Manage Learn to apply best practices and optimize your operations.

Online community aims to cut data center energy costs, save planet

Want to cut data center energy costs while saving the environment? The Green Grid and its online community aim to help.

Money is tight for many businesses, but demands on infrastructure just keep increasing. Add to that increased data center energy costs and it's a recipe for frustration in IT shops.

The Green Grid, a non-profit energy-efficiency group, hopes to help ease the burden with an environmentally friendly approach to data center resource management.

I'm Erin Watkins, and joining me this week are John Tuccillo, president and chairman of the Green Grid, and Brad Brech, IBM representative and Green Grid board member.

Tell us a little bit about the Green Grid and your responsibilities. John, we'll start with you.

John Tuccillo: I serve as the president and chairman of board for the Green Grid. The Green Grid is a global nonprofit organization that was assembled in 2007 to take on the challenges of resource efficiency in IT and in data centers. Today, the Green Grid has more than 200 member companies around the world, with a presence in Europe, Japan, a newly emerging presence in China and in North America.

Brad Brech : I'm a board member that represents IBM in the Green Grid and work as part of the great group of people in the consortium to try and help data centers around the world become more energy efficient.

There has been some harsh commentary on data center wastefulness lately. What are a few areas in the data center that need serious improvement, and what do you think has been a bit overblown?

John Tuccillo, president and chairman of the Green GridJohn Tuccillo

Tuccillo: I'm not sure I'd approach it from a specific section since some of the commentary that's been out in the media recently, in so much as I'd look at where the industry's come, where we still see opportunity for greater improvements and what some of the collaborative efforts around the world have been to essentially leapfrog the challenges as many times as we see them.

If you look at some of the work that the industry has done, for the first time ever you've got the industry looking at resource efficiency in both terms of economic return and environmental return. And through collaborative efforts -- through the Green Grid and others -- we've been making significant advancements in helping to improve those resource efficiencies.

Some examples of this could be that now we've got data centers that are being designed, built, commissioned and operated following globally accepted principles that have been developed by the Green Grid. Things such as how our friends at eBay, which is also a member company, recently designed and rebuilt an existing facility using the Green Grid's power usage effectiveness (PUE) metric as a target, but also using the data center maturity model (DCMM) to help guide the design and population of that data center, achieving some of the most significant PUE reporting in the industry.

Other examples include another member company, EMC, who actually uses PUE as a part of its reporting standards in their annual sustainability report. We've got other member companies such as Google that report their quarterly PUE on their website. We have another member company, Facebook, that is using one of our new environmentally structured metric called water usage effectiveness (WUE) as a way to quantify their water usage in an Oregon-based data center.

But it's not just in private industry where we're working to make these improvements. The industry's been working collaboratively with our friends in governments from around the world. That's one thing that perhaps isn't as well known amongst the global community. Private public partnerships between industries and in the United States, for example, between the Green Grid and the EPA and the DOA have really started to make significant improvements even within our Federal community in how they're managing, designing and building their data centers and how they're working to get the most productive return per unit of energy as possible.

They use themselves as examples of ways to offer training and education for all of the industry to make these improvements. We have similar relationships in Europe and Japan, and now in China, where the public private partnerships between government and industry are making significant inroads.

Brad Brech, IBM representative and Green Grid board memberBrad Brech

Brech: Yes, and John, just to add to that, one of the important things is that the Green Grid has done in this space is it's provided metrics, methods and help not only to what John pointed out -- our member companies -- to increase the efficiency of their data centers .

IBM is also a big user of PUE with all our data centers, but we use it with our customers so it flows out. Although everything doesn't turn overnight, think of the thousands and thousands of data centers around the world. I'm sure, like with everything else, there's a bell curve. The point is that the whole industry is moving forward. Efficiencies in data centers are rising and getting better. Are all of them to that point yet? Obviously not, but as a whole, the industry is moving forward.

The industry is working together to continue to move forward at a faster rate.

I'm hearing PUE and WUE. Are those both things the Green Grid has created, or were they existing terms?

Tuccillo: These are two of the three common metrics that were developed within the green grid. It's PUE, WUE and carbon usage effectiveness (CUE). These three metrics were developed within the collaborative framework that Brad mentioned. The way the Green Grid is structured, as these new bodies of work emerge within the organization, the "global community of collaboration" has equal opportunity across the industry to offer input, feedback, critique and balance for these bodies of work before they're launched by our global organization.

And that very open, transparent process is frankly one of the strengths of the organization for sure, but it also ensures that the work that we do has been developed by this cross section of the industry in a very productive manner.

It's really become a healthy dynamic. As Brad points out, it has an extension effect. As these resources, tools and metrics are being used with greater frequency by many of the large enterprise-class data centers, by many of the large IT partners, vendors and suppliers in the industry, that innovation extends into the utility companies, to the end-user community, government and academia -- it's a healthy cross section of the industry that then uses these resources for self-improvement and to help their peers improve.

But like Brad said, we're not done. We don't see ourselves being done any time soon. As an industry, we're collaborating on many new things to even further accelerate the resource efficiencies not only from the economic perspective but environmental as well. We've got lots of material that support those statements.

Brech: That's very good, John. Just going back, you mentioned metrics. What was important is, like with anything else, people don't improve unless they can measure and compare. The metrics provided a means for that to occur. With that, we have provided a lot of best practices. No two data centers are exactly alike, but most data centers can leverage the best practices that have been put forward and find the ones that work with them so that they can improve and measure their improvements.

Technology continues to change, so our job is never done. Hopefully two things happen: What we've learned by best practices encourages people to come up with new technologies to do things even more efficiently, and then we figure out how to alter and improve our best practices.

As John said, I don't think we're done by a long shot. It will be a continual effort as we find more and more ways to improve the situation.

Tuccillo: One of the things we've seen is what I call "tweaking." Because of the metrics and reporting mechanisms we've adopted, often what happens is we begin to see significant improvements by essentially tweaking either our IT infrastructure or facilities, shifting to that modular, scalable architecture that many have spoken of as ways to improve. And that comes from sharing.

A few years ago, the Green Grid conducted an energy audit of a United States Environmental Protection Agency data center. The EPA basically opened up the kimono and said "We want to make ourselves an example of how to improve."

They deliberately chose an older, midsized data center as their test case. We assembled a cross section of folks from pretty much every type of industry participant and, working with the EPA, went in and conducted the assessment.

The results of that came up with an average PUE for the type and age of data center. The Green Grid and the EPA made a number of recommendations, none of which required capital investment. It was just tweaking. How is it that, through their facilities layout and utilization of IT infrastructure, might they improve their energy efficiency? Or as we've come to look at it, how are they increasing their IT productivity per unit of energy?

The EPA not only executed the recommendations that the Green Grid made, but they went a little bit further through more tweaking. Other ways that they developed on their own also increased their utilization, and they have a heck of a success story as a result.

Look at the mandate that came out from the government a year ago in the United States -- the 25 points that the federal CTO and CIO supported in ways to improve energy efficiency in federal data centers.

When you boil it all down, it comes down to how you increase utilization, consolidate resources and quantify the intended performance. Those things only come about through this level of collaboration. By having the metrics in place, being able to illustrate where the opportunities are for improvement, have really helped to accelerate it for sure.

But that sharing of best-case scenarios allows people to interpret where there could be use from these best-case scenarios against individual business models.

Like Brad said, no two data centers are exactly alike, and the Green Grid doesn't want to put itself in a position of saying, "Thou shall determine productivity this way." What the best in industry can do is illustrate ways to quantify productive output from your IT and facilities infrastructure, which you can interpret to match with your particular business model for the same reasons you created a data center in the first place.

The federal government is the same way. They look at examples in private industry, actively engage with private industry and make significant improvements in their own IT resources and facility infrastructure and share that information within the industry.

Brad, I know you have to get going shortly. Is there any advice you want to add before you go?

Brech: I think the key advice is to continually look at what's going on in the data center, use the metrics to measure what's happening and then rationally take a look at best practices. Talk to other data center people, because they have a common bond in leveraging these metrics and working on these issues.

Just remember that every day there's something new going on and there's something new happening. Don't sit back; keep looking.

Stay tuned for part two of this podcast interview.

Dig Deeper on Data center design and facilities

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.