Stochastic surface modeling of deepwater depositional systems for improved reservoir models

High resolution reservoir models are often required for resource assessment and reservoir management. The reservoir models must reproduce geologically realistic heterogeneity that may affect predictions. An important heterogeneity in many deepwater depositional systems consists of facies variations controlled by geological events that lead to lobe geometries. A surface-based geostatistical scheme is presented to model high resolution lobes and their associated stacking patterns in a geometrically realistic manner. The lobes are constrained to match well data. Facies, porosity and permeability are then modeled within the lobe framework. Application to a reservoir with 36 wells demonstrates the practical applicability of the approach. This surface-based method may be readily extended to model other geometries in various reservoir settings.