| Keywords |
geothermometer, GeoT, iGeoT, optimization, modeling, saturation index, water chemistry, reservoir temperature |
| Abstract |
The computer program GeoT was previously developed to reconstruct the composition of deep geothermal fluids prior to degassing and mixing at shallow depths, and to estimate deep reservoir temperatures from the computed saturation indices of multiple minerals using complete fluid analyses collected from wells or springs. This multicomponent geothermometry approach, combined with numerical optimization, has shown promising results for fluids that have not fully equilibrated with reservoir minerals and/or that have been subject to dilution and/or gas loss. However, the assumption of fluid-mineral equilibrium on which all solute geothermometry methods rely, uncertainties in model input parameters (including thermodynamic data), and different optimization methods all introduce uncertainties in model results. Here we report on recent capabilities implemented into GeoT V2.0 to increase its range of applicability, and on various investigations of result uncertainty. GeoT was upgraded to allow equilibrium reaction of one or more minerals (e.g., calcite) when making temperature estimations, to take into account the effect of potential re-equilibration of fast-reacting minerals (e.g., calcite) along the fluid cooling path. An approach was also added into GeoT to automatically estimate (iteratively) steam weight fraction for degassed samples, assuming iso-enthalpic boiling from the estimated temperature at saturation pressure (pure water), down to a given sampling pressure. Finally, the optimization engine of iTOUGH2 was integrated into a new code version, iGeoT, thus allowing the estimation of any input parameter (and its uncertainty) necessary for the reconstruction of deep fluid composition without recourse to separate numerical optimization software. Various examples using both real and synthetic geothermal waters are discussed, including the effect of different optimization methods, objective functions, mineral-fluid reactions, and input thermodynamic databases on estimated deep reservoir temperatures. |