ESTIMATING SPATIAL DISTRIBUTIONS OF HETEROGENEOUS SUBSURFACE CHARACTERISTICS BY REGIONALIZED CLASSIFICATION OF ELECTROFACIES

Regionalized classification of electrofacies utilizes the statistical relationships between laboratory-determined hydrologic properties and field-measured geophysical properties to estimate spatial distribution s of porosity, permeability, and diagenetic characteristics. The metho d, illustrated with an application to the St. Peter Sandstone in the M ichigan basin, combines techniques for multivariate analysis and spati al estimation, Core plug and borehole geophysical data are clustered i nto electrofacies that reflect the hydrologic properties and diageneti c characteristics of the formation. Electrofacies characteristics then are used to assign a class membership probability at locations where only geophysical darn are available. Three-dimensional estimation of e lectrofacies occurrence is done by kriging datasets containing the pro bability of electrofacies membership at borehole locations. The discre tization and kriging geometry allow three-dimensional estimation of hy drologic parameters for a large region that incorporates meter-scale h eterogeneity. Finally, permeability and porosity are estimated at each grid location by probability-weighting. Because the electrofacies car ry information about both the hydrologic and lithologic properties, th e resulting spatial distributions provide an understanding of both the present-day flow characteristics and the extent of processes that con trol them.