Steve Mohr of Australia's Newcastle University has modelled the earth's fossil fuel reserves and come up with this massive study (warning: 13mb). PROJECTION OF WORLD FOSSIL FUEL PRODUCTION WITH SUPPLY AND DEMAND INTERACTIONS http://ogma.newcastle.edu.au:8080/vital/access/services/Download/uon:6530/SOURCE4?view=true Plenty of gems in the study, like current energy consumption being the equivalent of every person on earth having 90 slaves. But headlines are that best guess for a production peak in fossil fuel is 2016–18 (at a total production rate of 509–525 EJ/y). For the fossil fuels individually: Coal - 2019 (212–214 EJ/y). Oil - 2011-12 (179–188 EJ/y). Natural gas - 2019-2062 (143-157 EJ/y). Unconventional oil and gas will probably create a shoulder to the production curves, which will slow the fall-off the other side of their peaks, but will not change the peak dates themselves. The most optimistic scenario for oil (p136 - case 2 dynamic) does see a second oil production peak circa 2110 that about matches the conventional oil peak. However this would have to be all shale oil (tar sands would be gone by then). It is a little surprising that coal has so little time left (as China and India seemed to be rather relying on coal) and that gas will be such a big player in the final wash-up. On the other hand, it is what Mohr's group have learnt about Chinese coal consumption which is accelerating that peak. On the optimistic side, Mohr notes, it sure puts a concrete limit on the carbon footprint issue .