Have a question?
Ask us!
Ask MIT Climate

When we remove carbon dioxide from the atmosphere, how long does it stay locked away?

In theory, we can remove CO2 from the air for thousands or even millions of years, which in terms of addressing climate change is effectively permanent. But carbon removal can also fail.

 

November 24, 2025

Since humanity started burning fossil fuels for energy, we have added roughly one trillion tons of climate-warming carbon dioxide (CO2) to our atmosphere. One could imagine a very literal way to deal with the climate change that has resulted: take the CO2 back out. For many reasons, though, this kind of “carbon removal” is a very tall order.

Among those reasons is that sucking out the CO2 is only the first step.

“You want to keep the CO2 out of the atmosphere for at least, let’s say, hundreds of years,” says Ruben Juanes, a geophysicist and MIT professor of civil and environmental engineering.1 Shorter-term storage, on the scale of decades, may usefully slow the pace of warming and buy time for more permanent solutions, but ultimately still leaves us to deal with the same CO2 in the midst of ongoing climate change.

Why might carbon removal fail to keep CO2 out of the air over centuries? It depends on how that CO2 is stored.

Today, the large majority of carbon removal relies on “natural climate solutions.”2 These use trees and other plants to draw CO2 out of the atmosphere, storing the carbon in their tissues and the soils below. If these living storage vessels are lost, says Evan Fricke, an ecologist and research scientist in the MIT Department of Civil and Environmental Engineering, the carbon is destined to return to the air.

“Some of the prominent failures of forest-based natural climate solutions include attempting to plant trees in ecosystems where they don't naturally occur,” he says. “And the trees simply die or burn up.”

So in the worst case, natural carbon removal may last only years or decades. But it can also be effectively permanent. Earth, after all, is home to forests that are tens of millions of years old.

Successful carbon removal, Fricke says, takes good planning and continued care. That can mean planting the right vegetation in the right ecosystems for long-term survival. It can be a matter of conservation, making sure a new or restored forest is not later cut down for timber or development. And it can also be quite active. Certain farming practices, for example, can help soils store extra carbon. But it is widely assumed that, to prevent that carbon from leaking back out, farmers must keep up these practices year after year.3

Still, Fricke says a well-managed project can handle some reversals. Even if wildfire razes a forest, a healthy ecosystem can regrow. The key is to pursue large-scale projects in diverse ecosystems, and prepare to manage them for the long term. Planners can also hedge against risks with “buffer pools” of land beyond what they need to meet their carbon storage goals.

“A leaf will fall and decompose,” says Fricke. “A tree might burn. But a forest landscape can be resilient for millennia. And so, as we implement these projects at scale, we should expect reversals in certain areas, but if we're getting our science right, the numbers for the entire portfolio should be quite predictable.”

Other approaches to carbon removal keep their carbon out of the elements. In “carbon capture and storage” (CCS), for example, CO2 is pumped underground. A nearly-pure stream of CO2 is pressurized for easier transportation, then injected into carefully chosen geological formations.4

This, says Juanes, is not so different from the way nature stores carbon underground, as oil and natural gas. “These hydrocarbons have been underground for millions of years, over geologic time,” he says. Pressurized CO2 is very similar, chemically and physically—so “in principle, it can be sequestered essentially indefinitely.”

The trick is in the execution. Hydrocarbons, Juanes points out, take millions of years to form, and the surrounding rock has ample time to react. CO2 injection applies pressure a lot faster. “So it is really crucial to understand how that overpressure can affect the safety or security of storage over the long term,” he says. The wrong choice of geology, or a too-aggressive injection, could let captured CO2 escape.

The world’s first CCS project began off the coast of Norway in 1996. Over the 30 years since, MIT Climate was unable to find any documentation of a CCS project losing a significant amount of CO2 to the atmosphere,5 but small leaks have been detected.6

There have also been several false starts and warning signs. “Very little of what happens underground happens exactly as we expect,” Juanes says, “from prospecting for minerals, to storage of CO2, to production of hydrocarbons.” CCS operators try to map the subsurface as well as possible before injection starts, but we cannot know everything about the rock in which CO2 will be stored.

As a result, Juanes says, “some projects have experienced unexpected behavior underground, and have had to either ramp down or shut down.” In 2011, for example, the In Salah project in Algeria halted early after researchers found signs that CO2 injection was cracking the rock and deforming the surface.7 As critics have noted, this is part of a pattern of projects underperforming.8

One could see this as a good sign for the permanence of CCS. Operators have spotted problems in advance, and adjusted or shut down without major leaks. But the challenges of dealing with pressurized CO2 underground add to the risks and costs. Are there more stable options?

A few types of carbon removal “mineralize” CO2 into limestone and other solids. This can be done by injecting CO2 into the right geologic formations, where it reacts with the underground rock. Or it can be done aboveground through “enhanced rock weathering,” speeding up chemical reactions between certain rocks and the CO2 in the air.

“In principle, if we can turn the CO2 into a solid, that's the most permanent form of storage,” says Juanes. This carbon is almost surely safe over tens of millions of years, until the slow forces of volcanism vent it back into the air.

But, he adds, all kinds of carbon removal can be effectively “permanent” if done right. As we choose which ones to pursue, we’ll have to weigh much more than their duration. Natural climate solutions are relatively cheap and ready to deploy. CCS can in principle handle large volumes of CO2 quickly. And all forms of carbon removal must compete with not emitting CO2 in the first place, by far the most “permanent” solution of all.

“Each of these has its trade-offs related to risks of reversals, the scale angle, the cost angle,” says Fricke. “Anyone who says that one of these is the silver bullet is probably lying to you. If there was a clear winner that we could implement at scale, without costs, we would have done it already.”

 

Thank you to Donald Yates of Perth, Australia, for the question.

Submit your own question to Ask MIT Climate

Get the latest from Ask MIT Climate monthly in your inbox

Read more Ask MIT Climate

 

Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license (CC BY-NC-SA 4.0).
Footnotes

1 Juanes notes that 1,000 years is sometimes used as a shorthand measure of how long an effective carbon removal technique might keep CO2 out of the atmosphere, as in the Intergovernmental Panel on Climate Change’s Special Report on Carbon Dioxide Capture and Storage (2005). There is nothing magical about that number, and even the IPCC special report does not adopt or suggest it as an official target, but it may provide a helpful rough guide to thinking about the timescales relevant to the human experience of climate change. In other contexts, like in many of the carbon offset markets that let participants pay for projects that help address climate change, 100 years has become a frequently used benchmark.

2 Smith, Stephen, et al. The State of Carbon Dioxide Removal 2024 - 2nd Edition. https://doi.org/10.17605/OSF.IO/F85QJ.

3 See, e.g., Paul, Carsten, et al. “Carbon farming: Are soil carbon certificates a suitable tool for climate change mitigation?” Journal of Environmental Management 330 (2023). https://doi.org/10.1016/j.jenvman.2022.117142. And, Oldfield, E.E., et al. “Agricultural soil carbon credits: Making sense of protocols for carbon sequestration and net greenhouse gas removals.” Environmental Defense Fund and Woodwell Climate Research Center (2021). The description above is a greatly simplified story about the complex and incompletely understood process by which soils absorb and release carbon, and researchers have recently argued that some forms of agricultural soil carbon sequestration may persist for decades or longer without maintenance (see Dynarski, Katherine, Deborah Bossio, and Kate Scow. “Dynamic stability of soil carbon: Reassessing the ‘permanence’ of soil carbon sequestration.” Frontiers in Environmental Science 12 (2020). https://doi.org/10.3389/fenvs.2020.514701) and that soil carbon sequestration may usefully contribute to addressing climate change even over short timescales (see Leifeld, Jens. “Carbon farming: Climate change mitigation via non-permanent carbon sinks.” Journal of Environmental Management 339 (2023). https://doi.org/10.1016/j.jenvman.2023.117893).

4  The CO2 could be drawn from the atmosphere, through “direct air capture” machines that chemically strip CO2 from the air, but more often it comes from a power or industrial plant. One strategy, “bioenergy with CCS” or BECCS, takes a roundabout approach to carbon removal: Plants are grown, absorbing CO2 from the air; those plants are burned for energy; and the CO2 released when they’re burned is captured from the power plant and injected underground.

5 With some important caveats. Pipelines carrying CO2 for injection have leaked. In addition, most CO2 injected underground today is used for “enhanced oil recovery,” flushing hard-to-reach oil out of the ground, with CO2 storage only a secondary consideration. This practice has caused occasional “blowouts” in which oil and CO2 burst from a well. Pipeline leaks and blowouts can be fixed before significant volumes of CO2 are lost, but their safety risks can be very serious: A 2011 blowout near Tinsley, Mississippi, killed wildlife, and a 2020 pipeline leak near Satartia, Mississippi, led to dozens of hospitalizations. It is also the case that more than one CCS project has experienced internal leaks, where CO2 did not escape to the atmosphere but did end up in unexpected parts of the subsurface. In 2024, for example, the company Archer-Daniels-Midland paused its CO2 injections in Illinois after a corroded well allowed CO2 and brine into an unintended part of the formation, prompting concerns (fortunately not borne out) that CO2 would enter the groundwater and contaminate it.

6 See, e.g., Yang, Manping, et al. “Identification of CO2 leakage in an active CO2-EOR field, Songliao Basin, China.” Energy & Fuels 38 (2024). https://doi.org/10.1021/acs.energyfuels.4c03612. Yang, Changbing, et al. “Soil gas dynamics monitoring at a CO2-EOR site for leakage detection.” Geomechanics and Geophysics for Geo-Energy and Geo-Resources 3 (2017). https://doi.org/10.1007/s40948-017-0053-7. Wells, Arthur, et al. “The use of tracers to assess leakage from the sequestration of CO2 in a depleted oil reservoir, New Mexico, USA.” Applied Geochemistry 22 (2007). https://doi.org/10.1016/j.apgeochem.2007.01.002. There is reason to believe that enhanced oil recovery, as in the papers cited above, is more vulnerable to these leaks than CCS done purely for carbon storage, as oil fields may have abandoned, imperfectly sealed wells through which CO2 can escape.

7 Ringrose, Philip, et al. “The In Salah CO2 Storage Project: Lessons learned and knowledge transfer.” Energy Procedia 37 (2013). https://doi.org/10.1016/j.egypro.2013.06.551.

8 See, e.g., Robertson, Bruce and Milad Mousavian. “The carbon capture crux: Lessons learned.” Institute for Energy Economics and Financial Analysis (2022).

Aaron Krol
by Aaron Krol, MIT Climate Portal Writing Team
Ruben Juanes
featuring guest expert Ruben Juanes, MIT professor of civil and environmental engineering
Evan Fricke
and guest expert Evan Fricke, research scientist, MIT Department of Civil and Environmental Engineering
Related MIT Groups
More Resources for Learning
MIT Climate Portal: "What is the risk that CO2 stored underground after carbon capture will escape again?"
World Economic Forum: "Durability, not permanence: Why we must act on natural climate solutions now"
The State of Carbon Dioxide Removal (Report)
Topics
Atmosphere
Carbon Capture
Carbon Removal
Forests

Want to learn more?

Listen to this episode of MIT's "Today I Learned: Climate" podcast on storing CO2 underground.

LHF: Hello, and welcome to Today I Learned: Climate, MIT’s climate change podcast. I’m Laur Hesse Fisher. And today, we’re talking about storing carbon dioxide underground. Which is something companies are doing right now, today, to the tune of tens of millions of tons a year.

Why? Well, if we put CO2 into the atmosphere—say, by burning coal, oil and gas—it heats up our planet. So people have come up with ways to capture this CO2 from the smokestacks and exhaust streams of coal, gas and industrial plants, so that it can’t escape into the atmosphere. There is even technology that can pull CO2 out of the air around us, something that folks call “direct air capture”. Whether pulling it from a smokestack or from the air, companies compress that CO2 into a fluid, pump it underground, and voila! It can’t contribute to climate change. 

If you’re interested in how these technologies work – and its benefits and challenges –  you can check out our two episodes: TIL about carbon capture, and TIL about removing CO2 from the atmosphere.

But now you might be wondering—what happens to this liquified CO2 under our feet? Is it dangerous? You might ask us, as Barbara Ann W. of North Carolina did: could pumping CO2 underground cause earthquakes or contaminate drinking water? Or you might, like Christopher B. of the United Kingdom, ask us: is there a risk that CO2 stored underground will escape?

Today, we’re answering these questions with help from Prof. Brad Hager. He’s a geophysicist and Associate Director of MIT’s Earth Resources Laboratory.

BH: We actually have a lot of experience with fluids under pressure underground. I mean, oil and natural gas themselves are trapped in the subsurface for millions of years, until someone comes along and drills a hole and lets them out. So because of the oil and gas industry, we know a lot about conditions under which fluids are trapped stably underground.

LHF: One way to think about carbon storage is that it’s returning carbon to where we got it. The carbon was part of oil and gas snug below the earth, we drilled it out and burned it to make energy, and now we’re collecting it and pumping it down there again as liquid CO2.

Still, you can’t just drill a hole anywhere you like and start pumping CO2 into it. That really could cause leaks and earthquakes.

BH: The big thing which determines whether injecting fluids causes earthquakes is basically the geology that you're injecting into. You do not want to inject into an area that has active faults. And you don't want to inject into an area that has brittle rocks.

LHF: If you inject any kind of fluid underground, it’s going to raise the pressure in the surrounding rocks. And if you’re injecting near a fault line, which are areas more susceptible to earthquakes, that pressure might actually slide the earth on top of it around. It’s kind of like turning on an air hockey table: you add some pressure coming up from below, which moves around the puck.

BH: If the local rocks are brittle, like, say, sandstone or granite, they’re also more likely to crack. And just like injecting near a fault line, that could trigger an earthquake, and it could let the carbon dioxide leak back out.

LHF: That hasn’t been documented at any CO2 storage sites, but wastewater has leaked when oil and gas companies pump this wastewater underground, so we know it’s a real risk.

What you want, ideally, is some sort of underground formation that has room to take in a bunch of fluid without moving or cracking.

Rocks like sandstone and limestone are porous. If you inject CO2 into these, it can seep into the pores of these stones and stay there, kinda like water seeping into sand. But you don’t want the whole formation to be porous.

BH: You also want a “caprock”: a hard layer on top that seals in the CO2. The caprock should be solid, but a little malleable. Shale is often a good candidate.

And finally, you want to inject the CO2 quite deep, at least 3,000 feet. That will keep it at a high pressure, so it remains a dense fluid and doesn’t turn back into a gas. It’s also deeper than the aquifers that we use for drinking water.

LHF: Put it all together, and that’s an awfully specific list of requirements. You might be wondering: how do we even find these places?

BH: It requires a lot of study. But with fairly standard techniques that the oil and gas industry uses all the time. You’d start with seismic reflection studies to characterize the structures.

LHF: That means that geologists make a vibration at the Earth’s surface, using something like an air gun or a piston that hits the ground really fast. That creates a seismic wave that travels underground and then reflects back up, and with special equipment we can “listen” for what kinds of rocks are underground.

BH: And if that looks promising you’d drill some test holes to get some ground truth on those seismic images, so you can actually sample the rocks that are there and understand things like their porosity, their permeability, how easy or difficult it is for fluids to flow through them.

And there are a lot of places in the world where this seismic exploration and drilling has already been done in the search for oil.

LHF: Yeah, it turns out that oil and gas tend to be found in the same kinds of malleable rocks that are good for storing CO2. Sometimes we can just turn around and use those same places for storage.

In fact, the most common way that companies store captured CO2 today is by pumping it into active oil wells to help flush more oil out. This is something called “enhanced oil recovery,” and to be clear, it’s not a long-term climate solution, because it’s used to help push out more oil that’s going to be burned and put more climate pollution into the air.

But geologists have also scouted out a lot of formations that could be used for CO2 storage without the oil production.

BH: For instance, the Gulf of Mexico is an area which is very conducive to carbon storage. But there's been a lot of extraction activity already in the Gulf, so one would have to be careful not to inject fluids near abandoned wells where the CO2 might leak back out. You know, you want to make sure that you're not in an area which had been turned into Swiss cheese by previous drilling operations.

LHF: Now, I mentioned at the beginning of this episode that carbon storage is already going on. So another question we can ask is—have we caused any leaks, or earthquakes?

BH: It depends. So there's a place in the North Sea called Sleipner, where CO2 injection has been going on for about 30 years, since the mid 1990s, at a rate of a million tons a year. And the layer that they're injecting into is so porous and so permeable, that they don't even have to pump the fluid in. It basically runs in under its own weight. It’s been a real success story.

But then there was a site in Algeria, for example, where they were injecting carbon dioxide, but the rocks were very tight. They had difficulty getting the CO2 in and there was evidence that they actually started to fracture the rock. The caprock was very thick, so the CO2 didn’t end up leaking, but they did have to halt storage.

LHF: So far, there haven’t been any CO2 storage disasters. That’s great news, and a sign that this is a plausible climate solution. But the failed project in Algeria does underscore how important it is to have responsible management of these storage sites, to monitor them and to make adjustments once they get going.

And, unfortunately, there are companies that have not always been careful enough about the environmental risks like these. For instance, we could look to something that humans pump underground today in much greater quantities than CO2: wastewater.

BH: In a lot of places where oil is produced, water is mixed in along with the oil. So the oil that's coming out is not pure. And sometimes, like in Oklahoma, they're getting 10 times as much water out as they're getting oil. So the usual way of disposing of this wastewater is to drill a hole in the ground and inject it back underground. Right now there are hundreds of billions of gallons of wastewater injected every year.

LHF: So if we’re worried about causing earthquakes when we pump carbon underground, a good first question to ask might be: are we causing earthquakes now, when we dispose of all this wastewater underground?

BH: Earthquakes can be a big problem, but they’re a tractable problem. Most wastewater is injected without causing any earthquakes at all. There are places like Saudi Arabia, for example, where a lot of wastewater is injected, and earthquakes are not resulting.

But then there are places like Oklahoma where injecting this wastewater has led to earthquakes. 

LHF: Yeah, Oklahoma saw a surge of earthquakes in the 2010s, alongside a boom in the local oil and gas industry.

BH: We knew that earthquakes were happening in Oklahoma before people began drilling for oil and gas and pumping wastewater back in. So it was known to be an area of seismic risk. And unfortunately, some companies in Oklahoma were not very careful about where they injected the wastewater.

LHF: These kinds of events are preventable, but they do have to be prevented. And that means, when there’s a proposed carbon storage project under evaluation, it’s not just the geology we have to ask questions about. We also need to ask the kinds of questions we would pose for any big infrastructure project that might impact the environment. Like, does the company have a good track record? What regulations are in place, and are they well enforced? Can we get a third party to evaluate the safety risks?

BH: It's basically a management question of carrying out the storage responsibly. And I want to be clear that there are risks. But there are risks to everything, and the risks for continuing to emit carbon dioxide into the atmosphere without taking it out far outweigh the risks of putting the CO2 underground.

LHF: That’s it for our episode today. 

Do you have a question about climate change? Maybe we answered it as part of our Ask MIT Climate series. You can find out at climate.mit.edu. And if we haven’t, ask us! Leave us a voicemail message at 617 253 3566 or visit https://climate.mit.edu/ask. We release answers as episodes here on TILclimate as well as on the website. 

I’ve got to say, we love hearing from our listeners. It totally lights up our day. We would love to hear from you, too. Let us know who you are, what you’re working on, what you’re wondering about, and why you listen to the show. Send us an email at climate@mit.edu.

TILclimate is the climate change podcast of the Massachusetts Institute of Technology. Aaron Krol is our Writer and Producer. David Lishansky is our Sound Editor and Producer. Michelle Harris is our fact-checker. Sylvia Scharf is our Climate Education Specialist. The music is by Blue Dot Sessions. And I’m your Host and Executive Producer, Laur Hesse Fisher. 

Thank you Prof. Brad Hager for speaking with us; to Barbara Ann and Christopher for your questions; to Lindsay Fendt, for original reporting used in this episode; and to you, our listeners. Keep up the curiosity.
 

Keep exploring

Check out these related Explainers, written by scientists and experts from MIT and beyond.

View All

Related Pieces