An Integrated Framework for Supplementing Online Water Quality Monitoring in the Detection of Contamination Events in Water Distribution Networks

Surveillance Response Systems (SRSs) have been deployed in Water Distribution Networks (WDNs) to detect various contamination events. However, in WDNs, some contaminants may remain undetected by an SRS due to the specificity of online water quality monitoring (OWQM). To overcome this limitation, OWQM can be supplemented with additional datasets to enhance the detection capabilities of the SRS framework. These additional datasets are based on health-seeking behaviors exhibited by consumers after consuming contaminated water as well as customer complaints. In this research, we implement a set of Bayesian networks in a clustered network to fuse these alternate datasets (simulated using an ABM due to the limited information associated with real events) with traditional OWQM to determine the likelihood of an ongoing contamination event.