Computer‐generated optimal pumping schedule

If the demand forecast is accurate, a computerized pumping schedule can save a utility money by responding quickly to shifts in demand. In Pittsburgh, Pa., electricity costs are governed by a complicated tariff. A study of the city's water distribution system was carried out to determine whether pumping costs could be reduced through computer scheduling of pump operation. Analysis focused on a subsystem that includes six pumping stations and six reservoirs. Quadratic relationships between pumping rate and energy cost were used to determine the sequence and duration of pump operation. The software generates an hourly pumping schedule at each station. The program was run on the subsystem using historical demand and actual 1989 electricity bills. The comparison indicated the optimal solution could have saved 15 percent in energy costs. The computerized optimization package is being tested at one pumping station.