Shankar Narasimhan S

Urban water distribution networks (WDNs) are large scale complex systems with limited instrumentation. Due to aging and poor maintenance, significant loss of water can occur through leaks. We present a method for leak detection in WDNs using repeated water balance and minimal use of additional off-line flow measurements. A multi-stage graph partitioning approach is used to determine where the off-line flow measurements are to be made, with the objective of minimizing the measurement cost. The graph partitioning problem is formulated and solved as a multi-objective mixed integer linear program (MILP). We further derive an approximate method inspired by spectral graph bisection to solve the MILP, which is suitable for very large scale networks. The proposed methods are tested on large scale benchmark networks, and the results indicate that on average, flows in less than 3% of the pipes need to be measured to identify the leaky pipe or joint.

Water Distribution Networks (WDNs) are vulnerable to accidental or deliberate contamination. Such contamination can be detected and identified by deploying a network of sensors. If the sensor network detects the presence of a contaminant, it is also very important to take corrective response actions to minimize the effects of contamination on the population being served. One possible mitigation option is to prevent the contaminated water from reaching any customer, by shutting down the distribution network using shut-off valves placed in the WDN. The design problem considered in this work is to determine pipes where the shut-off valves can be optimally located such that it is possible to prevent the contaminated water from reaching any demand point, regardless of the source node from where the contamination has originated. We refer to this problem as the actuator network design problem. We map the problem of actuator design into a graph partitioning problem and the minimal set of actuators are identified for ensuring the total shutdown of the network.