A mean-weighted CVaR model for distribution company's optimal portfolio in multi-energy markets

Abstract

In transmission and distribution separated electricity markets, with the highly fluctuant price, distribution companies need to purchase electric power reasonably in several energy markets such as spot markets, long-term tolling agreements and forward contracts etc. to realize profit maximum and risk minimum. Conditional value-at-risk (CVaR) might measure risk efficiently, with only one kind of price distribution considered. In case of existence of more price distributions, e.g. logarithmic normal distribution, a method called weighted CVaR (WCVaR) to measure the purchasing risk of the distribution company is proposed in this paper. A mean-WCVaR model is built to measure the profit and risk in the purchasing process as a mathematical programming problem to derive the efficient frontier that indicates the optimal tradeoffs available to distribution company between expected revenue and purchasing risk in several energy markets. Lastly, a simulation case is carried out to prove the efficiency of the proposed model, which paves a new way for distribution company to determine the optimal purchasing strategies considering the risk.