My Green Technology Strategies students have been working out the CO2 emissions caused by computers. The first step in this is to understand the National Greenhouse Accounts (NGA) Factors. The latest set of these from the Department of Climate Change, are from June 2009. The National Greenhouse and Energy Reporting (Measurement) Technical Guidelines (June 2009) have several different ways to calculate emissions, but the factors use “Method 1” .
The most relevant factors for ICT equipment are those for "Indirect (scope 2) emission factors for consumption of purchased electricity from the grid". This is because in Australia, most computer and telecommunications equipment runs from mains electricity supplied over the grid. The factor is expressed in kg CO2-e/kWh and allows energy (kWh) to be converted to kg of CO2e. This assumes a given form of electricity generation and distribution method.
Australian electricity is generated mostly by burning coal. Different grades of coal used in different states result in different amounts of CO2. As an example, Victoria uses more polluting brown coal with a factor of 1.22, whereas NSW uses cleaner black coal with a factor of 0.89. South Australia uses a higher proportion of natural gas which is cleaner gving a factor of 0.77 and Tasmania uses hydro electricity, with much less CO2e at 0.23.
To demonstrate the calculations, my netbook, plugged into a large screen uses about 40 Watts of electricity. Assuming it is run 10 hours a day, this is: 40 x 10 = 400 Watt Hours per day.
Assuming the computer is run 5 days a week, this is: 400 x 5 = 2,000 Watt Hours per week or 2kWh per week.
If the computer is run 50 weeks a year, this gives: 2 x 50 = 100 kWh per year.
My computer is in the ACT most of the time, where, according to the Department of Climate Change, the Emission factor is 0.89 kg CO2-e/kWh.
So my computer causes: 100 x 0.89 = 89 kg CO2-e per year.
Of course this is an over estimate, as my computer is not on all the time. As it is a netbook, I turn it off and take it with me. Even when plugged in the computer switches to a low power while idle. So this is an upper bound on the energy use.
But the two biggest mistakes students make with such calculations are more fundamental: excessive precision and wrong units. I have deliberately chosen round numbers to make the demonstration calculations easy. Students tend to make calculations to dozens of decimal places. This level of precision is spurious as there are so many assumptions used in the estimating of the factors and power measurements. The assumptions used are likely to result in a figure which could be off by a factor of four: rather than using about 100 kg CO2-e per year, my computer might be using between 25 and 400 kg CO2-e per year.
The second problem is wrong units of conversion. Typically in a calculation while worrying about dozens of decimal places the student uses the wrong units, confusing kW with MW or (or using British energy units) ending up being wrong by a thousand or million times. A quick sanity check is required to see if the figure derived is reasonable.
As a cross check, I found an estimate of 0.58 Kg CO2e per day from a desktop computer in the Green Ration Book. Multiplying this by 5 days a week and 50 weeks a year gives 145 kg CO2-e per year. This is about 50% higher than my calculation, which seems reasonable for a higher power desktop computer (a figure to within four times would be near enough).