A dynamic energy benchmarking methodology on room level for energy performance evaluation. (October 2021)

• Record Type: Journal Article
• Authors: Zhou, Xuan; Mei, Yukun; Liang, Liequan; Fan, Zhubing; Yan, Junwei; Pan, Dongmei
• Is Part Of: Journal of building engineering. Volume 42(2021)
• Publication Date: 2021-10
• Subjects: Energy benchmarking on room level -- K-shape clustering algorithm -- Daily energy consumption pattern -- Energy performance evaluation -- Energy saving potential; Building -- Periodicals; 690.05
• DOI: 10.1016/j.jobe.2021.102837
• Languages: English
• ISSNs: 2352-7102

Abstract

Abstract: Building energy benchmarking is the basis of energy performance assessment. Most of existing energy benchmarking studies are aimed at building-level or subsystem-level energy performance evaluation, which cannot provide detailed energy performance information inside the building. This study aims to develop a dynamic room-level energy benchmarking method to evaluate energy performance of all rooms in office building. Taking a government office building as a case study, this paper analyses the influence of room types and energy consumption patterns on energy consumption, and proposes a K-shape clustering method to mine five typical energy consumption patterns in this building to describe occupancy behaviors of nearly 1000 rooms. The quartile method is used to establish energy benchmarks for different room types and energy consumption patterns. Finally, this study takes staff room as a case to evaluate the energy performance of rooms. Through the establishment of room-level energy benchmark, it is found that the total energy saving potential accounts for 24.69% of the total energy consumption of all the rooms. This method provides detailed information to help building managers and engineers understand energy consumption status on room level better and help them develop targeted energy-saving measures. Moreover, it should be noted that this energy benchmarking could be periodically updated with energy performance due to the increasing awareness of occupants for energy conservation. Highlights: A dynamic energy benchmarking method is developed to evaluate the energy performance of rooms. Five typical DECPs are mined by a K-shape clustering algorithm to describe occupancy behaviors. Energy saving potential of all rooms in a government office building is estimated quantitatively.