URTeC 1582380

Ambient Fracture Imaging: A New Passive Seismic Method Alfred Lacazette*, Jan Vermilye, Samuel Fereja, Charles Sicking, Global Geophysical Services Inc., Microseismic Services Division, 1625 Broadway St., Suite 1150, Denver, Colorado 80202 Copyright 2013, Unconventional Resources Technology Conference (URTeC)

This paper was prepared for presentation at the Unconventional Resources Technology Conference held in Denver, Colorado, USA, 12-14 August 2013.

The URTeC Technical Program Committee accepted this presentation on the basis of information contained in an abstract submitted by the author(s). The contents of this paper

have not been reviewed by URTeC and URTeC does not warrant the accuracy, reliability, or timeliness of any information herein. All information is the responsibility of, and, is

subject to corrections by the author(s). Any person or entity that relies on any information obtained from this paper does so at their own risk. The information herein does not

necessarily reflect any position of URTeC. Any reproduction, distribution, or storage of any part of this paper without the written consent of URTeC is prohibited.

Summary

Knowledge of discrete transmissive features such as faults or fracture zones prior to drilling offers substantial

benefits for all phases of operations from exploration to development planning and hydraulic fracture design. We

describe a new application of a relatively new passive-seismic method. The product is images of seismic activity

produced by discrete features such as faults and fractures in the reservoir prior to drilling. We will present data

acquired with passive seismic listening arrays in two unconventional reservoirs and a fractured carbonate reservoir.

An extensive body of literature shows that fracture/fault zones with high resolved shear stress correlate positively

with fracture transmissivity. In many cases, discontinuities with high resolved shear stress will be the most

microseismically active. This method therefore allows mapping of many hydraulically transmissive zones prior to

drilling or any other activity.

The passive seismic method (Tomographic Fracture Imaging™ or TFI) directly images seismically active hydraulic

fractures and natural fractures as complex surfaces and networks. Until recently, TFI has been used primarily to

image hydraulic fracture treatments. We report a new application: imaging ambient seismic activity both prior to

frac monitoring and by quiet-time monitoring using 3D reflection grids, i.e. by recording on a reflection grid or

dedicated passive seismic grid when shooting or vibrating is not in progress. The resulting images reveal

seismically active fracture and fault zones that correlate well with features illuminated by fracing or that are imaged

by 3D reflection seismic attributes. The presence and hydraulic transmissivity of some of these features have been

validated by independent measures including chemical tracers and pressure monitoring.

This work also considers drivers of ambient seismic activity including earth tides and epeirogenic movements.

Earth stress studies have established that the brittle crust is a self-organizing critical system in a state of frictional

equilibrium and hence is constantly on the verge of movement. The earth’s brittle crust is continually loaded by a

variety of forces although tectonic movements, isostacy and isostacy-related flexure appear to dominate. It has been

shown that stress or pressure changes of less than 0.01 atmospheres can stimulate seismicity. We present our work

to date on quantitative correlations between earth tides and ambient seismic activity imaged on large grids. The

results suggest that earth tides help stimulate release of stored elastic strain energy in the brittle crust.

Introduction

Natural fracture networks strongly influence the movement of hydrocarbons in the subsurface and hence the

productivity of wells and entire fields. The importance of natural fractures to productivity of many unconventional

reservoirs is widely recognized. Reflection seismic methods can provide bulk measures of velocity and velocity

anisotropy. Although these values have proven useful in many cases for fracture detection in advance of drilling,

they do not reveal the discrete fracture network directly and they do not indicate fracture systems that are likely to

be hydraulically transmissive. Similarly, high-resolution reflection seismic methods can reveal larger fractures, but

do not reveal the detailed network or the likelihood that the fractures will be productive. Other methods for

URTeC 1582380 Page 2

understanding subsurface fracture networks include borehole images, open- and cased-hole production logs,

chemical and radioactive frac tracers, well tests, interference tests, production data, and other methods that require

drilling and completing multiple wells, thus precluding field development planning that accounts for the natural

fracture network.

This work describes a new passive-seismic method that images subsurface fracture networks prior to drilling that are

likely to be hydraulically transmissive. The data required can be obtained inexpensively by listening to ambient

seismic emissions during standard 3D reflection surveys when shooting or vibrating is not in progress. The resulting

images allow potential sweet spots to be targeted first and allow field development planning to begin before starting

a drilling campaign. Ambient images can also aid in planning hydraulic fracture treatments. Ambient images of

natural fracture networks are less complete than images produced by hydraulic fracture monitoring, but are useful

for exploration and pre-drill development planning. More complete TFIs are obtained when monitoring a fracture

treatment because the fluid pressure pulse and poroelastic stress changes produced by the frac illuminate the

transmissive fracture fairways over a large area and provide more energy for imaging. However, many features that

are illuminated by fracture treatments appear in the ambient, pre-frac images.

This paper is intended to provide a brief overview of Tomographic Fracture Imaging and a progress report and

overview of our work-to-date on ambient TFI. More complete publications are in preparation. Here we summarize

the imaging method, explain why it works, and provide several examples of results that were validated with

independent data sets. We also present our work-to-date on correlation of bulk ambient seismicity with earth tides.

Method

Tomographic Fracture Imaging™ or TFI (Geiser et al, 2006; Geiser et al, 2012; Lacazette and Geiser, 2013; Sicking

et al, 2013) is a proprietary, patented and patent-pending technology. In brief, TFI is accomplished as follows.

1. Single (vertical) or multi-component passive seismic data is recorded continuously throughout the time

period of interest with a surface or shallow buried array. Uniform grid geometries are preferred, hexagonal

closest-packed grids are ideal.

2. The data is processed to remove noise, make static corrections etc. similarly to standard reflection seismic

processing. .

3. A velocity model of the study volume is developed using any available data such as sonic logs, stacking

velocities from seismic data, or VSP surveys. The velocity model is as detailed as possible. If sufficient

data are available, then a 3D spatially-varying, anisotropic velocity model is built. If data such as string or

perforation shots are available, then the initial model is fine-tuned using this data.

4. The study volume is divided into voxels. Using ray-tracing, a table of one-way travel times from every

voxel to every receiver is computed.

5. The traces are aligned in time for each voxel, and the semblance or other quantity (such as a measure of

energy) is computed for discrete time intervals to produce a five-dimensional data volume. The dimensions

of the data volume are the X, Y, Z coordinates of the voxels, the time-step, and the semblance or other

measure of cumulative seismic activity in each voxel during each time-step.

6. The computed volumes are analyzed, edited, and then combined to produce a single stacked depth volume.

The stacked volume may be a combination of data from minutes to hours and sometimes days.

7. The volume is clipped to eliminate low-level noise leaving sinuous clouds of high activity.

8. The Tomographic Fracture Image is generated by finding the central surfaces of the clouds. These central

surfaces represent the large, seismically-active fractures.

For hydraulic fracture monitoring, two slightly different proprietary workflows are used to produce two types of

TFIs. Near-well TFIs show the highest energy features connected to or near the perforations, which are the induced

fractures and most heavily stimulated natural fractures. Reservoir-scale TFIs show the total activity throughout the

imaged volume. For ambient imaging studies only the reservoir-scale workflow is used.

Other branches to this workflow include analysis of individual time steps to identify and locate microearthquake

(MEQ) hypocenters or perforation shots and producing movies of cumulative activity volumes from moving-sums

of time steps.

The method works because:

URTeC 1582380 Page 3

The earth’s brittle crust is a self-organized critical system (e.g. Leary, 1997) in a state of frictional equilibrium

due to pervasive fracturing (Zoback, 2007). Stress or pressure changes <0.01 atmospheres are sufficient to

stimulate seismicity (Ziv and Rubin, 2000). Earthquake magnitude vs. frequency distributions are linear on

log-log plots (magnitude is a logarithmic measure), so small earthquakes are many orders of magnitude

more abundant than large ones and the distribution appears to have no lower magnitude limit (Ziv and

Rubin, 2000). However, distinct earthquakes are not the only type of seismic activity. For these reasons,

the brittle crust is constantly emitting seismic energy at a low level in the absence of artificial stimuli.

Hydraulic fracture treatments stimulate activity both by reducing the normal stress on pre-existing natural

fractures, breaking new rock, and altering the present-day stress state. Note that a pressure pulse can be

transmitted through a fracture network without transmitting fluid (Lacazette and Geiser, 2013).

An extensive, 18 year old body of work demonstrates that in many/most cases resolved shear stress on a

fracture is correlated positively with its hydraulic transmissivity (e.g. Barton et al, 1995; Heffer et al, 1995;

Morris et al, 1996; Ferrill et al, 1999; Sibson , 2000; Takatoshi and Kazuo, 2003; Tamagawa and Pollard,

2008; Heffer, 2012; Hennings et al, 2012). Higher resolved shear stress is positively correlated with

seismic activity so that more microseismically active fractures are generally expected to be more

hydraulically transmissive. Note that the converse is not necessarily true.

Rock mechanics theory, field studies (e.g. Vermilye and Scholz, 1998) and experiment (e.g. Janssen et al,

2001) show that large fractures are embedded in clouds of smaller fractures and seismic emissions. Hence

the central surfaces of high activity clouds are the loci of large fractures.

TFI is extremely sensitive because it stacks total trace energy rather than simply looking for large, discrete

events. Stacking total trace energy captures energy from numerous microearthquakes (MEQs) too small to

discriminate as individual events and/or too small or indistinct to have separable P and S wave arrivals.

These small events represent much more energy that the relatively-rare large MEQs amenable to standard

seismological methods. Perhaps more importantly, stacking total trace energy captures energy from Long

Period, Long Duration (LPLD) events (Das and Zoback, 2011; Zoback et al, 2012), and perhaps other as

yet unclassified types of seismic activity.

Stacking over extended time periods both adds tremendous sensitivity and images surfaces. Sensitivity is

increased because random noise is canceled and spatially stable signal is stacked. Surfaces are imaged

because activity along large fracture surfaces is episodic and extended time periods are required to fully

image such surfaces.

In common with other methods employing seismic emission tomography, TFI stacks energy from large

numbers of traces which cancels noise and increases signal strength.

The method detects only seismically active features, which leads to the following caveats:

High resolved shear stress, and hence seismicity, does not always correlate positively with hydraulic

transmissivity.

Poroelastic stress waves induced by a hydraulic fracture treatment can cause fractures to become seismically

active without a direct fluid connection to the wellbore.

Weak features such as fault breccia zones may be very transmissive with little resolved shear stress or

associated seismic activity because they are unable to support substantial shear stress. We find that such

weak features are detected only sporadically along their length by this method and in some cases are not

detected at all.

The following sections provide examples of some ambient TFI results and brief discussions of independent data sets

that validate the results. Drivers of ambient emissions are discussed subsequently.

Example 1: Tight gas sand, southern West Virgina

Figure 1 shows some passive seismic results from a detailed reservoir characterization study of stacked

unconventional reservoirs carried out in the Upper Devonian section of southern West Virgina. Some aspects of this

study are described in Lacazette and Geiser (2013), Geiser et al. (2012), Moos et al. (2011), Franquet et al. (2011),

and Mulkern et al. (2010) and additional publications are in preparation. The TFI results were confirmed by

chemical and radioactive tracer data, wellbore images and other data (Geiser et al 2012; Lacazette et al, in prep).

URTeC 1582380 Page 4

The results show that pre-frac ambient seismic activity imaged major features revealed by frac monitoring. The

zone of intense fracturing in the southwest corner of the study area is prominent and the highly transmissive area

around the vertical well in the southwest corner of the diagram (west southwest of the horizontal wellhead) is

identified. See Geiser et al (2012), especially Figure 6, for additional detail. This study was blind, so that well and

other data was not known to personnel engaged in the TFI study until after delivery of the results.

Figure 1. Depth slices of ambient activity and TFIs at the level of a horizontal well. North is up. LEFT – Ambient

image of four hours of cumulative seismic activity in the Berea Sandstone Fm. RIGHT – TFIs from all 9 frac stages

in the horizontal Berea well shown at the center of each diagram. A subsequent horizontal well (red bar in SW)

confirmed that the area in the southwest is unusually heavily fractured and highly permeable.

Example 2: Gas shale, Pennsylvania

Figure 2 shows results from a fracture treatment of a well in an important gas shale formation. TFI and cumulative

activity sums show that stage 3 of the horizontal well (A in Figure 2) immediately connected to an older vertical

well (B in figure 2) through a preexisting natural fracture system. Well B was monitored for pressure and chemical

tracers during the fracture treatment.

Activity in the natural fracture system began immediately when stage 3 was pumped because increased fluid

pressure in the fracture reduced the normal stress and hence friction along the feature thus allowing slip. During

stage 4 activity was primarily confined to the area around well B. Pressure and chemical tracer anomalies were not

detected in well B until commencement of stage 5. Two stages of pumping were required to move sufficient fluid to

displace any preexisting fluid and account for leak-off into the rock matrix and natural fracture system.

The ambient data shown at the lower right of Figure 2 is heavily contaminated by noise from a nearby factory, but

some prominent features of the frac stage results were forecast by the ambient data. These features are the east-west

trending feature that extends east of the wellbore (feature 1 in Figure 2) and the prominent NW-SE trending fracture

zone just south of well B (feature 2 in Figure 2). Note that a fracture treatment provides much more energy than

ambient emissions so that noise was much less of a problem for fracture monitoring. Had less coherent noise been

present it is likely that more detail of the natural fracture network would have been resolved by the ambient data.

URTeC 1582380 Page 5

Figure 2. Depth slices from a Tomographic Fracture Imaging study of a horizontal shale well. All slices are at the

level of the wellbore. North is up. LEFT –10 minute semblance stack with horizontal treatment well (A) and offset

vertical producer (B) shown in gray, Stage 3 perforation depth range in red. Stack shows the time interval shortly

after breakdown when activity between the treatment well and vertical producer was declining. Blue = low

cumulative seismic activity, red = high. CENTER –Tomographic Fracture Image™ of stage 3. The fracture

connecting the wells is imaged clearly. Color shows relative intensity of cumulative seismic activity (red high, blue

low), as in image on left. TOP RIGHT – Detailed view showing combined reservoir-scale TFIs for stages 1-5

(color). BOTTOM RIGHT – Same as top right but with ambient TFI overlain in grayscale. Note that prominent

features in the ambient data were later activated during the fracture treatment, especially the large fracture zones

marked by 1 and 2. See text for discussion.

Example 3: Carbonate anticline

This section presents some highlights of an ambient imaging study performed on a carbonate anticline in the

Colombian Andes. A surface array of standard 10 Hz vertical component geophones was deployed over the study

area in a hexagonal closest-packed grid pattern and data was recorded continuously for four days. The array density

was approximately 87 receiver points/km2. This study was conducted blind, so that GMS personnel did not know of

the well results or even the existence of these approx. 50-year-old wells prior to delivery of the passive seismic

products. The study was conducted because a well-developed natural fracture system is required for production in

this play.

Structure in the study area is complex, as shown in Figure 3. An early thrust-fault-bend anticline, now largely

inactive, is cut by wrench faults that are currently active. The current tectonic regime appears to be reactivating

parts of the older thrust-fault-bend folds.

URTeC 1582380 Page 6

Figure 3. Diagrams show the four-day cumulative seismic activity sum (stacked semblance) in relation to the

geologic structure of the anticline. LEFT - Vertical section of interpreted PSDM reflection seismic data in grayscale

with overlay of cumulative activity in color. RIGHT – Horizon extraction of cumulative activity data overlain on a

structure contour map of the horizon with fault cuts. Line of cross-section at left is shown. Hot colors represent

high activity levels. Note the correspondence of activity with specific structural domains and in particular with the

late, currently active wrench fault that cuts the anticline (right-hand diagram).

Figure 4. Reservoir-level depth slice of the ambient TFI derived from the four-day activity sum shown in Figure 3.

Symbols indicate well productivity. 0 = no production. S = some production. E = Excellent production.

The role of earth tides

Earth stress studies show that the earth’s brittle crust is in a state of frictional equilibrium (Zoback, 2007). Recent

studies that monitor slow crustal movements with permanent GPS base stations show that even relatively stable

parts of the continental cratons are in constant motion and that crustal movement vectors have varying velocities and

directions. The resulting vector fields can be used to produce strain maps that correlate with earthquake abundance.

URTeC 1582380 Page 7

For examples see El-Fiky and Kato (2006), Tesauro et al (2006), and the U.S. National Geodetic Survey online

toolkit (see references for link). These observations are relevant because they show that even very slow strain rates

(as in Western Europe) keep the crust pervasively loaded with enough elastic strain energy to produce large

earthquakes. Because earthquake size vs. frequency scales linearly on log-log plots with no apparent lower limit,

subtle, widespread seismic activity is to be expected. In addition to tectonic movements, other forces including

isostacy and isostacy-related flexure (e.g. Jurkowski et al, 1984) and even seasonal water loading (Davis et al, 2004)

produce movements that contribute to the elastic strain energy stored in the crust. In summary, the earth’s brittle

crust is kept loaded with elastic strain energy from a variety of sources.

What causes this pervasive strain energy to be released in the form of subtle, pervasive ambient seismicity? This

question is important for commercial development of ambient seismic methods for several reasons, including the

information that it may reveal about fracture system properties and perhaps determining optimum times and methods

for sampling ambient activity. The answer to this question could have implications for fields beyond hydrocarbon

exploitation including earthquake forecasting, the tidal energy balance of the earth-moon system, and tectonics.

This section presents some of our work to date investigating the role that earth tides may play in ambient seismic

emissions. Our approach is to relate bulk measures of cumulative seismic activity to earth tides. The data presented

in this section are from the carbonate anticline study described here as Example 2. The volume of rock studied was

137.4 km3 (33.0 mi

3). We used the mean cumulative semblance value in discrete time intervals as a measure of the

bulk seismic activity within the study volume and compared it to displacement variables representing the earth tide

movements in the corresponding time period.

Earth tide movements for the study area were computed using software available from the U.S. National Geodetic

Survey (Figure 5). The north-south, east-west, and vertical components of movement have different periods and are

out-of-phase because of the complex and constantly changing angular relationships between the earth, sun, and

moon. Consequently, the average activity variable cannot be examined as a function of time. Instead we have

compared the average activity to the earth tide displacements, linear and angular velocities and accelerations in

terms of a vector that connects a point on the earth before and after a specified time period (Figure 6). We computed

velocities in the north-south, east-west, and vertical directions, the angular velocities with respect to these axes, the

total length of the displacement vector, the horizontal length of the displacement vector, and the horizontal angular

velocity (Figure 6). Corresponding accelerations were also computed.

Figure 5. Earth tides for the four-day survey period and the days before and after computed with software from the

U.S. National Geodetic Survey. The north-south, east-west, and vertical movement components are shown

separately. Note that the components are out-of-phase.

URTeC 1582380 Page 8

Figure 6. LEFT – Diagram illustrating the displacement variables computed from the earth tide data. The variables

represent the magnitude and orientation of a vector that connects the same point on the earth at the beginning and

end of a specified time period and hence represent velocities. Accelerations are also computed. RIGHT – The

displacement variables used for the regression shown in Figure 8 with the sensitivities, i.e. the predictive values of

the variables for the regression.

Experimentation with different time intervals showed that averaging over a 20 minute interval was required to reveal

structure in the data (Figure 7). Simple crossplots of such moving averages (e.g. Figure 7) indicate that there is

structure in the data and that multiple variables are likely required to predict bulk ambient activity from earth tide

movements.

Figure 7. Crossplot of the 20 minute moving averages of the absolute length of the total displacement vector

(Figure 6) vs. the mean activity measure in the study volume.

We performed the exploratory multivariate statistics presented here with the Transform™ software package

(www.transformsw.com). Non-parametric, non-linear multivariate statistical analysis (regression) was performed to

predict the activity variable from various combinations of the displacement variables. The results show that 40% -

50% of the variability of the mean ambient seismicity in the entire volume can be predicted by various combinations

of the displacement variables. Figure 8 shows the results of a run with all of the variables shown in the table in

Figure 6. Eliminating variables with low positive or negative sensitivities yields plots similar to Figure 8 but with

somewhat lower R2 values. A run with only the North, NorthAngle, Horizontal, and Total velocities explains

40% of the observed variability of the mean activity (i.e. has an R2 of 0.4).

URTeC 1582380 Page 9

We find these correlations remarkably high given that they represent an average of. 137.4 km3 (33.0 mi

3) of

structurally complex rock riddled with a complex fracture network (Figures 3 and 4). Furthermore, the complex

polyphase structure and active deformation in the study volume must be producing a very complex stress regime

within the structures. The four-day cumulative intensity of ambient activity is related clearly to distinct structural

domains (e.g. fold panels, see Figure 3). It is likely that different structural domains and faults and other fractures

with different orientations in each structural domain are stimulated differently by earth tide movements in different

directions. We suspect that separate analysis of homogeneous structural domains will yield stronger correlations

between earth tides and ambient seismic emissions. This work is ongoing. These results raise the possibility that

the seismic response to earth tide displacement components could represent a measure of natural fracture

orientations and other fracture network properties and perhaps even aspects of the fracture flow regime.

Figure 8. Predicted mean activity (vertical axis) vs. observed mean activity (horizontal axis) in 20 minute intervals.

See text for discussion.

Conclusions

We conclude that ambient seismic emissions can be imaged to reveal parts of hydraulically transmissive natural

fracture networks whose existence can be verified by independent data sets. We tentatively conclude that earth tides

measurably affect the intensity of ambient seismic emissions.

Acknowledgements

The authors gratefully acknowledge our clients (who choose to remain anonymous) for both permission to publish

the results presented here and especially for their insight and courage to work with a radical new technology. We

also acknowledge the work of the following Global Geophysical Services personnel: Gloria Eisenstadt for the

structural interpretation shown in Figure 3, Sean Boerner and Rohit Singh for their work on multivariate statistical

analysis of the earth tide data, and the numerous personnel of both Global and other companies who worked on the

acquisition and processing of the ambient and reflection seismic data presented here.

References

Barton, C.A., Zoback, M.D., Moos, D., 1995. Fluid flow along potentially active faults in crystalline rock. Geology

23 (8), 683-686.

URTeC 1582380 Page 10

Das, I., Zoback, M.D., July 2011. Long-period, long-duration seismic events during hydraulic fracture stimulation of

a shale gas reservoir. The Leading Edge 30 (7), 778-786.

Davis, J.L; Elósegui, P.; Mitrovica, J.X. and Tamisiea, M.E. (2004) Climate-driven deformation of the solid Earth

from GRACE and GPS: Geophysical Research Letters, v. 31, L24605, 4 p.

El-Fiky, G. and Kato, T. (2006) Secular crustal deformation and interplate coupling of the Japanese Islands as

deduced from continuous GPS array, 1996-2001: Tectonophysics, v. 422, no. 1-4 (August), p. 1-22.

Ferrill, D.A., Winterle, J., Wittmeyer, G., Sims, D., Colton, S., Argmstrong, A., Morris, A.P., 1999. Stressed rock

strains groundwater at Yucca Mountain, Nevada. GSA Today 9, 1–8.

Franquet, F.A., Mitra, A., Warrington, D.S., Moos, D., Lacazette, A., 2011. Integrated Acoustic, Mineralogy, and

Geomechanics Characterization of the Huron Shale, Southern West Virginia, USA. SPE 148411.

Geiser, P.A., Vermilye, J., Scammell, R., Roecker, S., 2006. Seismic used to directly map reservoir permeability

fields. Oil Gas J.. Dec. 11.

Geiser, P., Lacazette, A., Vermilye, J., 2012. Beyond “dots in a box”: an Empirical View of Reservoir Permeability

with Tomographic Fracture Imaging. First Break, 63-69.

Heffer, K.J., Fox, R.J., McGill, C.A., 1995. Novel Techniques Show Links between Reservoir Flow Directionality,

Earth Stress, Fault Structure and Geomechanical Changes in Mature Waterfloods. Society of Petroleum

Engineers Annual Technical Conference & Exhibition, Dallas, TX. Oct. 22-25, 1995. SPE 30711.

Heffer, K.J., 2012. Geomechanical mechanisms involving faults and fractures for observed correlations between

fluctuations in flowrates at wells in North Sea oilfields. In: Spence, G.H., Redfern, J., Aguilera, R., Bevan, T.G.,

Cosgrove, J.W., Couples, G.D., Daniel, J.-M. (Eds.), Advances in the Study of Fractured Reservoirs.

Geological Society, vol. 374. Special Publications, London. First published online August 28.

Hennings, P., Allwardt, P., Paul, P., Zahm, C., Reid Jr., R., Alley, H., Kirschner, R., Lee, B., Hough, E., 2012.

Relationship between fractures, fault zones, stress, and reservoir productivity in the Suban gas field, Sumatra,

Indonesia. AAPG Bull. 96 (4 (April)), 753-772.

Janssen, C., Wagner, C.F., Zang, A., Dresen, G., 2001, Fracture process zone in granite: A microstructural analysis,

Int. J. Earth Sci, 90, 46-49.

Jurkowski, G.; Ni, J. and Brown, L. (1984) Modern uparching of the Gulf coastal plain: Journal of Geophysical

Research, v. 89, p. 6247–6255.

Leary, P.C., 1997. Rock as a critical-point system and the inherent implausibility of reliable earthquake prediction.

Geophys. J. Int. 131, 451-466.

Lacazette, A., Geiser, P., 2013, Comment on Davies et al., 2012, Hydraulic fractures: How far can they go?. Marine

and Petroleum Geology 43, 516-518.

Moos, D., Vassilellis, G., Cade, R., Franquet, F.A., Lacazette, A., Bourtembourg, E., Daniel, G., 2011. Predicting

Shale Reservoir Response to Stimulation: the Mallory 145 Multi-well Project. SPE 145849.

Morris, A.P., Ferrill, D.A., Henderson, D.B., 1996. Slip tendency and fault reactivation. Geology 24, 275–278.

Mulkern, M., Asadi, M., McCallum, S., 2010. Fracture Extent and Zonal Communication Evaluation Using

Chemical Gas Tracers. SPE 138877.

Sicking, C., Vermilye, J., Geiser, P., Lacazette, A., 2013, Permeability field imaging from microseismic.

Geophysical Society of Houston Journal, (3(March)), 11-13.

Sibson, R.H., 2000. Fluid involvement in normal faulting. Journal of Geodynamics 29, 469–499.

Takatoshi, I., Kazuo, H., 2003. Role of stress-controlled flow pathways in HDR geothermal reservoirs. Pure and

Applied Geophysics 160, 1103–1124.

Tamagawa, T., Pollard, D.D., 2008. Fractured permeability created by perturbed stress fields around active faults in

a fractured basement reservoir. AAPG Bull. 92, 743-764.

Tesauro, M.; Hollenstein, C.; Egli, R.; Geiger, A.; Hans-Gert, K. (2006) Analysis of central western Europe

deformation using GPS and seismic data: Journal of Geodynamics, v.42, no. 4-6 (Nov.-Dec), p. 194-209.

U.S. National Geodetic Survey Geodetic Toolkit. Available at:

http://www.geodesy.noaa.gov/TOOLS/Htdp/Htdp.shtml

Vermilye, J.M. & Scholz, C.H. (1998) The process zone: A microstructural view of fault growth, Journal of

Geophysical Research, 103, B6, 12,223-12,237.

Ziv, A., Rubin, A.M., 2000. Static stress transfer and earthquake triggering: no lower threshold in sight? J. Geophys.

Res. 105 (B6), 13631-13642.

Zoback, M.D., 2007. Reservoir Geomechanics. Cambridge University Press, 464 pp.

Zoback, M.D., Kohli, A., Das, I., McClure, M., 2012. The Importance of SlowSlip on Faults during Hydraulic

Fracturing Stimulation of Shale Gas Reservoirs. SPE 155476.