Mechanistic understanding of the performance difference between methanol- and glycerol-fed partial denitrification anammox in tertiary moving bed biofilm reactors treating real secondary effluent.

Two pilot-scale tertiary moving bed biofilm reactor (MBBR) treatment trains were operated onsite for 371 days in a local wastewater treatment plant (WWTP) to compare their treatment performance and mechanistic difference when methanol and glycerol were used as carbon sources, respectively. Both trains were able to meet the tertiary effluent total inorganic nitrogen (TIN) requirement of < 3 mg/L, with 31.6% ∼ 46.3 % methanol savings or 30.9 % ∼ 43.8 % glycerol savings over full denitrification projected at influent dissolved oxygen in the range of 0∼3 mg/L. Very different nitrite provision mechanism was found between the two types of carbon sources, i.e., the nitrite sink by anammox through its outcompetition of dentification was the major source of nitrite provision mechanism for anammox bacteria when methanol was used as a carbon; while the rate differential between denitratation and denitritation was the major nitrite source when glycerol was used as a carbon. The cause of this mechanistic discrepancy can be ascribed to the dramatic different half-saturation constants between the two types of carbon sources (e.g., half saturation constant of glycerol was 1.7 times that of methanol). This study provided fundamental understandings that can be used to reconcile the controversy over whether methanol is suitable for partial denitrification anammox in low strength wastewater treatment.