for the case in Figure 4 is computed at the
fracture length of 109.7 ft. As the properties and the confining stresses are based on the real data, there is a change in the confining
stresses with the material properties and vice versa. This makes it difficult to distinguish between the effect of material parameters
and the confining stresses on the fracture height containment and the fracture opening reduction. In order to quantify these effects,
we have carried out two investigations by manipulating the material properties and the corresponding confining stresses. All the
other parameters i.e. Q = 40 bpm, h j /h f = 0.04 ft, h f = 25 ft, h0 = 22.5 ft are kept the same in these investigations. In the first
investigation, the hydraulic fracture is propagated through the offsets of different offset angles for two two different materials: a)
shale and b) limestone. The confining stresses are kept in a hydrostatic state with the magnitude equal to 6474 psi. The results in
Figure 12 (left) show that there is more height containment for the material with larger moduli. In contrast to that there is not a large
influence of the material parameters on the fracture opening reduction, see Figure 12 (right). In the case of 80◦ offset angle, fracture
was not able to propagate through the 80◦ offset.
In the second investigation, the hydraulic fracture is propagated through the offsets of different offset angles for two different cases
of confining stresses: a) lower confining stress of 6063 psi and b) higher confining stress of 6474 psi. In each case the confining
stresses are kept in a hydrostatic state of stress. The rock material is kept as the limestone for both cases. It is obvious from the results
in Figure 13 that there is more height containment for the higher confining stresses. In contrast to that there is not a large influence
of the confining stresses on the fracture opening reduction. This behavior is very much similar to the effect of material parameters.
Now we investigate the effect of each parameter in Table 2 separately. Let us first consider the effect of fluid injection rate on
the fracture height growth and the fracture opening reduction for the case of large fracture offset i.e. h j /h f = 0.04 ft, h f = 25 ft,
h0 = 22.5 ft. It can be seen in Figure 14 (left) that there is more height containment in the case of low flow rate as compared to the
high flow rate. The difference is more pronounced in the case of shale as compared to the case of limestone. The fractures with 90◦
offset angle and the fracture in shale with the 80◦ offset angle in the case of low rate of pumping were not able to propagate through
the fracture offset so their results cannot be compared. The fracture opening reduction can be seen in Figure 14 (right). There is no
clear trend in the fracture opening reduction when the high rate of injection is compared with the low rate of injection. The difference
8 SPE 168622

100
α = 90° α = 90°
18
°
α = 80 α = 80°
90
16 α = 70° α = 70°
°
α = 60 α = 60°
80
14

Reduction in Opening (%)

Height Containment (%)

70
12

60
10

50
8

6 40

4 30

2 20

0 10
0.19 0.2 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29 25 26 27 28 29 30 31 32
Time (sec) Fracture Length (ft)

Fig.
 11—Height containment and  opening reduction for Limestone with Q = 1 bpm, K = 0.1325 Large-offset
hj
= 0.04 ft, h f = 25 ft, h0 = 22.5 ft .
hf

18 70
Shale (Large Offset, High Rate, K = 0.2538) Shale (Large Offset, High Rate, K = 0.2538)
Limestone (Large Offset, High Rate, K = 0.0527) Limestone (Large Offset, High Rate, K = 0.0527)
16
60

14
Reduction in Opening (%)
Height Containment (%)

50
12

40
10

8
30

6
20

10
2

0 0
60 70 80 60 70 80
α (deg) α (deg)

hj
Fig. 12—Comparing the rock material for the same confining stress σv = σh = 6474 psi, Q = 40 bpm, = 0.04 ft, h f = 25 ft, h0 = 22.5 ft.
hf

in opening reduction for each material is within 2 to 5 percent range for each offset angle. In the case of small fracture offset in Figure
15 (right), same trends can be seen. There seems to be no effect of the pumping rate on the reduction in fracture offset opening. In
the case of limestone with the low injection rate, there is relatively more height containment as compared to the same material and
the higher injection rate.
The dimensionless fracture toughness K for all the test cases considered thus far is less than 3. This indicates the fracture propa-
gation in the viscosity dominated regime. Let us also investigate the cases where the fracture propagation is dominated by the rock
fracture toughness, i.e. the case of slow pressurization. In Figure 16 (right), it can be seen that, in the case of slow pressurization, the
fracture velocity decreases considerably at the fracture offset. There is a sudden peak in height containment due to this reduction in
fracture velocity, see Figure 17 (left). Once the fracture passes through the fracture offset the fracture velocity increases which results
in a decrease in fracture height containment. There is no clear trend in the fracture opening reduction, see Figure 17 (right) but the
fracture opening is reduced by 44 to 54 percent in both the cases of slow and fast pressurization. The injection rate for the toughness
dominated propagation is so low that it is not practicle to have this injection rate during the hydraulic fracture stimulation. Even for
this practically very low injection rate, the conclusion remains the same i.e., there is more height containment in the case of low flow
rate as compared to the high flow rate. There is no clear trend in the fracture opening reduction when the high rate of injection is
compared with the low rate of injection.
Let us now investigate the effect of the ratio h j /h f on fracture height growth and the pinching effect at the fracture offset. There is
a clear trend which can be seen in Figures 18 and 19 which indicates that there is more height containment in the case of large fracture
offsets as compared to the small fracture offsets. In terms of the pinching effect at the fracture offset, it seems that the pinching effect
is more pronounced in the small offsets as compared to the large offsets with the exception of 60◦ offset angle. But even in that
case the difference in the pinching effect at the fracture offset for small and the large offset is less than 5 percent. The trend is more
obvious for larger fracture offset angles that the pinching effect is more dominant in the case of smaller fracture offsets as compared
to the larger fracture offsets.
SPE 168622 9

18 70
Limestone (Low σv and σh) Limestone (Low σv and σh)
Limestone (High σv and σh) Limestone (High σv and σh)
16
60

14

Reduction in Opening (%)

Height Containment (%)

50
12

40
10

8
30

6
20

10
2

0 0
60 70 80 60 70 80
α (deg) α (deg)

hj
Fig. 13—Comparing the confining stress for the Limestone Q = 40 bpm, K = 0.0527, = 0.04 ft, h f = 25 ft, h0 = 22.5 ft.
hf
The effect of injection rate Q in the case of toughness dominated fracture propagation
20 80
Limestone (High Rate, K = 0.0527) Limestone (High Rate, K = 0.0527)
Limestone (Low Rate, K = 0.1325) Limestone (Low Rate, K = 0.1325)
18 Shale (High Rate, K = 0.2538) Shale (High Rate, K = 0.2538)
70
Shale (Low Rate, K = 0.6384) Shale (Low Rate, K = 0.6384)

16
60
Reduction in Opening (%)
Height Containment (%)

14

50
12

10 40

8
30

6
20
4

10
2

0 0
60 70 80 60 70 80
α (deg) α (deg)

hj
Fig. 14—The effect of injection rate Q in the case of large offset = 0.04 ft, h f = 25 ft, h0 = 22.5 ft.
hf

Conclusions

In this contribution we have applied a newly developed hydraulic fracture simulation code based on the eXtended Finite Element
Method (XFEM) to a practical problem of hydraulic fracture propagation through fracture offsets of given length and angles. We
have carried out a detailed parametric study of the problem of fracture offsets in order to evaluate the effect of each parameter on the
fracture height growth and fracture opening reduction at the fracture offset. The geometric parameters of the problem include fracture
offset angle,and the ratio of the fracture offset length to the length of the straight fracture. In addition to the geometric parameters,
the effects of different material parameters, confining stresses and the fluid injection rate are considered. Among all the parameters
considered, it was observed that the fracture offset angle has the most profound effect on fracture height growth and fracture opening
reduction at the fracture offset. The same dependence of opening reduction on the offset angle was observed by Jeffrey et al. (2009,
a) and Jeffrey et al. (2009, b) but the authors have attributed that to the difference in the confining stress normal to the main fracture
and the fracture offset. In this contribution we have kept a hydrostatic state of confining stresses, in spite of that we have observed
a strong dependence of height containment and fracture opening on the offset angle. This result shows that this dependence on the
offset angle is actually an effect of the considered geometry which will be magnified in the case of confining stresses with a deviatoric
component.
The results of the parametric study can be summarized as follows:
1. Height containment:

• Higher offset angle α results in more height containment.

• Larger material modulus results in more height containment.
• Higher confining stress results in more height containment (for the considered injection parameters).
• Lower injection rate results in more height containment. This is valid for the toughness dominated propagation as well as
for the propagation in the viscosity dominated regime.
10 SPE 168622

14 90
Limestone (High Rate, K = 0.0527) Limestone (High Rate, K = 0.0527)
Limestone (Low Rate, K = 0.1325) Limestone (Low Rate, K = 0.1325)
Shale (High Rate, K = 0.2538) 80 Shale (High Rate, K = 0.2538)
12 Shale (Low Rate, K = 0.6384) Shale (Low Rate, K = 0.6384)

70

Reduction in Opening (%)

Height Containment (%)

10
60

8
50

40
6

30
4

20

2
10

0 0
60 70 80 60 70 80
α (deg) α (deg)

hj
Fig. 15—The effect of injection rate Q in the case of small offset = 0.01 ft, h f = 95 ft, h0 = 31.5 ft.
hf

6180 80
Q = 0.005 BPM, K = 2.4 Q = 0.005 BPM, K = 2.4
Q = 0.001 BPM, K = 3.6 Q = 0.001 BPM, K = 3.6
6170
70
6160 Fracture Height (ft)
Inlet Pressure (psi)

6150 60

6140
50
6130

6120 40

6110
30
6100

6090 20
0 50 100 150 200 250 300 350 400 450 0 50 100 150 200 250 300 350 400 450
Time (sec) Time (sec)

hj
Fig. 16—Pressure response and fracture height growth in the case of slow pressurization α = 60◦ , = 0.04 ft, h f = 25 ft, h0 = 22.5 ft.
hf

40 54
Q = 40 BPM, K = 0.2538 Q = 40 BPM, K = 0.2538
Q = 1 BPM, K = 0.6384 Q = 1 BPM, K = 0.6384
Q = 0.005 BPM, K = 2.4 52 Q = 0.005 BPM, K = 2.4
35
Q = 0.001 BPM, K = 3.6 Q = 0.001 BPM, K = 3.6

50
30
Reduction in Opening (%)
Height Containment (%)

48
25

46
20

44

15
42

10
40

5
38

0 36
−1 0 1 2
10 10 10 10 25 30 35 40 45 50 55
Time (sec) Fracture Length (ft)

hj
Fig. 17—The effect of injection rate Q in the case of toughness dominated fracture propagation α = 60◦ , = 0.04 ft, h f = 25 ft, h0 = 22.5 ft.
hf

hj
• Higher ratio of results in more height containment.
hf

2. Reduction in fracture opening:

• Higher offset angle α results in more reduction in fracture opening at the fracture offset.
 SPE 168622 11

18 90
Limestone (Large Offset, K = 0.1325) Limestone (Large Offset, K = 0.1325)
Limestone (Small Offset, K = 0.1325) Limestone (Small Offset, K = 0.1325)
16 Shale (Large Offset, K = 0.6384) 80 Shale (Large Offset, K = 0.6384)
Shale (Small Offset, K = 0.6384) Shale (Small Offset, K = 0.6384)

14 70

Reduction in Opening (%)

Height Containment (%)

12 60

10 50

8 40

6 30

4 20

2 10

0 0
60 70 80 60 70 80
α (deg) α (deg)

hj
Fig. 18—The effect of fracture offset ratio in the case of low injection rate Q = 1 bpm.
hf

20 90
Limestone (Large Offset, K = 0.0527) Limestone (Large Offset, K = 0.0527)
Limestone (Small Offset, K = 0.0527) Limestone (Small Offset, K = 0.0527)
18 Shale (Large Offset, K = 0.2538) 80 Shale (Large Offset, K = 0.2538)
Shale (Small Offset, K = 0.2538) Shale (Small Offset, K = 0.2538)

16
Reduction in Opening (%) 70
Height Containment (%)

14
60

12
50
10
40
8

30
6

20
4

2 10

0 0
60 70 80 60 70 80
α (deg) α (deg)

hj
Fig. 19—The effect of fracture offset ratio in the case of high injection rate Q = 40 bpm.
hf

• There is not an appreciable effect of material moduli on reduction in fracture opening at the fracture offset.
• There is not an appreciable effect of confining stress on reduction in fracture opening at the fracture offset.
• There is not an appreciable effect of injection rate on reduction in fracture opening at the fracture offset. The fracture
opening is reduced by 45 to 55 percent for the toughness dominated propagation as well as for the viscosity dominated
propagation.
hj hj
• Generally, lower ratio of results in more reduction in fracture opening. For small offset angles, higher ratio of
hf hf
results in more reduction in fracture opening.
This study concludes that while each parameter affects the fracture height containment, the fracture opening reduction is dominated
by the fracture offset angle which is a geometric effect of the fracture offsets. These fracture offsets generally develop at the weak
interfaces between the bed boundaries. In this study we have not considered the effect of deviatoric confining stresses and the
frictional interfaces. In the absence of frictional interfaces all the energy used to open the main fracture is consumed to shear the
fracture offset. In reality, some of the energy used to open the main fracture is dissipated at the frictional interface. Some of the fluid
will also leak into the weak interfaces, this eventually further reduces the fracture opening at the fracture offset. In this way, this study
shows the lower bound in terms of the fracture height containment and opening reduction due to fracture offsets. Any consideration
of the confining stresses with a deviatoric component and frictional interface is going to magnify this effect.

References

Adachi, J. and E. Detournay (2002). Self-similar solution of a plane-strain fracture driven by a power-law fluid. International Journal
for Numerical and Analytical Methods in Geomechanics 26(6), 579–604.
Adachi, J. I., E. Detournay, and A. P. Peirce (2010). Analysis of the classical pseudo-3d model for hydraulic fracture with equilibrium
height growth across stress barriers. International Journal of Rock Mechanics and Mining Sciences 47(4), 625–639.
 12 SPE 168622

Athavale, A. and J. Miskimins (2008). Laboratory hydraulic fracturing tests on small homogeneous and laminated blocks. In The 42nd
U.S. Rock Mechanics Symposium (USRMS), June 29 - July 2, 2008 , San Francisco, CA, Number 08-067. American Rock Mechanics
Association.
Barree, R. and P. Winterfeld (1998). Effects of shear planes and interfacial slippage on fracture growth and treating pressures. In SPE
Annual Technical Conference and Exhibition, 27-30 September 1998, New Orleans, Louisiana, Number 48926. Society of Petroleum
Engineers.
Belytschko, T. and T. Black (1999). Elastic crack growth in finite elements with minimal remeshing. Internat. J. Numer. Methods
Engrg. 45, 601 – 620.
Beugelsdijk, L., C. dePater, and K. Sato (2000). Experimental hydraulic fracture propagation in a multi-fractured medium. In SPE Asia
Pacific Conference on Integrated Modelling for Asset Management, 25-26 April 2000, Yokohama, Japan, Number 59419. Society of
Petroleum Engineers.
Boyer II, C. M., P. Stubbs, and F. Scherer (1986). Measurement of coalbed properties for hydraulic fracture design and methane
production. In SPE Unconventional Gas Technology Symposium, Number SPE 15258. Society of Petroleum Engineers. SPE Un-
conventional Gas Technology Symposium, 18-21 May 1986, Louisville, Kentucky.
Chudnovsky, A., J. Fan, Y. Shulkin, J. W. Dudley, W. B. Nichols, and G. K. Wong (2001). Hydraulic fracture containment in layered
media, experiment and computer simulation. In 38th U.S. Symposium on Rock Mechanics (USRMS), Number ARMA 01-0273.
American Rock Mechanics Association.
Chuprakov, D., O. Melchaeva, and R. Prioul (2013a). Effective and Sustainable Hydraulic Fracturing, Chapter 8: Hydraulic Fracture
Propagation Across a Weak Discontinuity Controlled by Fluid Injection, pp. 157–182. InTech.
Chuprakov, D., O. Melchaeva, and R. Prioul (2013b). Injection-sensitive mechanics of hydraulic fracture interaction with discontinu-
ities, ARMA 13-252. In 47th US Rock Mechanics Geomechanics Symposium. American Rock Mechanics Association.
Detournay, E. (2004). Propagation regimes of fluid-driven fractures in impermeable rocks. International Journal of Geomechan-
ics 4(1), 35–45.
Fisher, M. K. and N. R. Warpinski (2012). SPE Production & Operations 27(1), 8–19.
Fries, T. and T. Belytschko (2010). The extended/generalized finite element method: An overview of the method and its applications.
Internat. J. Numer. Methods Engrg. 84, 253 – 304.
Gordeliy, E. and A. Peirce (2013a). Coupling schemes for modeling hydraulic fracture propagation using the XFEM . Computer
Methods in Applied Mechanics and Engineering 253(0), 305 – 322.
Gordeliy, E. and A. Peirce (2013b). Implicit level set schemes for modeling hydraulic fractures using the XFEM. Computer Methods
in Applied Mechanics and Engineering 266(0), 125 – 143.
Gu, H. and E. Siebrits (2008). Effect of formation modulus contrast on hydraulic fracture height containment. SPE Production &
Operations 23(2), 170–176.
Hanson, M. E., R. J. Shaffer, and G. D. Anderson (1981). Effects of various parameters on hydraulic fracturing geometry. SPE
Journal 21(4), 435–443.
Jeffrey, R. and A. Bunger (2009). A detailed comparison of experimental and numerical data on hydraulic fracture height growth
through stress contrasts. SPE Journal 14(3), 413–422.
Jeffrey, R., A. Bunger, B. Lecampion, X. Zhang, Z. Chen, A. van As, D. Allison, W. De Beer, J. Dudley, E. Siebrits, et al. (2009).
Measuring Hydraulic Fracture Growth in Naturally Fractured Rock. In SPE Annual Technical Conference and Exhibition, pp. SPE
124919.
Jeffrey, R., A. Bunger, B. Lecampion, X. Zhang, Z. Chen, A. van As, D. Allison, W. de Beer, J. Dudley, E. Siebrits, M. Thiercelin,
and M. Mainguy (2009). Measuring hydraulic fracture growth in naturally fractured rock. In SPE Annual Technical Conference,
Number SPE 124919. Society of Petroleum Engineers.
Jeffrey, R., X. Zhang, and M. Thiercelin (2009). Hydraulic fracture offsetting in naturally fractured reservoirs: Quantifying a long-
recognized process. In SPE Hydraulic Fracturing Technology Conference, Number SPE 119351. Society of Petroleum Engineers.
Kresse, O., X. Weng, D. Chuprakov, R. Prioul, and C. Cohen (2013). Effective and Sustainable Hydraulic Fracturing, Chapter 9:
Effect of Flow Rate and Viscosity on Complex Fracture Development in UFM Model, pp. 183–210. InTech.
Miskimins, J. L. and R. D. Barree (2003). Modeling of hydraulic fracture height containment in laminated sand and shale sequences. In
SPE Production and Operations Symposium, 22-25 March 2003, Oklahoma City, Oklahoma, Number 80935. Society of Petroleum
Engineers.
 SPE 168622 13

Moës, N., J. Dolbow, and T. Belytschko (1999). A finite element method for crack growth without remeshing. Internat. J. Numer.
Methods Engrg. 46, 131 – 150.
Palmer, I. D. and H. Carroll Jr. (1983). Three-dimensional hydraulic fracture propagation in the presence of stress variations. SPE
Journal 23(6), 870–878.
Peirce, A. P. and E. Detournay (2008, June). An implicit level set method for modeling hydraulically driven fractures. Computer
Methods in Applied Mechanics and Engineering 197(33-40), 2858–2885.
Teufel, L. W. and J. A. Clark (1984). Hydraulic fracture propagation in layered rock: Experimental studies of fracture containment.
SPE Journal 24(1), 19–32.
Thiercelin, M., J. Roegiers, T. Boone, and A. Ingraffea (1987). An investigation of the material parameters that govern the behavior
of fractures approaching rock interfaces. In 6th ISRM Congress, August 30 - September 3, 1987 , Montreal, Canada, Number
6CONGRESS-1987-049. International Society for Rock Mechanics.
Van Eekelen, H. (1982). Hydraulic fracture geometry: Fracture containment in layered formations. SPE Journal 22(3), 341–349.
Warpinski, N. R. (1991). Hydraulic fracturing in tight, fissured media. Journal of Petroleum Technology 43(2), 146–151, 208–209.
Wu, H., A. Chudnovsky, J. Dudley, and G. Wong (2004). A map of fracture behavior in the vicinity of an interface. In Gulf Rocks
2004, the 6th North America Rock Mechanics Symposium (NARMS), June 5 - 9, 2004 , Houston, Texas, Number 04-620. American
Rock Mechanics Association.
Zhang, X. and R. G. Jeffrey (2008). Reinitiation or termination of fluid-driven fractures at frictional bedding interfaces. Journal of
Geophysical Research-Solid Earth 113(B8).
Zhang, X., R. G. Jeffrey, and M. Thiercelin (2007). Deflection and propagation of fluid-driven fractures at frictional bedding interfaces:
A numerical investigation. Journal of Structural Geology 29(3), 396–410.
Zilian, A. and T. Fries (2009). A localized mixed-hybrid method for imposing interfacial constraints in the extended finite element
method (XFEM). Internat. J. Numer. Methods Engrg., DOI: 10.1002/nme.2596.