Structural evolution, petrophysics, and large-scale permeability of faults in sandstone, Valley of Fire, Nevada

This study of faulted aeolian Aztec sandstone, southern Nevada, elucidates poorly understood details of fault flow characteristics and geometry. The faults formed by shearing along joint zones and comprise a network of two predominately strike-slip fault sets with opposite slip sense. An outcrop- to kilometers-scale conceptual model for the evolution of the strike-slip fault network is presented whereby the network forms by linking of first generation faults via Mode I splay fractures, which are subsequently sheared to form a second generation of faults that have slip sense opposite to the prior generation. At the outcrop-scale at least five hierarchical generations of structures are identified. The final geometry of the fault network is dictated by the characteristic splay fracture kink-angle.; Cross-fault flow characteristics are quantified using detailed petrophysical analysis. Petrophysical data indicate that fault rock permeability is significantly lower than host rock permeability, and that the faults will act as lateral barriers with respect to reservoir production time-scales. The petrophysical data also show that fault rocks are capable of sealing small to moderate hydrocarbon columns with respect to geologic time-scales, assuming adequate continuity of the fault rock over large areas of the fault.; Large-scale permeability characteristics of the faults are quantified using numerical flow simulation techniques that utilize idealizations of detailed field maps. The computed fault zone permeabilities are strongly anisotropic in all cases. Permeability enhancement of nearly an order of magnitude (relative to the host rock) is observed for the fault-parallel component, while fault-normal permeability might be two orders of magnitude less than the host rock permeability.; The numerical flow models are shown to be highly sensitive to the chosen boundary conditions. No-flow boundary conditions sufficiently capture the global flow characteristics of upscaled fault regions. Periodic boundary conditions break connectivity between both high and low permeability features, which tends to result in erroneous upscaled permeabilities. Globally upscaled regions insufficiently predict transport problems. Transport predictions are improved by a step-wise method of removing the through-going high-permeability features from the fine model, upscaling to a coarse grid, and then explicitly representing the high-permeability features in the coarsened model for flow simulations.