Stochastic Generation Planning Considering Fuel Management

The introduction of renewable energy whose power outputs fluctuate depending on the weather causes uncertainties in power generation operations. To prevent effects by the uncertainties such as surplus or shortage of fuel tank and maximize the operational profitability, power generation companies require economic fuel risk management. As this risk management, a new power generation planning that maximizes the expected profitability under the uncertainties modeled by numerous scenarios is developed. To reduce the computation time to optimize the plan, the proposed method only considers the effective scenarios affecting the results of the optimization and applies progressive hedging which is a fast optimization method. To evaluate the performance of the proposed method, a weekly plan, which assumes 56 uncertainty scenarios and 27 generators, is constructed. As a result, we confirmed that the computation time can be reduced by 92% compared to a method which only applies the optimization solver, while mitigating the error of the objective function at 0.42%. In three - month plan assumed 14 uncertainty scenarios, the proposed method can optimize less than an hour with parallel processing in a computer.