Evaluation of Glycerol and Waste Alcohol as Supplemental Carbon Sources for Denitrification
Supplemental carbon has been successfully added and implemented at biological nutrient removal treatment plants all around the world in order to reach low nitrogen discharge limits. Although, methanol has been the most prevalent external electron donor used due to its low cost and effectiveness, many utilities are moving away from it due to cost volatility, safety issues, and hindered performance in cold weather conditions. Many sustainable and alternative sources are being researched, such as glycerin-based products (Rohrbacher et al., 2009), sugar-based waste products (Pretorius et al., 2007), and effluents from food and beverage industries (Swinarski et al., 2009).
Four 22-L sequencing batch reactors (SBRs) were utilized to investigate four different supplemental carbon sources: 100% reagent grade methanol, 100% reagent grade glycerol, bio-diesel glycerol waste, and an industrial waste alcohol. These reactors were operated at 20ï¿½"C with a 15 day solids retention time. Intensive profiles were carried out three times a week to monitor performance and collect data to calculate COD consumption: nitrate-nitrogen denitrified (C: N) ratios. The glycerol and bio-diesel glycerol waste reactors performed similarly as they both exhibited significant and consistent nitrite accumulation during the entire experiment. Based on reactor restart, nitrite accumulation was evident and significant within two days after startup and consistent for all further operation. Rapid nitrate to nitrite reduction coincident with COD uptake was also observed. The two glycerol reactors demonstrated an increased carbon demand over time. The commonly reported hypothesis that activated sludge transitions from a generalist population of ordinary heterotrophic organisms (OHO) that use substrate, glycerol in this case, less efficiently, producing low yields and slow growth rates, to a specialist population that use glycerol more efficiently, with higher yields and slightly faster growth rates, was verified. This is known as the generalist-specialist theory. While this hypothesis appears to be supported from an overall analysis of the data, the actual mechanism seems to be intracellular glycerol storage coincident with rapid nitrate to nitrite denitrification, followed by slow nitrite reduction to nitrogen gas. This can possibly lead to degradation of the internally stored glycerol in the aerobic zones of the following cycle, implying a significant economic impact with glycerin addition. Although this has not been investigated further, it is believed that the presence of glycogen-accumulating organisms (GAOs) could be responsible for this intracellular storage of glycerol resulting in partial denitrification and accumulation of nitrite.
The methanol and waste alcohol reactors also performed similarly to each other and neither of these reactors exhibited any nitrite accumulation upon carbon addition. The specific denitrification rate (SDNR) of the waste alcohol was slightly higher and increased more rapidly than for the methanol reactor. The C: N for these two reactors was comparable, and methanol was close to the expected value of 4.8 g COD utilized/ g nitrate-N denitrified. The C: N for the waste alcohol during steady state operation was somewhat higher than expected. The waste alcohol exhibited an ï¿½"alcoholicï¿½" odor upon addition to the reactors during startup, but this issue diminished as the biomass became acclimated to the waste alcohol.
Both industrial waste alcohol and glycerol can be considered viable alternatives to methanol; however, glycerol supplementation for denitrification can be problematic. If the glycerol dose is not optimized, then partial denitrification is observed and will lead to nitrite in the effluent, causing an increased chlorine demand for plants applying chlorine for disinfection. This is thought to occur due to energy limitations resulting from carbon storage and thus, using glycerol at treatment plants performing biological phosphorus removal (BPR) or enhanced biological phosphorus removal (EBPR) might see inefficient removal due to selective carbon utilization by polyphosphate-accumulating organisms (PAOs), or due to competition between PAOs and GAOs. Although denitrification of nitrate to nitrite occurs more quickly with prolonged glycerol addition, it also results in an increased carbon demand which causes a significant impact economically.