How Carbon Capture Technology Works

Chart 1: CO2 concentration by source. Capturing from flue gas (4–33%) is orders of magnitude easier than from ambient air (0.042%). Source: N43 and Hermes, compiled from IEA and IPCC data.

01 The Carbon Problem

The planet has a stock-and-flow problem. Since the beginning of the industrial era, human activity has transferred roughly 1.7 trillion tons of carbon from underground fossil reservoirs into the atmosphere, primarily as carbon dioxide. The concentration of CO2 in the atmosphere has risen from about 280 parts per million in 1750 to over 420 parts per million today — a level not seen in at least three million years. The flow continues: global annual CO2 emissions from fossil fuels and industry are around 37 billion tons, and despite two decades of climate policy, that number has not yet peaked. The stock drives warming; the flow determines how fast the stock grows.

Carbon capture technology addresses both. Point-source capture — removing CO2 from the exhaust gases of power plants, cement factories, and steel mills — reduces the flow. Direct air capture — pulling CO2 from ambient air — addresses the stock. Both are thermodynamically costly, because CO2 is a stable, low-energy molecule. Separating it from a gas mixture requires energy proportional to how dilute the CO2 is, which is why capturing from a concentrated flue gas is always cheaper than capturing from air. The entire engineering of carbon capture is a search for chemical and physical processes that minimize this thermodynamic penalty.

02 Post-Combustion Capture: The Amine Process

The dominant industrial method for capturing CO2 from flue gas is chemical absorption using aqueous amine solutions, typically monoethanolamine (MEA). The process is conceptually simple: flue gas is passed through an absorber column where it contacts the amine solution. The amine reacts with CO2 at temperatures around 40–60°C, forming a carbamate compound. The CO2-rich solution is then pumped to a stripper column where it is heated to about 100–140°C, which reverses the reaction and releases concentrated CO2 gas. The regenerated amine solution is recycled back to the absorber. The concentrated CO2 is compressed — typically to 100–150 bar — for transport and storage.

The amine process is mature — it has been used in the gas-processing industry for decades to remove CO2 from natural gas streams before pipeline transport. What is newer is applying it to the massive, lower-pressure, variable-composition flue gases of coal and gas power plants. The primary energy penalty is the heat required for regeneration: the stripper reboiler typically consumes 20–30% of the plant's steam output, which is energy that would otherwise generate electricity. This "energy parasite" effect means a coal plant with post-combustion capture produces roughly 25–35% less net electricity per ton of coal burned. Alternative solvents — piperazine-promoted amines, amino acid salts, water-lean solvents — and alternative processes — membrane separation, cryogenic separation, calcium looping — are all under active development to reduce this penalty.

03 Direct Air Capture

Direct air capture (DAC) is the thermodynamically punishing end of carbon removal. Ambient air contains CO2 at about 420 parts per million — 0.042% — which is roughly 300 times more dilute than flue gas from a coal plant. The second law of thermodynamics sets a minimum energy requirement for separating CO2 from air, and the minimum work is about 0.5 gigajoules per ton of CO2. In practice, DAC plants consume 8–12 gigajoules per ton, mostly as heat for sorbent regeneration. This is the irreducible cost of moving a gas from a very dilute stream to a concentrated one.

Two engineering approaches have reached commercial-scale demonstration. The first, developed by Climeworks, uses solid sorbent filters that adsorb CO2 from air drawn through large fans. The filters are then heated to 80–100°C to release concentrated CO2 — a temperature low enough to be supplied by waste heat or low-grade geothermal energy. Climeworks operates the Orca plant in Iceland (4,000 tons/year capacity, commissioned 2021) and the larger Mammoth plant (36,000 tons/year, commissioned 2024), both powered by the Hellisheidi geothermal field. The CO2 is mineralized in basalt formations through the Carbfix process, where dissolved CO2 injected into reactive basalt forms carbonate minerals within years — effectively turning CO2 into rock.

The second approach, developed by Carbon Engineering (now owned by Occidental Petroleum's subsidiary 1PointFive), uses a liquid potassium hydroxide solution to absorb CO2 from air. The resulting potassium carbonate is processed through a calcium-looping cycle that produces a stream of concentrated CO2 and regenerates the absorbent. The process operates at higher temperatures — the calciner reaches about 900°C — and is designed for large-scale deployment. Occidental's first commercial DAC plant in Ector County, Texas, designed for 500,000 tons/year, is the largest DAC facility yet built. The video embedded above covers this facility and its backers.

Chart 2: Global CCS capacity by project stage, 2010–2030 (projected). The gap between operational and planned capacity illustrates the scale-up challenge. Source: N43 and Hermes, compiled from Global CCS Institute data.

04 Storage: Geology and Mineralization

Captured CO2 must go somewhere. The two principal destinations are geological storage and mineralization. Geological storage involves injecting supercritical CO2 — compressed to a density similar to liquid, at pressures above 73.8 bar — into deep porous rock formations sealed by impermeable caprock. The most suitable formations are depleted oil and gas reservoirs (which held hydrocarbons for millions of years, demonstrating seal integrity), deep saline aquifers in sandstone, and unmineable coal seams. The Sleipner project in the North Sea, operated by Equinor since 1996, has injected over 20 million tons of CO2 into the Utsira sandstone formation at a depth of 1,000 meters. It is the longest-running commercial CCS project in the world and the proof that large-scale geological storage is technically feasible.

Mineralization is a different chemistry. The Carbfix process, developed in Iceland, dissolves CO2 in water and injects it into basalt formations. Basalt is rich in calcium and magnesium silicates, which react with dissolved CO2 to form stable carbonate minerals — calcite, magnesite, dolomite — within the rock pore space. The reaction is relatively fast: tracer studies at the Hellisheidi pilot showed that over 95% of injected CO2 was mineralized within two years. Mineralized carbon is the most permanent form of storage — it is literally rock, not a gas held in place by pressure and caprock. The limitation is that the process requires large volumes of water (about 25 tons of water per ton of CO2) and reactive basalt formations, which are abundant in Iceland but less common in the sedimentary basins where most point-source emissions occur.

05 Enhanced Oil Recovery: The Uncomfortable Partner

Approximately 80% of the CO2 captured globally each year is used for enhanced oil recovery (EOR), not for climate-motivated storage. In EOR, CO2 is injected into depleted oil reservoirs to mobilize residual oil that cannot be produced by primary or waterflooding methods. The CO2 mixes with the trapped oil, reduces its viscosity, and displaces it toward production wells. Some of the injected CO2 remains trapped in the reservoir; some returns with the produced oil and is re-injected. The net storage is real but the purpose is oil production, and the climate benefit is ambiguous: a ton of CO2 stored through EOR may enable the production of two to three barrels of oil that would otherwise stay in the ground.

This entanglement between CCS and the oil industry is not incidental. The companies with the subsurface engineering expertise, the pipeline infrastructure, and the regulatory experience for CO2 injection are predominantly oil and gas operators. Occidental Petroleum's acquisition of Carbon Engineering was motivated in part by the potential to use DAC-generated CO2 for EOR, producing what the company markets as "net-zero oil." Whether this represents genuine climate mitigation or an elaborate form of carbon accounting is a subject of ongoing debate among policy analysts and climate scientists. The technology is the same; the framing and the accounting determine the climate outcome.

06 The Cost Frontier

The cost of carbon capture is the variable that will determine its role in climate strategy. Post-combustion capture at a coal power plant currently costs approximately $40–60 per ton of CO2, including capture, transport, and storage. For natural gas plants, the cost is higher — roughly $50–70 per ton — because the flue gas is more dilute. Direct air capture is far more expensive: current costs are estimated at $600–1,000 per ton for first-generation plants, with a projected trajectory toward $100–200 per ton as the technology scales and engineering improves. These numbers are the reason CCS deployment remains modest relative to the scale of the climate problem.

The economic case for CCS depends entirely on a carbon price or equivalent policy mechanism. Without a price on CO2 emissions — whether through a carbon tax, an emissions trading system, or a subsidy for removal — there is no financial incentive for an industrial facility to capture and store CO2. The European Union's Emissions Trading System, with carbon prices that reached over €100/ton in 2023, has begun to make CCS economically viable for certain industrial applications. The United States' 45Q tax credit, expanded under the Inflation Reduction Act of 2022, provides up to $85 per ton for point-source capture and storage, and up to $180 per ton for DAC with storage. These are the most generous CCS subsidies in the world, and they have triggered a wave of project announcements. Whether the announced projects will be built, on time and on budget, is the question that will determine whether CCS becomes a meaningful climate technology or remains a perennially emerging one.

07 Limits, Critiques, and the Path Forward

Carbon capture faces criticism from two directions. From one side, it is accused of being too slow and too expensive to matter at climate-relevant scale. The 40 million tons per year of global CCS operational capacity in 2024 represents about 0.1% of annual emissions. Even the most aggressive deployment scenarios project that CCS could handle 10–15% of mitigation by 2050, with the remainder coming from electrification, renewables, efficiency, and demand reduction. From the other side, it is accused of being a moral hazard — a technology that extends the life of fossil fuel infrastructure by making it appear compatible with climate goals, thereby delaying the transition away from fossil fuels entirely.

Both critiques have merit, and both can be true simultaneously. The steel, cement, and chemical industries produce process emissions that cannot be eliminated by electrification alone — the CO2 comes from the chemistry of the process, not just from burning fuel. For these hard-to-abate sectors, CCS may be the only route to deep decarbonation available within the timeframe the climate requires. For power generation and transport, the case for CCS is weaker, because alternatives exist. The pragmatic view is that carbon capture is a targeted tool for specific emissions sources where no alternative is available, not a general-purpose substitute for emissions reduction. The scale of deployment will depend on cost reduction, policy support, and public confidence in the integrity of geological storage. The technology works. The question is whether the economics and the politics will follow.