Geochemical drivers of manganese removal in drinking water reservoirs under hypolimnetic oxygenation

Abstract

Manganese (Mn) is a naturally occurring contaminant commonly found in drinking water supplies. In lakes and reservoirs, water authorities increasingly use in situ treatment by hypolimnetic oxygenation (HOx) systems to remove metals such as Mn from the water column. HOx systems introduce dissolved oxygen (DO) to the bottom waters (hypolimnion) to promote oxidation and subsequent removal of metals from the water column. Previous laboratory studies have shown the importance of individual geochemical drivers (pH, alkalinity, mineral surfaces) on Mn oxidation, but few studies have examined the influence of these drivers of Mn removal in concert. In this study, we conducted field monitoring and laboratory experiments to examine how pH, alkalinity and the presence of mineral particles influence Mn removal at two drinking water reservoirs in southwest Virginia, both with HOx systems: Falling Creek Reservoir (FCR) and Carvins Cove Reservoir (CCR). Both reservoirs have had historical issues with elevated (>0.05 mg/L) Mn concentrations during seasonal stratification (May–October). Watershed geology contributes to differences in pH and alkalinity between the reservoirs, with FCR having lower historical medians of hypolimnetic pH and alkalinity (6.6 and 18 mg/L CaCO₃, respectively) than CCR (7.2 and 62 mg/L CaCO₃, respectively).

Results of laboratory experiments examining the influence of pH on Mn removal showed substantial Mn loss within 14 days only under high pH (10) conditions. Mn removal did not occur at pH 6 or 8 over the same 14-day period. In experiments with pH 10 and alkalinity >70 mg/L CaCO₃, near-total Mn removal occurred within 2 h. Mn removal occurred concurrently with precipitation of microscopic (<5 μm) particles, followed by formation of macroscopic (>100 μm) particles. Particles of both size classes were identified as Mn oxides (MnOx). These observations suggest that increasing pH and alkalinity promotes Mn oxidation and subsequent removal from solution. Results of experiments with pH 10 and alkalinity >70 mg/L CaCO₃ suggest that heterogeneous oxidation by MnOx partially drives rapid Mn removal. Thus, initial formation of MnOx creates a positive feedback loop that can enhance additional Mn loss. In experiments using water collected from FCR and CCR, we observed rapid Mn removal in unfiltered water (0.002–0.05 d⁻¹) but no significant removal of Mn in filtered water. These results, in combination with results of analysis of particles collected from reservoir water, suggest that minerals present in the water column likely catalyze MnOx formation. Together, our experimental results suggest that heterogenous oxidation is an important process of Mn removal, while pH and alkalinity variations of the range expected in natural freshwaters contribute less to differential Mn removal. The formation of MnOx particles during in situ oxygenation, as well as the presence of suspended minerals that occur naturally in water columns, play an important role in promoting Mn oxidation and should be accounted for in Mn removal treatment strategies.