Coupled Stratigraphic Evolution and Fluid Flow in Sedimentary Basins

Matthew Wolinsky, Duke University

Photo of Matthew Wolinsky

 

Sediment denuded from mountain belts ultimately accumulates as stratigraphy in sedimentary basins. In these basins, hydrocarbons (oil and gas) form and are stored (in reservoir rocks). My research focuses on stratigraphic evolution in passive (continental) margin basins, (e.g. offshore US east coast). Complex stratigraphy forms in response to variations in sediment transport processes driven by sea level variations over geologic time. Basinward decline in transport “energy” produces distinct boundaries between depositional environments (e.g. shoreline/shelf-break for sand/mud), where deposition is concentrated, and simulation of stratigraphic evolution over long timescales must account for movement of these boundaries. Since most sediment transport models are aimed at much shorter event timescales (e.g. a single storm or flood), models must be scaled up to compute effective sedimentation/erosion rates over stratigraphic timescales.

Rapid deposition leads to pressurization of pore water inside the growing sedimentary deposit, which can lead to slope instability, and low permeability shales (“mudstones”) can allow overpressures to survive over long timescales. Thus risks of offshore oil-well blowouts and tsunamigenic landslides are intricately linked with the depositional history of a basin. An exciting part of my research is the co-evolution of stratigraphy and pore-pressure over large space and time scales. Previous models have focused on small areas such as the continental slope, using a fixed domain. However, over the long timescales associated with pore pressure evolution, migration of the slope occurs, making such approaches unrealistic. A dynamic meshing technique I developed overcomes this limitation, allowing arbitrary migration of localized depocenters. Another issue is the high aspect ratio of continental margins, which leads to ill-conditioning of pore-pressure diffusion unless vertical node density is small. I developed a sub-grid parameterization of hydraulic parameters to allow this without loss of accuracy.

Abstract Author(s): Matthew A. Wolinsky