Study Of Electricity Purchasing Strategies For Large Power Consumers By Two-stage Stochastic Programming Model

In electricity market, large power consumers can purchase electrical energy in terms of the long-term contract, spot and spinning reserve market against their self generators. The study of optimal purchase strategies for large power consumers is very active research topic in China’s electric power markets now, which has been widely attention. This paper based on the linear partial information(LPI) theory studied the problem of purchase strategies for large power consumers under uncertain electricity market.First, the linear partial information theory and two-stage stochastic programming are presented. According to the specific conditions of purchase electricity market reform, the two-stage stochastic programming with linear partial information on the probability distribution(LPI) is employed to model the purchasing strategies for large power consumers. The maximization of the minimal expected value(MaxEmin) is applied to define the second stage recourse function, and the solving method is given. With the historical data of California electricity market, the results show that the established model in this paper is feasible and effective. In order to consider the market risk, the paper based on the theory of Markowitz, use the variance as the measure of risk. A two-stage stochastic programming model with minimum electricity purchasing cost and risk is built. A modified L-shaped algorithm is designed to solve the mathematical problem. Finally, historical data from the Californian electricity market is served for demonstrating Markowitz’s mean-variance model can greatly reduce the risk of electricity purchasing of large power consumers.