Greenhouse gas emissions per kg of fossil fuel

The reason they can do that for coal and natural gas is because the ratio of carbon and hydrogen are known. Anthracite is nearly pure carbon. Natural gas is methane. (Though some places it's actually a mix of methane, ethane, butane, and propane, but the chart is dealing with methane.) But for arbitrary "fossil fuel", in particular for petroleum, it's more problematic. Gasoline has a different hydrogen fraction than diesel, which is different from jet fuel, which is different from heating oil. So it's not easy to say for generic "fossile fuel" how much carbon gets burned per unit fuel. Which fossil fuel? When it comes to other air pollutants, it's even worse. Some coal and petroleum has a lot of sulphur; some has nearly none. But some power plants have scrubbers that take sulphur gases out of the exhaust; others don't. The extent to which nitrogen oxides are created depends a lot on the exact nature of the combustion plant, but also on pollution controls. One of the things a catalytic converter in a car does is to break down NOx. Likewise the emission of uncombusted hydrocarbons. That's another thing that a catalytic converter gets rid of in a car, and big power plants try to burn completely, because exhausted hydrocarbons are wasted fuel. So unfortunately, the only real answer to your question is, "It depends." What fuel? Burned how? In what kind of facility?

well, which fossil fuel = the global average mix for energy production, as stated in my data source. I know the correct answer is "it depends" but even an order of magnitude guess for C02 alone would be helpful.

Awesome sgt sandwich, thanks. I tried out these numbers (62kg coal/38kg gas) with the EPA conversion factors I linked in that .pdf. It came out slightly less - 251kg of CO2 instead of 331kg - but I think theirs is a measurement whereas yours is an upper bound. Given that they are in the same ballpark I am confident enough to use this number. (oh also, the IEA statistic has a similar breakdown for electricity generation - not quite equivalent since it is global and not just US - but again in the ballpark). So incidentally, would you believe manufacturing one PC accounts for about 600kg of CO2 emissions?

Which PC? The tower workstation I used to own certainly must have been quite different than the budget laptop I later bought. "Measure with a micrometer, mark with a pencil, cut with an axe." Ever hear that? If you're trying to do calculations, you should try to use the same kind of resolution all the way through. If you're using rough estimates for the amount of fuel consumed (e.g. 100 kg of undefined fossil fuels, which is a suspiciously round number) then it's unimportant whether that is 75% or 73% carbon. You're already operating in a realm where at best you're going to be within a factor of three or four of the true answer, so there's no point in trying to be more accurate than that.

100kg was just an example. The actual figure comes from a study in http://books.google.com/books?hl=en&lr=&id=t0WgI5w1clsC&oi=fnd&pg=PA41&dq=environmental+impact+in+the+production+of+personal+computers&ots=dkPCkaONjg&sig=iNhrkA7ZmyI1Zayu9iCVFgaZAqM. It's for a desktop PC. Unfortunately the figures are a little old -- for laptops they have certainly changed -- but they are the best ones I have been able to find, and believe me I have looked very hard for them. In any case, you will notice that my answer (600kg) has only one significant figure. I appreciate the advice Steven, but it's really not necessary.