Better constraining silica-enthalpy mixing models in a setting of two separate (karst and non-karst) dilution regimes

Abstract

It would be possible to elucidate an ambiguity associated with the “silica-enthalpy mixing models” geothermometric technique, provided that an independent estimate was obtained concerning the enthalpy and the dissolved SiO2 content of the hot end-member involved in a mixture between two liquid-phase groundwaters. This would allow one to ascertain whether or not, prior to mixing with the cold parent-water, the hot end-member had undergone a boiling process. The specificities of the Băile Herculane geothermal system (South Carpathians, Romania) provides an opportunity to solve this uncertainty, in contrast to the majority of the so-far investigated natural settings. Advantage has been taken, in this respect, of the fact that a hot brackish Na-Ca-Cl-type parent-water was found to be involved at this site in two separate mixing settings: one which involved only a diluting fluid that was, itself, chloride-rich, and another one which additionally involved, as a third end-member, karst freshwater. By exploiting the two distinct, and converging, linear mixing trends (obtained from the conservatively-behaving natural tracers Cl− and Na+, as well as reasonably tight correlations between the chloride concentration and the contents of both heat and SiO2), the concerned hot brackish end-member contents of heat (enthalpy) and of conservative solutes were quite accurately assessed. This result enabled independent validation of the “no steam loss” hypothesis, out of various alternative scenarios proposed by the “silica-enthalpy mixing models” technique.