Development of Tool-Based Strategy and Control Action for Biological Intrusion Through Localized Leakage in Water Distribution Network Using EPANET and ML

There is always a question of modeling the biological intrusion in the water distribution system, as various uncertainties are associated with limited hydraulics and quality inputs. The research paper addresses solutions to such complexities by approaching various tool-based methods for the real-time water distribution network (WDN). The existing intermittent water supply system of Subhanpura zone located in Vadodara city, India, is modeled and simulated in EPANET by applying rule-based controls on operating valves. Hydraulic models follow the observational pressure-driven approach (PDA) and emitter-based demand-driven approach (DDA). A classification model is developed from simulated results of demand, pressure, and residual chlorine (RC) by applying conditional probabilities based on the junction’s RC value. The results are treated with artificial neural network (ANN) to verify the intrusions and tested for validation accuracy of PDA and DDA. The novelty of the research is the identification of a tool-based decision-making methodology and the development of a control strategy to avoid intrusions at leakage spots. Strategy is achieved by modified rule-based controls that allow sustaining pressure limits within the WDN. The system provides comparatives that can locate the leakage and address the probable biological intrusions in WDN.