An optimal bidding and scheduling method for load service entities considering demand response uncertainty

• Published in: Applied energy 2022-12, Vol.328, p.120167, Article 120167
• Main Authors: Han, Rushuai, Hu, Qinran, Cui, Hantao, Chen, Tao, Quan, Xiangjun, Wu, Zaijun
• Format: Article
• Language: English
• Subjects: Aggregation deviation; Bi-level scheduling model; Energy storage system; Load serving entity; Online learning
• ISSN: 0306-2619; 1872-9118
• DOI: 10.1016/j.apenergy.2022.120167

With the rapid development of demand-side management technologies, load serving entities (LSEs) may offer demand response (DR) programs to improve the flexibility of power system operation. Reliable load aggregation is critical for LSEs to improve profits in electricity markets. Due to the uncertainty, the actual aggregated response of loads obtained by conventional aggregation methods can experience significant deviations from the bidding value, making it difficult for LSEs to develop an optimal bidding and scheduling strategy. In this paper, a bi-level scheduling model is proposed to maximize the net revenue of the LSE from optimal DR bidding and energy storage systems ESS scheduling by considering the impacts of the uncertainty of demand response. An online learning method is adopted to improve aggregation reliability. Additionally, the net profit for LSEs can be raised by strategically switching ESS between two modes, namely, energy arbitrage and deviation mitigation. With Karush–Kuhn–Tucker (KKT) optimality condition-based decoupling and piecewise linearization applied, this bi-level optimization model can be reformulated and converted into a mixed-integer linear programming (MILP) problem. The effectiveness and advantages of the proposed method are verified in a modified IEEE RTS-24 bus system. [Display omitted] •The quantitative relationship between aggregation deviations and incentives is established.•An online learning method is adopted to mitigate the load aggregation deviation.•Energy storage systems are strategically switched between energy arbitrage and DR’s deviation mitigation modes.•The impact of DR uncertainty on the locational marginal price and demand response bidding is analyzed.