Most drinking water production plants use rapid sand filters for the removal of manganese from groundwater. The start-up of manganese removal on newly installed sand media is slow, taking several weeks till months. Reducing this period in order to prevent the loss of water during this phase has become an issue of concern. In this study pilot and bench scale experiments were conducted to investigate the mechanism, influence of operational conditions (e.g. filtration rate, manganese loading) and measures that enhance manganese removal capacity of the sand media. Other filter media were investigated with the objective of finding suitable substitutes for the sand. The development of the adsorptive/catalytic coating the sand media in a pilot plant was very slow, notwithstanding the relatively high pH of 8. Low manganese concentration and more frequent backwashing resulted in a longer start up period of the manganese removal. It can not be excluded, that nitrite has a negative effect as well. Measurements in the lab and bench scale tests show that the rate of adsorption/oxidation of manganese in the top layer of the filter bed is too low to explain the complete manganese removal in the filters. It is likely that the adsorptive catalyst in the top layer has partly been covered with ferric hydroxide. From the Freundlich's isotherms determined for 6 different filter media, a crushed medium consisting of mainly manganese dioxide and some silica, iron and aluminium (Aquamandix) followed by iron oxide coated sand, containing some manganese, demonstrated highest adsorption capacities at pHs 6 and 8. These materials can therefore be potential substitutes for sand in situations of slow start-up of manganese removal.
- filter media
- © IWA Publishing 2009