Manganese removal by media filtration: Release and complexation

An investigation of manganese release from manganese oxide coated filter media, observed in a water treatment plant (WTP), was performed. The study involved collection of the ground water (GW) and the full scale green sand (GS) filter media, 16S rRNA gene sequencing of the extracted DNA from freshly sampled GW and GS, laboratory column experiments with the collected GS and GW followed by 16S rRNA gene sequencing of the DNA from GS used in the laboratory column experiments. In 16S rRNA gene sequencing of the DNA collected from the GW and GS, Geobacteraceae, Burkholderia ferrariae, Galionella feruginea, Ochrobactrum, Mesorhizobium, Pseudomonas and Methylobacrum were found as the most dominant species. Among the sequenced microorganisms, ∼20% for the GW and ∼45% for the GS microorganisms were manganese reducers. In laboratory column experiments, a significant level of manganese release was observed from the column with adequate carbon source and non-autoclaved media. The manganese release continued as long as the carbon source was available and no chlorine was added. 16S rRNA gene sequencing results, performed with the DNA from GS before and after the application pre-filter chlorine, showed a significant decrease in the manganese reducers (40% as compared to 97% for the sodium acetate fed non chlorinated situation). The molecular analysis results are in excellent agreement with the laboratory column experiment observation and strongly suggest that the manganese release can occur in GW treatment plants due to the activity of manganese reducing microorganisms; a high level of solution phase free chlorine can limit the release. Possibe manganese-NOM interaction/complexation in raw, filter influent and filter effluent water samples collected from several WTPs was investigated via the high performance size exclusion chromatography (HPSEC) method with continuous UV detection coupled with an inductively coupled plasma mass spectrometer (ICP-MS). The HPSEC-UV254-ICPMS results showed two major UV 254 peaks and one manganese peak at different HPSEC elution times suggesting no significant NOM-manganese complexation. In laboratory column experiments with full scale filter media, some NOM adsorption was observed for media grains having a larger aluminum to manganese coating ratio, however no correlation could be demonstrated regarding how different molecular size fraction of NOM may impact manganese removal by the oxide coated media