Spatial and temporal variations of manganese concentrations in drinking water.

The objective of this study was to assess the variability of manganese concentrations in drinking water (daily, seasonal, spatial) for eight communities who participated in an epidemiological study on neurotoxic effects associated with exposure to manganese in drinking water. We also assessed the performance of residential point-of-use and point-of-entry devices (POE) for reducing manganese concentrations in water. While the total Mn concentrations measured during this study were highly variable depending on the location (< 1-2,700 μg/L), daily or seasonal variations were minimal. Flushing the tap for 5 minutes did not significantly reduce total manganese concentration for 4 out of 5 sampling locations. The efficiency of reverse osmosis and ion exchange for total Mn removal was consistently high while activated carbon provided variable results. The four POE greensand filters investigated all increased (29 to 199%) manganese concentration, indicating deficient operation and/or maintenance practices. Manganese concentrations in the distribution system were equal or lower than at the inlet, indicating that sampling at the inlet of the distribution system is conservative. The decline in total Mn concentration was linked to higher water residence time in the distribution system.