Energy storage system scheduling for peak demand reduction using evolutionary combinatorial optimisation

This paper is concerned with finding an optimal energy storage system (ESS) schedule for peak demand reduction and load-levelling given only the information certainly available to and controllable by the Distribution Network Operator (DNO) which are the substation demand profile information and the DNO-owned ESS parameters. Methods such as set-point control are usually suboptimal and can create new peaks. Other methods require more parameters and actions than the DNO can fully control to form the basis for optimisation and can be computationally complex. The method presented in this paper uses simple heuristics to find possible optimal operation points for the ESS and improves the solutions found using genetic algorithm optimisation. A case study is presented showing a UK distribution network with a peak capacity violation which is resolved using the method and the results are compared to a closed-loop set-point control method.