Applications of Ensemble Kalman Filter for characterization and history matching of SAGD reservoirs

Abstract

Steam-assisted gravity drainage (SAGD) is the most robust thermal recovery process that has unlocked western Canadian heavy oil and bitumen reserves into economical recovery. The prime challenges in SAGD heavy oil developments and well planning in the Northern Alberta formations are: characterizing the reservoir heterogeneity and identifying the potential steam barriers that may interfere with the recovery process. If characterized earlier, the field development plans could be efficient and effective. In SAGD projects, temperature sensors at several depths within observation wells are available for monitoring steam chamber growth. Characterization using data available from these real-time sensors and dynamic production data integrated in a closed-loop could be a probable solution. Ensemble Kalman filter (EnKF), a state and parameter estimation technique, has shown good promise for reservoir characterization using dynamic production data in conventional reservoirs. For the above discussed problem, constrained based adaptive EnKF approach was implemented.