Stabilisation of Biological Phosphorus Removal from Municipal Wastewater

The biological phosphorus removal (BPR) from wastewater has developed considerably during the last decades and is applied in many present wastewater treatment plants (WWTP) all over the world. The process performance and the control of the BPR are under the influences of daily and seasonal variations of the influent wastewater concentrations and are not yet always guaranteed. Even though the scientific knowledge and practical experience has reached a high level of understanding of the involved key-processes it is still necessary to apply chemical precipitation of phosphorus during the time periods, where the complete BPR can not be achieved. The understanding of the main phenomena involved into such failure of BPR and the development of operational or control strategies to overcome these deficiencies are the main areas of investigation of this thesis. Investigations of the failure of BPR have been performed on an alternating pilot plant, receiving municipal wastewater. The pilot plant is equipped with an automatic measurement system based on the flow injection analysis (FIA) principle. Continuos analysis of the ammonium (NH4-N), nitrate (as NOx-N) and phosphorus (PO4-P) was performed in all important places of the plant. Based on literature studies and investigations of the available pilot plant measurement data experimental designs were developed to produce operational conditions where the BPR failed. The process was investigated during periods of low influent concentrations and increased hydraulic load, with subsequent re-establishment of normal conditions. A process disturbance of this type results in an increase in the phosphate concentration level in the effluent, shortly after the wastewater returns to normal strength. During the first part of the thesis it was examined if an extended version of the activated sludge model No. 2 (ASM2) including denitrification by phosphate accumulating organisms (PAO) can be calibrated and validated on the existing system. A set of parameter was determined through a simple evolutionary parameter estimation strategy. This parameter set was used during the whole investigation. The prediction of the dynamic model was used for development of new operational and control strategies. Based on the model simulations an external carbon source addition (ECSA) was designed in order to overcome the BPR process failure. With the help of such addition, the internal carbon storage compounds could be maintained at a high level, indicated by poly-hydroxyalcanoate (PHA) measurements and the accumulation of phosphorus in the effluent could be avoided. Experimental investigations imply that the ECSA together with a reduction of aeration time during periods of low organic influent concentration could improve and stabilise the BPR. The identification of a minimum PHA level, necessary to ensure complete BPR, was however not possible. The failure of BPR was sometimes observed even when comparatively high internal PHA concentrations were present. The experiments where therefore further investigated with help of the model. The study indicates that a PHA limitation is not the only factor affecting the recovery of the BPR process during such periods. The actual amount of PAO present in the system during and after such disturbance can play a role in the deterioration of BPR. In the last part of the thesis the obtained understanding through the operational investigations was used for the design of a model predictive control (MPC) strategy for the BPR system under low influent concentrations. The MPC strategy is based on the addition of an ECSA keeping the internally stored carbon products at a high level. A simple modelbased observer estimating the unmeasured influent soluble organic or internal PHA concentrations is used. This observer improves the robustness operation of the addition significantly. The MPC control performance is compared to a simple feed-forward control design, which also is based on the estimated organic influent concentrations. It is shown that the simple feed-forward as well as the more advanced MPC strategies can improve the BPR without major interference with the other biological processes. With increasing complexity of the strategy the amount of external added carbon could be significantly reduced while the effluent PO4-P concentration could be kept below the required concentration of 1mgP/L.