Comparative Thermal and Energy Analysis of a Hybrid Cooling Data Center With Rear Door Heat Exchangers

Author:

Gao Tianyi1,Samadiani Emad1,Sammakia Bahgat1,Schmidt Roger2

Affiliation:

1. Binghamton University-SUNY, Binghamton, NY

2. IBM Corporation, Poughkeepsie, NY

Abstract

Data centers consume a considerable amount of energy which is estimated to be about 2 percent of the total electrical energy consumed in the US, and their power consumption continues to increase every year. It is also estimated that roughly 30–40 percent of the total energy used in a data center is due to the thermal management systems. So, there is a strong need for better cooling methods which could improve the cooling capacity and also reduce energy consumption for high density data centers. In this regard, liquid cooling systems have been utilized to deal with demanding cooling and energy efficiency requirements in high density data centers. In this paper, a hybrid cooling system in data centers is investigated. In addition to traditional raised floor, cold aisle-hot aisle configuration, a liquid-air hybrid cooling system consisting of rear door heat exchangers attached to the back of racks is considered. The room is analyzed numerically using two CFD based simulation approaches for modeling rear door heat exchangers that are introduced in this study. The presented model is used in the second section of the paper to compare the hybrid cooling system with traditional air cooling systems. Several case studies are taken into account including the power increases in the racks and CRAC unit failure scenarios. A comparison is made between the hybrid cooling room and a purely air cooled room based on the rack inlet temperatures. Also in this study, total energy consumption by the cooling equipment in both air-cooled and hybrid data centers are modeled and compared with each other for different scenarios. The results show that under some circumstances the hybrid cooling could be an alternative to meet the ASHRAE recommended inlet air temperatures, while at the same time it reduces the cooling energy consumption in high density data centers.

Publisher

American Society of Mechanical Engineers

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3