Engineered Geothermal Systems (EGS) have garnered significant attention as a possible source of geographically disperse, carbon-free energy without the environmental impact of many other renewable energy sources. However, a significant barrier to the adoption of EGS is the uncertainty in whether a suitably high heat extraction rate can be economically attained at a particular site. Several mechanisms, which may cause reduced efficiency or effective project failure, have been theorized. Among these, the possibility of local stress changes in the fractured rock mass along with coupling between adjacent fractures, or sets of fractures, may lead to situations where closure of large parts of the fracture network, and the coincident aperture increase of other areas, causes a reduced heat rate. Here we provide details of a coupled fluid-solid mechanics approach (using the Livermore Distinct Element Code, LDEC, and the Non-isothermal Flow and Transport simulation code, NUFT) to determine how different network topologies and flow conditions can enhance or depreciate heat flow rate.
Short-circuiting in a rock mass, where the permeability is governed primarily by the permeability of the fractures (and, accordingly, the impermeability of the matrix), can have serious deleterious effects on the heat extraction rate achievable in hot, dry rock reservoirs. Pilot projects (e.g., Rosemanowes, Hijiori, and Ogachi ) have demonstrated that short-circuiting is a major issue that needs to be addressed for successful EGS projects. Though stimulation options, such as fracture mineralization and deposition technologies, can be used to selectively decrease permeability in parts of the flow network (especially close to the wellbore), these technologies have seen little field use and provide only relatively coarse control of the permeability field. Prevention or avoidance are also options, and these require a more nuanced understanding of the role of fluid flow and geomechanics on the permeability field within the reservoir. We will demonstrate a method to couple fluid transport and solid mechanics of the rock matrix with a reservoir modeling tool for determining far-field stress, thermal, and hydraulic properties. Together, these provide an approach to estimate the sensitivity of the heat extraction rate to changes in system parameters. First, a fracture and mesh generation package developed internally is used to generate synthetic realizations of fracture networks from a stochastic description, which is derived from analyses of field data. For coupled hydro-geomechanical simulation, the Livermore Distinct Element Code (LDEC) is used to explicitly model the fully coupled discrete fracture network (DFN) and geomechanics problem in two and three dimensions. Finally, heat extraction rates are estimated using an anisotropic homogenization of the DFN onto a regular Eulerian mesh and estimated via the Non-isothermal Flow and Transport simulation code (NUFT).
Analytical solutions for the modeling of fractures and the associated deformation for given pressure boundary conditions are available, and DFN models (either Cartesian-aligned or arbitrarily oriented disks) have been implemented in reservoir-scale codes. Somewhat more recently, arbitrarily tessellated fracture network geometries have been developed  and extended [3, 4] for networks with static permeability (i.e., decoupled from geomechanics).