Coupled DFN-Based Hydro-Mechanical Modeling of Hydraulic Stimulation in the Utah FORGE Reservoir

Abstract

Seismic and fiber-optic monitoring at the Utah FORGE site during the 2024 stimulation and circulation tests has enabled a refined characterization of the subsurface fracture network. The resulting discrete fracture network (DFN) provides improved constraints on fracture distribution, orientation, and connectivity. This study incorporates this refined DFN to perform a coupled hydro-mechanical analysis of the 2024 hydraulic stimulation using the discrete element method (DEM) based software XSite. To constrain fracture mechanical and hydraulic properties, a systematic sensitivity analysis is first conducted using Stage 1 of the 2022 hydraulic stimulation. Key DFN, hydraulic, and mechanical parameters are varied to identify parameter combinations that best reproduce both the observed injection pressure response and the recorded microseismicity. The sensitivity analysis demonstrates that the system’s response to fluid injection is highly influenced by the DFN realization and that the best match to field measurements is obtained by assuming an initially impermeable DFN with moderate shear strength which generally reproduces the extent of microseismic events and injection pressure history. The calibrated modeling framework is then applied to stimulation Stages 8–10 of the 2024 campaign to assess the influence of DFN geometry and fracture hydraulic behavior on pressure response, fracture activation, and flow path development. The results confirm that an initially strong and impermeable DFN provides a better match to field observations, whereas weak and permeable DFN scenarios overestimate the spatial extent of fracture reactivation due to their diffuse response. While the updated DFN captures aspects of the observed stimulation behavior, particularly shear failure extent, more realistically in Stages 8 and 9, discrepancies remain near Stage 10, indicating missing fracture pathways or uncertainties in local fracture properties. These findings highlight the importance of integrating geophysical data into DFN-based models to enhance their predictive capability and optimize reservoir development strategies in naturally fractured geothermal systems.