A Case Study of Scaling Up 2D Geostatistical Models to a 3D Simulation Model

Abstract

ABSTRACT This paper reviews the procedure used in a recent simulation study in which results from two-dimensional (2D) cross-sectional geostatistical models were scaled up to a three-dimensional (3D) simulation model. Geostatistics (which included fractals) were used in the cross-sectional models to incorporate the effects of fine-scale heterogeneities in the coarsely-gridded, full-field model. Areal and vertical variograms were developed from well log data and used in the realizations. The geostatistical models were modified after the conditional simulations to preserve the integrity between each geostatistical model and the known geologic model. For this waterflood simulation study, the scale-up process involved matching the flood-front velocities and bypassed/trapped oil saturation values in the geostatistical models to those properties in the coarsely-gridded models. Although the scale-up procedure was specific to this reservoir study, the overall process can be modified for application in other studies.