An Energy Correction

by John Lansdale

I recently wrote an article about energy use in computers. It was based on verifiable information gathered from various web sources, mostly Wikipedia Orders of Magnitude (Energy) (1), which I’d put into a spreadsheet so I could extend and compare. I was hypothetically worried there would be enough energy in the world to power a billion geeks for the next 50 years. You know, so there wouldn't need to be a world war and all the poor people in China can be fed and we won't live in poverty in a bleak, pollution-scarred landscape and all, sustainability.

The numbers spoke for themselves. When talking of sustainability, efficiency in computers makes little difference. Even if using super low power 10 watt laptops. Automobiles and heating use far more energy. Any efficiency campaign has to start there. In energy economics I noted the brain using 20 watts x 24 hours = 413 calories of energy per day was a more significant topic.

Surprisingly someone at techsoup(2), Microsoft's Mindshare project (3) (Microsoft’s outreach to local user groups, we belong to it) has written an article on a similar topic and came up with a different conclusion. They compared an Energy star (109 watts active, 58 off) to a conventional (115 active 81 off mode) computer. They published:

Assuming that you shut down your computer for 16 hours a day (versus keeping it active 24-7), you could save $0.16 and 2.63 pounds of carbon dioxide per day, or 960 pounds per year. A car emits about 2,666 pounds of carbon dioxide per year, meaning that the energy you will save over the course of the computer’s life (3 years = 2,880 lbs) will equal a bit more than taking a car off the road for a year.

Somehow they implied this was economically significant and environmental. The amount of money saved by not using your PC at all (strangely not using Energy Star as they described), 16 cents / day = $58 / year seems reasonable. But their number for the car's carbon dioxide emission is way off.

According to a recent US Department of Energy report (4) average miles traveled per vehicle is 12,041 with an energy intensity in gallons per 1000 miles 49.5 or 20.2 miles per gallon. A burned gallon of gasoline creates about 20 pounds of CO2 (5). So an average US car would burn 12,041 / 20.2 = 992 gallons of gas x 20.37 = 20,200 pounds of carbon dioxide.

Techsoup author's conclusion should have been:

... A car emits about 20,200 pounds of carbon dioxide per year, meaning that the energy you will save over the course of the 7 computer’s lives (21 years = 20,160 lbs) will equal a bit more than taking a car off the road for a year.

Please though, do your own calculation. Use good, official data and be careful with the detail.

References

(1) Wikipedia Orders of magnitude (energy)
http://en.wikipedia.org/wiki/Orders_of_magnitude_(energy)

(2) Microsoft's techsoup - Reduce Your Computer's Power Consumption in Minutes
http://blog.techsoup.org/node/361

(3) Microsoft Mindshare - faq
https://www.microsoft.com/mindshare/faqs.mspx

(4) Household Vehicles Energy Use: Latest Data & Trends
http://www.eia.doe.gov/emeu/rtecs/nhts_survey/2001/

(5) Department of energy Carbon Dioxide weight per Gallon of Gas Calculation
http://www.fueleconomy.gov/feg/co2.shtml

The way this weight is calculated is interesting. When gas burns one carbon ( C ) atom is combined with two oxygen’s (O₂) from the air to create water (H₂O) and carbon dioxide ( CO₂ ). The atomic weight of Carbon is 12 and Oxygen 16 so the weight of carbon dioxide is 12 + (2 * 16) = 44. The relative weight of carbon dioxide to carbon in gasoline is 44/12 = 3.7.A gallon of gas weighs 6.3 pounds which is 83% carbon or 5.5 pounds of carbon. Then 3.7 * 5.5 = 20.37 pounds.

Note that petroleum consumption is energy use. CO₂ emission and energy efficiency is the same topic.

DacsGear!
Mugs and more, visit CafePress to order