Abstract：

Fractured reservoirs are becoming increasingly important in reservoir engineering. However, due to the complexities of the fracture systems, it is difficult to predict well performance in fractured reservoirs. The dominant numerical models to deal with fractured reservoirs are introduced. These include the equivalent continuum model, the dual media model, the discrete fracture model and the embedded discrete fracture model. History matching is an essential process to improve the model predictability by using production data. A variety of automatic history matching methods have been developed, such as the gradient-based method, the evolutionary algorithm, the artificial neural network method and the ensemble Kalman filter method.Microseismic data can provide useful information to constrain modeling of the hydraulic fracture distribution, and can be used jointly with production data to improve the model prediction.

Key words：fractured reservoir; automatic history matching; microseismic monitoring

Received： 2019-07-20

Corresponding Authors： xueliang@cup.edu.cn

Cite this article：XUE Liang, WU Yujuan, LIU Qianjun, LIU Yuetian, WANG Jun, JIANG Long, CHENG Ziyan. Advances in numerical simulation and automatic history matching of fractured reservoirs. Petroleum Science Bulletin, 2019, 04: 335-346.

URL：