Probabilistic sizing and scheduling co-optimisation of hybrid battery/super-capacitor energy storage systems in micro-grids

Abstract

While established deterministic capacity planning models for single-component energy storage systems exist, little attention has been given to probabilistic sizing of hybrid energy storage systems (ESSs) using swarm-based meta-heuristic algorithms. This highlights two key research opportunities, namely: (1) studying the impact of preserving model-inherent characteristics and optimising daily system dispatch on narrowing reality gaps in hybrid ESS designs, and (2) the optimal integration of hybrid ESSs into grid-connected micro-grids based on their applications, with potentially significant financial implications for model designs. In response, this paper introduces a novel probabilistic hybrid ESS capacity planning optimisation model based on a state-of-the-art meta-heuristic algorithm. To demonstrate the effectiveness of the model within a community micro-grid scheme, a case study of an eco-village in Aotearoa New Zealand is presented. The simulation results indicate a ∼4 % and ∼36 % premium above the deterministic results respectively in the most likely case and worst-case probabilistic scenarios. On the other hand, the best-case stochastic estimate of the life-cycle cost of the hybrid ESS is found to be ∼39 % lower than that of the deterministic modelling. Additionally, the economics of temporal energy arbitrage using the battery bank is investigated, indicating that at the current capital cost of stationary LiFePO4 batteries and the present fixed feed-in-tariff (NZ$0.08/kWh), it is not economically viable to cycle the storage for arbitrage reasons alone. In conclusion, this paper highlights the critical need to incorporate probabilistic optimisation techniques and emphasises the importance of sizing and scheduling co-optimisation when designing hybrid ESSs for integration into grid-connected micro-grids.