Fluid-flow effects of compartmentalized distribution of compaction bands in an aeolian sandstone in three dimensions

It is well established that the depositional architecture of dunes (cross-bed orientations and dune boundaries) controls the formation and orientation of compaction bands (CBs) in three dimensions, resulting in a compartmentalized distribution of low-angle bed-parallel CBs and high-angle CBs in the Jurassic aeolian Aztec Sandstone, exposed in the Valley of Fire State Park, Nevada, USA. In this study, we used two idealized configurations to represent the characteristic compartmentalization of CBs in three dimensions and performed flow computations to investigate the fluid-flow effects of these configurations. The results suggest that the upscaled permeability of the compartmentalized compaction-band arrays is influenced significantly by the permeability of the compaction-band sets and their orientations and distributions. In particular, in Configuration A, which represents a combination of both the high-angle and bed-parallel domains, upscaled permeability in the direction normal to the dune trend is controlled primarily by the high-angle CB domain. In contrast, in Configuration B, in which the high-angle CB domain is distributed within a limited portion, upscaled permeability in the direction normal to the dune trend is controlled by both the high-angle CB domain and the bed-parallel CB domain. The upscaled permeability in the direction parallel to the dune trend in Configuration A is controlled by both high-angle CBs and bed-parallel CBs, but in Configuration B it is controlled mainly by bed-parallel CBs. In comparison, in both configurations A and B, upscaled permeability in the vertical direction is controlled primarily by the bed-parallel CB domain. The orientation of the major permeability component in both configurations A and B remains almost unaffected by the variation in permeability of CB sets, presumably because the preferred flow path crosses a minimum number of CBs. In contrast, the plunges of the minor permeability component changes significantly in Configuration A, but remains nearly the same in Configuration B. This suggests that the interplay between the spatial distributions of CB sets and their permeability exerts significant influence on the orientation of the minor permeability component in Configuration A, whereas the permeability of bed-parallel CBs has dominant control on the minor permeability component in Configuration B. The difference between the magnitudes of the major principal permeability component and the minor principal permeability component (permeability anisotropy) is about a factor of 2.5, 2.2 and 1.9 in Configuration A for cases 1, 2 and 3, respectively, and about a factor of 5.1, 3.4 and 1.3 in Configuration B for cases 1, 2 and 3, respectively. The parametric study implies that the range of potential variations in the permeability values in flow models would generally yield similar results for the major permeability component, but may yield different results for the minor permeability component. The results presented in this study clearly demonstrate that the compartmentalized distribution of CBs exerts strong influences on fluid flow through aeolian sandstone.