EVALUATION OF ALTERNATIVE ELECTRON DONORS FOR DENITIFYING MOVING BED BIOFILM REACTORS (MBBRs)
Moving bed biofilm reactors (MBBRs) have been used effectively to reach low nutrient levels in northern Europe for nearly 20 years at cold temperatures. A relatively new technology to the US, the MBBR has most typically been used in a post-denitrification configuration with methanol for additional nitrate removal. Methanol has clearly been the most commonly used external carbon source for post-denitrification processes due to low cost and effectiveness. However, with the requirement for more US wastewater treatment plants to reach effluent total nitrogen levels approaching 3 mg/L, alternative electron donors could promote more rapid MBBR startup/acclimation times and increased cold weather denitrification rates.
Bench-scale MBBRs evaluating four different electron donor sources, specifically methanol, ethanol, glycerol, and sulfide (added as Na2S), were operated continuously at 12 Â°C, and performance was monitored by weekly sampling and insitu batch substrate limiting profile testing. Ethanol and glycerol, though visually exhibited much higher biofilm carrier biomass content, performed better than methanol in terms of removal rate (0.9 and 1.0 versus 0.6 g N/m2/day.) Maximum denitrification rate measurements from profile testing suggested that ethanol and glycerol (2.2 and 1.9 g N/m2/day, respectively) exhibited rates that were four times that of methanol (0.49 g N/m2/day.) Sulfide also performed much better than any of the other three electron donors with maximum rates at 3.6 g N/m2/day and with yield (COD/NO3-N) that was similar to or slightly less than that of methanol. Overall, the yield and carbon utilization rates were much lower than expected for all four electron donors and much lower than previously reported; indicating that there could be advantages for attached growth versus suspended growth processes in terms of carbon utilization rates. The batch limiting NO3-N and COD profiles were also used to find effective Ks values. These kinetic parameters describe NO3-N and COD limitations into the biofilm, which affect the overall denitrification rates. Compared to the other electron donors, the maximum rate for methanol was quite low, but the estimated Ks value was also low (0.4 mg/L N). This suggests high NO3-N affinity and low mass transfer resistance. The other three electron donors estimated higher Ks values, indicating that these biofilms have high diffusion resistance.
Biofilm process modeling is more complex than for mechanistic suspended growth, since mass transfer affects substrate to and into the biofilm. Simulating the bench-scale MBBR performance using BioWin 3.0, verified that Î¼max and boundary layer thickness play key roles in determining rates of substrate utilization. Adjustments in these parameters made it possible to mimic the MBBRs, but it is difficult to determine whether the differences are due to the MBBR process or the model.