Global trends, performance metrics, and energy reduction measures in datacom facilities

被引:26
作者
Isazadeh, Amin [1 ]
Ziviani, Davide [2 ]
Claridge, David E. [1 ]
机构
[1] Texas A&M Univ, Sch Mech Engn, Energy Syst Lab, College Stn, TX 77840 USA
[2] Purdue Univ, Sch Mech Engn, Ray W Herrick Labs, W Lafayette, IN 47907 USA
关键词
Data center; Power density; Performance metric; Edge computing; Decentralized data center; Workload migration; Renewable energy; High temperature data center; Waste heat recovery; Energy storage; PHASE-CHANGE MATERIAL; AIR-SIDE ECONOMIZERS; CHANGE MATERIALS PCMS; WASTE HEAT-RECOVERY; DATA CENTERS; THERMAL-CHARACTERISTICS; LOOP THERMOSIPHON; COOLING SYSTEMS; STORAGE; PIPE;
D O I
10.1016/j.rser.2023.113149
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Over the last decade, the demand for data center and network services has increased dramatically. To meet this demand, global average rack power density has risen from 2.4 kW/rack in 2011 to 8.4 kW/rack in 2020 with the aid of technological advancements. About 36% of global data centers have racks above 30 kW/rack. Average Power Usage Effectiveness (PUE) dropped from 2.5 in 2007 to 1.65 in 2013, but has been almost flat since then, at 1.59 in 2020. The depletion of finite fossil fuel reservoirs, adverse impact of greenhouse gas emissions on global warming, recyclability of electronic waste, and the rising cost of energy are pushing the datacom industry toward more energy-efficient and sustainable technologies including renewable energy, highly efficient elec-tronics cooling, waste heat recovery, passive cooling systems, and energy storage, which are broadly reviewed in this paper. Use of renewable energy resources through power purchase agreements rose by 346% from 2.4 GW in 2017 to 10.7 GW in 2020. Air-side economizers can reduce PUE by 30-50% and water-side economizers by 10-30%. The server/rack outlet is an ideal spot for waste heat recovery and can be coupled with a multi-stage heat pump to save significant energy if coupled to a district heating system. Active and passive deployment of thermal energy storage can save more than 20% of cooling energy and reduce temperature fluctuations by more than 60%. Edge computing, decentralization, and virtualization can lead to lower latency, higher scalability, and reliability during failure.
引用
收藏
页数:34
相关论文
共 215 条
[1]  
Abedin AliH., 2011, OPEN RENEWABLE ENERG, V4, DOI [DOI 10.2174/1876387101004010042, 10.2174/1876387101004010042]
[2]   A comparison of heat transfer enhancement in a medium temperature thermal energy storage heat exchanger using fins [J].
Agyenim, Francis ;
Eames, Philip ;
Smyth, Mervyn .
SOLAR ENERGY, 2009, 83 (09) :1509-1520
[3]   Review of blockchain-based distributed energy: Implications for institutional development [J].
Ahl, Amanda ;
Yarime, Masaru ;
Tanaka, Kenji ;
Sagawa, Daishi .
RENEWABLE & SUSTAINABLE ENERGY REVIEWS, 2019, 107 :200-211
[4]  
Ahmed A., 2018, E3S WEB C, V32, DOI [10.1051/e3sconf/20183201001, DOI 10.1051/E3SCONF/20183201001]
[5]  
Ahuja N, 2012, P IEEE SEMICOND THER, P104, DOI 10.1109/STHERM.2012.6188833
[6]   Potential of data centers for fast frequency response services in synchronously isolated power systems [J].
Al Kez, Dlzar ;
Foley, Aoife M. ;
Ahmed, Faraedoon W. ;
O'Malley, Mark ;
Muyeen, S. M. .
RENEWABLE & SUSTAINABLE ENERGY REVIEWS, 2021, 151
[7]  
Alsop T, 2019, LEADING METHODS USED, V2021
[8]   Blockchain technology in the energy sector: A systematic review of challenges and opportunities [J].
Andoni, Merlinda ;
Robu, Valentin ;
Flynn, David ;
Abram, Simone ;
Geach, Dale ;
Jenkins, David ;
McCallum, Peter ;
Peacock, Andrew .
RENEWABLE & SUSTAINABLE ENERGY REVIEWS, 2019, 100 :143-174
[9]  
[Anonymous], 2014, Cisco visual networking index: Global mobile data traffic forecast update, 2014-2019
[10]  
[Anonymous], 2016, P 15 INTERSOCIETY C, DOI DOI 10.1109/ITHERM.2016.7517694