上海交通大学学报

上海交通大学学报 ›› 2025, Vol. 59 ›› Issue (9): 1327-1337.doi: 10.16183/j.cnki.jsjtu.2023.528

• 新型电力系统与综合能源 • 上一篇    下一篇

计及㶲效率的低碳数据中心算力综合能源系统规划及运行优化

林嘉瑜1, 韩俊涛1,2, 王永真1,2(), 韩恺1,2, 韩艺博1,2, 李健1   

  1. 1 北京理工大学 机械与车辆学院, 北京 100083
    2 北京理工大学 重庆创新中心, 重庆 401120
  • 收稿日期:2023-10-20 修回日期:2023-12-17 接受日期:2024-02-06 出版日期:2025-09-28 发布日期:2025-09-25
  • 通讯作者: 王永真,副教授,博士生导师,电话(Tel.):010-68912510;E-mail:wyz80hou@bit.edu.cn.
  • 作者简介:林嘉瑜(2002—),本科生,从事新能源科学与工程相关研究.
  • 基金资助:
    国家自然科学基金青年科学基金(52006114)

Capacity Planning and Operational Optimization for Low-Carbon Data Center Integrated Energy System Considering Exergy Efficiency

LIN Jiayu1, HAN Juntao1,2, WANG Yongzhen1,2(), HAN Kai1,2, HAN Yibo1,2, LI Jian1   

  1. 1 School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100083, China
    2 Innovation Center in Chongqing, Beijing Institute of Technology, Chongqing 401120, China
  • Received:2023-10-20 Revised:2023-12-17 Accepted:2024-02-06 Online:2025-09-28 Published:2025-09-25

RichHTML

7

PDF (PC)

2154

补充材料

1

摘要：

随着数字经济的快速发展,数据中心(DC)的能耗及其碳排放显著增加.近些年,数据中心算力综合能源系统(DC-IES)的构建已经成为全球净零排放趋势下数据中心节能减排的重要趋势之一.为助力低碳DC-IES规划与构建,基于热力学第二定律㶲的“品质”分析法,综合考虑能源转化设备㶲效率随设备负载率变化的特性,构建计及设备动态㶲特性的低碳DC-IES容量配置及运行优化能量和经济的多目标规划优化模型,揭示了不同目标下DC-IES 的能流分布特性.结果表明,计及动态设备㶲效率的优化方案相较恒定设备㶲效率的优化方案, 㶲损失率、经济成本、碳排放量分别降低2.6%、1.9%、4.8%,存在明显优势.同时,相较于经济最优方案,多目标优化方案能够显著降低DC-IES的碳排放量及㶲损失率,降低幅度分别为22.72%和20.73%;相较于㶲损失率最小和碳排放量最低方案,多目标优化方案下DC-IES经济成本分别降低54.54%和60.78%;与仅使用电网供电的方案相比,将DC视为综合能源系统的多目标优化方案能够降低40.97%的碳排放量.

关键词: 数据中心, 综合能源系统, 㶲效率, 算力, 能效

Abstract:

With the rapid development of the digital economy, the energy consumption and carbon emissions of data centers (DCs) have significantly increased. In recent years, the construction of data center integrated energy systems (DC-IES) has emerged as one of the critical trends in energy conservation and emission reduction for DCs under the global net-zero emission initiative. To support the planning and construction of low-carbon DC-IES, this paper proposes a multi-objective optimization model for capacity allocation and operational planning of DC-IES, integrating energy and economic considerations with a focus on low-carbon performance. Based on the “quality” analysis method of exergy from the second law of thermodynamics, the model proposed comprehensively accounts for the dynamic exergy efficient characteristics of energy conversion devices under varying load conditions, revealing the energy flow distribution characteristics of DC-IES under different objectives. The computational results indicate that compared with the optimization scheme assuming constant equipment efficiency, the scheme considering dynamic equipment efficiency reduces energy loss rate, economic cost, and carbon emissions by 2.6%, 1.9%, and 4.8%, respectively, demonstrating clear advantages. Moreover, compared with the economically optimal scheme, the multi-objective optimization scheme significantly reduces carbon emissions and energy loss rate of the DC-IES by 22.72% and 20.73%, respectively. Furthermore, compared to the scheme scenarios with the minimum exergy loss rate and lowest carbon emissions, the multi-objective optimization scheme reduces economic costs by 54.54% and 60.78%, respectively. Compared with the scheme relying solely on grid electricity supply, the multi-objective optimization scheme that regards the DC as an integrated energy system can reduce carbon emissions by 40.97%.

Key words: data center (DC), integrated energy system (IES), exergy efficiency, computational power, energy performance

中图分类号: