Mapping spontaneous biological phosphorus removal in a membrane bioreactor process without the anaerobic condition : Investigating the effect of alternative external carbon sources

Abstract: Phosphorus removal in wastewater treatment is commonly achieved by chemical precipitation, enhanced biological phosphorus removal (EBPR) or a simultaneous combination of the two. A defined anaerobic condition is widely regarded as the critical element to sustain EBPR. However, this study demonstrates that EBPR is indeed occurring in a 4.5 m3/h membrane bioreactor (MBR) pilot plant without a defined anaerobic condition. Although designed for chemical precipitation alone, a low average Fe/P molar ratio (iron dose/phosphorus removed) of 0.9 ± 0.1 suggests that EBPR could be contributing to a simultaneous phosphorus removal. Weekly flow-proportional grab samples through the process showed a phosphate (P) release between the primary anoxic tanks, followed by a larger uptake in the aerobic tanks. In laboratory batch tests with limited acetate supply in the presence of nitrates, the anoxic P-release began and then abruptly stopped whilst the nitrate concentration continued to decrease. This could be explained by denitrifiers out-competing PAOs for soluble substrate since a large P-release occurred when excess acetate was supplied in the presence of nitrates. It is therefore unlikely that PAOs are operating in the pilot despite the presence of nitrates as was concluded in a study with similar spontaneous EBPR observations. Instead, it is suggested that EBPR is enabled by intermittent anaerobic conditions in the primary anoxic tanks due to low nitrate concentrations (< 1 mg NO3/l) recirculating back after post-denitrification. The external carbon source supplied to the pilot was changed from glycerol to ethanol to assess any effect on the spontaneous EBPR. After ethanol had been supplied for 30 days, increased P-release rates were observed in batch tests supplied with ethanol (0.1 to 0.4 mg P/g VSS∙h) and batch tests supplied with acetate (8.6 ± 0.4 to 10.3 ± 0.4 mg P/g VSS∙h). However, the overall consumption of glycerol was less than that of ethanol, whilst the total phosphorus removal and the Fe/P ratio remained similar whether ethanol or glycerol was supplied to the pilot plant. Should operators wish to avoid any possible spontaneous P-release in the post-denitrification step, methanol is recommended as the external carbon source when considering laboratory P-release results, past research and operation costs.