Stochastic optimization for energy storage allocation in smart grids in the presence of uncertainty

A key subject in the study of smart grids is to accommodate uncertainty in various contexts, including planning and operation of electricity grids in the presence of distributed generation from renewable energy sources, stochastic demand patterns, and varying network configurations. The impact of uncertainty on the solution of different problems formulated in the optimal power flow framework calls for stochastic programming paradigms in the form of two- or multi-stage problems, or optimization programs with chance constraints. In this chapter, we focus on the problem of optimally siting and sizing the energy storage systems in a distribution network. These devices are recognized as good candidates to tackle different issues, such as voltage/frequency regulation, minimal curtailment of renewable generation, peak shaving, or others. For the sizing problem, a scenario-based approach is adopted to cope with uncertain demand and generation profiles at the different buses of the network. A novel scenario reduction technique is presented to make the resulting stochastic optimization problem computationally tractable. A heuristic strategy based on network voltage sensitivity analysis is adopted to deal with the combinatorial nature of the energy storage siting problem. The overall procedure is tested on a IEEE benchmark network, highlighting good performance on a realistic case study.