Multiscale modeling of reservoir systems using geostatistical methods

Abstract

Identifying factors and quantifying the processes, influencing the extent of reservoir rocks and the distribution of their heterogeneities are fundamental prerequisites to improve the characterization of reservoir connectivity and its dynamic functioning. In this paper, we propose a multiscale approach based on combined geological–geophysical investigations, geostatistics, and 3D geological modeling to build a structural model of faulted-reservoir systems, at regional and local scales within the Jeffara basin, southeast Tunisia. The regional modeling procedure is calibrated by outcrops, a Digital Elevation Model (DEM), boreholes, and well data logs. The geological processing, itself, is used as a tool to extend the database from a variety of documents such as geological maps, geological cross-sections, outcrop descriptions, and especially the wealthy geological knowledge. The original and extended data are then coded and stored in a common georeferenced database. This provides a detailed input for the geostatistical modeling procedures that enabled to precisely capture the varying extent and shape of the hydrogeological units at regional scale. Geophysical data, available along the Jeffara plain, were added as a complement to the Jeffara geodatabase and used to establish the fault network and to build a local architectural model of the coastal aquifer system, based on time-to-depth conversion using kriging with external drift. The results have allowed (i) a major update of the geological configuration of the Jeffara basin, (ii) more precision on the geometries and extent of the stratigraphic sequences and (iii) an accurate prediction of the main characteristics of water reservoirs, i.e. their occurrence, thickness, facies and dynamic properties variation at the basin and reservoir scale. Such multiscale modeling provided effective tools for better understanding the hydrostratigraphic setting, the hydrodynamic functioning of neighboring aquifers, and to assist water resources management.