Stochastic unit maintenance scheduling model for a power producer considering risk in power market

This paper proposed a day-based stochastic unit maintenance scheduling (UMS) model for power producers in power market. Its objective is to maximize the benefits over the entire scheduling periods, considering the fluctuating electricity prices. The current deterministic UMS model was discussed and the stochastic model was introduced to optimize outage scheduling. In the proposed scheme, the potential benefit of power producers mainly includes expected energy-selling profits in spot markets as well as maintenance cost at each period. It is extended from the pool-based market to a hybrid market containing bilateral contract and energy market. In order to evaluate the related risk from the uncertain energy prices, a framework for the Latin Hypercube Sampling Monte Carlo (LHS-MC) simulation was adopted, which outperformed the ordinary Monte Carlo method. According to its characteristics, the stochastic model can be decomposed into external and internal parts, solved by Genetic Algorithm (GA) and linear programming method respectively. Compared with the current models, the model takes into consideration of the influences of the uncertain electricity prices and lowers the associated risk. Numerical examples on a four-unit producer were utilized to demonstrate its usefulness and effectiveness.