The results of computer simulation of the El Dorado surfactant-polymer EOR pilot project, Butler County, Kansas indicated that conventional data from the project and other data in the public domain were not adequate for geologic, reservoir and process characterizations in a complex numerical simulation. As used by GURC in geologic characterization, and by INTERCOMP in process characterization and input into the CFTE simulator, the collective body of field and chemical data and related assumptions necessary for simulator input was not sufficient to predict how the chemical flood would behave in the Admire 650-foot sandstone reservoir. Based upon this study, a comprehensive body of data requirements for EOR simulation is defined in detail. Geologic characterization includes descriptors for rock, interwell and intrasystem correlations; reservoir characterization includes descriptors for fluid/rock, production, and flow rate properties; process characterization includes descriptors for chemical properties, interactions and functions. Reservoir heterogeneity is a principal problem in EOR simulation. It can be overcome within reasonable economic limits by successive orders of descriptors from: microscale (rock), achieved through borehole and core analyses; to macroscale (interwell), achieved through multiple borehole correlations; to megascale (intrasystem), achieved through extrapolation of rock and correlative well data into a generic depositional model that more »