國立高雄大學圖資館 |

Language: English

Back

Liquid Versus Air: Life Cycle Carbon of Cooling Down AI Data Centers /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Liquid Versus Air: Life Cycle Carbon of Cooling Down AI Data Centers /William J Hassel.
Author:	Hassel, William J.,
Description:	1 electronic resource (136 pages)
Notes:	Source: Masters Abstracts International, Volume: 86-07.
Notes:	Advisors: Leighton, Mark; Gloria, Thomas P.
Contained By:	Masters Abstracts International86-07.
Subject:	Sustainability.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31768092
ISBN:	9798302148377

Liquid Versus Air: Life Cycle Carbon of Cooling Down AI Data Centers /
Hassel, William J.,

Liquid Versus Air: Life Cycle Carbon of Cooling Down AI Data Centers /William J Hassel. - 1 electronic resource (136 pages)

Source: Masters Abstracts International, Volume: 86-07.

The rise of AI and cryptocurrency technologies has led to exponential growth in data centers globally. As these facilities consume substantial energy and contribute to significant greenhouse gas emissions, implementing sustainable cooling technologies in data center facilities is increasingly crucial. Preliminary findings suggested that applying liquid cooling systems within these data centers holds significant potential to markedly decrease energy consumption, greenhouse gas emissions, and water usage (Manganelli et al., 2021). Initial estimations indicate that data center facilities incorporating liquid cooling mechanisms could reduce both embodied and operational carbon emissions by up to 50% throughout their life cycle compared to those employing conventional air cooling methods. This adaptation could result in near-zero water usage and decrease the physical size of data center buildings, further reducing carbon emissions and capital expenditure, potentially achieving cost neutrality over the facility's life cycle.I primarily sought to answer three questions: How do liquid and conventional air cooling methods compare in terms of carbon emissions when evaluated through a comprehensive LCA? How does adopting liquid cooling influence data center facility design and resource utilization? Considering both upfront and operational costs, what are the cost implications of transitioning from traditional air cooling to liquid cooling? I hypothesized that data centers with liquid cooling systems will demonstrate substantial reductions in embodied and operational carbon emissions and water use, and despite higher upfront costs, liquid cooling systems will prove cost-neutral over their life cycle. To test these hypotheses, I analyzed the environmental and cost implications of data center cooling technologies through a comparative life cycle assessment (LCA) of liquid and traditional air cooling systems. The LCA adhered to the ISO standards (14040, 14044, 14067, and 15686), encompassing all stages of the data center's life cycle, reporting Global Warming Potential in kgCO2e. Material inventory data from equipment manufacturers and data center operators were aggregated and anonymized to create representative models of data centers using both cooling methods. The liquid cooling LCA models achieved up to 50% reductions in embodied and operational carbon emissions compared to conventional air cooling systems. Liquid cooling also enabled near zero water usage and a 50% smaller data center building size, which resulted in embodied carbon reductions and less resource use throughout the life cycle. My research findings present a compelling case for approaching data center cooling solutions with a holistic life cycle perspective, identifying the most environmentally sustainable and cost-effective cooling method, and providing recommendations for reducing data centers' environmental impact. The outcomes should be significant for data center operators, designers, and construction managers, enabling more sustainable and cost-effective design strategies for future data centers.

English

ISBN: 9798302148377Subjects--Topical Terms:

376755
Sustainability.
Subjects--Index Terms:

AI

Liquid Versus Air: Life Cycle Carbon of Cooling Down AI Data Centers /
LDR:04599nmm a22004693i 4500 001 676414
005 20250420214402.5
006 m o d
007 cr|nu||||||||
008 250901s2025 miu||||||m |||||||eng d
020 $a 9798302148377
035 $a (MiAaPQD)AAI31768092
035 $a AAI31768092
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Hassel, William J., $e author. $0 (orcid)0009-0001-8157-0413 $3 990250
245 1 0 $a Liquid Versus Air: Life Cycle Carbon of Cooling Down AI Data Centers / $c William J Hassel.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2025
300 $a 1 electronic resource (136 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 86-07.
500 $a Advisors: Leighton, Mark; Gloria, Thomas P.
502 $b A.L.M. $c Harvard University $d 2025.
520 $a The rise of AI and cryptocurrency technologies has led to exponential growth in data centers globally. As these facilities consume substantial energy and contribute to significant greenhouse gas emissions, implementing sustainable cooling technologies in data center facilities is increasingly crucial. Preliminary findings suggested that applying liquid cooling systems within these data centers holds significant potential to markedly decrease energy consumption, greenhouse gas emissions, and water usage (Manganelli et al., 2021). Initial estimations indicate that data center facilities incorporating liquid cooling mechanisms could reduce both embodied and operational carbon emissions by up to 50% throughout their life cycle compared to those employing conventional air cooling methods. This adaptation could result in near-zero water usage and decrease the physical size of data center buildings, further reducing carbon emissions and capital expenditure, potentially achieving cost neutrality over the facility's life cycle.I primarily sought to answer three questions: How do liquid and conventional air cooling methods compare in terms of carbon emissions when evaluated through a comprehensive LCA? How does adopting liquid cooling influence data center facility design and resource utilization? Considering both upfront and operational costs, what are the cost implications of transitioning from traditional air cooling to liquid cooling? I hypothesized that data centers with liquid cooling systems will demonstrate substantial reductions in embodied and operational carbon emissions and water use, and despite higher upfront costs, liquid cooling systems will prove cost-neutral over their life cycle. To test these hypotheses, I analyzed the environmental and cost implications of data center cooling technologies through a comparative life cycle assessment (LCA) of liquid and traditional air cooling systems. The LCA adhered to the ISO standards (14040, 14044, 14067, and 15686), encompassing all stages of the data center's life cycle, reporting Global Warming Potential in kgCO2e. Material inventory data from equipment manufacturers and data center operators were aggregated and anonymized to create representative models of data centers using both cooling methods. The liquid cooling LCA models achieved up to 50% reductions in embodied and operational carbon emissions compared to conventional air cooling systems. Liquid cooling also enabled near zero water usage and a 50% smaller data center building size, which resulted in embodied carbon reductions and less resource use throughout the life cycle. My research findings present a compelling case for approaching data center cooling solutions with a holistic life cycle perspective, identifying the most environmentally sustainable and cost-effective cooling method, and providing recommendations for reducing data centers' environmental impact. The outcomes should be significant for data center operators, designers, and construction managers, enabling more sustainable and cost-effective design strategies for future data centers.
546 $a English
590 $a School code: 0084
650 4 $a Sustainability. $3 376755
650 4 $a Information technology. $3 184390
653 $a AI
653 $a Carbon emissions
653 $a Cryptocurrency technologies
653 $a Data centers
653 $a Life cycle
653 $a Liquid cooling
690 $a 0640
690 $a 0474
690 $a 0489
690 $a 0800
710 2 $a Harvard University. $b Extension Studies. $3 966837
720 1 $a Leighton, Mark $e degree supervisor.
720 1 $a Gloria, Thomas P. $e degree supervisor.
773 0 $t Masters Abstracts International $g 86-07.
790 $a 0084
791 $a A.L.M.
792 $a 2025
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31768092