Bayesian Long-Short Term Memory for History Matching in Reservoir Simulations
Type
Conference PaperKAUST Department
Earth Science and Engineering ProgramEnergy Resources & Petroleum Engineering
Physical Science and Engineering (PSE) Division
Energy Resources and Petroleum Engineering Program
Ali I. Al-Naimi Petroleum Engineering Research Center (ANPERC)
Date
2021-10-19Online Publication Date
2021-10-19Print Publication Date
2021-10-19Permanent link to this record
http://hdl.handle.net/10754/672970
Metadata
Show full item recordAbstract
History matching is critical in subsurface flow modeling. It is to align the reservoir model with the measured data. However, it remains challenging since the solution is not unique and the implementation is expensive. The traditional approach relies on trial and error, which are exhaustive and labor-intensive. In this study, we propose a new workflow utilizing Bayesian Markov Chain Monte Carlo (MCMC) to automatically and accurately perform history matching. We deliver four novelties within the workflow: 1) the use of multi-resolution low-fidelity models to guarantee high-quality matching, 2) updating the ranges of priors to assure convergence, 3) the use of Long-Short Term Memory (LSTM) network as a low-fidelity model to produce continuous time-response, and 4) the use of Bayesian optimization to obtain the optimum low-fidelity model for Bayesian MCMC . We utilize the first SPE comparative model as the physical and high-fidelity model. It is a gas injection into an oil reservoir case, which is the gravity-dominated process. The coarse low-fidelity model manages to provide updated priors that increase the precision of Bayesian MCMC. The Bayesian-optimized LSTM has successfully captured the physics in the high-fidelity model. The Bayesian-LSTM MCMC produces an accurate prediction with narrow uncertainties. The posterior prediction through the high-fidelity model ensures the robustness and precision of the workflow. This approach provides an efficient and high-quality history matching for subsurface flow modeling.Citation
Santoso, R., He, X., Alsinan, M., Kwak, H., & Hoteit, H. (2021). Bayesian Long-Short Term Memory for History Matching in Reservoir Simulations. Day 1 Tue, October 26, 2021. doi:10.2118/203976-msSponsors
We would like to thank CMG Ltd. for providing the IMEX academic license, and UQLab for the software license. We would also like to thank KAUST and Saudi Aramco for the support of this work.Publisher
SPEae974a485f413a2113503eed53cd6c53
10.2118/203976-ms