The Green Grid › Carbon_Usage_Effectiveness_White

Posted In:

Metrics & Measurements

02 December, 2010

Download

Editor:
Christian Belady, Microsoft

Contributors:
Dan Azevedo, Symantec
Michael Patterson, Intel
Jack Pouchet, Emerson
Roger Tipley, Hewlett-Packard

In this paper, The Green Grid proposes the use of a new metric to complement the series of metrics that TGG has introduced in the past few years, which includes power usage effectiveness (PUE), data center energy productivity (DCeP), energy reuse effectiveness (ERE), and others. This white paper introduces the carbon usage effectiveness (CUE) metric and addresses data center–specific carbon emissions, which is emerging as an extremely important factor in the design, location, and operation of these facilities today and in the future. CUE, combined with PUE, enables data center operators to quickly assess the relative sustainability of their data centers, compare the results, and determine if any energy efficiency and/or sustainable energy improvements need to be made.

RE: Will anybody share data towards public 'carbon footprinting' of facilities?

Hi Francine, Thanks for your post. The Green Grid has a liaison with the Open Data Center Alliance, which is doing some work directly related to carbon footprint. You may want to visit their website at http://www.opendatacenteralliance.org/ for more information. Best regards, The Green Grid Administration

Posted at 05:37 PM on May 31, 2012 by The Green Grid Administration

Will anybody share data towards public 'carbon footprinting' of facilities?

A very interesting paper - thank you. We've been doing some work at Mastodon C (www.mastodonc.com) to provide estimates of the CUE of different public cloud options. Public cloud is an area where CUE is especially relevant, given that users are usually free to move across different locations and providers rather than being tied into an owned and operated data centre. Given that public cloud providers don't offer comprehensive or standardised data on their footprints, we're currently using climate and eGrid energy grid data to infer estimates of the footprints. I believe this to be a reasonable approach, given that based on my own modelling, 60% of the variation between a 'good' and a 'bad' CUE is likely to be related to the primary power source, and 20% is related to the external climate of the location, which are both publicly available information (the remainder determined by physical design, occupancy, backup power source, etc). We would be extremely interested in working with any data centre or public cloud providers who would like to participate in a public and open effort to provide ratings information: our current models are, I believe, pretty good estimates, but are all based 'outside the walls' so could clearly be improved. I believe that this is a particularly important issue for the public cloud, given the buyers' mobility and the dramatic differences in CUE between providers and locations; enhanced transparency could lead to extremely significant global carbon reductions.

Posted at 06:20 AM on May 11, 2012 by Francine Bennett, Mastodon C

Re: "Equivalent" CO2

Thank you for your thoughtful question. Yes, there are numerous gasses associated with CO2e, or as we listed in our formula CO2eq with eq denoting equivalent. The EPA along with other global reporting organizations have developed a model for combining the individual gasses into a single CO2e number based upon their relative ‘climate impact’ versus CO2. Most utilities will report their CO2e data as a single number. We recommend you use the CO2e number and not attempt to combine the individual gas data yourself. We will examine our descriptor and add appropriate footnotes/references in the next revision. Again, thank you for bringing this to our attention.

Posted at 09:47 AM on June 30, 2011 by Jack Pouchet

Re: "Equivalent" CO2

Posted at 05:33 PM on June 29, 2011 by kavi\admin

"Equivalent" CO2?

I've seen a utility report their emissions as "CO2 Equivalent." I see in the EPA eGrid that they break out NOx, SO2, CO2, CH4, N20. How do I convert from the eGrid break-outs to "CO2 Equivalent?" Will the CUE metric be updated to account for greenhouse impact of non-CO2 emissions?

Posted at 07:32 AM on June 28, 2011 by Edward H

Re: Pounds vs kg

Hello Andrew, Thanks for your comment. The Green Grid continues to work closely with the EPA, Dept. of Energy and other government/industry organizations around the world to help make data centers more energy efficient, sustainable and cost-effective. We’ve recently placed a strong emphasis on helping organizations realize the environmental impact of their data centers and expect the increased visibility around CUE and TCE to serve as an industry catalyst for even more robust efforts to improve sustainability within data center facilities. As you point out, there are some strong similarities between the TCE metric and The Green Grid’s CUE metric, as well as some notable differences. While you’ve highlighted the different units of measurement used for TCE and CUE, it’s also important to distinguish that the TCE metric establishes the CO2 output emission rate of electricity used by the facility, which could be interpreted to exclude the purchase of chilled water or not. However, the CUE metric establishes the CO2 output emission rate of total data center energy, which includes purchased water, and all energy sources used on site. Again, we appreciate your feedback as we encourage greater collaboration across the industry and look to develop data center metrics that are widely adopted in the U.S. and abroad.

Posted at 01:27 PM on February 11, 2011 by Dan Azevedo

Pounds vs kg

The easiest to obtain source of normalized carbon data for grid sourced electricity in the US is the EPA’s eGrid as the paper references. While the use of kg (vs. pounds CO2 emitted) has greater international acceptance, US eGrid data is provided in pounds. For this reason it was decided to simplify conversions for US users by using pounds CO2 emitted in what we have been using as the TCE Metric since 2007. It appears that this ‘new' sustainability metric is largely the old TCE concept represented in kg. See: http://www.datacenterdynamics.com/Media/MediaManager/Technology_Carbon_Efficiency.pdf It is great to see this concept finally gaining greater awareness/acceptance. If (when?) the US begins regulating carbon, it will be interesting to see if government regulation is enforced in imperial or metric units for greater international acceptance and trade considerations.

Posted at 07:17 AM on December 17, 2010 by Andrew Cook

Re: Carbon emissions

Hello John, As the metric is recommended as an end user tool to drive improvement in your environment respectively, if you have or can obtain more relevant data then what has been recommended, by all means please do so. It is more important to be consistent in the approach and calculation if using the metric for individual improvements over time, regardless if it the source calculations are based on utility or state data.

Posted at 04:27 PM on December 08, 2010 by Dan Azevedo

Carbon emissions

The paper suggests using regional information (for example, from the EIA) for carbon emission data. Would you agree that using information from the relevant utility within the state would be preferable - assuming the utility publishes that information? EIA publishes average data for each state. I have found that different utilities within the same state may have significantly different generation mixes. This can result in significantly different carbon footprints.

Metrics & Measurements

White Paper

Comment Policy

WP#32-Carbon Usage Effectiveness (CUE): A Green Grid Data Center Sustainability Metric

Share

RE: Will anybody share data towards public 'carbon footprinting' of facilities?

Will anybody share data towards public 'carbon footprinting' of facilities?

Re: "Equivalent" CO2

Re: "Equivalent" CO2

"Equivalent" CO2?

Re: Pounds vs kg

Pounds vs kg

Re: Carbon emissions

Carbon emissions

Comment On This Document

Related Content