Shale Gas Production Forecasting using Reservoir Simulation with Hydraulic Fracture Mapping
This dissertation presented a hybrid Embedded Fracture Dual Porosity (EF-DP) model for shale gas production forecasting. We presented a new correlation between the microseismic magnitude and the shape factor coefficients in the classic dual-porosity model accompany a procedure to calibrate this coefficient with real production data. The novel contribution was the employment of the shape factor in the dual-porosity model to characterize a relatively complex small scale fracture network, which numerically integrates with the large scale fractures geometry in the EDFM model to forecast shale gas well production. A constrained result of using microseismic data to measure the length and direction of the large scale fractures was put into the embedded discrete fracture model (EDFM) to integrate the large scale fracture into a corner point grid. The EF-DP model considered stimulation data such as total fluid volume for hydraulic fracturing, flow rate, wellhead pressure, sand concentration, and proppant size. To verify the credibility of this model, we performed two sets of parameter sensitivity analyses for the large scale fractures and the small scale fractures. We used two sets of real-world shale gas production data for history matching and successfully used the EFDP model to quantify analysis the impact of frac-hit on a shale gas producing well. Parameter sensitivity analysis confirmed that enhanced small scale fracture permeability could effectively increase production, mainly by strengthening far-field reservoir drainage volume. According to the application results, we found that the EFDP model was effectively and accurately predict shale gas production, and quantitatively evaluate the impact of frachit between multiple wells. The refracturing candidate selection results guided further well completion strategy improvement. Microseismic-based approaches provide a robust fracture network model, further reservoir modeling calibration and simulation studies can reveal invaluable information about the active stimulated zone.