Simple near well models for fast generation of dynamic GOR

This report presents the use of auto regressive models as simplified near well models for fast online generation of rate dependent gas oil ratios (GOR) used in offline and real time production optimization and reservoir management. The main aim of this work is to investigate the use of simple linear and non-linear models applied to oil field data. Fast generation of GOR may improve reservoir optimization schemes, production optimization and allocation calculations, as this enable continuous access to GOR estimates. The main contribution of this work is the application of modified auto regressive models with exogenous variables (ARX) in order to generate GOR estimates for use in production optimization. The models are planned to be applied in the commercial reservoir optimization software.

Oppdragsgiver: Center for Integrated Operations in the Petroleum Industry