Improving reservoir history matching of EM heated heavy oil reservoirs via cross-well seismic tomography

Abstract

Enhanced recovery methods have become significant in the industry's drive to increase recovery rates from oil and gas reservoirs. For heavy oil reservoirs, the immobility of the oil at reservoir temperatures, caused by its high viscosity, limits the recovery rates and strains the economic viability of these fields. While thermal recovery methods, such as steam injection or THAI, have extensively been applied in the field, their success has so far been limited due to prohibitive heat losses and the difficulty in controlling the combustion process. Electromagnetic (EM) heating via high-frequency EM radiation has attracted attention due to its wide applicability in different environments, its efficiency, and the improved controllability of the heating process. While becoming a promising technology for heavy oil recovery, its effect on overall reservoir production and fluid displacements are poorly understood. Reservoir history matching has become a vital tool for the oil & gas industry to increase recovery rates. Limited research has been undertaken so far to capture the nonlinear reservoir dynamics and significantly varying flow rates for thermally heated heavy oil reservoir that may notably change production rates and render conventional history matching frameworks more challenging. We present a new history matching framework for EM heated heavy oil reservoirs incorporating cross-well seismic imaging. Interfacing an EM heating solver to a reservoir simulator via Andrade’s equation, we couple the system to an ensemble Kalman filter based history matching framework incorporating a cross-well seismic survey module. With increasing power levels and heating applied to the heavy oil reservoirs, reservoir dynamics change considerably and may lead to widely differing production forecasts and increased uncertainty. We have shown that the incorporation of seismic observations into the EnKF framework can significantly enhance reservoir simulations, decrease forecasting uncertainties and cope with the growing nonlinearity caused by the heating process for efficient and accurate reservoir forecasting.