Controls on fluviolacustrine reservoir distribution and architecture in passive salt-diapir provinces: Insights from outcrop analogs

Abstract

Fluviolacustrine strata host significant hydrocarbon volumes in basins characterized by syndepositional growth of passive salt diapirs. An understanding of salt-sediment interaction is critical to the prediction of reservoir distribution and architecture in these strata. Large-scale stratal geometries and thickness changes resulting from salt movement are commonly apparent on seismic data, but to date, there are few predictive models for facies architecture at subseismic, reservoir scale. This article uses a high-quality outcrop data set of fluviolacustrine strata in an exhumed salt basin (Upper Triassic Chinle Formation, Paradox Basin, Utah) as an analog for improved understanding of subsurface data sets of similar structural and sedimentological setting. Salt-sediment interaction in the Chinle Formation is expressed by localized lateral variations in stratigraphic thickness, angular stratal relationships, and changes in facies architecture. Based on these criteria, there is evidence for salt-sediment interaction across a series of syndepositional salt structures, including anticlines above buried salt pillows, salt walls exposed at surface, and salt-withdrawal minibasins. Stratigraphy and facies architecture across these structures reflect the following controls: regional subsidence, localized differential accommodation space, and localized paleogeomorphology. Both localized controls were driven by syndepositional salt movement, which exhibited subtle spatial and temporal variations during the deposition of the Chinle Formation. The outcrop data set is used to develop generic predictive models of facies distributions and architectures resulting from different conditions of regional tectonic subsidence and/or fluvial energy. Analysis of stratigraphic expansion across syndepositional passive diapirs suggests that the outcrop-derived models are applicable to many subsurface data sets.