Thermal Performance Analyses and Optimization of Data Center Centralized-Cooling System
Siqi Wang,Rang Tu,Xianzhong Chen,Xu Yang,Kun Jia
DOI: https://doi.org/10.1016/j.applthermaleng.2022.119817
IF: 6.4
2022-12-12
Applied Thermal Engineering
Abstract:In this study, the factors that influence the power consumption of a data center centralized-cooling system were investigated from both the indoor thermal environment and chilled water mass flow rate perspectives to maximize the free cooling utilization potential. First, a cooling load model is proposed and verified. Subsequently, based on a typical centralized-cooling system that combines free cooling devices and chillers, mathematical analyses were conducted on its thermal performance, considering different climate regions and heat dissipation densities related to the IT equipment. The results indicated that up to 83.5 % and 75.3 % of the annual cooling demand could be recovered by free cooling in Harbin and Beijing, respectively. However, during actual operations, this value is reduced because of the mixture of cold and hot air and inappropriate chilled water flow rates. To achieve the lowest annual or daily power consumption of the cooling system, the chilled water flow rates were optimized. Under optimal conditions, Power Usage Effectiveness (PUE) can be improved by 4.5 to 30.6 % compared to the worst case scenario under discussion.
energy & fuels,engineering, mechanical,thermodynamics,mechanics