1 Comment

Summary:

Simply moving to the cloud isn’t an environmental strategy. Now, says Eirikur Hrafnsson of GreenQloud, the conversation has shifted to how we can make the cloud more sustainable and resilient in a 24/7 connected and data-hungry world.

big cloud
photo: Editor B

It’s no big secret these days that moving to the cloud is no longer a sufficient environmental policy. Sure, it may reduce paper usage, underutilized hardware and localized energy costs, but those things alone aren’t a business necessity yet.

More recently, the conversation has shifted from merely moving to the cloud in order to achieve efficiency to making the cloud more sustainable and resilient. What we’ve learned over the last several months, from various research reports and extensive media coverage, about the amount of CO2 generated from the IT industry’s exponentially growing online data and applications is alarming.

McKinsey’s 2007 forecast that by 2020 the data industry will represent over 4% of the total global CO2 emissions is more plausible than ever. Considering that most of the world’s data centers are still powered by majority fossil fuel-based energy grids, the question remains: what is the best way for the IT industry to limit dangerous CO2 emissions and foster an environment of resilience?

Greenpeace made waves across the IT industry this spring by publishing its “How dirty is your data” report, which launched a media firestorm and summer-long naysaying campaigns from the likes of Amazon, Google, Microsoft, Apple and Facebook. This data wound was reopened at the end of September in an eye-opening article by James Glanz of the New York Times entitled “The Cloud Factories: Power, Pollution and the Internet,” that explored how data centers and the cloud use, and misuse, power. Whether you agreed with the article or not, there is no denying that it reignited a passionate call for cloud and data center service providers to lead the IT industry to be more energy efficient and lower global carbon emissions caused by online data.

In response to such concerns and public scrutiny, many big data companies have turned to evangelizing their PUE (power usage efficiency), while others have stepped up their purchasing of Carbon Credits as a means of offsetting their data carbon load. I question, if not altogether disagree, with this approach.

Given the current numbers, Information and Communications Technology (ICT) cannot be sustained through carbon offsets or today’s PUE standards alone. While advancements in hardware longevity and efficiency will inevitably go a long way to helping to make the cloud more sustainable, the resulting carbon emissions that are generated from the industry’s reliance on fossil fuel-based energy grids to power and cool servers, as well as the underutilization of available clean renewable energy, is far greater than the amount of carbon offsets currently generated for environmental renewal and protection programs through the purchase of carbon credits.

This statement is supported by a number of reports and studies both directly and indirectly related to fossil fuel availability. This week, for example, an article came out that looked at why energy regulations and the election matter. Also, the UN recently allowed carbon-burning coal plants to qualify for carbon credits if the coal power plant is viewed as critical to a particular region. The world knows we are running out of fossil fuel. In his talk at the Structure:Europe this week, Tate Cantrell said about 130 years left of coal at current production. The World Coal organization actually has it at 112 years:

We’ve been here before, in the 1970s during the global oil crisis. We also know, from the New York Times article and others, that data centers are increasing straining power grids and the more online services that are available to businesses and the public, the more Jevons paradox comes into play. Finally, as fossil fuels become less available, prices will rise (again, see 1970s oil crisis) and energy will become increasingly, if not prohibitively, expensive, which means communities, municipalities and critical life services will be competing with data service providers for that energy.

In a recent TechTarget article, Phil Nail, CTO of AISO, a solar-powered hosting company based in Riverside County, Calif., estimated that “cooling costs can account for 30% to 40% of the power load of many data centers.” This may be the reason many companies and the industry as a whole are beginning to look beyond carbon offsets and setting sights on developing supplemental programs that splice renewable energy solutions into existing strategies while also sourcing geographically strategic locations, with both cooler climates and cheaper renewable energy resources.

In much the same way the IT industry has become global through online infrastructure, companies are now looking beyond their back yard for the most effective, efficient and renewable resources to increase resilience. Both Facebook and BMW, for example, have very publicly revealed their plans to put more data in Nordic territories and capitalize on the renewable energy resources available in the North to power their data.

At our company, as a way to better gauge the IT industry’s adoption of green power, we developed a Green Power Usage Effectiveness (GPUE) metric— a spin on PUE calculations — to measure the CO2 emissions created for each usable KWh for IT equipment and to enable the industry to view their performance against this metric

The North is not unique in its efforts toward more energy resilience. On the other side of the Equator, Brazil has demonstrated energy resilience initiatives in particular with ethanol, wind and biomass energy to compensate for hydroelectric power shortages during seasons of low rainfall. Like Iceland, Brazil’s hydroelectric grid grew significantly following the global oil crisis of the 1970s.

Both countries have since focused strongly on liberating themselves from oil dependency by developing renewable energy resources, with Iceland focusing predominantly on hydroelectric and geothermal energy as well as carbon recycling. According to a recent Bloomberg article, Brazil expects to see $235 billion in renewable energy investments over the next decade. Whether intended or not, the focus on renewables has positioned both countries as attractive solutions for Big Data’s emerging energy and CO2 crisis.

We will see a lot of discussion and innovation in the coming months and years, focused on the best approach to achieving resilience, from an environmental standpoint and to meet the demand for data accessibility in a 24/7 connected world. More effectively utilizing current renewable and non-renewable energy, cultivating efficiency in data management and constantly innovating in these areas, as we are beginning to see now, will go a long way towards fostering resilience. It won’t hurt to have a few watchdog reminders now and then to keep the IT industry on its toes.

Eiríkur Hrafnsson is the co-founder and chief global strategist of GreenQloud, a public compute cloud headquartered in Iceland. 

  1. The new environmental book, Green Illusions, shows the importance of addressing Jevons Paradox as a first step in combating our broader energy and environmental challenges. The author argues that Jevons Paradox is a better project to pursue than new energy technologies, which have many negative side effects and limitations.

    This book also details a hidden “boomerang effect” that is undermining green energy investments.

    Here’s part of the book: http://books.google.com.ph/books?id=uVIh-5BMofEC&pg=PA174&lpg=PA174&dq=Green+Illusions+jevons&source=bl&ots=4NEEyQqL0l&sig=o3o0SmjeUYFu_pQYLSkHs-3fcp8&hl=en&sa=X&ei=03mOUMuvGsiwiQfd44GQAg&ved=0CBoQ6AEwAA#v=onepage&q=Green%20Illusions%20jevons&f=false

    Share

Comments have been disabled for this post