Virtual outcrop models of the Moab Fault to assess uncertainties in sub-surface fault seal prediction
The Moab Fault is a world-class example of a normal fault that cuts a layered sandstone-shale sequence of the Paradox Basin, southeastern Utah. The fault is segmented and is best exposed in a series of canyons that trend sub-perpendicular to the fault strike. Clean, high quality sandstones crop out in the footwall of the Moab Fault, whilst inter-bedded sands and shales are preserved in the hanging wall. As a result, the Moab Fault has been studied by both academic and industry-based geoscientists to understand the geometry and kinematics of fault linkage, the nature and distribution of fault-related damage within reservoir-quality sandstones and the role of fault rocks (e.g. shale smears) in providing membrane seals. In particular, studies by Foxford et al. (1998) and Davatzes & Aydin (2005) have provided detailed field descriptions of the nature and distribution of fault rocks and fault-related damage along the Moab Fault, and their impact on hydrocarbon flow.
Predicting the impact of faults on fluid flow in the subsurface depends on accurate knowledge of fault throw and sequence properties (e.g. volume of shale in the faulted sequence). A key issue is to determine the impact of uncertainties in fault throws and host sequence properties on such calculations. We have addressed this problem by developing two digital, 3D structural models of the Moab Fault: (1) a course scale model based upon regional digital elevation models and published geological data (maps, stratigraphic columns and lithological descriptions); and (2) a fine scale model based upon interpretations of terrestrial laser scan (LiDAR) data obtained from the canyons that cut the Moab Fault. The interpretations of the laser scan data have been ground-truthed against structural data (e.g. fault and bedding plane orientations) and lithological descriptions collected in the field. Fault polygons have been constructed from the modelled fault / horizon intersections in order to estimate throw variations along the Moab Fault, and hence calculate shale gouge ratios for both the course and fine scale models. The computed shale gouge ratios are compared with the actual fault rock distribution (in particular, shale content) mapped by Davatzes & Aydin (2005). The greatest differences between the course and fine scale models arise from uncertainties in fault linkage patterns and the way in which fault throws are partitioned between different segments of the Moab Fault. Our results emphasise that better understanding of fault geometry and linkage can lead to significant improvement in the prediction of the likely sealing capacity of fault systems.