Research paperCitation2020

Cooling architecture selection for air-cooled Data Centers by minimizing exergy destruction

Energy (Elsevier BV) · Rohit Gupta, Sahar Asgari et al.

A paper comparing four air-cooled data centre architectures with CFD and first/second-law (exergy) analysis, showing that the chiller dominates irreversibility.

On this page

Overview

This paper by Rohit Gupta, Sahar Asgari, Hosein Moazamigoodarzi, Souvik Pal and Ishwar K. Puri was published in Energy (Elsevier, volume 201, 2020, article 117625). The paper addresses cooling architecture selection on the grounds that about 33% of overall data centre energy consumption is attributed to cooling infrastructure.

The method combines CFD with first- and second-law (exergy) analysis. The four architectures compared are in-row cooling, rack-mountable cooling units, underfloor air delivery and overhead delivery. Each is served by a vapour-compression chilled-water plant. Exergy destruction is broken down by component (airspace, CRAH or cooling-unit heat exchangers, chiller) and linked to a dimensionless group of Peclet number and dead-state temperature ratio.

The main results are these: the chiller contributes the largest share of total exergy loss, about 55-60%, across all geometries and scenarios. Airspace exergy loss is up to 20% of the total for underfloor delivery, from hot/cold mixing and plenum pressure drop, versus about 5% for the other three architectures. Rack-mountable cooling units have up to 23% lower overall exergy loss than the underfloor scheme.

Increasing cooling-unit airflow progressively raises chiller exergy loss as well as fan power, so the authors recommend exergy-aware coordinated control of CRAH fans and chillers. The study is a comparative analysis, and its findings are limited to the four architectures and scenarios examined. It nevertheless shows that close-coupled architectures waste less exergy than raised-floor delivery and that chiller irreversibility is the main area for improvement.

Key points

  1. About 33% of overall data centre energy consumption is attributed to cooling infrastructure.
  2. The chiller contributes the largest share of total exergy loss, about 55-60%, across all geometries and scenarios.
  3. Airspace exergy loss is up to 20% of the total for underfloor delivery versus about 5% for the other three architectures.
  4. Increasing cooling-unit airflow raises chiller exergy loss as well as fan power.
  5. The authors recommend exergy-aware coordinated control of CRAH fans and chillers.
  6. Rack-mountable cooling units have up to 23% lower overall exergy loss than the underfloor scheme.
  7. Architectures compared: in-row, rack-mountable, underfloor and overhead delivery.

Using it in design

  • When comparing room, row and rack cooling architectures, report chiller, airspace and heat-exchanger losses separately.
  • Show that over-supplying airflow costs chiller energy as well as fan energy.
  • Prefer close-coupled or contained delivery where raised-floor mixing losses are high.
  • Point to thermally driven chillers or heat reuse as the main way to cut chiller irreversibility.

Monthly newsletter · free

Keep your reading list current

Once a month: new library sources, changes to Turkish regulations and upcoming data centre events.

SoSo CAD

Check cooling capacity alongside your design

Review zone heat loads, available capacity and the capacity balance after the largest unit is lost in SoSo CAD.

Request a demo
Zone heat loads and capacity balance in the SoSo CAD cooling workspace

Who should read it

Mechanical designers choosing cooling architecture, energy modellers and data centre efficiency researchers.

Related standards

  • ASHRAE TC9.9 Thermal Guidelines

Publication details

Authors
Rohit Gupta; Sahar Asgari; Hosein Moazamigoodarzi; Souvik Pal; Ishwar K. Puri
Publisher
Energy (Elsevier BV)
Year
2020
Volume
201
Pages
117625
Document language
English
Source type
Scientific
Access
Citation only
Licence
Licence conditions not established — confirm the conditions with the publisher
Original licence note

Read from the authors' preprint in UCL Discovery. No explicit open licence could be verified for that copy, and Crossref lists only Elsevier TDM licences for the published version, so this is a citation only.

Citation

Rohit Gupta; Sahar Asgari; Hosein Moazamigoodarzi; Souvik Pal; Ishwar K. Puri (2020). Cooling architecture selection for air-cooled Data Centers by minimizing exergy destruction. Energy (Elsevier BV). 201, 117625. https://doi.org/10.1016/j.energy.2020.117625

All content and rights belong to the respective authors and publishers. SoSo CAD provides only the citation, its own summary and a link to the source, and is not responsible for the sources' content.