An alternate oxic-anoxic process automatically controlled. Theory and practice in a real treatment plant network.

A simple mathematical model of an alternate oxic-anoxic process has been elaborated. It enables us to optimise the cycle time on the basis of maximum nitrates concentration in the effluent and the desired nitrogen removal performance. At the same time the model can be employed to verify the impact of the variations of flow rate and influent characteristics as well as the operational parameters of the process. Actually, the model confirms the process efficiency but its feasibility in real plants needs a local or remote process control. To verify these theoretical conclusions a real wastewater plant (700 PE) has been upgraded in an alternate oxic-anoxic process. It was implemented with software able to elaborate the data of dissolved oxygen concentration and oxidation reduction potential. Moreover, the evaluation of the flexing points was performed to manage mixer and blowers. A one-year experience of plant management allowed us to obtain very high nitrogen removal. However, the performances were different during wet or dry weather periods. The statistical analysis of probe signals evaluation confirmed the capability of the control device to detect the flexing points during the anoxic phase (70-94%). On the other hand, the capability of detecting the DO signal was lower, in particular when the oxygen demand was similar to the amount of supplied oxygen. The hourly variations of flow rate and mass loading determines different conditions for starting the anoxic phase: over aeration, over loading and the equivalence of oxygen demand and supply, are the main factors determining the blowers stopping.