Use of Channel Information Update and Discrete Cosine Transform in Ensemble Smoother for Channel Reservoir Characterization

Abstract

Ensemble Kalman filter (EnKF) is one of the powerful optimization schemes for production data history matching in petroleum engineering. It provides promising characterization results and dependable future prediction of production performances. However, it needs high computational cost due to its recursive updating procedures. Ensemble smoother (ES), which updates all available observation data at once, has high calculation efficiency but tends to give unreliable results compared with EnKF. Particularly, it is challenging to channel reservoirs, because geological parameters of those follow a bimodal distribution. In this paper, we propose a new ES method using a channel information update scheme and discrete cosine transform (DCT). The former can assimilate channel information of ensemble models close to the reference, maintaining a bimodal distribution of parameters. DCT is also useful for figuring out main channel features by extracting out essential coefficients which represent overall channel characteristics. The proposed method is applied to two cases of 2D and 3D channel reservoirs and compared with EnKF and ES. The method not only provides reliable characterization results with clear channel connectivity but also preserves a bimodal distribution of parameters. In addition, it gives dependable estimations of future production performances by reducing uncertainties in the prior models.