Integrating sustainability indicators in the monitoring and control strategy of WWTPs – Moving towards a smart digital water society
Vasilaki V.1 Volcke, E.I.P.2, van Loosdrecht3, M.C.M., Katsou E.1, 1Brunel University London, UK, 2Ghent University, Belgium, 3Delft University of Technology, The Netherlands
(free)
Abstract
Greenhouse gas (GHG) sensors have only recently been widely used to provide information on the sustainability of the biological processes in wastewater treatment plants (WWTPs). There is still a gap in integrating this knowledge into WWTPs operation and real-time control. In this study, multivariate statistical techniques are applied to the online data collected from long-term real-field N2O monitoring campaigns in order to:
i) gain a better understanding on the dynamic behaviour of N2O emissions,
ii) explain the combined effect of the online operational variables on N2O emissions,
iii) identify disturbances in the process and their effect on N2O emissions and
iv) visualize the range of the operating parameters that optimize N2O emissions in the target biological processes.
A statistical methodological approach is developed applying changepoint detection techniques to identify changes in the behaviour of N2O fluxes, combined with hierarchical k-means clustering and PCA, to provide insights on N2O emissions patterns and generation pathways. The information was used to develop back propagation artificial neural network models simulating N2O emissions for each sub-period. The applied methodology can provide an alternative to the prediction of N2O emissions at different wastewater treatment processes using long term historical data.
