Injecting CO₂ underground? That sounds like sweeping a problem under the rug.
It’s a bit more complicated than that. They aren’t just looking for some sort of empty cavern to dump a giant cloud of gas into. They want microscopic cracks in rocks and impermeable layers that can prevent the CO₂ from migrating back up.
This is called carbon capture and storage (CCS). A piece of equipment is used to separate CO₂ from the exhaust of an industrial source. The captured CO₂ is compressed and transported to its location. It is injected into a deep geological formation via a well at a suitable location.
The U.S. Department of Energy calls storage the last step in this process. Though it might seem straightforward, it’s the geology that determines whether or not a given project will be able to store enough CO₂ and prevent it from escaping.
One site for storage could be a deep saline formation, a porous rock filled with water that’s way too salty to drink. A good reservoir might be sandstone. The spaces between grains can act as a reservoir for fluids. But these spaces also need to be well connected so that CO₂ can travel away from the injection well.
A layer of caprock, a much less permeable rock, is needed above the reservoir. Think of this as a lid. Shale can be a good seal since it has very small pores that inhibit fluid flow.
Old oil or gas reservoirs that are now depleted could also be used. Their geology has effectively trapped hydrocarbons for millions of years. This is a sign that the trap could work. Even if there’s an old poorly sealed well, it could be a conduit in what might otherwise be very impermeable rock.
Deep underground, CO₂ behaves differently. The high pressure and temperature deep underground can put CO₂ into a supercritical state. CO₂ in this state has the density of a liquid, but can flow through pore spaces. As explained by the National Energy Technology Laboratory, this means a lot more CO₂ can fit underground compared to the same amount of CO₂ at the surface.
Initially a lot of CO₂ will rise up through the reservoir. It’s more buoyant than the salty water around it. But the caprock will prevent it from rising up. This is called structural trapping.
The CO₂ as it moves through the reservoir can have small amounts stranded in between grains. Since the droplets are surrounded by water they’re not easily able to move anymore. This is called residual trapping.
Some of the CO₂ can be dissolved into the salty water. Once dissolved, it’s not as likely to rise up. Longer term, dissolved carbon can react with minerals in the rock to form a solid carbonate mineral.
That last part links climate tech to the natural carbon cycle. CO₂ dissolved in rainwater makes a weak acid. As water weathers rocks, it releases dissolved components, which can eventually form carbonates. In mineral storage, these reactions are used underground in a controlled way.
A material that might be particularly useful is basalt. This dark volcanic rock has minerals with a high concentration of calcium or magnesium. These can react with the dissolved CO₂ to help form stable carbonates.
Researchers in Iceland, the CarbFix project, injected CO₂ dissolved in water into basalt. A 2016 paper in Science found that over 95% of the injected CO₂ was able to mineralize in less than two years. This surprised many scientists since it’s often thought that mineral trapping would take much longer.
Basalt isn’t perfect for everywhere, though. Dissolving CO₂ needs water. There also needs to be a reasonable rate that the rock will accept fluids. Plus, you’ve got to worry about transporting CO₂ over long distances. So it’s possible that the geology looks promising, but the project is impractical.
There’s a lot of theoretical storage available. The USGS national assessment found about 3,000 metric gigatons of technically accessible mean storage capacity in onshore U.S. and state waters. But each reservoir will have its own pressure limitations and fault history.
There are risks. Injection will increase the pressure underground. It can push salty water away from the well. Unexpected fractures could alter the movement of the CO₂. All these risks need to be assessed before injection takes place.
Hence, monitoring is an essential part of the storage process and is not optional. Environmental Protection Agency’s Class VI program includes characterization and carbon dioxide plume modeling requirements for a project. Tracking of movement and protection of drinking water sources is necessary.
The technology is already moving from theory to practice. In April 2026, EPA issued the first Class VI permit in Kansas. The permit allows for injection of carbon dioxide from ethanol production into the Arbuckle formation, over 3,400 feet deep. And the permit mandates monitoring for 50 years after injection stops.
Of course, CCS is no substitute for cutting emissions. It takes energy and costs money to capture the carbon. But some industrial emissions are hard to get rid of because the CO₂ isn’t just from fuel combustion. It’s a product of the chemistry of the process. A big one is cement manufacturing.
According to the International Energy Agency, the amount of carbon capture in operation and construction increased according to their latest 2026 project update. So did storage capacity. But there are still many planned projects facing delays.
The simplest answer is that geology can help address climate change. But only if the geology is good and the project is well monitored. We need a porous reservoir to take the carbon dioxide. We need a reliable caprock to hold the carbon dioxide, and maybe some of that carbon gets chemically sequestered in stone over time.
Rock can appear still and eternal from the surface. But it can be dynamic and reactive deep down when it’s full of fluids. So the subsurface is not just down there under our feet. It might be a part of dealing with the carbon that’s already in the air above our heads.




