Prediction of Reservoir Parameters Variation Based on Time-Lapse Seismic Survey

Abstract

Fluid saturation and pressure are two of most important reservoir parameters during oil and gas production scheme adjustment. A method to compute the change of fluid saturation and pressure with multi-parameters regression was presented based on time-lapse seismic inversion data. Rock physical models of unconsolidated sand rock reservoirs were determined according to the real field’s conditions to analyze how seismic attributes change with variation of reservoir parameters. The radial basis function artificial neural network which was trained by this model was used to predict saturation and effective pressure. The predicted results are of high consistency with reservoir numerical simulation, which provide valuable reference for reservoir dynamic monitoring.