SmartPlanet and ZDNet colleague Andrew Nusca just reported the results of a new Green IT confab, which provided estimates that data center power management can cut the IT energy drain by more than $1 billion at a single site.
Representatives of The Green Grid — a consortium that seeks to improve energy efficiency in data centers and business computing ecosystems worldwide — presented the results of a recent assessment of a mid-tier data center operated by the EPA, and said more than $1.1 billion could be saved by greening up the data center:
Anything that can promote Green IT and low-impact data centers is great stuff. But I’d love to see a bigger-picture study on the degree of power consumption that IT is helping us to avoid. For example, how many physical retail stores have not been built, and do not operate, due to e-commerce? How many automobile trips and additional office space is no longer necessary due to telecommuting and remote work? Perhaps, we’ll find, for every kWh IT consumes, it saves x number of kWhs. I’m not aware of data on this.
Studies, such as that published by the Lawrence Berkeley National Laboratory back in 2007 (PDF download available here), estimated that electricity used by server computers doubled between 2000 and 2005, and now gulp down about 45 billion kilowatts per hour – equivalent to the amount of power used by the entire state of Mississippi in 2005. Additional devices such as storage, network equipment, and client front-ends were not included in the calculations.
It’s no secret that data centers – especially those run by social networking leaders such as Google, Yahoo!, Microsoft, and Amazon – are the new industrial behemoths of the 21st Century. These providers are even taking care to make sure that they build their latest and greatest data centers within range of hydroelectric power sources.
Other vendors have been jumping on the “green” bandwagon, and many across the industry are pressing data center operators and solutions providers to go green, and better conserve energy resources. These are all noble goals, and there’s no doubt it’s in the best interest of companies to save on power costs.
However, most studies on IT energy consumption don’t to take into account the overall savings in electricity and power usage as a result of mass computerization. The author of the Lawrence Berkeley study, Jonathan Koomey, admits right up front that the study “only assesses the direct electricity used by servers and associated infrastructure equipment. It does not attempt to estimate the effect of structural changes in the economy enabled by increased use of information technology, which in many cases can be substantial.”
These structural changes can be substantial indeed, and it would be interesting to see studies on how these impacts are being felt.
Consider these potential areas of direct and indirect energy savings and you quickly get a very long list. Together, we’re probably talking about more resources saved than that used up by servers. Consider just one aspect: It’s generally assumed – and documented on a case-by-case basis – that e-business has drastically reduced the amount of paperwork moving inside and between organizations. Or, looking at the proliferation of online news and information, which many people now turn to, instead of hard-copy newspapers, magazines, or journals. How much paper, and therefore trees, has been saved as a result of Websites and PDF files? How has this savings translated into reduced tree harvesting and less energy consumed for the physical delivery of such documents?
Consider the ways IT has helped in energy conservation by society at large:
- The e-commerce channel is now a strong part of many businesses, and it can be assumed that to some degree, it has replaced some bricks and mortar construction and management, and all the energy consumption that goes with that. How many stores and shopping centers have not been built because of the replacement of sales through online channels?
- Oil companies have been able to cut back on costly, time- consuming, sometimes environmentally risky drilling exploration because they can now model geologic environments with sophisticated tools running on large systems. How much disruption to the environment has been averted because of such modeling?
- There’s the less frequent travel required, since collaborative tools and platforms enable teams to work virtually across the globe. How much have we saved in jet fuel for trips that no longer take place?
- Online and distance learning have brought campuses right to students’ homes, cutting down on travel to and from physical campuses. Likewise, telecommuting enabled through IT has cut down on work commutes. How much fuel has been saved for these trips to work and school?
- There have been industry-specific gains. For example, insurance companies leverage mobile technologies to cut down on trips made by claims adjusters out in the field - a savings that alone along ought to add up to quite a few barrels of oil.
Another area worth further study, not directly related to energy savings and environmental impact, is the economic opportunity and revitalization is now a possibility to distressed communities, and no longer limited to a few fortunate regions. E-business and networking have opened up new channels of knowledge and opportunity to people that have long been denied such access.
Think about remote rust-belt communities or Native American reservations that for years have been cut off, by geographic distance, from the mainstream of business. Residents have had to leave their communities to seek employment or attend school in other parts of the country. Now, it’s possible to build businesses, and attend the nations’ best universities, without leaving these communities. Will these distressed communities begin to see less outflow of human capital as a result of this connectivity to the world?
As we employ greener technologies to bring down IT and data center energy consumption, it would be very eye-opening to see more data on the positive impact virtualized environments are having on our resources. In turn, we need to be able to measure not only how ramping up IT by x percent will consume x kilowatts of power, but also how is being saved on a wider level in terms of fuel costs and natural resources.
