The Study Of Phosphorus Bioavailability In Effluents From Advanced Nutrient Removal Treatments

Abstract

Because phosphorus (P) is the main limiting nutrient in the majority of surface waters, developing protective and cost efficient P source control plans is crucial for reducing eutrophication risk. Currently, it is assumed all of the P discharged from Wastewater Treatment Plants (WWTP) is bioavailable. This study used standard algal bioassays to determine the extent to which the various forms of P in WWTP effluent are available for algal growth and how the mineralization kinetics of dissolved P should best be represented in water quality models. The effluents from a pilot plant with various alum-based process were tested. The bioassay indicated that percent bioavailable P (%BAP) declined as P removal level increased (r<super>2</super> = 0.98) and only 7 ± 4% of the P was bioavailable in the final effluent. The chemical speciation and biological uptake experiments for 21 selected P containing compounds showed that in a majority of cases (81%) these species did not follow the classic assumption that Soluble Reactive P (SRP) is a representation of bioavailability. A new classification scheme is proposed to link the connection between bioavailability and operational chemical measures. After characterizing the P, both chemically and biologically, in effluents from 14 advanced nutrient removal facilities with a wide range of phosphorus removal technologies, a regression model was derived between the operational categories and bioavailability. This showed a strong statistical association between BAP and Total Reactive P (TRP) (r<super>2</super> ≈ 0.81), with a BAP/TRP ratio of 0.61 ± 0.24, suggesting TRP could be used as a conservative predictor of BAP. Furthermore, this study indicated that the bioavailability and P speciation varies greatly from one treatment process to another, while in most cases the majority (> 60%) of the effluent P is recalcitrant for algal growth. Finally, the mineralization kinetics of dissolved P in effluents from tertiary process was assessed by bioassays. Model fitting results showed two-pool model and three-pool models fit the experimental data very well with r<super>2</super>> 0.9 and the mineralization rate determined in these first-order decay models could be seamlessly incorporated into existing Total Maximum Daily Loading (TMDL) models without structural modifications. This study also provided a scientific basis to consider the importance of recalcitrant P in tertiary effluents in future modeling practices.