Improving wind power utilisation under stormy weather condition by risk‐limiting unit commitment. Issue 15 (28th September 2018)

• Record Type: Journal Article
• Authors: Cai, Ximing; Qin, Zhijun; Hou, Yunhe
• Is Part Of: IET renewable power generation. Volume 12:Issue 15(2018)
• Page Start: 1778
• Page End: 1785
• Publication Date: 2018-09-28
• Subjects: power grids -- linear programming -- stochastic programming -- power generation dispatch -- power generation economics -- integer programming -- wind power plants -- power generation scheduling -- wind turbines -- load shedding; stormy weather condition -- risk‐limiting unit commitment -- power grid resilience -- extreme weather events -- wind power generation -- wind turbines -- sharply reduced generation resources -- power supply -- reluctant wind curtailment -- wind stormy event -- risk‐limiting constraints -- wind power utilisation -- day‐ahead unit commitment -- real‐time load shedding -- multistage mixed‐integer linear programming model -- commercial solvers -- time‐horizon -- quick‐start generating units; Renewable energy sources -- Periodicals; 333.79405
• DOI: 10.1049/iet-rpg.2018.5447
• Languages: English
• ISSNs: 1752-1416

Abstract

Abstract : Power grid resilience has been a major concern as extreme weather events become more frequent recently. Particularly, wind power generation has constituted a significant body of generation portfolio worldwide. Under stormy weather conditions, wind power generation would be curtailed due to the over‐speed protection of wind turbines. Subsequently, the sharply reduced generation resources would significantly endanger the reliability of power supply. In this study, the authors propose to apply the risk‐limiting methodology to improve the efficient utilisation of wind power prior to reluctant wind curtailment on the advent of wind stormy events. The proposed method integrates three programs, namely, day‐ahead unit commitment (UC), hourly‐ahead UC, and real‐time load shedding into one single model along with risk‐limiting constraints. The first program provides the baseline of hourly dispatch; whereas, the latter two programs serve as recourse means while wind stormy event unfolds. The proposed model is cast into a multi‐stage mixed‐integer linear programming model and solved by commercial solvers. Illustrative examples demonstrate that the proposed method can reduce the total dispatch cost over the time‐horizon including the wind stormy event, by postponing the timing of proactively starting quick‐start generating units and shedding loads, in contrast to conventional two‐stage stochastic dispatch methods.