A Numerical Simulation Approach for Shale Fracture Network Characterization Using Hybrid EDFM Method

Abstract

Abstract Efficient and economical production in shale reservoirs requires specific knowledge and techniques to characterize the fracture network, which consists of both the large hydraulic fractures and the natural fractures. Current research mainly focuses on the hydraulic fracture part, while research on characterizing the natural fractures is relatively rare due to the complexity and uncertainties of the natural fracture distribution in reservoirs. This paper proposed a numerical method for characterizing this complex fracture network in shale reservoirs which handles both the hydraulic fracture and the natural fractures at full dimensions using a combination of microseismic data, geostatistical information, fracking plan, and reservoir simulation. We use microseismic clouds to generate mappings of hydraulic fractures and choose the most appropriate set using the fracking schedule and flowback simulation. The natural fractures are treated using a fractal probabilistic model calibrated with geostatistical information including fracture length, density, and intensity. Both the hydraulic and fractal natural fractures mentioned above are then modeled using embedded discrete fracture model (EDFM) and are coupled to create the hybrid fracture network and applied with reservoir simulation grids. Numerical simulation is performed to investigate further into the fracture network, and history matching is applied to reduce the uncertainty by select the most appropriate fracture network from multiple realizations. Sensitivity analysis is also performed to analyze the controlled natural fracture parameters on total production. We further used our approach to study a horizontal well performance in the Permian Basin and compared our simulation results with traditional fracture characterization and model methods. Our established work has shown the best agreement with history production data compared with randomized fracture model and the multimedium method. This provides a novel method to characterize fracture networks in shale and may be used to guide production optimizations and plan management.