1. INTRODUCTION

Although hydraulic fracturing has been widely used for

decades (Roberts, 1866, Bugbee, 1942, Clark, 1949), and

the technology to implement and interpret the induced

fractures has been continuously evolving (Yost, 1988,

NRC, 2001, Fri, 2006, NETL, 2011, Trembath et al.,

2012, Saldungaray & Palish, 2012), many aspects are still

not understood. Specifically, this includes the fracture

geometry and its interaction with natural features such as

existing fractures, bedding planes, and various

heterogeneities. The objective of this research is to gain a

fundamental understanding of the hydraulic fracturing

processes in shales through controlled laboratory

experiments, in which the mechanisms underlying

fracture initiation, propagation, and interaction with

geologic features in the rock are visually captured and

analyzed. Once these fundamental processes are properly

understood, methods that allow one to induce desired

fracture geometries in reservoirs can be developed.

Extensive work has been done at MIT to study fracture

initiation, -propagation, and coalescence (Reyes, 1991,

Bobet, 1997, Wong, 2008, Miller, 2008, Morgan, 2015,

Gonçalves da Silva, 2016, AlDajani, 2017, AlDajani,

2018, Gonçalves da Silva, 2018). These studies were done

on prismatic specimens with two pre-existing artificial

fractures (flaws) without and with the influence of

hydraulic pressure (Figure 1). Specimens were subjected

to uniaxial or biaxial compressive loading, and fracture

initiation and propagation mechanisms (tensile & shear)

were captured using a high-speed camera and a high-

resolution camera while simultaneously measuring the

stress-strain behavior. These experiments were

conducted on different materials: gypsum (artificial

material), different marbles (metamorphic rock), granite

(igneous rock), and different shales (sedimentary rock).

Figure 1 – Testing progression to study the fracture initiation,

propagation, and coalescence in rocks under various loading

conditions. a) Uniaxial or biaxial loading to failure. b)

Constant uniaxial load and pressurizing flaw to failure. c)

Constant biaxial load and pressurizing flaw to failure.

Modified from Gonçalves da Silva (2016).

ARMA 19–1552

Isotropic versus Anisotropic Stress Field Effects on

Hydraulic Fracture Mechanisms in Opalinus Shale

AlDajani, O.A.1, Germaine, J.T.2, & Einstein, H.H.1 1Massachusetts Institute of Technology, Cambridge, Massachusetts, USA 2Tufts University, Medford, Massachusetts

Copyright 2019 ARMA, American Rock Mechanics Association

This paper was prepared for presentation at the 53rd US Rock Mechanics/Geomechanics Symposium held in New York, NY, USA, 23–26 June 2019. This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and critical review of the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any position of ARMA, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of ARMA is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 200 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgement of where and by whom the paper was presented.

ABSTRACT: Although hydraulic fracturing has been widely used for decades, and the technology to implement and interpret the induced fractures has been continuously evolving, many aspects are still not understood. Specifically, this includes hydraulic

fracture initiation and propagation mechanisms and the effect of stress state and rock fabric. The objective of this study was to

determine the differences between hydraulic fracturing under isotropic and anisotropic stress conditions.

The rock used in this study is Opalinus Shale prepared into prismatic specimens with a pre-existing artificial fracture (flaw) in the

middle. Different external biaxial stresses are applied to simulate in-situ stress conditions followed by hydraulic pressurization of

the flaw until failure. Internal flaw pressure is measured throughout the pressurization and fracturing process. High-speed and high-

resolution cameras are used for visual analysis.

Two experiments are presented, discussed in detail and compared: 1- a specimen with a bedding plane orientation of 30° relative to

horizontal is subjected to a vertical stress of 3 MPa and a lateral stress of 1 MPa (anisotropic stress). 2- a specimen with the same

bedding plane orientation of 30° is subjected to biaxial isotropic stresses of 2 MPa (isotropic stress). The results show that the

combination of rock fabric and stress state affect the initiation and propagation of hydraulic fractures in shale. This adds to

fundamental knowledge on how fractures behave and may provide insight into strategic hydraulic fracture treatments for field

applications.

Hydraulic fracture geometries and mechanisms are

affected by stress state (Perkins & Kern, 1961, Simonson

et al., 1978, Cleary, 1980, Warpinski et al., 1982), rock

fabric (Fisher & Warpinski, 2012, Suarez-Rivera et al.,

2013), and other factors (Daneshy, 1978, Teufel & Clark,

1981, Biot et al., 1983, Blaire et al., 1989). In any given

petroleum reservoir, the stress state can vary spatially,

even along a single wellbore, due to the complex geologic

structures such as salt domes and/or folds (Zoback, 2010).

Rock fabric, i.e. bedding plane orientation, natural

fractures, and localized heterogeneities, affect the local

stress field and may influence the propagation of the

hydraulic fractures, and thus play an important role in

dictating the complexity of the induced fractures. In this

paper, we specifically investigate the effect of the biaxial

(quasi true triaxial) stress regime on the produced

hydraulic fractures in shale. The two stress states

investigated are isotropic stress (𝜎1 = 𝜎3) and anisotropic stress (𝜎1 ≠ 𝜎3).

1.1. Specimen Mineralogy

The rock used in this study is Opalinus Shale from Mont

Terri, Switzerland, which often has distinct alternating

layers as shown in Figure 2.

Figure 2 - Image of Opalinus Shale showing two distinct

alternating layers, a dark clay-rich layer & a light quartz- and

carbonate- rich layer.

The mineralogy was measured using X-ray diffraction

and is presented in Table 1.

Table 1 – Bulk mineralogy analysis results of Opalinus Shale

core sample from X-ray diffraction.

Mineral %

Quartz 33.0

K-Feldspar 3.4

Plagioclase 1.8

Calcite 5.2

Dolomite 0.7

Siderite 1.8

Anatase 0.4

Pyrite 0.9

Muscovite 2.6

Chlorite (Tri) 2.3

I+I/S-ML 32.1

Kaolinite 15.8

1.2. Mechanical Properties

The mechanical properties were measured through

unconfined compression tests and are presented in Table

2. Although these results were for other Opalinus Shale

cores, they fall within range of the extensive mechanical

properties tested and presented by Bock (2001).

Table 2 – Mechanical properties of intact (no flaw) Opalinus

Shale prismatic specimens subjected to unconfined

compression tests (AlDajani, 2017). UCS, MPa E, MPa ʋ

load ⊥ to bedding 17.26 1327 0.33

load ∥ to bedding 5.76 1947 0.26

1.3. Specimen Preparation

Prismatic specimens are prepared by dry cutting cored

borings with various bedding plane orientations and a pre-

cutting a flaw in the middle. The intricate preparation

techniques used are described in AlDajani (2017) and are

meant to preserve the shale’s chemical and mechanical

integrity from in-situ conditions until testing.

In the previously mentioned studies, most experiments

were conducted on specimens with double flaws to

determine their interaction. In this study, only a single

vertical flaw is cut to investigate the effect the stress state

has on hydraulic fractures emanating from a single source,

and observe the interaction of the produced hydraulic

fractures with the rock fabric. The specimen dimensions

and loading configuration are shown in Figure 3. Note

that the specimens used in this study have a bedding plane

orientation of 30° from horizontal.

Figure 3 – Schematic of a prismatic specimen subjected to a

constant biaxial load and a pressurized prefabricated flaw to

induce hydraulic fractures to study fracture mechanisms.

2. EXPERIMENTAL SETUP

The novel and unique hydraulic fracture experimental

setup introduced by Morgan et al. (2017) and described in

detail by AlDajani (2018) is shown in Figure 4.

2.1. Testing Setup & Procedure

The main challenge and advantage of this setup was the

ability to induce hydraulic fractures in an externally

loaded rock specimen and be able to capture the fracture

mechanisms in detail. This involved the flaw

pressurization device (Figure 5).

Figure 5 – Three-dimensional rendering of updated flaw

pressurization device components (isometric view) showing

transparent polycarbonate window and larger flaw seal with

front injection needle inserted into the flaw.

This device, its components, and operation were

described in detail by AlDajani (2018). When placed into

the load frame, it allows one to apply external stresses

onto the specimen and pressurize the flaw to produce

hydraulic fractures while simultaneously capturing visual

and acoustic observations of the detailed fracture

mechanisms.

After applying the external biaxial stresses, the device

allows one to apply stress on the specimen face by

clamping the front viewing window to the rear steel plate,

i.e. creating a true triaxial stress state around the flaw.

This stress (𝜎2) was determined to be approximately 2 MPa by controlling the clamping screws’ torque and

measuring the force on a load cell in place of the

specimen. The same torque, and hence intermediate

stress, was uniformly applied for all experiments.

Computer control is used to apply loads at specified rates

in a synchronized fashion. Specifically:

• In the isotropic test, 2 MPa are simultaneously applied in the axial and lateral directions at a rate of

3.5 MPa/min, following an isotropic compressional

stress path, and then to hold 2 MPa biaxially.

• In the anisotropic test, 1 MPa was applied simultaneously in the axial and lateral directions at a

rate of 3.5 MPa/min, initially following an isotropic

compressional stress path. After the lateral stress was

held at 1 MPa, the axial stress was increased at the

same rate until 3 MPa is achieved.

Figure 4 – Schematic of the experimental setup used in this study. The load frame subjects a constant biaxial stress on to the specimen

and then the pressure volume actuator (PVA) injects fluid into the pre-cut flaw. The fractures are observed using a high-speed

camera, a high-resolution camera, and an acoustic emission system, simultaneously. The fluid pressures were measured in the PVA

as well as internally in the flaw. Modified from Morgan et al. (2017).

The stress paths for these two tests are shown in p-q space,

where 𝑝 =1

2(𝜎1 + 𝜎3) and 𝑞 =

1

2(𝜎1 − 𝜎3).

Figure 6 – Stress paths for the two tests in p-q space where 𝜎1 is the axial load and 𝜎3 is the lateral load.

After loads are applied, the flaw is initially saturated by

pumping fluid from the pressure-volume-actuator (PVA)

through the tubing into the flaw (see Figure 4), and out

through the flaw pressure measurement needle (see

Figure 5). Once saturated, a pressure transducer is

attached to this needle to close the system. The flaw is

then pressurized at a constant injection rate of 1.33

mL/min for both tests. Pressure and volume

measurements are taken at the PVA, and internal flaw

pressure is measured with the pressure transducer probing

the flaw. The fluid injected is hydraulic oil to prevent the

hydration of clays, and has a dynamic viscosity of

approximately 4 cP.

Imagery acquisition was done with a high-speed (HS)

camera at 1,000 frames per second (fps) on a 1 megapixel

(MP) sensor and a high-resolution (HR) camera at 0.5 fps

on a 20 MP sensor. The HS camera is manually triggered

to capture the failure of the specimen, i.e. the end of the

test. The HR camera captures time lapses of the test from

beginning to end. The acoustic acquisition system

samples data at 5 MHz from 8 acoustic sensors, which are

spring-loaded in specialized platens surrounding the

specimen. Acoustic observations are not discussed in this

paper.

To show practical relevance, the concept of this

experiment is similar to bringing the rock to in-situ stress

conditions, the saturation phase is analogous to drilling

mud in the wellbore, and the pumping phase follows what

is done in field operations to induce hydraulic fractures.

In our tests, the following data are acquired to be

analyzed:

a. HS video b. HR images c. Internal flaw pressure d. PVA pressure and volume e. Acoustic emissions (not discussed in this paper)

The imagery is then analyzed, and what is visually

captured is analyzed and drawn into sketches for a clearer

graphical representation of what happened.

3. RESULTS AND DISCUSSION

Recall that two stress states investigated: anisotropic and

isotropic (Figure 7).

Figure 7 – Schematics of tested load configurations. a)

Anisotropic stress state acting on a specimen with 30° bedding

planes. b) Isotropic stress state acting on a specimen with 30°

bedding planes.

3.1. Anisotropic Stress State (𝜎1 = 3 𝑀𝑃𝑎, 𝜎2 ≈2 𝑀𝑃𝑎, 𝜎3 = 1 𝑀𝑃𝑎)

The data collected from the anisotropic hydraulic

fracture experiment are shown in Figure 8. The

internal flaw pressure is the red curve, and the

injected volume is the blue curve. The black

triangles indicate where sketches were taken for

image analysis. A sketch is taken from the HS or

HR images when a significant event occurs such as

fracture initiation or a specific interaction of the

hydraulic fracture with features in the rock.

Figure 8 – Pressure and volume data acquired for the

anisotropic stress hydraulic fracture test. The internal flaw

pressure is the red curve, and the injected volume is the blue

curve. The black triangles indicate where sketches were taken

for image analysis.

The fracture progression is shown in Figure 9,

where each sketch number corresponds to the image

of the specimen at the denoted

pressures/times/volumes in Figure 8 (designated by

the labeled black triangles).

Figure 9 – Sketches of fracture progression throughout

pressurization of the flaw in the anisotropically loaded

specimen. Sketch numbers refer to numbered black triangles on

pressure curve in Figure 8. Fracture initiation is denoted with

red letters. The flaw seal boundary is indicated by the rounded

square. (T) indicates opening in tension, and subscript bp

indicates propagation along a bedding plane.

The final sketch (Sketch 10) is enlarged and shown

in Figure 10.

Figure 10 – Final sketch (Sketch 10 in Figure 8) of the

anisotropic stress hydraulic fracture test.

As shown in Figure 8, fluid is injected at a constant

flow rate, and the internal flaw pressure response is

measured. The maximum pressure was 4.32 MPa,

but the first fracture A(T)bp initiated at the top flaw

tip at 3.49 MPa (Sketch 1in Figure 9) and started

propagating along the bedding plane. A second

fracture B(T)bp initiated at a bedding plane

intersecting the middle of the flaw at 3.92 MPa

(Sketch 2) while A(T)bp continued to propagate. As

propagation of A(T)bp and B(T)bp continued, C(T)bp

branched from A(T)bp (Sketch 3). At this point,

A(T)bp arrested and propagation continued along the

intersecting bedding plane C(T)bp (Sketch 4). The

non-linear pressure response between sketches 3

and 4 reflects the dilation as a result of fracture

propagation while the drastic pressure drop shortly

afterwards is due to the fracture reaching the seal

boundary (Sketch 5). Despite that, the fractures

continued propagating until they reached the

specimen boundaries (Sketch 10 in Figure 10). The

pressure record shows that the pressure to initiate

fractures is greater than that needed to propagate

fractures, as was established by Irwin (1956) and

Feng & Gray (2017). An image taken at the end of

the test is shown in Figure 11 which corresponds to

the final sketch in Figure 10.

Figure 11 – Image of hydraulically fractured specimen

subjected to anisotropic stress at the end of the test.

3.2. Isotropic Stress State (𝜎1 = 𝜎2 ≈ 𝜎3 =2 𝑀𝑃𝑎)

The data collected from the isotropic stress hydraulic

fracture experiment are shown in Figure 12. The plotted

curve colors and symbols are the same as in Figure 8

(section 3.1.).

Figure 12 – Pressure and volume data acquired for the isotropic

stress hydraulic fracture test. The internal flaw pressure is the

red curve, and the injected volume is the blue curve. The black

triangles indicate where sketches were taken for image analysis.

As shown in Figure 12, fluid is injected at a constant rate,

and the internal flaw pressure response is measured. The

maximum pressure in this test was 7.70 MPa. It is worth

noting that no fractures were detected in the image

analysis prior to the drastic pressure drop, which is a sign

that a hydraulic fracture(s) has propagated past the seal

boundary. The likeliest explanation was that the

fracturing started on the rear face of the specimen before

becoming visible on the front (imaged) face. The front

and rear faces of the specimen were photographed after

the test and they show good correspondence (Figure 13).

The earlier fracturing at the rear is supported by the wider

wet region around the fractures on the rear face, indicating

longer exposure to the injected fluid.

Figure 13 - Images of the front and rear face (mirrored for

comparison) of the hydraulically fractured specimen taken after

the test. The fractures show good correspondence and are likely

to have started on the rear face before appearing on the front.

Regardless of this fact, the visual observations on the

imaged face of the rock throughout the test were analyzed

as this test still shows the entire fracture behavior of this

specimen under isotropic loading.

The fracture progression is shown in Figure 14, where each sketch number corresponds to the image of the

specimen at the denoted pressures/times/volumes in

Figure 12 (designated by the labeled black triangles).

Figure 14 – Sketches of fracture progression throughout

pressurization of the flaw in the isotropically loaded specimen.

Sketch numbers refer to numbered black triangles on pressure

curve in Figure 12. Fracture initiation is denoted with red

letters. The flaw seal boundary is indicated by the rounded

square. (T) indicates opening in tension, and subscript bp

indicates propagation along a bedding plane.

The final sketch (Sketch 6) is enlarged and shown in

Figure 15.

Figure 15 – Final sketch (Sketch 6 in Figure 12) of the isotropic

stress hydraulic fracture test.

The first crack to initiate is A(T)bp (Sketch 1 in Figure 14),

which is a bedding plane that opened in tension due to the

hydraulic pressure. By Sketch 2, A(T)bp had propagated

to the seal boundary. By Sketch 3, tensile fracture B(T)

initiated near the middle of the flaw and started

propagating across bedding layers, while A(T)bp had

propagated well past the seal boundary. Afterwards, the

two fractures continued propagating to the boundaries of

the specimen, with A(T)bp simply along the same bedding

plane, and B(T) across bedding layers. An image taken at

the end of the test is shown in Figure 16 which

corresponds to the final sketch in Figure 15.

Figure 16 – Image of hydraulically fractured specimen

subjected to isotropic stress at the end of the test.

3.3. Discussion First and foremost, there is one consistent behavior

among the two tested loading conditions. As the flaw is

pressurized, the first thing to occur is that a bedding plane

near the flaw tip, where the highest tensile stresses exist,

opens in tension. Propagation continues along this

bedding plane. This is strong evidence that the rock fabric

influences hydraulic fracture initiation and propagation.

Though tensile strength of the bedding planes was not

directly measured, it is evident that they are planes of

weakness in shales. In other words, the hydraulic

fracturing process is controlled by stress concentration

and fabric.

This is quite interesting, especially in the anisotropically

loaded test. Given the stress state applied, one would

expect, theoretically, the hydraulic fractures to propagate

in the direction of 𝜎1, i.e. vertically in this testing configuration. This was shown experimentally by

AlDajani (2018) where specimens with horizontal

bedding and a single vertical flaw were subjected to

uniaxial stress, which can be regarded as a special case of

anisotropic biaxial stress. These experiments all produced

hydraulic fractures propagating in the uniaxial direction.

Stress concentration is more extreme in the uniaxial case,

and the horizontal bedding planes did not provide

initiation locations. This confirms the effect of a

combination of fabric and stress concentration effect.

Other observations indicate that there is a combination of

the effects of stress concentration and rock fabric:

• The second hydraulic fracture occurs at and propagates along a bedding plane in the anisotropic

test, but propagates through the matrix in the isotropic

test. Furthermore, the second fracture in the isotropic

stress propagated horizontally with no preferential

direction.

• The maximum pressure reflects how much pressure is required to propagate a fracture to the seal boundary,

and it is significantly higher in the isotropic test.

One initial explanation is that in an anisotropic stress

field, the stress concentrations around the flaw are more

extreme than in an isotropic stress field. However, further

work is necessary to determine how stress concentration

and rock fabric interact.

4. SUMMARY & CONCLUSIONS

The objective of this paper was to study the differences

between hydraulic fractures produced in shale specimens

subjected to isotropic and anisotropic stress conditions.

The hydraulic pressure testing setup at MIT allowed real-

time analysis of fracture initiation and propagation by

utilizing a transparent flaw pressurization device and

imagery equipment.

The anisotropic stress test applied and held the following

external stresses: σ1 = 3 MPa, σ2 ≈ 2 MPa, σ3 = 1 MPa, and the flaw was pressurized at a constant flow rate of

1.33 mL/min. This resulted in hydraulic pressure opening

up bedding planes and propagating along them.

The isotropic stress test applied and held the following

external stresses: σ1 = σ2 ≈ σ3 = 2 MPa, and the flaw was pressurized at the same constant flow rate of 1.33

mL/min. This resulted in one hydraulic fracture

propagating along a bedding plane and the other across

bedding layers.

For both stress conditions, the first hydraulic fracture to

initiate was at the flaw tips where the tensile stress is

highest and at the intersection with a bedding plane. The

fracture continued to propagate along the same bedding

plane. Thus, the rock fabric has a strong effect in dictating

fracture initiation and propagation. However, the

characteristics of the secondary hydraulic fracture and the

pressures were different between the two tests.

While the combined effect of stress state and fabric may

explain the differences in hydraulic fracture initiation and

propagation, more work is needed to fully understand

their interaction.

The results from such experiments can be very insightful

to interpret fracture complexity in past field operations or

in the planning stage for future treatments, where the

stress state can vary spatially, even along the same

wellbore. They can also be used as a validation tool for

theoretical and numerical models.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the support of this

research by TOTAL in the context of the Multi-scale

Shale Gas Collaboratory project. We not only received

financial support, but were also helped through many

constructive discussions with our technical contacts. We

also would like to acknowledge the Underground

Research Laboratory in Mont Terri, Switzerland which

provided the shale used for this study. The authors

also acknowledge the help in designing and fabricating

from Stephen W. Rudolph in the Department of Civil

and Environmental Engineering at MIT. The authors also

thank their colleagues Dr. Stephen P. Morgan and Bing

Q. Li for their help running the experiments.

REFERENCES

1. AlDajani, O. A., Morgan, S. P., Germaine, J. T., & Einstein, H. H. (2017, August). Vaca Muerta Shale–Basic

Properties, Specimen Preparation, and Fracture Processes.

In 51st US Rock Mechanics/Geomechanics Symposium,

San Francisco, California, USA.

2. AlDajani, O. A., Germaine, J. T., & Einstein, H. H. 2018, August 21. Hydraulic Fracture of Opalinus Shale Under

Uniaxial Stress: Experiment Design and Preliminary

Results. In Proceedings of the 52nd US Rock

Mechanics/Geomechanics Symposium held in Seattle, WA,

USA, 17–20 June 2018. American Rock Mechanics

Association.

3. Biot, M. A., Medlin, W. L., & Masse, L. (1983). Fracture penetration through an interface. Society of Petroleum

Engineers Journal, 23(06), 857-869.

4. Blair, S. C., Thorpe, R. K., Heuze, F. E., & Shaffer, R. J. (1989, January 1). Laboratory Observations Of The Effect

Of Geologic Discontinuities On Hydrofracture

Propagation. American Rock Mechanics Association.

5. Bobet, A. (1997). Fracture coalescence in rock materials: experimental observations and numerical

predictions (Doctoral Dissertation, Massachusetts

Institute of Technology).

6. Bock, K. (2001). Rock mechanics analyses and synthesis: RA experiment. Rock mechanics analyses and synthesis:

Data report on rock mechanics, Mont Terri Technical

Report 2000-02. Brugge, Belgium: Q+S Consult.

7. Bugbee, J. M. (1943, December 1). Reservoir Analysis and Geologic Structure. Society of Petroleum Engineers.

doi:10.2118/943099-G

8. Clark, J. B. (1949, January 1). A Hydraulic Process for Increasing the Productivity of Wells. Society of Petroleum

Engineers. doi:10.2118/949001-G

9. Cleary, M. P. (1980, January). Analysis of mechanisms and procedures for producing favourable shapes of

hydraulic fractures. In SPE Annual Technical Conference

and Exhibition. Society of Petroleum Engineers.

10. Daneshy, A. A. (1978, February 1). Hydraulic Fracture Propagation in Layered Formations. Society of Petroleum

Engineers. doi:10.2118/6088-PA

11. Feng, Y., & Gray, K. E. (2017). Discussion on field injectivity tests during drilling. Rock Mechanics and Rock

Engineering, 50(2), 493-498.

12. Fri, Robert W. "From Energy Wish Lists to Technological Realities." Issues in Science and Technology 23, no. 1

(Fall 2006).

13. Gonçalves da Silva, B. M. (2016). Fracturing processes and induced seismicity due to the hydraulic fracturing of

rocks. (Doctoral dissertation, Massachusetts Institute of

Technology).

14. Gonçalves da Silva, B. M., & Einstein, H. (2018). Physical processes involved in the laboratory hydraulic fracturing of

granite: Visual observations and

interpretation. Engineering Fracture Mechanics, 191, 125-

142.

15. Irwin, G. R. (1956). Onset of fast crack propagation in high strength steel and aluminum alloys (No. NRL-4763).

NAVAL RESEARCH LAB WASHINGTON DC.

16. Miller, J. T. (2008). Crack coalescence in granite (Master's Thesis, Massachusetts Institute of

Technology).

17. Morgan, S. P. (2015). An experimental and numerical study on the fracturing processes in Opalinus

shale (Doctoral dissertation, Massachusetts Institute of

Technology).

18. Morgan, S. P., Li, B. Q., & Einstein, H. H. (2017, August). Effect of injection rate on hydraulic fracturing of Opalinus

clay shale. In 51st US Rock Mechanics/Geomechanics

Symposium. American Rock Mechanics Association.

19. National Energy Technology Laboratory (NETL). (2011, March). Shale Gas: Applying Technology to Solve

America's Energy Challenges. Retrieved March 12, 2016,

from http://www.netl.doe.gov/technologies

/oil-gas/publications/brochures/

20. National Research Council (NRC). 2001. Energy Research at DOE: Was It Worth It? Energy Efficiency

and Fossil Energy Research 1978 to 2000. Washington,

DC: The National Academies Press. doi: 10.17226/10165.

21. Perkins, T. K., & Kern, L. R. (1961, September 1). Widths of Hydraulic Fractures. Society of Petroleum

Engineers. doi:10.2118/89-PA

22. Reyes, O. M. L. (1991). Experimental study and analytical modelling of compressive fracture in brittle

materials (Doctoral dissertation, Massachusetts Institute

of Technology).

23. Roberts, E. A. (1866). U.S. Patent No. 59,936. Washington, DC: U.S. Patent and Trademark Office.

24. Saldungaray, P. M., & Palisch, T. T. (2012, January 1). Hydraulic Fracture Optimization in Unconventional

http://www.netl.doe.gov/technologies/oil-gas/publications/brochures/http://www.netl.doe.gov/technologies/oil-gas/publications/brochures/

Reservoirs. Society of Petroleum Engineers.

doi:10.2118/151128-MS

25. Simonson, E. R., Abou-Sayed, A. S., & Clifton, R. J. (1978, February 1). Containment of Massive Hydraulic

Fractures. Society of Petroleum Engineers.

doi:10.2118/6089-PA

26. Suarez-Rivera, R., Burghardt, J., Stanchits, S., Edelman, E., & Surdi, A. (2013, March 26). Understanding the

Effect of Rock Fabric on Fracture Complexity for

Improving Completion Design and Well Performance.

International Petroleum Technology Conference.

doi:10.2523/IPTC-17018-MS

27. Teufel, L. W., & Clark, J. A. (1981). Hydraulic-fracture propagation in layered rock: experimental studies of

fracture containment (No. SAND-80-2219C; CONF-

810518-7). Sandia National Labs., Albuquerque, NM

USA.

28. Trembath, Alex, Jesse Jenkins, Ted Norhaus, and Michael Shellenberger. Where the Shale Gas Revolution Came

From. Rep. N.p.: Breakthrough Institute, 2012. Print.

29. Warpinski, N. R., Schmidt, R. A., & Northrop, D. A. (1982). In-situ stresses: the predominant influence on

hydraulic fracture containment. Journal of Petroleum

Technology, 34(03), 653-664.

30. Wong, L. N. Y. (2008) Crack coalescence in molded gypsum and Carrara marble (Doctoral Dissertation,

Massachusetts Institute of Technology).

31. Yost, Albert B. “Eastern Gas Shales Research,” Morgantown Energy Technology Center, 1988,

http://www.fischerhtropsch.org/DOE/_conf_proc/MISC/C

onfh89_6103/doe_metch89_6103h2A.pdf

32. Zoback, M. D. (2010). Reservoir geomechanics. Cambridge University Press.

http://www.fischerhtropsch.org/DOE/_conf_proc/MISC/Confh89_6103/doe_metch89_6103h2A.pdfhttp://www.fischerhtropsch.org/DOE/_conf_proc/MISC/Confh89_6103/doe_metch89_6103h2A.pdf