reservoir geomecanics

346 Reservoir geomechanics
with faults (Antonellini and Aydin 1994), may be planes of lower permeability than
the matrix rock. As such phenomena are associated with formations of high porosity
and permeability formations, the degree to which compaction bands and shear bands
affect reservoir permeability is unclear. Sternlof, Karimi-Fard et al. (2006) discuss the
effects of compaction bands on reservoir permeability. Faults and shear bands in highly
porous sands can cause permeability reduction due to the communition and porosity
reduction associated with shearing, but the damage zone adjacent to faults and shear
bands is an area of enhanced permeability (Antonellini and Aydin 1994). This can result
in a situation where cross-fault flow is impeded but flow parallel to the fault plane is
enhanced in a permeability halo surrounding the fault. This halo would result from
a damage zone adjacent to the fault that consists of numerous, relatively small critically
stressed faults. A similar phenomenon is likely associated with large displacement
faults that are frequently associated with a relatively impermeable fault core, consisting
of ultra-fine grained cataclasite (Chester, Chester et al. 2005) that result from pervasive
shearing. In this case too, the fault may be relatively impermeable to cross-fault
flow but flow parallel to the fault may be appreciably enhanced in the damage zone
surrounding it.
For the cases in which the critically-stressed-fault hypothesis is most likely to be
applicable (brittle rocks with low matrix permeability), it is worth considering the
implications for permeability anisotropy in a highly fractured medium. As noted in
Chapter 5, the implications of the critically-stressed-fault hypothesis for permeability
anisotropy in a fractured reservoir are markedly different from those that arise assuming
that maximum permeability is parallel to S Hmax because it is controlled by Mode I
fractures. This is easily seen in the idealized view of the orientations of critically
stressed faults in different tectonic regimes presented in Figure 5.1. Ascan be seen in
that figure, in a normal faulting environment (row 2), if there are conjugate sets of normal
faults present (as theoretically expected), the direction of maximum permeability will
be subparallel to S Hmax (similar to what would be seen with Mode I fractures, row 1)
but the dip of the normal faults will also have an effect on permeability anisotropy. In a
strike-slip faulting environment (row 3), conjugate faults would cause flow to be greatest
at directions approximately 30 ◦ to the direction of S Hmax , significantly different from
what is expected for Mode I fractures. In reverse faulting environments (row 4) flow
along critically stressed conjugate faults would be maximum parallel to the direction
of S hmin , orthogonal to the direction of S Hmax .
While the idealized cases shown in Figure 5.1 are helpful in a general sense, there
are many places around the world characterized by normal/strike-slip faulting stress
states (S V ∼ S Hmax > S hmin )orreverse/strike-slip faulting (S Hmax ∼ S hmin > S V ) which
make the idealized cases shown in Figure 5.1 overly simplified because multiple fault
sets are likely to be active. In other words in a normal/strike-slip stress state, one might
observe any number of the sets of active faults that are shown in the idealized stereonets
for normal and strike-slip faulting in Figures 5.1b,c. An analogous situation is true for