| Abstract |
High-precision gravity measurements can be a useful indicator of large scale changes in fluid mass within a geothermal reservoir related to production, reinjection or phase changes. In New Zealand, repeat microgravity surveys have been performed at a number of producing geothermal fields over the last 50 years. However, this data is not routinely incorporated into reservoir models used to manage these fields. The goal of this paper is to determine a methodology for calibrating a reservoir model against changes observed in microgravity, and to apply this to a high-temperature geothermal field. We created a shallow reservoir model (to -900 masl) from a TOUGH2 heat and fluid flow model of a geothermal field which had been calibrated using well enthalpy and pressure data. Gravity changes were calculated at measurement locations using density values produced by the shallow reservoir model. The calculation was run multiple times using the PEST inverse modelling software, comparing calculated gravity with measured gravity at each iteration. A number of parameters were varied to find the best-fit model, for example fluid flux at the base of the model, the extent of deep fluid flow, and rock properties including porosity, permeability, fracture volume and fracture spacing. The initial numerical model, which was calibrated only with well enthalpy and pressure data, captures first-order features detected by microgravity measurements. However, the average misfit in the microgravity data could be reduced by more than half by changing the bulk rock permeability and porosity, and the fluid enthalpy and flow rate at the base of the model. Although this represents a significant reduction in misfit, the model still does not replicate all the features of the gravity data. Therefore while repeat gravity surveys are most sensitive to changes in the shallow subsurface, they may also be influenced by large-scale changes in the deeper reservoir. |