Day-Ahead Rolling Window Optimization of Islanded Microgrid with Uncertainty

Abstract

The integral variability of electricity demand and intermittency of renewable energy resources (RERs) pose special challenges in the operation of islanded Microgrid (MG). The uncertainty associated with load and generation data further magnifies the problem resulting in huge energy curtailments in real time thus making MG operation expensive. Hence, this paper proposes a stochastic 24-hour ahead rolling window optimal energy scheduling framework to minimize the amount of lost load and lost generation in islanded MG while considering the demand response (DR) potential of heating, ventilation air-conditioning (HVAC) loads, end user thermal preferences and storage systems. These thermal characteristics are modeled via two-capacity building model. The proposed methodology is formulated as a stochastic linear programming problem. The effectiveness of the framework is validated by simulation for the entire winter season using realistic data. Ultimately, the results are used to calculate and control loss of load probability (LOLP) and loss of load expectation (LOLE) in islanded MG at minimum cost.