| Abstract |
Management of fluid reinjection is of critical importance for maintaining geothermal reservoir performance. Reinjection has posed a problem for portions of the Hellisheidi Geothermal Field, southwest Iceland, where a number of wells are drilled into active faults. The Hellisheidi Geothermal Field is located in the southern part of the Hengill Area, an active volcanic system consisting of Mt. Hengill and fracture/fault zones to the north- and south-west (Sæmundsson, 1967; Franzson et al., 2005). Injection tests have resulted in swarms of small earthquakes and with the injectivity of the wells exhibiting a high dependence on temperature of the reinjected water. Strongly coupled thermo-hydro-mechanical effects on fractures in the fracture-governed reservoir likely explain the temperature dependent injectivity. In order to better understand the temperature dependent behavior of the reservoir under reinjection conditions over short- and long-term operations, we are developing numerical simulations of reservoir behavior using the numerical code FALCON (Podgorney et al., 2010). FALCON is a hybrid finite element-discrete element model where the physics of local-scale rock deformation and fracture propagation are solved using discrete element methods while continuum multiphase fluid flow, heat transport, and large-scale geomechanics are solved using finite element methods. All governing equations are solved in a fully implicit, fully coupled manner. In this approach, the continuum equations are solved on an underlying finite element mesh with evolving stress and temperature dependent porosity and permeability for each element. This paper summarizes some of the reinjection behavior observed in the field, describes the conceptual model for the reinjection behavior in the reservoir, and documents the numerical implementation of the conceptual model into the FALCON simulator. Results of simulations are compared to pressure, temperature, and injection data from the reservoir, with details of the governing equations and constitutive relations required to reproduce the observed field data. Simulations results are also compared to analytical calculations of changes in fracture aperture necessary for the observed injection behavior. |