Optimization of real-time operation in water supply systems incorporating demand charges

The cost and availability of electrical energy have made operating cost in water supply systems an important priority. The optimal pump operation has presented some challenges in research, Problems are due to complicated tariff, especially demand charges; size of and difference among the systems; and discrete pump discharges.;Computer software for the model is verified by applying it to the City of Pittsburgh's Water Supply System. A simulation run on part of the system is carried out (for winter and summer months) in a test year. Historical water demands are used to determine what would have been the optimal pump schedules. Results show that the optimal solution could have saved 20 percent of the actual billing. The software may be used in other systems by developing a similar data base, and updating some modules. The model is most effective if the demand forecast is accurate.;In this study, an optimization model is developed for generating the pump schedule for one or more pump stations in order to minimize the costs of pumping while satisfying water demand and system constraints. The model utilizes decomposition of the system (into several subsystems), and of the planning period (into operational periods). Progressive optimality is applied to solve dynamic programming formulation. The pump discharges are discretized and rearranged by heuristic methods in order to reduce number of pump switching. In real-time operation, the long-term model is used to estimate the billing monthly demand. This demand is incorporated in the short-term model which can obtain the optimal daily schedule in a relatively short time.