Quantifying magnitude of potential phosphorus removal and recovery using plant-wide modeling

Minimizing the release of nutrients from municipal Water Resource Recovery Facilities (WRRF) and runoff is vital to reducing nutrient pollutants in waterways. In the United States, municipal WRRF implement phosphorus (P) removal and recovery processes to limit nutrient releases leading to eutrophication and hypoxic zones. These systems vary in their potential for removal of P versus recovery of P, total P removal, and their cost of implementation at the WRRF. While models do exist for commonly used P removal processes, little work has been done to develop a general heuristic framework aimed at determining what type of P removal process is best fit for WRRFs of varying throughput, influent compositions, and treatment configurations. This work develops a generalized plantwide process model used to assess the feasibility of operating various types of nutrient recovery systems at a WRRF. The model compares the effectiveness of plant configurations in P removal and potential recovery as struvite. Full-scale WRRFs are modeled and simulated using hydromantis GPS-X software. The nutrient removal configurations considered are activated sludge (AS), activated sludge with ferric chloride chemical precipitation (AS_CP), modified Bardenpho enhanced biological phosphorus removal (EBPR), and side-stream struvite precipitation (EBPR_FBR). State input variables (total P, total nitrogen, Ortho-P, flow rate) were randomly sampled using Monte Carlo Latin Hyper Cube sampling to create potential treatment scenarios. Calculated and operational variables were based on state variables and used as design variables to optimize configurations for P removal. The model outputs supplied data regarding the P removal, sludge production, and uncontrolled struvite production in order to compare the effectiveness and operational functionality. This plantwide modeling assessment demonstrates a meaningful design space to compare the magnitude of P removal and recovery across configurations and influent characterization. The AS_CP plant demonstrated reliable P removal and TP effluent quality with a sludge production increase of 3.3-5.5 times other configurations. EBPR and EBPR_FBR showed a slight increase in P removal, but require further study regarding the state, calculated, and operational variables intended to optimize the performance. The EBPR P removal demonstrated dependence on influent TP and TKN and a carbon excess in the EBPR train, indicating the need for an alternative approach to optimizing the process. Further work addressing the uncertainties associated with plant wide modeling of the configurations assessed will provide data to best inform decision making regarding P removal technologies. A generalized plantwide process model that considers various types of nutrient recovery schemes allows for cost-effective implementation at municipal WRRFs working to manage P release to the environment.LimitedAuthor requested closed access (OA after 2yrs) in Vireo ETD syste