Our recent post containing information about Green ICT issues at universities generated a lot of interest. Here is another insight: it is laboratory facilities are the big energy consumers on campuses.

A 2015 University of North Carolina at Chapel Hill purchasing policy illustrates how labs compare to other buildings. "These 'plug loads' vary widely but on average consume 30% of the electricity used in laboratory buildings and 20?25% of the electricity used in administrative buildings."

2014

A UK report says that imaging and audio-visual equipment can be significant contributors to a university's ICT energy consumption. This is not surprising given that these technologies moved from analog to digital over a decade ago and are now firmly embedded in ICT infrastructures.

2010

Carbon and Computers in Australia, commissioned and published by the Australian Computer Society, ranks "Education and Training" as second out of 20 economic sectors in terms of total ICT carbon emissions and tops in terms of ICT emissions per employee. There is no reason to believe this is not typical of educational institutions in other developed countries.

2009

The University of Wisconsin at Madison reported that while computer and laboratory equipment accounted for only ~8% of total energy consumption*, an analysis of electricity consumption by facility type** revealed that lab buildings consumed ~60% of the institution's electricity. (Classroom buildings accounted for only about ~15% of electricity. (Link removed as of December 2015.)

Universities will sometimes exempt research labs from energy-saving measures like occupancy-based HVAC practices. The "Occupancy Schedules" table at the University of North Carolina policy was an example, but the policy had disappeared by late 2015.

Another is that no university holds lab directors and senior researchers accountable in any meaningful way for their labs' energy consumption or carbon footprints. (Please post a comment if you know of an exception.)

* HVAC energy consumption was separately reported. Assuming a PUE of 2.0, the gear would actually be responsible for ~16% of energy consumption. Also, this figure does not appear to include the energy consumption associated with the university's extensive telecommunication infrastructure.
** These figures appear to exclude electricity consumption for residence buildings.