Identification of possible contamination sources using reverse hydraulic simulation

In case of a contamination in a water distribution system the water quality sensors should ring the alarm bells. Once a contamination is detected by one or more sensors, the immediate question is what is the source or sources of pollution. Assuming the network's hydraulics are known, this paper describes a method of using a reverse hydraulic and quality simulation to identify all of the networks nodes that can "reach" a specific set of sensors at a given time. Using real time SCADA data a hydraulic simulation of the system is performed up to the detection times and the hydraulic simulation results are reversed. Then, a theoretical water quality simulation is performed with tracers injected at the node associated with the sensors that detected the contamination. The algorithm tracks and records the tracers upstream to find all possible contaminating nodes. By using a superposition technique for all possible contaminating nodes, the algorithm can find the most likely set. The algorithm may suggest the location of the contamination source while providing information regarding safe areas in the network. This methodology is fast enough to be used in real time and simple to implement. The methodology is demonstrated through a simple example application.