ABSTRACT

During the last few years, retrenchment of the oil industry from costly exploration programs has redirected activities toward the enhanced exploitation of known reservoirs. As a result, research and developments have focused on advanced and detailed reservoir description. A significant aspect of reservoir description is the transformation of qualitative geological characterization to the appropriate, numerical, reservoir grid block models, critically controlling reservoir performance forecasting.

This paper presents a two-step methodology for transforming detailed geological description of reservoir heterogeneity to engineering-useable models, incorporating all available information and interpretations. The first step consists of the simulation of the reservoir lithofacies, conditional to all available information and data. This simulation is performed using a novel technique, generating alternative images of reservoir heterogeneity at any desirable scale. The second step consists of the modelling of reservoir rock properties combining the simulated images of the reservoir lithofacies with geostatistical techniques, quantifying geological characteristics, such as continuity, trends, and anisotropies. Of particular importance is the determination of effective block permeability, which is performed using a new technique. Accordingly, block horizontal and vertical permeabilities are considered as spatial power averages of point support permeability values, which are generated within each block using the geological-statistical characteristics of the reservoir and available core data.

The two-step methodology and corresponding techniques described above are demonstrated in an application at the Crystal Viking Field, south-central Alberta. Crystal is a substantial and fully characterized reservoir, producing since the early eighties.

End_of_Record - Last_Page 210-------