Hydraulic Fracturing Design in Geothermal Wells: A Utah FORGE Case Study

Unlike traditional hydraulic fracture design, fractures in geothermal injection wells must intersect production wells. The primary objective of this paper is to design perforation clusters and fractures in geothermal injection and production wells to optimize inter-well connectivity and promote the formation of complex fracture networks. A secondary objective is to evaluate the role of natural fractures on the geometry of the fracture network created. A stochastic discrete fracture network (DFN) was generated to represent natural fractures at the Utah FORGE site based on a detailed geological characterization of the site. The propagation of multiple hydraulic fractures was modeled using a displacement discontinuity method (DDM) to account for the interaction of the hydraulic fractures with natural fractures. We evaluated key parameters that control the fracture network geometry: cluster design (number of clusters and spacing), perforation design (number in each cluster), pumping schedule, fluid rheology and proppant loading. The number of fractures that intersect the production well was estimated based on the simulation results. The optimum design was one in which the largest proportion of fractures intersected the injection well. Perforation and fracture designs in geothermal wells were shown to be quite different than in oil and gas wells. Our results show that stress shadow effects limit the number of fractures intersecting the production well. Fractures either grow tall or grow laterally to avoid the stress shadow of adjacent fractures. Among the perforation designs tested, extreme limited entry perforation designs not only made the fluid and proppant distribution more uniform but also resulted in smaller stress shadow effects. We showed that an optimum cluster spacing is needed to minimize stress shadow effects but also maximize the energy recovery from multiple fractures. Tapered perforation designs improved fracture uniformity and enhanced inter-well connectivity. A reduced ramp-up rate resulted in a more uniform fracture length and height distribution. Alternate stage fracturing (the Texas-Two-Step method) significantly improved inter-well connectivity. Results from fiber optic data used for monitoring real-time fracture propagation were consistent with the simulation results and verified the optimized hydraulic fracturing designs. We show for the first time that pumping schedules, perforation cluster design, cluster spacing, and fracture sequencing can play a crucial role in ensuring good connectivity between EGS well pairs. We also show how fracture designs in geothermal wells can be optimized. Observations from fiber optic measurements confirm many of the design recommendations.