RESERVOIR MODELING OF INTRAPARASEQUENCE HETEROGENEITY BY MIMICKING SEDIMENTARY PROCESSES: A CASE STUDY FROM THE KENILWORTH MEMBER, BLACKHAWK FORMATION

Abstract:

Wave-dominated shallow-marine siliciclastic successions form some of the most prolific oil and gas reservoirs worldwide. When modeling this reservoir type, it is common practice to use parasequence boundaries to define zones and model facies associations within these zones as shore-parallel belts. In contrast to this relatively simple approach, a growing number of studies are documenting intraparasequence-scale heterogeneities that can affect flow behavior. Some of these heterogeneities, for example, clinoforms, are difficult to capture with current reservoir modeling algorithms. The newly developed Geometric Pretraining for Articulation Prediction in Three-Dimensional Shapes (GEOPARD) algorithm is a rule-based, process-mimicking modeling tool that was designed to capture the geometry and intraparasequence-scale heterogeneity typical of these reservoirs. The aim of this studyis to use GEOPARD to model the well-known outcrops of the K4 Parasequence of the Kenilworth Member, Utah, to test its ability to reproduce reservoir geometries and heterogeneities and better understand how the algorithm mimics geological processes. A down-dip-oriented section of the outcrop was created by mapping flooding surfaces, facies associations, and bedset boundaries on virtual outcrops. GEOPARD was then used to model the succession, mimicking different sets of processes interpreted to have had an impact in the record of the studied succession, and its results were compared with the outcrop section. Analysisof the approach taken in this study shows that the algorithm uses its built-in rules and its input variables to capture the direct effects of processes on the geometry of the shoreline and its changes in time. The algorithm's output is a grid with three-dimensional bedsets and facies associations within them that can be conditioned to well or log data. This study highlights the importance of this new type of approach to reservoir modeling that has the potential to provide fastand robust models for many types of reservoirs with complex internal heterogeneity.