Page 256 - Fundamentals of Gas Shale Reservoirs

236 PASSIVE SEISMIC METHODS FOR UNCONVENTIONAL RESOURCE DEVELOPMENT

The simple explanation for these apparently incompatible data throughout a much larger part of the reservoir volume
observations is that frac fluid pressure primarily induced slip than borehole data. Consequently, passive seismic data can
on pre‐existing fractures by reducing friction. As discussed identify stress compartments where fracture treatments will
in Section 10.2.2.3, reactivating even poorly oriented behave differently. Focal mechanism solutions can identify
fractures requires less energy than generating new fractures. stress changes caused by a fracture treatment during a frac­
Pre‐existing fractures include vertical joints and wrench ture treatment (e.g., Neuhaus et al., 2012; Williams‐Stroud
faults. Vertical joints are present in most strata, typically as et al., 2012a, b).
orthogonal sets. Vertical wrench faults and joints commonly Focal mechanism solutions and the relationship between
develop in active fold‐thrust belts due to repeated local inter­ focal mechanism solutions and earth stress are described in
change of Smin and Sint during thrust‐belt propagation. In Section 10.3.2. Again, a focal mechanism is a description of
summary, reservoirs in active contractional faulting tectonic the strain produced by an MEQ; although the strain axes are
regimes are full of vertical fractures that are easily reacti­ often equated with the stress axes, this is not necessarily
vated when frac fluid pressure reduces friction on these true. Consequently, a population of MEQ focal mechanisms
fractures. Consequently, effective hydraulic fracture stimu­ can be used to provide an overall estimate of the strain in a
lation may be possible even in apparently unfavorable tec­ volume of rock by simple averaging. Twiss and Unruh
tonic regimes. However, treatment pressures equal to Sv can (1998) argue that such averages are good estimates of bulk
only occur if horizontal fractures are being driven, which strain, but provide poor stress estimates if substantial
should be visible in passive seismic results. Horizontal material rotation occurs. Such material rotation is not a
fracturing is more common in soft rocks rich in clays and consideration during hydraulic fracture treatments. On the
organic matter. time and volume scales of hydraulic fracture treatments,
infinitesimal strain of the total rock volume is a good
assumption.
10.6.2 Interpreting Reservoir Stress from Focal In addition to simple averaging methods, a variety of
Mechanisms
methods used by structural geologists to evaluate popula­
10.6.2.1 Introduction Reservoir stress, rock, properties, tions of fault‐slip data are equally applicable to evaluating
and natural fractures are the main influences on hydraulic populations of MEQ focal mechanisms produced by
fracture propagation. Pure hydraulic fractures run perpen­ hydraulic fracture treatments. However, the outcrop or core
dicular to Smin and the stress state controls reactivation of data used by structural geologists is better constrained
natural fractures by the fracture treatment. Consequently, because the true fault orientation is known. These methods
knowledge of reservoir stress is critical for frac design. are reviewed by Allmendinger et al. (1989) and Marrett and
Borehole data are an excellent source of reservoir stress Allmendinger (1990). Such methods are implemented in a
information and is especially valuable as a supplement to variety of shareware and low‐cost commercial software
passive seismic data. Acoustic borehole images are the best packages that can be downloaded for free or purchased
and most reliable source of borehole stress data because they online. Such packages include the popular Stereonet and
provide detailed, unambiguous images of breakout and FaultKin programs by R. Allmendinger of Cornell University.
image drilling‐induced fractures. Commercial versions of In this section, we will discuss simple, practical stress
these logs include UBI (Schlumberger), CBIL (Baker Atlas), interpretation methods that users can apply easily with
and CAST (Halliburton). Crossed‐dipole shear‐wave logs such software and provide some practical examples.
can provide useful stress information, but are strongly influ­ Interested readers should note that more sophisticated
enced by borehole rugosity and rock features such as cross‐ methods are available for estimating the stress axes from
bedding and natural fractures. Oriented caliper data, populations of earthquake focal mechanisms (e.g., Arnold
including caliper data from electrical borehole imagers, pro­ and Townend, 2007; Gephart and Forsyth, 1984). These
vide only a crude constraint on borehole breakout and cannot methods are applicable to microseismic data (e.g., Urbanic
reliably differentiate breakout from key seat and other types et al., 1993). The uncertainties in inverting focal mecha­
of borehole rugosity. Electrical borehole images are not a nisms for stress are described by Abers and Gephart (2001).
reliable breakout indicator in conductive rocks, especially
shales. Induced fractures in oriented core can provide excel­ 10.6.2.2 Reservoir Stress
lent stress information (Kulander et al., 1979, 1990). Various Averaging Focal Mechanisms and Fracture Complexity
commercial mini‐frac logging tools can provide quantitative Moment‐tensor averaging methods are common in earth­
stress estimates. quake seismology (e.g., Jost and Herrmann, 1989) and, as
The downside of borehole stress data is that the borehole previously discussed, are readily applied using various
samples form only a small part of the reservoir. The stress shareware or low‐cost software packages intended for
state in a reservoir can vary on all scales from borehole‐ to structural geologists. In the examples that we present here,
field‐scale. MEQ focal mechanism solutions provide stress we use the method of Tibi et al. (2013). The average