Global warming has become a very "hot topic" (pun intended) in the Information Technology industry.
Rising energy costs are gaining the attention of IT Directors, CIOs and CFOs. That, coupled with an increased environmental awareness, are driving significant change in how IT is managed and how IT is used.
Simply stated, IT plays two roles with respect to sustainability: part of the problem, and part of the solution.
On the problem side of the ledger, the scope and scale of IT infrastructure results in some alarming statistics. It is estimated that data centres consume 1.5 percent of all of the electricity produced worldwide. Unfortunately, half of the electricity that is used for computation is required for cooling the computers in the data centres. Add in the fact that there are probably a greater number of CPUs outside the data centre and we quickly get close to the (vague) estimate that the IT industry power consumption has a greater CO2 impact than the global airline industry.
Not only do we need to be concerned about the environmental impacts of power consumption, but the sheer number of computers that exist require us to consider the environmental impacts of the manufacturing processes and, of course, the safe handling of the e-waste generated as we replace systems on a typical three or four year cycle. And the problem doesn't stop there. It is estimated that each knowledge worker uses computers and printers to generate about 1,000 pages per month. For the University of Calgary that means on a yearly basis we print a stack of paper about 30,000 feet high. If there are 40,000 universities the size of the U of C then the stack of paper generated by universities each year would reach the moon!
Each time I do these calculations I ask someone else to confirm the numbers. I am continually startled by the math. How can 40,000 universities generate so much paper that they could reach the moon? I oscillate from a frightening picture of stacks of paper reaching the moon and throwing it out of orbit, to ideas of how to I could attain the next X-Prize by stacking one sheet of paper on another ;-)
Back to the "carbon footprint" issue. There seems to be a great many ways to calculate the carbon impact of any particular consumption model. I was surprised by how many carbon footprint calculators there are on the web and more so, how widely the results vary. But regardless of the tool, it seems apparent that an information-intensive organization such as the university is consuming a lot of resources in the creation and dissemination of information. Roughly, in an extremely non-scientific analysis, I conclude that the university approximately carries an IT carbon footprint equivalent to that of sum of the faculty/staff across the rest of their lives. If people are really as concerned as they suggest in some forums, this could lead to some interesting dialogue during the next round of union or faculty association negotiations.
Fortunately, there are ways we can begin to address some of these issues. New server and cooling technologies can significantly reduce the power consumption in the data centre. Simple changes in the way we manage PCs -- such as blanking the screen instead of fancy graphical screen savers -- can significantly reduce the power consumption. We can insist that our vendors use manufacturing approaches that take into consideration environmental issues. We can deploy more efficient printing technologies or implement processes that remove the need for the paper-based document altogether. There are many viable options, if only we take the time and effort to factor environmental issues into how we purchase, use and dispose of technology.
But where I think IT gets really interesting is on the solution side. Not only can we reconfigure IT to reduce the impact of IT on the environment, but we can also use IT to reduce the impact of our other activities. For example, we can use IT to better monitor our environmental footprint; we can use IT to ensure we are only providing the services -- such as light and heat -- where we need it, when we need it; we can use IT to collaborate across distance and time and reduce the need for travel. There are many opportunities to use IT to help create a sustainable world.
Now, we need a model to help us determine how to evaluate these environmental opportunities against our simple return on investment calculations!
Great post. Eco friendliness is certainly the new buzz in the IT industry and, for good reason. However, I believe the UofC can do a lot more than considering "blanking the screen of PCs" to save power.
I think it's time for us to re-think how we do "desktop computing" and should seriously consider eliminating the PC in favor of thin clients. No, this is not a step backwards to the days of old terminal style computing, rather it is a step forward into a world where ones' desktop is accessible from anywhere on campus, resides in a secure data center on enterprise class systems, is resilient to power outages and data corruption and is always available. It's an architecture where *one* administrator can manage thousands of desktops, where applications can be deployed within minutes to thousands of end users and -best of all- where an end users never loose all their files due to a hard-drive failure. This kind of solution isn't new and it's alread been in use for years around campus - some of our Sun Rays are 6 years old.
Let's focus on power: Assume 5000 campus PCs consume on average 80 watts of power each (they are not used all day & excludes monitors). Operating 24x7x365 @ 9 cents / KWH costs ~$63/yr/PC -or- $315K for the campus. If we instead opt for thin clients whom consume < 4 watts on average (excludes monitor and Data Center servers) then the cost is ~$3/yr/client -or- ~$16K. Even adding $30K of power for the servers is no match for the PCs.