World Energy Analysis for Stabilizing Atmospheric Carbon Dioxide

Current world “primary power” consumption is 12 terawatts (TW), 85% of it from fossil fuels. This is 16 times what it was a century ago, when world population was only a quarter as great. During that century fossil fuel combustion increased atmospheric carbon dioxide from 275 to 370 ppm (parts per million). Stabilizing it at 450 ppm “could be needed to forestall coral reef bleaching, thermohaline circulation shutdown, and sea level rise from disintegration of the West Antarctic Sheet.” This “could require Herculean effort” -- and 25 TW of emission-free power by midcentury. Stabilizing atmospheric carbon dioxide at 550 ppm, which requires 15 TW of emission-free power by midcentury, is regarded as “a major challenge.”

These are the assessments of Martin Hoffert and a consortium of 17 colleagues (Science, 298, 981-987 (1 Nov 02)), who note that technologies to produce these amounts of emission-free power don't exist today and go on to examine ways to achieve atmospheric carbon dioxide stabilization. All of them except decarbonization and geoengineering relate to energy sources. Their first approach is improved efficiency, another way to think of “conserving” energy in the sense of “using” less of it to accomplish a given task. After chronicling a list of efficiency improvements, including “ultra fuel-efficient cars” that travel 68 miles on a gallon of gas, the authors lament that SUVs have driven the US car/light truck fleet mileage to a 21-year low of 20 miles per gallon. Add to this that more than 80% of increased petroleum usage comes from Asia, and the need becomes apparent to pursue decarbonization strategies, at least as a bridge to developing sufficient nuclear and renewable energy.

The authors observe at the outset that “all renewables suffer from low areal power densities.” This is especially true of the photosynthesis which generates biomass, which weighs in at only 0.6 W/m2 and requires more than 10% of Earth's land surface to generate 10 TW of “primary power,” an area “comparable to all of human agriculture.” Photovoltaics and wind do better at 15 W(electric)/m2, but a square array 470 km by 470 km is required to generate 10 TW (3.3 TW(electric) equivalent), much more than the 3 km2 installed between 1982 and 1998. “A massive (but not insurmountable) scale-up is required to get to 10 to 30 TW equivalent.” This could be facilitated by installing photovoltaic arrays on solar power satellites: 660 of them the size of Manhattan beaming power to a 10 km x 13 km surface rectenna could generate 10 TW. Surface renewables, because of their intermittency, require storage systems. Hydrogen fuel cells could fill this need, especially if they become the means for energizing transportation. But Protein Exchange Membrane cells require platinum catalysts, and “a 10 TW hydrogen flow rate would require 30 times as much as today's annual world platinum production.” Moreover, a new power grid would be needed to handle electricity from decentralized renewable power sources.

In addition to problems of waste disposal and weapons proliferation plaguing nuclear fission, the authors observe that there is also not enough uranium. Current estimates and power reserves of uranium underground can generate only 60-300 TW-years of primary energy, and extracting all the uranium dissolved in water flowing into the oceans can provide only 2 TW/yr. Though the authors have hope for the future of nuclear fusion, they do not see it contributing to the stabilization of carbon dioxide emissions by 2050. The only realistic midcentury stabilization of carbon dioxide emissions which the authors see from nuclear sources is breeder reactors, which have now been dropped by France, Germany, and Japan as well as the US. They prefer breeding ²³³U from ²³²Th, because thorium is three times more abundant than uranium and ²³³U is more difficult to separate from uranium than ²³⁹Pu is. This breeding would be facilitated by neutrons from fission breeders which “could support perhaps 10 satellite burners [fission reactors] whereas a fission breeder supports perhaps one.”