Leak analysis in pipeline systems by means of optimal valve regulation

Estimation of water loss at small leaks and openings in different areas of a pipe network is handled by means of an iterative two-step procedure. The first step is estimation of the parameters in the network simulation model. It is achieved with the maximization of a likelihood function, equal to the sum of the squares of the difference between the observed water heads and flow rates and their computed values in a given number of nodes and pipes. The estimated parameters include the loss factors and the loss exponent. They provide the water loss per unit time in each area of the network, called zone. The second step is optimization of the openings of the valves included in the network. It is constrained by the minimum pressure head allowed in the nodes. Closure of the valves enhances the different response of the network with respect to the water loss and the load in each node. The robustness of the procedure is tested with numerical experiments by relaxing most of the assumptions made in the model. The water loss estimation in a real network with fictitious measures also shows the robustness of the procedure.