Fluoride in thermal and non-thermal groundwater: Insights from geochemical modeling

Abstract

High fluoride (F) groundwaters (>1 mg/L) have been recognized as a water quality problem for nearly a century and occur in many countries worldwide. The affected aquifers can be sedimentary, metamorphic or igneous rocks, but the process giving rise to high-F concentrations has been studied with geochemical modeling and an examination of the rock sources. The association of high-F with silicic igneous rocks such as granites and rhyolites results from magmatic differentiation (fractional crystallization, fractional melting, and crustal assimilation) wherein F is enriched in the liquid phase because of its incompatibility in the mafic minerals that crystallize early during cooling. Further development of F-rich groundwaters occurs during the evolution of Na-HCO3 waters because of removal of Ca through ion-exchange and calcite precipitation, thereby raising the F concentration from minerals like fluorite and fluorapatite to maintain solubility equilibrium. Increasing temperatures enhance this effect because of the retrograde solubility of calcite. From geochemical modeling using the PhreeqcI code, the primary variables controlling F concentrations are DIC (dissolved inorganic carbon), salinity (ionic strength), PCO2, and temperature. Complexing is also important but plays a more secondary role. Considering these variables, an improved set of plotting parameters, F/Cl vs. HCO3/Cl, are shown to be effective in interpreting groundwater analyses. This approach is demonstrated by examining case studies from the Black Creek aquifer, South Carolina, USA, the Madison regional aquifer, midwestern USA, the Mizunami Underground Research Laboratory, Japan, New Zealand thermal waters, the San Luis Valley groundwaters, Colorado, USA, and the Aquia aquifer, Maryland, USA.