Resilience enhancement of distribution network under typhoon disaster based on two-stage stochastic programming. (15th May 2023)

• Record Type: Journal Article
• Authors: Hou, Hui; Tang, Junyi; Zhang, Zhiwei; Wang, Zhuo; Wei, Ruizeng; Wang, Lei; He, Huan; Wu, Xixiu
• Is Part Of: Applied energy. Volume 338(2023)
• Publication Date: 2023-05-15
• Subjects: Typhoon disaster -- Distribution network -- Wind field model -- Fragility model -- Stochastic programming; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2023.120892
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: A novel two-stage SMILP investment decision-making framework to enhance distribution network resilience is proposed. Multiple uncertainties introduced by typhoon, distribution line damage, and load fluctuations are taken into account. The real-world data of typhoon "Mangkhut" in Guangdong, China is collected to model wind field uncertainty. Intentional islanding constraint is notably added in the second stage. The proposed model combines distribution network long-term resilience planning with short-term post-event recovery. Abstract: The reliability of power supply in distribution network is vulnerable to extreme weather events such as typhoon. Pre-event preparation can effectively mitigate the deterioration of system resilience. Therefore, we propose a distribution network resilience enhancement decision-making framework which is formulated as a two-stage stochastic mixed-integer linear programming (SMILP) model. The first stage invests coordinately in four strategies, including hardening lines, installing distributed generators (DG), allocating mobile emergency generators (MEG), and deploying switches, etc. And the goal at the first stage is to minimize the investment cost of resilience enhancement strategies. The objective at the second stage is to ensure the minimum expected recourse operation cost of the comprehensive strategies for all typical scenarios. The proposed model can combine distribution network long-term resilience planning with short-term post-event recovery. Furthermore, in order to address the uncertain problems of wind field, line damage and load fluctuations under typhoon disaster, this paper proposes the following solutions. For wind field uncertainty, a detailed wind field model considering time transition, sea-land transition and extreme value distribution is established to deal with wind speed prediction. For line damage uncertainty, an improved stress-strength interference model is set up. And for the load fluctuation uncertainty, scenario generation using load random multipliers is applied to address load uncertainty. The proposed framework is tested in IEEE 33-bus distribution system using historical data from 2018 super typhoon "Mangkhut" in China and demonstrates the SMILP model can significantly reduce the post-event expected recourse operation cost and meanwhile improve the distribution network resilience.