State estimation in water distribution networks using graph-theoretic reduction strategy

Steady state analysis of flows and pressures in a water distribution network is a major concern for hydraulic engineers involved in the monitoring and control of water distribution systems. Most of the hydraulic network solution strategies currently available are meant for obtaining flows/pressures given a minimum number of measurements. If more measurements than the minimum required are available, then the redundant measurements can be exploited to obtain a better estimate of the state of the network, taking into account noise in these measurements. In this paper, a state estimation technique for well instrumented water distribution networks is described. The proposed method uses graph-theoretic concepts to reduce the dimensionality of the problem and thereby achieves significant computational efficiency. Applicability of the proposed state estimation procedure is demonstrated on realistic urban water distribution networks. It is also shown that the proposed method can be used for state estimation even when some of the demands are not measured.