Data Assimilation Using Ensemble Kalman Filter ( EnKF)

Wisam Shaker
Supervisor: Prof. Zhangxing (John) Chen


This presentation shows that a single model does not represent the range in uncertainty. History match optimizers can get too focused on finding the “best” model. The EnKFsucceeded to estimate the permeability distribution close to the same of the known truth case. No correlations were found between pressure, temperature and rate mismatch quality components. A significate reduction in the uncertainty range has been gained in the range of 50-70% when the production data is available every 3 months. The updated ensemble is a good representative of the reservoir uncertainty range.