Origin of sodium bicarbonate groundwaters, Southern Hills Aquifer System, USA by silicate hydrolysis

Abstract

Groundwaters dominated by sodium-bicarbonate represent an important end-member continental water type. The Southern Hills aquifer system in southwestern Mississippi and southeastern Louisiana, USA provides an excellent field laboratory for studying the origin and evolution of Na–HCO3 groundwaters in siliciclastic aquifers. Our research, which is based on establishing spatial variations in water chemistry, mass balance relations, and thermodynamic considerations, supports the hypothesis that the process of generating Na–HCO3 groundwaters in the Southern Hills aquifer system is driven by irreversible dissolution-precipitation reactions primarily involving silicate minerals, not by calcite dissolution and cation exchange, as has been proposed by others. The evidence supports the following scenario: oxidation of organic carbon produces organic acids which react with detrital plagioclase feldspars having an average bulk composition of approximately Ab0.70An0.30. In the up-gradient part of the system, the incongruent dissolution of plagioclase produces a kaolinite phase and releases dissolved silica, Na, Ca, and HCO3 into solution. As the silica concentration and the sodium/hydrogen ion activity ratio in the waters increase down gradient, the waters become saturated with respect to a (Na,Ca)-smectite. As plagioclase hydrolysis continues as a result of production of carbonic acid, there is a progressive increase in (Na+/H+) and (Ca2+/(H+)2) activity ratios and now a progressive decrease in dissolved H4SiO40 as a result of buffering by kaolinite-smectite. It is likely that some Ca is preferentially removed from solution over Na by adsorption on newly-formed smectite. In the most distal parts of the aquifer system the high bicarbonate alkalinity and high pH result in the waters becoming saturated with respect to calcite, and the precipitation of calcite is another probable sink for Ca. Using dissolved bicarbonate as a reaction progress variable provides a useful technique for evaluating controls on groundwater compositions. It is also clear that the evolution of groundwater compositions in this and similar systems cannot be understood without understanding the controls on dissolved silica.