Mark H. Holtz, Dr. Hongliu Zeng, Paul Knox, and Mike De Angelo
Bureau of Economic Geology
John A.& Katherine G. Jackson School of Geosciences
The University of Texas at Austin

ABSTRACT
The construction of a 3-D geocelluar reservoir model and the population of its cells with petrophysical properties are best accomplished by integrating all available information. Because the information collected and described comes from many disciplines, however, it is at several different scales. Typically the data are 1-D logs rich in vertical, small-scale, petrophysical data but having very limited lateral extent. The 3-D seismic, on the other hand, is rich in lateral, larger scale data that have no direct measurement of the petrophysical properties. Using reservoir engineering data, we can calculate petrophysical properties from fluid flow characteristics representing a pressure compartment scale. Additionally, many other properties needed in determining the petrophysical character of a reservoir are not measured at all but instead are inferred from other information.

Complete Integration of geophysical, geologic, and reservoir engineering data results in the development of the most plausible model, completely populated by petrophysical properties. The 3-D seismic volume yields not only the structural architecture but also a large-scale distribution of shale volume that is based on neural network modeling of seismic and wireline data. Large-scale, seismically derived shale volume can be combined with a small-scale 3-D shale volume constructed from well control and geologic depositional facies description. Functions of the interrelationships between all salient petrophysical properties, including porosity, permeability, irreducible water saturation, capillary pressure, and residual gas saturation, are developed so that reservoir volumetrics can be calculated and then compared with material balance volumes. The 3-D model is then optimized to give the most geologically plausible sand-shale distribution while maintaining a petrophysical property distribution that is consistent with pressure and production data. Model optimization also includes the distribution of residual gas saturation values on a cell-by-cell basis so that material balance to aquifer influx is accounted for.