Tracking the information about your manuscript
Communicate with the editorial office
Query manuscript payment status Edit officeCollecting, editing, reviewing and other affairs offices
Managing manuscripts
Managing author information and external review Expert Information Expert officeOnline Review
Online Communication with the Editorial Department
Page Views
Journal Overview
REFRIGERATION AND AIR-CONDITIONING
Launched in 1990, Monthly
Indexed by:
China Academic Journal Comprehensive Evaluation Database
China Journal Full-text
Database:
Chinese Core Journal (Select) Database
Chinese Science and Technology Journal Database
Annual Report for Chinese Academic Journals Impact Factors
Telephone: 010-83560065，83510099
Domestic Unified Serial Number (CN):11-4519/TB
International Standard Serial Number (ISSN):1009-8402
Research progress on green and low-carbon technologies for cooling systems in intelligent data centers: from carbon source to carbon sink
Wei Zuyuan;Wang Ningbo;Tian Bo;An Linyi;Dong Zijun;Shao Shuangquan;Amidst the burgeoning development of information technologies such as big data and artificial intelligence, data volume and computing power have become the core competitive advantages of the digital economy. Concurrently, the rapid expansion of data centers has led to a sharp increase in their operational energy consumption and carbon emissions. Cooling systems, as high-energy-consumption facilities, constitute a key factor exacerbating carbon emissions. Consequently, advancing energy efficiency in cooling systems is of paramount significance for implementing the “carbon neutrality” strategy and achieving sustainable development in data centers. This paper proposes an iterative upgrade pathway for data center cooling technologies: “high-carbon→low-carbon→zero-carbon→negative-carbon”, which is elaborated sequentially to provide reference for the construc-tion and research of sustainable data centers. In the low-carbon phase, liquid cooling technology, leveraging its high-efficiency heat transfer, replaces traditional air cooling, reducing cooling energy consumption and improving Power Usage Effectiveness(PUE). Progressing toward zero-carbon necessitates integrating natural cooling sources and renewable energy to curb carbon emissions at the source. The breakthrough to negative-carbon relies on waste heat recovery, converting low-grade thermal energy into resources such as district heating. This approach not only avoids new carbon emissions but also generates carbon removal benefits, transforming data centers into computing power and thermal energy suppliers and carbon sink contributors.
Carbon footprint and carbon reduction strategies for computer room air conditioner
Cheng Liquan;Wu Junfeng;Qu Boyi;Dai Lin;Wang Shouchuan;Wei Sheng;Cheng Yongqiang;The carbon footprint of the computer room air conditioner of a data center is analyzed using life cycle assessment method. The analysis results show the main contribution of greenhouse gas emissions comes from the electricity consumption and refrigerant emissions of the use and maintenance stage, accounting for 94.24% and 5.36% respe-ctively. The carbon emissions of other stages account for a small proportion in the entire life cycle and this part can be basically ignored. Through sensitivity analysis, it can be concluded that energy consumption of the operation stage is the main influencing factor of carbon emissions from computer room. The use of natural cold sources and increasing the return air temperature for computer room air conditioner can effectively reduce the carbon emissions. Using refrigerants with lower GWP can also reduce carbon emissions, but its impact on the carbon emissions from computer room air conditioner in the entire life cycle is less. The carbon emissions of production and transportation processes account for a relatively small proportion in the entire life cycle, which can be basically ignored.
The integration development of evaporative cooling and liquid cooling in the era of computing power
Huang Xiang;Li Tong;Chu Junjie;With the increasing demand of computing power, the problem of high energy consumption in data center is increasingly prominent. At the same time, with the training iteration of general large model greatly pulling the demand for intelligent computing power, and the proportion of intelligent computing power is also increasing. At present, the evaporative cooling technology has been widely used in data centers, and the intelligent computing data centers with higher demand for cooling are also constantly promoting the rapid development of liquid cooling technology. This paper introduces the application status of evaporative cooling technology, the cooling form of intelligent computing data center and the liquid cooling technology, and proposes the fusion development of evaporative cooling and liquid cooling to reduce the energy consumption of data center and provide reference for the construction and research of data center with high heating density.
Impact of thermophysical properties of coolants on cooling performance in single-phase immersion cooling server
Sun Ying;Against the backdrop of the commercialization wave of AI technologies, the surging demands for large-model training and inference have led to exponential growth in data center energy consumption. Traditional cooling technologies can no longer meet the requirements for heat dissipation and energy efficiency, urgently calling for advanced cooling solutions to break the deadlock. With its advantages of high heat dissipation efficiency and energy conservation, single-phase immersion liquid cooling technology has emerged as a crucial development direction for data center heat dissipation. As the core medium of this technology, the performance evaluation of coolants and the changes in their physical properties exert far-reaching impacts on the heat dissipation effect and operational reliability of the system. This paper conducts an investigative evaluation of several common coolants in the market and uses CFD numerical simulation to analyze the influence of physical property changes of coolants on the cooling system. The results show that: under low flow rates(1-4 m3/h), fluorocarbons exhibit better cooling performance(e.g., the average GPU temperature of SF10 is 5.7 ℃ lower than that of BINGYI 797); under high flow rates, hydrocarbons surpass them(e.g., BINGYI 797 is 3.5 ℃ lower than SF10). Among the physical property parameters, density shows the largest impact on average GPU temperature, followed by thermal conductivity, dynamic viscosity, and specific heat capacity. Density and thermal conductivity could reduce the average GPU temperature by 9.8 ℃ and 9 ℃ respectively. An increase in dynamic viscosity leads to a temperature rise of 6.7 ℃; while specific heat capacity has a smaller impact(only reducing the temperature by 2.7 ℃).
Annual dynamic energy consumption analysis of cooling and heating composite system in data centers under variable IT loads
Zhou Feng;Yao Kewen;Zhang Xuebin;Chen Mingfei;Ma Guoyuan;With the rapid upgrade of computing power and expansion of infrastructure in data centers, their power consumption has experienced exponential growth, and energy conservation and consumption reduction has been identified as a key research direction. In practical operation, two major challenges have been encountered: IT equipment does not operate at a constant power, as the real-time utilization rate of processors and network traffic load show dynamic fluctuations; meanwhile, the annual dynamic coupling of the actual integrated cooling and heating system needs to be considered for full-system optimization. To address these challenges, an cooling and heating composite system for data centers under variable IT loads is constructed in this paper, which targets the annual dynamic IT loads of data centers. Taking the Inner Mongolia Autonomous Region as a case study, an optimized cold storage scheduling scheme covering valley periods and part of flat periods is investigated for the integrated cooling and heating system. Compared with the original scheme, the energy consumption of water pumps, water chillers and heat pumps in the optimized integrated system has been decreased by 4.04%, 8.95% and 0.26% respectively, while that of cooling towers has been increased by 8.78%. Moreover, the total energy consumption has been reduced by 5.31%, and the annual operating cost has been saved by 9.51%. This study can provide support and reference for the energy consumption optimization and scheduling scheme selection of integrated cooling and heating composite systems in data centers.
