How much carbon can we remove?

A new Nature study sets a “prudent planetary limit” for underground CO₂ storage in sedimentary basins at around 1,460 GtCO₂.The authors arrive at this number by layering conservative exclusions: ruling out deep and shallow storage, excluding far offshore and contested areas, applying buffers for seismicity, aquifers, protected zones, and urban areas. They also exclude in-situ mineralization in rock, which offers storage potential orders of magnitude larger than the article’s proposed limit.

Critics argue that the result is therefore very conservative. Deep storage is shown to work, and in-situ mineralization has begun scaling beyond pilots. The paper provides a risk-filtered lower bound for safe underground storage, not an upper limit for planetary storage.

The ~1400 Gt CO₂ number is still staggering, it could hold all historical oil emissions and the emissions from burning all proven oil reserves. Then we have other CDR methods, which are not constrained by geologic pore space, including biochar, enhanced rock weathering, and ocean alkalinity enhancement. There is a limit to how much can be removed per year by these methods, but not a cumulative limit. This means that CDR is not fundamentally “stock-limited”, removing carbon today with the methods listed above does not chip away at a static, limited capacity.

The authors of the Nature study raise a fair point that pore space used today, albeit in large supply, to abate ongoing fossil fuels won’t be available tomorrow for legacy removals. But scaling storage infrastructure now will help build capacity to address legacy emissions later. If we don’t develop it now, we won’t have the capability to use CDR and storage infrastructure to draw down historic emissions later, bringing temperatures down. 

Companies and countries must not simply assume that large-scale, cheap CDR will be there when needed. Costs could be higher, projects slower, and politics messier than expected. That doesn’t mean CDR or CCS shouldn’t be used. It means strategies should stay flexible and technology-neutral. Plan, yes, but invest in the plans. Build the infrastructure. What is not acceptable is to assume CDR will be cheap and abundant, and then do nothing to make it real.