The most commonly applied strategies for optimal water quality sensor placement in drinking water distribution systems are aimed at contamination early warning systems. These strategies aim to minimize the number of people affected in case of a deliberate contamination of drinking water in the distribution system, and provide a valuable tool. A number of factors which are usually not taken into account, including the response strategy to the identification of a contamination event, the fallibility of sensors and changes in network configuration (valve manipulation) and operation, may affect the results of these strategies. Since the quickness and effectiveness of a response is generally also a function of the location of the contamination event (both source and first detection), knowledge on the response strategy should also be part of the sensor placement optimization methodology. Hydraulic models generally play a central role in the optimization of sensor placement. The validity of their computations strongly depends upon accurate and up to date information on the network, which is often not fully available (e.g. unregistered valve status changes). Therefore, a sensor network configuration which is somewhat robust to these issues is desirable. Besides contamination early warning systems, there are several other reasons for placing water quality sensors in distribution network, including process control and monitoring, regulatory monitoring, etc. These require a different approach to optimization of the sensor network in terms of sensor locations. In this paper, we demonstrate the application of different sensor location optimization strategies in drinking water distribution networks, with aims such as minimization of the number of people affected, maximization of distribution network coverage, optimization of sensor network robustness and optimization of contamination source identification. We present and compare results of these different approaches applied to hydraulic models of a real drinking water distribution network in the Netherlands.
van Thienen, Peter, "Alternative Strategies For Optimal Water Quality Sensor Placement In Drinking Water Distribution Networks" (2014). CUNY Academic Works.