The use of a novel algorithm to analyse a sequential storage problem in a refrigeration plant

Abstract

The minimum-cost operating strategy for an idealized refrigeration plant having cogeneration capacity was previously analysed using a linear programming model. It is shown here that the addition of chilled-water storage can reduce energy costs by substantially altering the optimum operating strategy. A novel algorithm that uses a sequence of linear programming optimizations to analyse the dynamic behaviour of a store, is discussed. The method avoids approximations inherent in usual dynamic programming techniques, gives useful insights into the problem structure, and offers significant savings in computational time. The economic advantages of storing chilled water in the model plant are discussed, and a target figure for store capital cost is derived.