ABSTRACT
The reliability and accuracy of in-situ ion selective electrode and ultraviolet (NOx) probes have been investigated at four different treatment plants with different operational conditions. This study shows that the mentioned probes tend to compromise their accuracy and trending stability at lower NOx of <1.0 mg N/L, which if used as a measuring variable for PI feedback controller for denitrification (biological reduction of nitrate to nitrogen gas), would cause overfeeding the external carbon source. In-situ Clark-type N2O sensors, recently introduced for industrial scale use (Unisense Environment) could potentially open a new horizon in the automation of biological processes and particularly denitrification. To demonstrate the applicability of such probes for automation, two in-situ N2O probes were used in two treatment plants in parallel with NOx-N probes. The effects of operational conditions such as COD/N ratios and the correlation between NOx and N2O were investigated at those plants. N2O production at non-detect dissolved oxygen concentrations and pH of 7-7.2 were found to be a function of influent nitrogen load or the ratio of COD/NINFLUENT. Finally, using an N2O probe as a proxy sensor for nitrates is proposed as a measured variable in the PI feedback in the automation of the denitrification process with a NOx set point of <1.2 mg N/L).