structural geology, propagation mechanics and - Stanford School of ...

and Pollard 2000; Lothe et al. 2002), rendering them impediments to fluid flow under
saturated conditions (Figure 5.2).
Sternlof et al. (2004) have shown that anastomosing arrays of subparallel, subvertical
CBs abundantly exposed in the Aztec sandstone (Figure 5.3) would significantly reduce
bulk permeability and induce permeability anisotropy at outcrop scales. Part of a
widespread Jurassic æolian system that includes the Navajo and Nugget sandstones
(Marzolf 1983; Blakey 1989), the 1,400-m-thick Aztec is a medium-grained sub-arkose
that is typified by large-scale cross stratification (Bohannon 1983) and comprises an
exhumed analog for aquifers and reservoirs in similar lithologies. The pervasive CB
arrays, which cut across depositional bedding as the oldest structural fabric present,
formed in response to compression associated with the Cretaceous Sevier Orogeny (Hill
1989; Eichhubl et al. 2004; Sternlof et al. 2005). If present in an active aquifer or
reservoir, these band arrays would cause substantial fluid-flow effects at scales relevant
to production and management (Sternlof et al. in review).
The specific hydraulic impacts exerted by CBs depend strongly on their porosity and
permeability relative to the surrounding sandstone matrix, as well as on geometrical
factors such as average band thickness, spacing and connectivity (Sternlof et al. 2004).
Their limited thickness, however, renders direct measurement of CB permeability
difficult and potentially inaccurate under even optimal conditions (Antonellini and Aydin
1994). Given the generally limited availability of potentially unconsolidated core samples
from active aquifers and reservoirs, problems proliferate. Methods for estimating
permeability from epoxy-impregnated thin-sections using computational techniques
provide an alternative (e.g. Berryman and Blair 1987; Adler et al. 1990; Bakke and Øren
1997; Blair et al. 1996).
In this paper, we evaluate a new physics-based algorithm for estimating permeability
from thin section images (Keehm et al. 2004). In order to simulate the realistic situation
of limited subsurface sample supply, we apply the method to a single, representative thin
section from the Aztec (Figure 5.4) and compare the resulting estimates of CB and matrix
permeability to available measurements reported for the Navajo and Aztec sandstones
(Antonellini and Aydin 1994; Flodin et al. 2005).
127