Statistical identification of monod-kinetic parameters from on-line measurements

Tune series analysis of on-line monitored ammonia and nitrate concentrations from full-scale wastewater treatment plants operated according to an alternating scheme makes the identification of Monod-kinetic expressions possible. The models presented in the present context only include kinetic parameters which have shown to be significant in a statistical sense. Estimates of kinetic parameters for the nitrification and denitrification processes are obtained by applying these models to the time series of ammonia and nitrate concentrations. In this paper, the concept of statistical identification which depends on the two conditions of theoretical and practical identification, is described. Experiences from estimating time series models of the nitrification and denitrification processes with data from two wastewater treatment plants are discussed. It appears that the dynamic of the biological processes on a full-scale plant is strongly varying. The proposed models are suitable for on-line control, because the states of the plant are continuously updated as new information from the on-line sensors becomes available.