IMPACT OF PREFERMENTATION ON WASTEWATER AND BIOKINETIC PARAMETERS FOR BIOLOGICAL NUTRIENT REMOVAL

Introduction The ability of a prefermenter to enhance EBPR performance in BNR systems through increased VFA is well established in the literature (WEF, 1998; McCue et al., 2003; McCue et al., 2004). Less well understood are other potential benefits of prefermentation upon BNR performance. This study focused upon changes in wastewater and biokinetic parameters due to prefermentation, including RBCOD, the maximum specific growth rate for autotrophs and inert COD fractions. Using these experimentally determined values changes in BNR performance with and without prefermentation were modeled using Biowin by replacing standard default values rather than using included prefermentation modules or other existing prefermentation software. Later results will be compared to existing modules. Results and Discussion Prefermentation was found to significantly increase the RBCOD in both COD-limited and Plimited wastewaters. RBCOD values were high even prior to prefermentation due to a highly septic Florida wastewater (and 70mg/L plus of the RBCOD was in the form of VFAs). Typically it is assumed prefermentation will have no benefit for septic wastewaters but prefermentation was beneficial for COD-limited wastewaters in this study, with the RBCOD:TP ratio being more important in predicting the benefit of prefermentation than the absolute value of RBCOD or VFA. The influent RBCOD:TP ratio is now thought to be more predictive for phosphorus removal than the more established VFA:TP ratio, with a ratio of 11 or 12:1 implying a sufficient RBCOD:TP ratio (WERF, 2006). VFA:TP ratio requirements vary widely in the literature with values from 3:1 to 16:1 being cited (WERF, 2006; Metcalf and Eddy, 2003; Daigger et. al., 1993). Our data was consistent with improvement of EBPR for increasing RBCOD:TP ratios up to 12:1, with no improvement at higher ratios. Prefermentation increased RBCOD by 23% and 29%, with P-values of 0.0001 and 0.002 for COD and P-limited wastewaters, respectively. In this study prefermentation increased not only VFA, but also RBCOD. It is known that RBCOD can be fermented to VFA, but the conversion of SBCOD to RBCOD remains an area of disagreement. The observations here confirm that prefermenation can convert non-RBCOD (e.g. SBCOD) to RBCOD. It has not been experimentally confirmed that this transformation occurs in the anaerobic zone of BNR systems, even though this is assumed in most current dynamic BNR models. However there is at least anecdotal evidence from lab, pilot, and full scale systems that this conversion is not significant in many, if not all, BNR anaerobic zones (Wentzel, personal communication). If it is confirmed by later work that only prefermentation can increase the RBCOD of a wastewater (and thus the potential P removal), while anaerobic zones can, at least in some cases, only convert pre-existing RBCOD to VFA this could have an impact on process design decisions and dynamic modeling of BNR systems.The effluent phosphorus concentration for the PAS train (3.4 mg/L) was significantly lower than that of the CAS train (4.7 mg/L) during the COD-limited phase (α value of 0.05). However there was no significant difference between the PAS (0.8 mg/L) and CAS (1.0 mg/L) trains for the P-limited phase. Anaerobic P releases were significantly increased by prefermentation in both phases however. The effects on P can be readily understood since prefermentation increased the RBCOD:TP ratio of the influent. However prefermentation also had other effects on wastewater inert fractions, and on autotrophic growth rate. Prefermentation increased μAmax by 9% for COD-limited wastewater (P-value of 0.023) and by 4% for P-limited wastewater (P=0.07). The inert soluble COD fraction (sum of SI and Sp) was reduced from 11% of total COD (CTo) to 7% (P-value of 0.08) for COD-limited wastewaters with prefermentation and from 12% to 8% (P-value of 0.08) for P-limited wastewaters. The systems were modeled with Biowin and the simulation was similar to the observed differences in the pilot system.