Evolution of permeability in fracture networks: Geophysical and geochemical effects in enhanced geothermal systems

Dominant mechanisms for permeability change in hydrothermal fracture networks are driven by the combined action of thermal, hydrologic, mechanical, and chemical forcings. Potential mechanisms include, but are not limited to, thermomechanical deformation, mineral reaction, shear dilation, and chemical-mechanical creep. While some effort has been devoted to examining each of these processes individually, magnitudes of relative interaction remain poorly constrained at small- and largescale. In this work, a numerical simulator is used to model these processes at reservoir scale. Permeability and porosity are modified as fractures dilate or contract under the influence of pressure solution creep, thermo-hydro-mechanical compaction/dilation, and bulk mineral reaction in a deformable, dual-porosity medium. Simulations focus on a prototypical enhanced geothermal system as cold (70°C) water is injected at geochemical disequilibrium within a heated reservoir (250°C). For an injector withdrawal doublet, separated by 500m, the results demonstrate the strong influence of mechanical effects in the short term (several days), the influence of thermal effects in the intermediate term (<1 month at injection), and the prolonged and long-term (>1 year) influence of chemical effects, especially close to injection. Differences are examined between small scale, frequent fractures and large scale, more widely spaced fractures. Permeability increases more for widely spaced fractures as the thermal strain is distributed on fewer fractures per unit volume of reservoir. This in turn results in a lowering of fluid temperature in the reservoir as fluid throughput increases (under constant pressure drive) and as thermal diffusion length to the fracture concomitantly increases. A contact area based model for pressure solution creep is retrofit into the simulator and results indicate potential importance for pressure solution at reservoir scale. However, an equilibrium simplification is incapable of examining long term compaction trends, and a kinetic based form may be necessary to reproduce these large scale behavior.