How Much Carbon Do We Need to Capture?

Plans to end global warming hinge on driving net greenhouse gas emissions to zero (plus or minus a few gigatonnes). It's not going well. CO₂ emissions hit an all-time high last year, and for the first time average temperatures on Earth rose 1.5 C above preindustrial levels. To limit warming to 2 C, massive amounts of carbon dioxide will have to be sucked out of the atmosphere and locked away, according to the Intergovernmental Panel on Climate Change (IPCC).

There are old and new ways to do this. The old methods-growing more and bigger trees in temperate and tropical forests, stuffing more carbon into soils-can be cheap, but they have limits. Forests burn, die from disease, or get cut down, releasing some of the carbon they store. Microorganisms eventually break down much of what's in the soil. Both are hard to audit and constrained by available land. Another option-pulverized minerals spread on fields-can solidify airborne carbon. But like trees and soils, these approaches require a lot of land to sink a tonne of carbon.

Enter the machines: Several companies are now deploying high-powered fans or pumps that chemically isolate CO₂ from air or seawater and then pipe it to systems that inject it underground.

But direct air capture (DAC) systems consume a lot of energy and reagents that currently produce toxic by-products. To make a significant dent in global warming, all known removal methods-both conventional and novel-will probably have to scale up until their unwanted consequences limit further expansion.

What would it take to scale DAC to many billions of tonnes a year? Let's take a look.

Where to Bury the CO2

Geologists have identified ample reservoirs that could hold many trillions of tonnes of injected CO₂ underground for centuries. Around 51 megatonnes of CO₂ are already stored each year, and announced plans would scale that up sevenfold over the next decade. DAC operations contribute a minuscule amount to that total; essentially all carbon capture and storage operations today inject CO₂ recovered from fossil-fuel production or use, not from the atmosphere. But the same kind of injection infrastructure could be used for DAC as it ramps up and fossil-fuel use declines.

A Slow Start to a Monumental Task

To hold warming below 2 C, a total of 525 to 755 gigatonnes of carbon dioxide will need to be removed from the atmosphere by 2100, according to a 2018 review study by an international research group led by the Mercator Research Institute on Global Commons and Climate Change. That total will increase if global emissions do not start falling quickly.

From 2019 to 2023, just 9 Gt of CO₂ was removed, 99.9 percent through managed forestry.

All novel methods, such as DAC, biochar, and enhanced rock weathering, remove less than 2 million tonnes a year, according to the 2024 State of Carbon Dioxide Removal study. This accounted for only about 0.004 Gt of CO₂ from 2019 to 2023. Of that amount, just 0.00001 Gt was removed by DAC.

Scaling to Gigatonnes a Year

IPCC scenarios that aim to limit warming to 2 C depend on DAC and other novel removal methods ramping up fast in the decades ahead to reach 6 to 12 Gt of CO₂ each year by century's end-enough to offset residual greenhouse emissions from industry and agriculture. That's on top of 2 to 5 Gt removed annually by soils and forests. CO₂ capture and injection systems are so energy intensive that some DAC facilities have reportedly struggled to offset their own carbon footprints. Researchers in Saudi Arabia recently estimated that the world would have to make an additional 4.4 terawatts of carbon-free electricity and heat to sustain 10 Gt of CO₂ per year of DAC-significantly more than all clean energy consumed in 2024.