Day-ahead risk averse market clearing considering demand response with data-driven load uncertainty representation: A Singapore electricity market study

Yuanzheng Li,Jingjing Huang,Yun Liu,Tianyang Zhao,Yue Zhou,Yong Zhao,Chau Yuen
DOI: https://doi.org/10.1016/j.energy.2022.123923
IF: 9
2022-04-01
Energy
Abstract:Demand response program is being implemented in the National Electricity Market of Singapore, which boosts the flexibility of demand side to actively participate in the real-time electricity market. Meanwhile, it is also significant to implement such a program in the day-ahead market, since generation companies could arrange their generating plans and load providers are able to adjust their hourly purchasing schedules. However, uncertain factors should be considered in the demand response program of the day-ahead market, such as the uncertain electricity load. Regarding the issue, this paper proposes a day-ahead bidding and clearing framework considering demand response with uncertain and correlated nature of electricity loads. To this end, a data-driven Dirichlet process mixture model is introduced to represent the load uncertainty, which might bring about the economic risk. To further reduce such a risk, a worst-case conditional value at risk is integrated into our proposed framework, and a WCVaR based two-step risk averse market clearing model is proposed. Finally, we conduct numerical studies based on the Singapore electricity market. Numerical studies demonstrate the outperformance of Dirichlet process mixture model for the load uncertain representation, and also verify that the worst-case conditional value at risk based market clearing model could effectively reduce the economic risk while maximizing the social welfare.
energy & fuels,thermodynamics
What problem does this paper attempt to address?