Conventional design and operation of enhanced biological phosphorus removal (EBPR) focuses on providing volatile fatty acids (VFAs) from the collection system and sometimes from primary sludge fermentation to phosphate accumulating organisms (PAOs) in an influent anaerobic zone. Alternately, sidestream EBPR (S2EBPR) diverts a portion of the RAS to a longer hydraulic retention time (HRT) sidestream reactor where biomass and particulate COD is fermented endogenously to generate the VFAs required. Successful S2EBPR operation hinges on the sidestream zone producing enough VFA to support healthy PAO populations in the HRT available.
Clean Water Services incorporated S2EBPR into a recent secondary expansion but did not observe the process to improve EBPR stability over conventional EBPR operation. In order to investigate possible reasons, the apparent fermentation rate (AFR) at three CWS facilities was measured. The results showed that there is variably in the AFR across time and season in the same facility, between different facilities and based on the method used. The CWS testing also highlighted an integral parameter used to estimate the AFR from batch testing results: the ratio of VFA removed to orthophosphate released (the P-release ratio). Long term measurements of the P-release ratio show that it is also highly variable over time.
Polyhydroxyalkanoate (PHA) measurements were also performed during the bench scale AFR testing and during full scale S2EBPR operation. The goal is to determine if shifts in carbon storage may impact S2EBPR process performance. The results indicate strong differences in the type of PHA stored between conventional EBPR and S2EBPR. The possible reasons for this shift will be explored but it is as yet uncertain how carbon storage changes relate to differences in process performance.
This presentation will provide an overview of how apparent fermentation rate testing fits into design and operational decision making and will discuss the overall methodology and approach. CWS results will be compared against original design assumptions and against wider industry results. The importance of the P-release ratio and possible sources of variability over time will also be summarized.