The Effect of 'Selector' Design Features On Sludge Densification and Settleability In Mainstream WWTPs Operating With EBPR

Granulation has been gaining interest as a biological treatment process to increase biomass concentration and sludge settleability for enhanced nutrient removal. However, granulation technique has been limited to batch-type reactors. Yet, due to simplicity most processes still operate as continuous flow activated sludge (CFAS) making it challenging to achieve complete granulation; instead, partial granulation or densification occurs. There is still a desire to increase biomass concentration in CFASs for large scale wastewater treatment plants (WWTPs) (Yewei Sun, 2019). It is hypothesized that full-scale CFAS with enhanced biological phosphorus removal (EBPR) and low sludge volume index (SVI) contains densified sludge. Introduction of an anaerobic phase and feast-famine alternation promotes the proliferation of slow-growing microorganisms including the phosphate accumulating organisms (PAOs) which take up organic matters and store it as polyhydroxyalkanoate (PHAs). Anaerobic feeding and sufficient starvation time to ensure sludge stability and prevent filamentous outgrowth tend to stimulate microbial aggregation and phosphorus removal. This matter will create adequate substrate and oxygen gradients. Other environmental conditions such as adequate food to microorganism (F/M) ratio, large height to diameter ratio of reactor, and short settling time selects for dense microbial aggregates (Britt-Marie Wilén, 2018). In a study by Wei et al., it was shown that densified sludge, with a range of 0.5% to 80% of granules, has been detected in different EBPR plants based on their kinetic and metabolic selector features such as anaerobic F/M ratio, unmixed inline fermentation, and influent soluble COD fraction (Stephany P. Wei, 2020). Additionally, Martin et al., completed a survey of 39 WWTPs reporting 27 of the 39 plants experienced SVIs lower than 60 mL/g. Four of these facilities were further investigated in which all four included the feast-famine selector that is typical of EBPR (Kelly Martin, 2016). This study reviews the possibility of sludge densification on successful EBPR in mainstream CFAS processes to evaluate the correlation of SVI with phosphorus removal and F/M ratio and the effect of the densification 'selectors.' Further, lower SVI values indicate a quick settling mixed liquor that stays settled and thickens at the bottom of the clarifier thus allowing increase of the solids loading rate and potentially the surface overflow rate to the clarifier without negatively impacting clarification performance (i.e., solids-liquid separation). In turn, an increased clarifier capacity due to improved settling creates the conditions for potentially increasing flow and loadings to the bioreactor. A full-scale study is being conducted to evaluate the impact of applying 'selectors' to achieve sludge densification and successful EBPR. Data will be collected from multiple full-scale WWTPs operating EBPR. Currently, only data from Bonnybrook Wastewater Treatment Plant D (Secondary D) and Pine Creek Wastewater Treatment Plant (PCWWTP), Calgary, Alberta is available to be included in the abstract; however, data from multiple WWTPs will be detailed in the complete manuscript and presentation. The City of Calgary operates two full-scale BNR plants that require little to no chemical dosing for phosphorus precipitation: Bonnybrook Wastewater Treatment Plant (BBWWTP) Secondary D and PCWWTP. The design of PCWWTP and Secondary D has implemented three 'selector' features to encourage sludge settleability: (1) the standard anaerobic zone contributing to high PAO growth, (2) feast-famine control strategy by step-feeding the front end of the bioreactor, and (3) nitrified mixed liquor recycle to a strict anoxic cell. Secondary D was recently commissioned in Fall 2021 and consists of two BNR trains, Bioreactors 11 and 12, each designed to treat an annual average capacity of 47 MLD (12.4 MGD). The entirety of BBWWTP must comply with a monthly average total phosphorus discharge limit of 1.0 mg/L and the City of Calgary total river phosphorus loading objective of 240 kg/day. PCWWTP and Secondary D bioreactors have four zones: pre-anoxic, anaerobic, anoxic, and aerobic. Figure 1A and 1B illustrates the process configuration of Secondary D and PCWWTP, respectively. To evaluate the impact of feast-famine conditions and anaerobic F/M ratio as selector features the effluent quality and operational parameters were closely monitored. To check the possibility of densification, stirred SVI (SSVI), mixed liquor volatile suspended solids (MLVSS), and total suspended solids (TSS) were analyzed. Moreover, daily composite COD samples were collected, as well as nitrogen and phosphorus monitoring. Though Secondary D and PCWWTP operate a similar configuration, there is variability in sludge settleability between the facilities. However, the differences in SSVI may be attributed to specific operational features of each plant. The anaerobic 'selector' is standard in EBPR processes. This zone promotes high PAO growth contributing to phosphorus removal. Moreover, the step-feed design of the bioreactor creates feast-periods in the front end followed by aerobic famine-periods. In the famine zones, concentrations are relatively low where the secondary effluent TCOD concentration average is 45 mg/L. Initial application of the 'selector' features and specific operational parameters has been promising in Bioreactor 11 and 12 as the average SSVI-30 observed over a three-month period is 67 mL/g and 70 mL/g, respectively. The low SSVI observed in both trains suggested that there may be densified sludge. In comparison, the average SSVI-30 over a three-month period at PCWWTP is 89 mL/g. This value is approximately 20% higher than what is being observed at Secondary D with a similar configuration. In addition, Secondary D trains were previously supplemented with fermented supernatant (FSU) high in volatile fatty acids (VFA) to the anaerobic cell. The average TCOD concentration in the FSU stream in the past three-months was 2,335 mg/L, and the average VFA concentration was 513 mg/L. Yet as of the optimization plan, the FSU was reduced and completely cutoff in Bioreactor 12 only. This matter has reduced the F/M ratio from an average 1.8 gTCOD/gMLVSS to 1.3 gTCOD/gMLVSS. The F/M ratio in anaerobic and aerobic cells in relation to the SSVI-30 are displayed in Figure 2. The decrease in the F/M ratio has shown a very gradual increase in the SSVI-30 from an average of 70 to 75 mL/g which may show the effect of substrate availability on densification. Agreeing with literature (Kelly Martin, 2016), the low SVIs and densification may be attributed to the feast-famine strategy. Such a feast/famine strategy is known to select for PHA-storing organisms. F/M ratio will be compared with other full-scale facilities to determine if the 'selector' features applied has improved sludge settleability, or if it is the result of alternatives. This research is significant as it reviews the impact of different 'selectors' on sludge densification, and phosphorus removal, in full-scale WWTPs operating EBPR. Moreover, the full presentation will review the impact of the different 'selector' features and specific plant operational parameters to identify those parameters that have the biggest impact on improving sludge settleability while maintaining efficient phosphorus removal.