Page 232 - Fundamentals of Gas Shale Reservoirs

212 PASSIVE SEISMIC METHODS FOR UNCONVENTIONAL RESOURCE DEVELOPMENT

Contained joints are joints that are restricted to individual Figure 10.1, that is, they do not slip perfectly perpendicular
mechanical layers. For example, brittle rocks such as to Sint (e.g., Reches, 1983; Reches and Dieterich, 1983).
limestones, cherts, or sandstones interbedded with shales The planarity of newly formed faults is also subject to con­
may be jointed, while the softer shale interbeds are not. trol by stress ratios as is the case with joints.
Joints are an especially important feature of many uncon­
ventional reservoirs because they are generated by the 10.2.2.3 Natural Fracture Reactivation during Hydraulic
high fluid pressures that develop during maturation of self‐ Fracture Treatments The issue of natural fracture reacti­
sealing reservoirs. These relationships are especially well‐ vation during hydraulic fracture treatments is important for
documented in the Appalachian Basin of the United States many reasons including frac design, reservoir simulation,
(Engelder and Lacazette, 1990; Engelder and Whitaker, and interpreting passive seismic data. Current research is
2006; Engelder et al., 2009; Lacazette and Engelder, 1992; showing that natural fracture reactivation is commonly the
McConaughy and Engelder, 1999). Because unconventional most important mechanism of hydraulic fracture stimula­
reservoirs are organic rich, self‐sealing, and mature, they com­ tion. Natural earthquakes and most MEQs produced by
monly have well‐developed joints developed by poroelastic hydraulic fracturing result from slip on preexisting natural
natural hydraulic fracturing. fractures (e.g., Doe et al., 2013; Eisner et al., 2010b; Moos
Natural fractures are grouped into sets of fractures with et al., 2011; Williams‐Stroud et al., 2012). As we will see in
a common type, orientation, and other characteristics. Thus the discussions on MEQ focal mechanisms and stress
we may speak of joint sets, conjugate fault sets, and so on. inversion (Sections 10.3.2 and 10.6.2.2), this has bearing on
The complete natural fracture network in a reservoir (i.e., the interpretation of reservoir stress.
all the fracture sets) is referred to as the fracture system. Reactivation of existing fractures requires much less energy
than generating new fracture surface area. Consequently,
10.2.2.2 More about Natural Fracture Orientations reactivation of existing natural fractures is common, espe­
Natural fractures form in response to stresses that prevail at cially farther from the wellbore where fluid pressures are
the time that they form. If the fractures are formed very lower. Natural fractures are reactivated during a hydraulic
recently in geologic time or if they are forming now in the fracture treatment by two mechanisms: hydrojacking and
reservoir, then the natural fracture orientations may be geo­ hydroshearing, both of which generate seismic energy.
metrically related to the neostress. However, in most reser­ Hydrojacking occurs when the fracture is forced open by the
voirs all, or at least some, of the natural fractures formed frac fluid. Hydroshearing occurs when the frac fluid pressure
many millions of years earlier during ancient tectonic events reduces the normal stress and hence the friction on natural
so that the natural fracture orientations reflect the paleostress fractures allowing them to slide under the influence of the
orientations. Such fractures generally do not have a regular reservoir stress. The stresses that produce sliding are gen­
geometric relationship to the neostress orientations and mag­ erally a combination of the neostresses and stresses induced
nitudes, but may coincidentally have such a relationship. The by the fracture treatment itself. Hydrojacking and hydros­
natural fracture system in the Marcellus Shale Fm. in the hearing generally occur together. Das and Zoback (2013b)
United States is a famous example of such a coincidence present compelling passive seismic and other evidence for
(Engelder et al., 2009). An additional complexity is that reactivation of even poorly oriented fractures during
ancient fractures are often reoriented and/or reactivated by hydraulic fracture treatments.
tectonic events (such as folding and faulting) that postdate The common and useful assumption that reactivated
their original formation. In summary, it is not valid to draw fractures slip in a plane perpendicular to Sint is not strictly
conclusions about natural fracture orientations based only on true. Like newly formed faults, preexisting fractures slip
the neostress state, without independent knowledge of the under the influences of triaxial stresses so that fault movement
time and conditions of formation of the natural fractures. may occur in a plane that is not perpendicular to Sint.
Note that the planarity of a natural fracture, and hence the Figure 10.2 shows the simulated effect of natural fracture
orientation variability of a set, is strongly dependent on the reactivation during a hydraulic fracture treatment. The images
conditions of fracture formation. For example, joints that are map views of the simulated natural fracture population
develop under conditions where Smin is much smaller than in the Marcellus Shale. Also shown are the simulated
the other two stresses are very planar, whereas joints that hydraulic fractures that took proppant, colored by stage.
form when the stresses are more equal are irregularly shaped The simulation indicated that little new rock was broken by
and tend to hook into each other during propagation. The the frac. Instead, natural fracture reactivation was the pri­
common approximation, as shown in Figure 10.1, that faults mary mechanism of stimulation. The fracture population is
form as planar features that slip perpendicular to Sint is use­ a stochastic simulation that was calibrated with a passive
ful, but is not strictly correct. In reality, new faults form in a seismic image of the fracture population in the reservoir.
triaxial stress state so that new faults form as four orthorhombic This imaging process is discussed in Section 10.5.5. (See
sets none of which are oriented exactly as shown in Lacazette et al. [2014] for details of this study.) Look at