Without proper removal at a water treatment facility, the soluble manganese (Mn) concentration can reach and exceed the Secondary Maximum Contaminant Level (SMCL) of 0.05 mg/L in the water distribution system. At this level, soluble Mn can be oxidized to solid Mn-oxide particulates, leading to water discoloration events and resulting in numerous consumer complaints. Manganese-laden water can severely stain fixtures and laundry as well as increase turbidity and foul tastes. A major discoloration event can cause a decrease in consumer confidence in the quality of water provided to their taps. Currently, there is no other treatment alternative available that can remove soluble Mn with the high efficiency of the "natural greensand effect.

Therefore, researchers are developing ways to effectively create the natural greensand effect in a post-filtration sorptive contactor for application at water treatment facilities. The process of adsorption and oxidation of Mn onto oxide-coated media grains in the contactor will be used for the removal of soluble Mn. However, small media grains, such as sand or anthracite, could produce prohibitive head loss in a sorptive contactor. The focus of this research project was to show that Mn could be effectively removed via adsorption onto larger media (2.0-6.4mm) at hydraulic loading rates of 16-24 gpm/ft², thus producing less head loss and furthering the development of soluble Mn sorptive contactors to be implemented in water treatment facilities.

Research was conducted by executing laboratory- and pilot-scale experiments using columns packed with oxide-coated media. Three types of media were used: large grain "torpedo sand," pyrolucite granules, and small gravel. Before being packed into the columns, the torpedo sand and gravel media was coated with an oxide coating using a technique previously developed by Merkle (1995). Manganese uptake capacity was determined for each media type prior to use and after a number of contactor column experiments were completed. Water samples were collected during the experiments and analyzed for soluble Mn concentration. The Mn removal profile was determined by taking water samples at a certain time and at various depths in the media bed.

Experiments were conducted to determine the removal profile of the media types under different operating conditions. Hydraulic loading rate, influent Mn concentration, influent free chlorine concentration, and pH were the operational parameters varied. The effect of these parameters of the Mn removal profile was evaluated.

Although each media type was able to remove some percentage of soluble manganese from the applied water, pyrolucite media was the most effective media, often providing approximately 80-90% removal of initial manganese concentration. The removal performance of the large-sized media beds was affected by operational parameters as expected from knowledge of prior research. The contactor media beds also provided adequate soluble manganese removal under conditions available at the water treatment facility as determined from the pilot-scale experiments conducted at the Blacksburg-Christiansburg-VPI Water Authority

Another important and complementing facet of this research was the development of a proven model that would predict soluble Mn removal performance of various oxide-coated media types and the development of recommendations that could be used for implementing and operating such post-filtration sorptive contactors. A model was developed from first principles for the prediction of soluble Mn removal and fitted to the experimental data. The predictive model showed that removal performance depended on the specific surface area of the contactor media, HLR, and the mass transfer coefficient.

Recommendations for the operation of a sorptive contactor containing large oxide-coated media include an applied hydraulic loading rate of 16-24 gpm/ft² with an initial free chlorine concentration of 1.0-2.0 mg/L and a slightly alkaline pH of 7.0-8.0. Greater hydraulic loading rates are recommended to provide capital cost savings due to the decreased contactor footprint required. Alkaline pH is recommended for improved Mn removal. Facilities with a slightly acidic pH due to enhanced coagulation practices should consider adjusting the pH of the finished water for corrosion control prior to the Mn removal contactor for improved Mn adsorption performance.