Agustin Arguello Scotti, Christian Haug Eide, Ingrid Aarnes, Ragnar Hauge, Jacob Skauvold, John Howell

Wave dominated, shallow marine siliciclastic successions form some of the most prolific oil and gas reservoirs worldwide, where individual parasequences comprise the main flow units. Within parasequences, these reservoirs have been traditionally modelled as simple shore-parallel facies belts, given their reputation as simple reservoirs with subtle spatial heterogeneity in petrophysical properties. However, there are many documented cases in which intra-parasequence scale heterogeneities can have a relevant impact in reservoir behavior. Some of these heterogeneities, for example clinoforms associated with bedset boundaries, are difficult to capture with current reservoir modelling algorithms. The newly developed GEOPARD algorithm provides a rule-based approach that can capture the intra-parasequence scale heterogeneity typical of this type of reservoir. In this paper, we use the very well-known outcrops of the wave-dominated succession of the K4 parasequence of the Kenilworth Member, Utah, USA, to test the ability of the algorithm to reproduce geometries of reservoir interest. Having previous studies as reference, facies and bedset boundary geometries are mapped using virtual outcrops. The GEOPARD algorithm is then presented and used to match the geometries mapped in the outcrop, and the results are used to compare model and outcrop in detail. Most importantly, the logic used to mimic geological processes with the rule-based approach of the presented algorithm is discussed. The results and discussion of this study highlight the importance of this new approach to reservoir modelling that has the potential to provide fast and robust models for many different types of reservoirs beyond the shallow marine realm.