Incorporating a suitable level of heterogeneity into reservoir simulation models is a key prerequisite for accurate prediction of production rates and final recoveries. Commonly-used numerical techniques impose Gaussian models for spatial and multivariate statistics in the well conditioning data and in the undrilled reservoir volume. These methods often require complex variographic procedures in order to quantify the spatial heterogeneity of a given reservoir attribute. Such Gaussian-based approaches have inadequately characterized permeability extrema that dominate flow behavoir in certain reservoirs.

This study utilizes a newly introduced method, Levy-stable fractal simulation, for the distribution of reservoir properties at a former gas storage area in the Kuparuk River Field. The main reservoir zone in the gas injection region is a diagenetically and mineralogically complex clastic unit (C-sand). Because of these complexities, the C-sand unit presents difficulties in the lateral modeling of large changes in petrophysical properties observed in near-vertical wells. Prior efforts at modeling the movement of reservoir fluids, particularly gas, in the C-sand have met with limited success. Previous reservoir descriptions have been insufficiently detailed to adequately represent the inherent heterogeneity. The Levy-stable technique employs automatic calibration with log and core data for the interwell modeling of the complex spatial characteristics of reservoir properties within the C-sand. Levy-stable fractal simulations preserve the sharp jumps in reservoir properties observed at stratigraphic boundaries and within reservoir sub-zones.

An upscaled dynamic simulation model built to test reservoir mechanisms predicts gas movement out of the injection area along layers with correlated high permeabilities. This predicted gas movement is consistent with offset well behavior and historic reservoir pressure data. Thus, areas in the former gas injection region are predicted to have zones of relatively high oil saturation that have been bypassed by migrating gas, representing potential targets. Wells drilled in areas immediately adjacent to the modeled region support the predictions of the detailed geologic description and provide confirmation of simulated dynamic behavior.

AAPG Search and Discovery Article #90928©1999 AAPG Annual Convention, San Antonio, Texas