Disinfection By-Products and microbial occurrence in an Italian Drinking Water Distribution Network: cluster analysis for sampling point selection

Drinking Water Distribution Networks (DWDNs) should be managed balancing chemical and microbiological risks. Our goal is to develop a model linking microbiological and chemical parameters to the disinfection operating parameters as support management tool. Since the case study we focused on comprises 27 drinking water treatment plants (DWTPs) feeding a unique DWDN, the number of sampling points to be monitored for the model is huge. To reduce the numerosity of the sampling points without losing information, we performed a cluster analysis to identify groups of DWTPs with similar characteristics, to be considered as representative. Data were preprocessed to identify two values for each DWTP, proxy of the microbial and chemical risks: (i) probability of indicator microorganisms’ enumeration, and (ii) median of total trihalomethanes (TTHMs). Four clusters were identified, from which one DWTP was selected. Then, we planned the monitoring campaign, that will start in mid-June and last for 1-year.