Microgrid energy management: how uncertainty modelling impacts economic performance

Abstract

Short-term electricity prices are key economic input to model the optimal operation of grid-connected microgrids. In competitive electricity markets, these prices are not known in advance, and need to be forecasted. Price forecasts, however, have uncertainty, and thus, their errors will impact economic gains. Three main approaches have been employed in the literature to mitigate the uncertainties associated with price forecasts, i.e. rolling horizon optimisation techniques, interval optimisation and scenario-based methods. In this study, we investigate the economic values of using these approaches, as well as the combination of them, in the operation of microgrids. This is to inform microgrid operators on how to determine which approach should be adopted under different circumstances. Therefore, we first implement point, interval and scenario forecasting models for electricity market prices. The generated forecasts are then fed into a deterministic, robust and stochastic optimisation models for operation scheduling of a typical microgrid. The changes in total energy costs of the microgrid are then evaluated. The simulation results show that while the performance of different methods depends on the volatility of market prices, the model with point price forecasts and a rolling horizon operation scheduling either outperforms other methods or does equally well.