SUPPLEMENTAL METHANOL OPTIMIZATION FOR ENHANCED PERFORMANCE AND KINETICS IN A STEP-FEED BNR REACTOR

New York City Department of Environmental Protection and Metcalf & Eddy are researching effects of supplemental carbon on step-feed biological nutrient removal reactors with respects to improve total nitrogen removals. Based on historically low methanol kinetics (0.04 – 0.08 mg N/mg VSS/d) as compared to literature cited methanol kinetic rates (0.08 – 0.10 mg N/mg VSS/d) and rates observed with acetate (0.12 – 0.25 mg N/mg VSS/d) and considering the significantly lower costs of methanol, further investigations into optimizing methanol for enhanced denitrification is warrented. Five feed configurations were studied including (B1 and D1 addition, D1 addition, C1, D1 addition, A6, C1, D1 addition and A6, C1, D1 addition at elevated influent TN loadings) to identify optimum feed locations to overcome anoxic volume limitations that may be causing the low kinetic rates. Addition of methanol to A6, C1 and D1 exhibited performance (at both the normal and elevated influent TN loadings) similar to that of B1 and D1 addition but at a lower Supplemental Carbon to Denitrified Nitrogen ratio (3.1:1 - 4.3:1 normal N and 5.8:1 respectively). Additionally, the specific kinetic rates increased to levels similar to that of acetate. The increase in kinetics occurred when methanol was added to a zone in the pilot that had little to no wastewater carbon, thereby implying that there may be a specific anoxic methanol degrading bacteria that may face competition from other heterotrophic denitrifiers when methanol is fed in anoxic zone containing other carbon types. Thus, kinetic limitations caused by limited anoxic volumes can be lessened as seen by comparison of kinetics and performance of the various feed patterns; however, further optimization requires the separation of the effects methanol and wastewater carbon have on denitrification. This requires which the use of an ASM model fitted with equations that describe the processes related to methanol. Such a model will be employed in the next phase of optimization along with continued pilot testing including a GC-based method to deliver pilot methanol profiles.