Ricerc@Sapienza

The biological denitrification process is extensively discussed in scientific literature. The process requires anoxic conditions, but the influence of residual dissolved oxygen (DO) on the efficiency is not yet adequately documented. The present research aims to fill this gap by highlighting the effects of DO on the specific denitrification rate (SDNR) and consequently on the efficiency of the process. SDNR at a temperature of 20◦C (SDNR20◦C) is the parameter normally used for the sizing of the denitrification reactor in biological-activated sludge processes. A sensitivity analysis of SNDR20◦C to DO variations is developed. For this purpose, two of the main empirical models illustrated in the scientific literature are taken into consideration, with the addition of a deterministic third model proposed by the authors and validated by recent experimentations on several full-scale plants. In the first two models, SDNR20◦C is expressed as a function of the only variable food:microrganism ratio in denitrification (F:MDEN), while in the third one, the dependence on DO is made explicit. The sensitivity analysis highlights all the significant dependence of SDNR20◦C on DO characterized by a logarithmic decrease with a very pronounced gradient in correspondence with low DO concentrations. Moreover, the analysis demonstrates the relatively small influence of F:MDEN on the SDNR20◦C and on the correlation between SDNR20◦C and DO. The results confirm the great importance of minimizing DO and limiting, as much as possible, the transport of oxygen in the denitrification reactor through the incoming flows and mainly the mixed liquor recycle. Solutions to achieve this result in full-scale plants are reported.