THE IMPORTANCE OF MODELING METAL UPTAKE AND RELEASE IN THE EBPR PROCESSES

A cation uptake and release model implemented in BioWin™ (v2.1) was applied to simulate the fate of phosphorus and metal cations (Mg2+, K+and Ca2+) during batch uptake and release tests as well as full-scale EBPR plants. Simulation results during batch uptake and release tests and full-scale EBPR plants were compared with those obtained via experimental measurements. The BioWin™ model was able to predict final effluent quality close to field observed data. The model predicted anoxic P uptake and secondary P release in the last anoxic zones at plants where nitrate was likely depleted. However, a discrepancy was frequently found for both dynamic batch simulation and steady-state full-scale simulation at all plants studied. A noticeable difference was found between simulated phosphorus and the metal cation concentrations in the various anaerobic, anoxic and aerobic zones across both the EBPR processes and those measured in the field. The model did not provide accurate predictions for full-scale plant performance when using the P/HAc ratio determined from batch uptake and release tests. pH changes predicted by the model were typical for BNR plants (slight increase in pH in anaerobic zones) but did not agree with the measured values for the batch tests. This directly affected all pH sensitive reactions, such as potential precipitation of calcium phosphate which likely occurred during batch tests for the two plants studied. During batch uptake and release tests and across different zones in the full-scale EBPR processes, the model was able to simulate the trend of two metal cations (Mg2+, K+) associated with phosphorus, but not as well for Ca2+. The stoichiometric ratio of metal cations to phosphorus determined from batch tests varied among different plants, suggesting that there are hereditary factors from each plant that affect metal cation requirements. It is necessary to use site-specific metal cation/P stoichiometry and limit metal concentration in the BNR model for assessing the effects of metal cation limitations on P removal performance.