Scenario-Based Data Assimilation Framework to Improve Production Estimates for Geologically Complex Geothermal Reservoirs

Abstract

Geothermal energy has the potential to decarbonize heating, cooling, and power production. However, efficient and sustainable exploitation is challenging due to the limited data, which restricts our ability to characterize and quantify the multi-scale, hierarchical geological structures of geothermal reservoirs. This study proposes a scenario-based data assimilation framework that enables efficient modelling of multiple geological scenarios and is coupled with flow and heat transfer simulations for uncertainty analysis. We demonstrate the framework on a synthetic but geologically consistent low-enthalpy geothermal case, where heat is produced from a doublet in a channelized fluvial sandstone reservoir. Using the open-source Rapid Reservoir Modelling (RRM) tool, we efficiently create multiple geological scenarios with different reservoir architectures constrained by wellpath facies data. Reservoir properties for each scenario are modelled using geostatistics to generate a geologically plausible and sufficiently diverse ensemble of realizations. These are subjected to heat and flow simulations using the open-source Delft Advanced Research Terra Simulator (open-DARTS) to quantify uncertainties in temperatures and pressures. Finally, ensemble smoother with multiple data assimilation (ESMDA) is employed to assimilate temperature and pressure profiles at the injection well, monitoring borehole, and production well across all ensemble realizations for the different geological scenarios. Based on the deviation of modelled and observed well temperature and pressure profiles, the framework falsifies geological scenarios with poor data assimilation performance that is unlikely to reflect the actual reservoir architecture, and identifies plausible scenarios that yield more reliable production forecasts. The workflow effectively constrains geological uncertainties and improves forecast reliability in geothermal systems.