Open Access

ARTICLE

Comprehensive Index Evaluation of the Cooling System with the Level Loop Thermosyphon System in Different Computing Hub Nodes in China

Li Ling^*, Danhao Song, Qianlong Hu, Zihao Xiang, Zeyu Zhang

College of Civil Engineering, Hunan University of Technology, Zhuzhou, 412007, China

* Corresponding Author: Li Ling. Email:

(This article belongs to the Special Issue: Advanced Analytics on Energy Systems)

Energy Engineering 2025, 122(8), 3309-3328. https://doi.org/10.32604/ee.2025.065824

Received 22 March 2025; Accepted 23 May 2025; Issue published 24 July 2025

Abstract

Rack-level loop thermosyphons have been widely adopted as a solution to data centers’ growing energy demands. While numerous studies have highlighted the heat transfer performance and energy-saving benefits of this system, its economic feasibility, water usage effectiveness (WUE), and carbon usage effectiveness (CUE) remain underexplored. This study introduces a comprehensive evaluation index designed to assess the applicability of the rack-level loop thermosyphon system across various computing hub nodes. The air wet bulb temperature T_a,w was identified as the most significant factor influencing the variability in the combination of PUE, CUE, and WUE values. The results indicate that the rack-level loop thermosyphon system achieves the highest score in Lanzhou (94.485) and the lowest in Beijing (89.261) based on the comprehensive evaluation index. The overall ranking of cities according to the comprehensive evaluation score is as follows: Gansu hub (Lanzhou) > Inner Mongolia hub (Hohhot) > Ningxia hub (Yinchuan) > Yangtze River Delta hub (Shanghai) > Chengdu Chongqing hub (Chongqing) > Guangdong-Hong Kong-Macao Greater Bay Area hub (Guangzhou) > Guizhou hub (Guiyang) > Beijing-Tianjin-Hebei hub (Beijing). Furthermore, Hohhot, Lanzhou, and Yinchuan consistently rank among the top three cities for comprehensive scores across all load rates, while Guiyang (at a 25% load rate), Guangzhou (at a 50% load rate), and Beijing (at 75% and 100% load rates) exhibited the lowest comprehensive scores.

---

Keywords

---