Stochastic very short-term economic dispatch for wind power operation using flexible ramp reserve

摘要

Uncertainty of renewable generation is a major challenge in power system operation. Without scheduling enough fast reserves, the load shedding or wind curtailments are expected under rapid deviations of renewable generation. Therefore, the flexible ramp (FR) reserves are scheduled in short-term horizons to compensate the uncertainties of renewable resources and load demand. In this paper, a very short-term economic dispatch (ED) model is developed to handle the wind and load uncertainties using FR reserve. FR products refer to rapid 10-minute spinning reserves that are provided by the fast-response generating units and energy storage (ES) devices to handle the load and wind uncertainties in both upward and downward directions. The proposed model minimizes the costs of generation, contingency reserve, FR reserves, and charging of battery ES. Technical constraints of generating units and the transmission network limits are considered. The future scheduling hour is divided into six 10-minute intervals and the generation levels and required reserves are optimized for the subsequent receding 10-minute intervals. Point estimate (PE) method is utilized to reduce the wind generation and load uncertainty scenarios. The efficacy of PE method is compared with Monte Carlo technique. In addition to credible scenarios of wind and load uncertainties, some regional worst-case scenarios are considered in each electric area to minimize the amount of FR underestimation. The proposed ED model is formulated as a stochastic mixed integer linear programming (MILP) model that is solved using CPLEX algorithm in GAMS for the IEEE 118-bus and 300-bus test systems.