• Energy Research
• natural sciences close

A two-stage adaptive robust optimization is developed for pre-disturbance scheduling in microgrids (MGs) for handling uncertainties associated with electricity market prices, renewable generation, demand forecasts, and islanding events. The objective is to produce a reliable and optimal solution for MG operation that minimizes operational costs and the risk/failure in islanding events. In the literature, the uncertainty sets associated with islanding events cover a full scheduling period which results in a sub-optimal solution. In this paper, uncertainty sets corresponding to islanding events are modeled based on reliability/resiliency-oriented indexes of the MG/grid to achieve a more accurate/reliable solution. Besides, the Benders decomposition algorithm which is used to handle uncertainties in solving the optimization problem is time-consuming. Therefore, the column-and-constraint generation (C&CG) decomposition strategy is adopted to make the problem computationally tractable. Further, the uncertainty budget parameters are clarified to balance the conservatism and optimality (cost minimization) of the robust solution in uncertainty sets. The effectiveness of the proposed framework is evaluated and discussed by using a set of numerical studies with different scenarios in an MG. The simulations show that the proposed framework reduces operational costs by using the precise analysis of uncertainty budgets and a change in scheduling periods.