Modeling Injection-Induced Fracture Propagation in Crystalline Rocks by a Fluid–Solid Coupling Grain-Based Model

Hydraulic fracturing, which determines geothermal resource productivity, is one of the critical technical components in the construction of hot dry rock (HDR) reservoirs. Although the mechanical coupling between solid and fluid has been included in algorithms based on the discrete element method (DEM) generally employed to investigate hydraulic fracturing, crystalline rocks of reservoirs are mostly treated as homogeneous isotropic models without considering their petrographic texture. By combining a grain-based model (GBM) and a modified fluid–solid coupling algorithm, a novel hydro-GBM is constructed in this study to analyze the hydraulic fracturing response of polycrystalline rocks. Moreover, acoustic emission (AE) events during fracturing are extracted to describe the characteristics of hydraulic-fracturing-induced seismicity. Under in-situ conditions with high differential stress, the propagation direction of hydraulic fractures is mainly perpendicular to the direction of minimum in-situ stress, with little influence by material heterogeneity and fluid viscosity; under near-hydrostatic in-situ stress conditions, the microcracks along the mineral boundaries increase remarkably, and the fracture pattern tends to be complex, especially when a low-viscosity fluid is injected into the rock. From the Gutenberg-Richter type relationship between the AE event numbers and the moment magnitudes, it is found that large induced seismic events increase with in-situ stress and with fluid viscosity. In summary, the proposed hydro-GBM can well reproduce the propagation behavior of hydraulic fractures influenced by material heterogeneity, and the research results reveal the interactions between petrographic texture, in-situ stress, fluid viscosity, and hydraulic fracturing characteristics, which will provide a valuable reference for on-site reservoir stimulation.