A non-intrusive probabilistic multi-energy flow calculation method and its application in operation risk analysis of integrated energy systems. (December 2022)

• Record Type: Journal Article
• Authors: Dong, Bo; Li, Peng; Yu, Hao; Ji, Haoran; Li, Juan; Wu, Jianzhong; Wang, Chengshan
• Is Part Of: Sustainable energy technologies and assessments. Volume 54(2022)
• Publication Date: 2022-12
• Subjects: Integrated energy system -- Probabilistic multi-energy flow calculation -- Sparse polynomial chaos expansion -- Operation risk analysis -- Correlated uncertainties; Renewable energy sources -- Periodicals; Energy development -- Technological innovations -- Periodicals; Electric power production -- Periodicals; Energy storage -- Periodicals; 333.79
• DOI: 10.1016/j.seta.2022.102834
• Languages: English
• ISSNs: 2213-1388

Abstract

Highlights: An sPCE-based non-intrusive probabilistic multi-energy flow model is established. A solving algorithm based on Bayesian compressive sensing method is proposed. Three violation risk indices based on probabilistic multi-energy flow are constructed. A CVaR-based risk analysis method is developed for integrated energy systems. The proposed method reduces the computation time by 96% with an acceptable accuracy. Abstract: With the deep coupling of electricity, heat, and gas systems, the uncertainties in renewable energy sources and loads significantly impact the energy flow distribution of integrated energy systems. Probabilistic multi-energy flow calculations considering these uncertain factors have become essential for risk analysis, optimal management, and operational control. However, it is still difficult to efficiently and accurately deal with the diverse and large numbers of correlated random variables. This paper proposes a non-intrusive probabilistic multi-energy flow calculation method and explores its application in the operation risk analysis of integrated energy systems. The probabilistic multi-energy flow model is established considering the uncertainties and correlations of renewable energy sources and loads. The proposed model is solved within the sparse polynomial chaos expansion framework based on Bayesian compressive sensing. Thus, the probabilistic density functions of the risk indices of each subsystem can be obtained. On this basis, the conditional value-at-risk method is employed for the operation risk analysis. The feasibility and advantages of the proposed method are verified using a typical integrated energy system test case.