Abstract

A major problem with carbonates such as the San Andres Formation is the generally poor correlation between porosity and permeability. This is the result of the varied facies that controlled the subsequent diagenesis and dolomitization. In addition to this problem, is the very limited amount of permeability data that is available from cores and how to extrapolate that information to the entire reservoir. Using GRIDSTATS (a Texaco, in-house three-dimensional geostatistical program), variograms of the permeability were done to evaluate and to quantify the vertical and horizontal spatial variability. The porosity logs were normalized and converted to permeability using a linear transform from the regression of the crossplot of core porosity versus permeability. Using the GRIDSTATS program, the “hard” data (core permeability), “soft” data (transformed normalized porosity logs to permeability), and the correlation coefficient (variance observed in the core data) were kriged to produce a three-dimensional distribution of permeability. The ability to include spatial variability and data variance and to distribute this three-dimensionally provides a very powerful tool to evaluate the reservoir. GRIDSTATS ability to generate cross sections quickly and easily through the three-dimensional model has been used to evaluate completions, pattern conformity, daily fluid production vs. kh, drill depths, and horizontal laterals from existing vertical injection wells. In addition, the software was used to calculate OOIP using the three-dimensional grids of porosity and permeability, indicating an increase of 25% over previous estimates.