To the Editor

The use of air capture of carbon dioxide as a method for mitigating climate change is gaining traction in the debate surrounding geoengineering1,2,3. The potential of air capture is generally assessed in terms of monetary cost, estimated to be about a few hundred dollars per ton of carbon1,4, leading to the suggestion that the costs of air capture are equal to, or even more favourable than, the costs of conventional mitigation efforts4.

However, the ultimate physical constraint on air capture is the energy required to power the systems. At present it takes 30–60 GJ to chemically remove one ton of carbon from the atmosphere5,6,7, comparable to the 40–70 GJ of energy generated during the production of one ton of emissions. Thus, using today's technology, it takes at least ∼0.5 J of energy to capture the emissions generated in producing 1 J of fossil fuel energy. If the energy for capturing carbon comes from fossil fuels, then at least a third of society's fossil fuel energy would have to be diverted to air capture to eliminate all emissions. Alternatively, it would require an increase in the total production of fossil fuel energy of at least a third.

Neither option is impossible, but neither is definitively attainable, and both would be expensive and painful for the public. And by no means is either option obviously easier or cheaper than conventional mitigation. Technological breakthroughs may decrease the amount of energy required for air capture, but even capture energies of 20 and 10 GJ per ton of carbon require significant diversion of energy to air capture.

Renewable energy could be an option. Given typical estimates of mid-twenty-first-century emissions of 15 billion tons of carbon per year, more than 14 TW of renewable energy would be required to capture these emissions. This is comparable to total world energy consumption today. Although it may be possible to build this much renewable energy, would it make any sense to use it for air capture? It may be more prudent to simply replace the fossil fuel energy with renewable energy, which would lead to the about the same reduction in emissions.

There are clear benefits to air capture, and further research into this approach is clearly warranted. But at present the enormous amount of energy required for air capture represents a fundamental constraint to this approach, and it may preclude the possibility of using it to offset the majority of human emissions. Future analyses of air capture must explicitly take energy constraints into account.