Carbon dioxide removal actively pulls CO2 from the atmosphere and stores it. The IPCC says we need 6-10 billion tonnes per year by 2050. Current engineered capacity is under 0.01% of that. Here is every major method, what it costs, and who is buying.
Carbon dioxide removal (CDR) is any process that actively extracts CO2 from the atmosphere and stores it in a durable form. Unlike emission reduction, which prevents new CO2 from entering the atmosphere, removal reverses existing concentrations. It is the difference between turning off a tap and draining the bathtub.
The atmosphere currently contains approximately 422 parts per million (ppm) of CO2, up from 280 ppm before industrialization. Every method of limiting warming to 1.5°C modeled by the IPCC requires both drastic emission reductions and large-scale carbon removal. Reductions slow the inflow. Removal addresses the stock already accumulated.
CDR methods range from biological (planting trees, restoring wetlands, building soil carbon) to engineered (machines that filter CO2 from ambient air, minerals that chemically bind atmospheric carbon, biomass that is pyrolyzed into stable char). The methods differ enormously in cost, permanence, scalability, and co-benefits. Understanding those differences is essential for evaluating which pathways can deliver gigatonne-scale removal.
The terminology matters. "Carbon capture" typically refers to point-source capture at industrial facilities. "Carbon removal" or "CDR" refers to atmospheric extraction. Carbon credits for removal represent a verifiable tonne of CO2 pulled from the air and stored, while avoidance credits represent emissions that were prevented. The market increasingly treats these as fundamentally different products, and prices reflect the distinction: removal credits cost 10-100x more than avoidance credits because the climate service they deliver is more direct and more measurable.
Global CO2 emissions reached approximately 37.4 billion tonnes (37.4 GtCO2) in 2024, according to the Global Carbon Project. Even in aggressive decarbonization scenarios, residual emissions from sectors like agriculture, cement, long-haul aviation, and industrial heat persist for decades. These are emissions that current technology cannot economically eliminate.
The math is direct. To limit warming to 1.5°C, the remaining carbon budget from 2024 is approximately 275 GtCO2 (IPCC AR6, for a 50% probability). At current emission rates, that budget is exhausted before 2032. Even if emissions fall steeply, the accumulated atmospheric CO2 continues to drive warming. CDR is the only mechanism that reduces the atmospheric stock.
The IPCC's Sixth Assessment Report analyzed hundreds of climate pathways. Every pathway that limits warming to 1.5°C with limited or no overshoot includes CDR at scale. The median requirement: 6 gigatonnes of CO2 removal per year by 2050, rising to 10+ GtCO2/year by 2100. Some pathways model even higher requirements if emission reductions are slower than projected.
This is not a niche concern. CDR is now an integral component of climate strategy at the level of the IPCC, national governments, and major corporations. The question is no longer whether removal is needed, but which methods can scale, at what cost, and how fast.
Carbon removal methods split into two broad categories: nature-based approaches that harness biological processes, and engineered approaches that use technology and chemistry. A third hybrid category combines both. Each method has a distinct profile of cost, permanence, scalability, and co-benefits.
Afforestation and reforestation. Growing new forests or restoring degraded forests. Trees absorb CO2 through photosynthesis and store it in biomass and soil. Cost: $10-50/tCO2. Permanence: decades to centuries, contingent on the trees surviving. Reversal risk is significant: wildfires, drought, pest outbreaks, or future land-use decisions can release stored carbon. Global potential: 0.5-3.6 GtCO2/year, but land availability competes with food production and biodiversity.
Soil carbon sequestration. Practices like regenerative agriculture, cover cropping, and no-till farming increase the organic carbon stored in soil. Cost: $10-50/tCO2. Permanence: highly variable (years to decades), dependent on continued management practices. If a farmer reverts to conventional tillage, stored carbon can be re-released. Global potential: 0.5-5 GtCO2/year (estimates vary widely). Co-benefits include improved soil health, water retention, and crop resilience.
Coastal and marine ecosystems. Mangroves, seagrass beds, and salt marshes (blue carbon ecosystems) sequester carbon at rates 2-4 times higher per hectare than terrestrial forests. Permanence depends on ecosystem protection. Restoration cost: $10-100/tCO2. Total potential is limited by available coastline but the per-hectare intensity is high.
Direct air capture (DAC). Machines use chemical sorbents or solvents to filter CO2 from ambient air, then compress and inject it into geological formations for permanent storage. Climeworks operates the largest DAC facility (Mammoth, Iceland) at approximately 36,000 tCO2/year capacity. Cost: $400-600+/tCO2 at current scale, with projections of $100-200/tCO2 at scale by 2040-2050. Permanence: 10,000+ years in geological storage. Energy-intensive: requires 1,500-2,500 kWh per tonne of CO2 captured, which must come from clean sources to achieve net removal.
Biochar. Biomass waste (agricultural residues, forestry waste, food waste) is heated in the absence of oxygen (pyrolysis), converting it into a stable carbon-rich solid. At production temperatures above 550°C, the resulting biochar is stable for 500+ years. Cost: $131-164/tCO2e via Puro.earth-certified credits. Co-benefits: soil amendment (improved water retention, nutrient availability, microbial habitat), waste valorization, and potential energy co-product (syngas). Biochar is currently the largest engineered CDR pathway by volume of credits issued.
Enhanced rock weathering (ERW). Crushed silicate minerals (typically basalt) spread on agricultural land react with CO2 and rainwater, converting atmospheric carbon into stable bicarbonates that eventually wash into the ocean and are stored for 100,000+ years. Cost: $80-200/tCO2 (grinding and transport dominate costs). Co-benefits: soil de-acidification, nutrient release (calcium, magnesium, potassium). Companies like UNDO, Lithos, and Eion are deploying at pilot to commercial scale across thousands of hectares.
Bioenergy with carbon capture and storage (BECCS). Biomass is burned for energy, and the CO2 released during combustion is captured and stored geologically. The net effect: atmospheric CO2 was absorbed by the biomass during growth, then permanently stored after combustion. Cost: $100-200/tCO2. Concerns: large land and water requirements for biomass cultivation, competition with food production, and supply chain emissions. BECCS was heavily featured in early IPCC scenarios but faces growing skepticism about scalability.
Ocean alkalinity enhancement (OAE). Adding alkaline minerals to the ocean increases its capacity to absorb CO2 from the atmosphere. The ocean already absorbs approximately 25% of annual CO2 emissions. Enhancing alkalinity accelerates this natural process. Cost: $50-150/tCO2 (estimated). Still in early pilot stages. Concerns: potential impacts on marine ecosystems, measurement and verification challenges. Companies like Planetary Technologies and Vesta are conducting field trials.
Not all stored carbon stays stored. Permanence is the duration for which removed CO2 remains out of the atmosphere. It is the single most important variable distinguishing CDR methods, and the primary reason that removal credit prices span a 60x range.
A tree planted today absorbs CO2 as it grows. If that tree burns in a wildfire in 30 years, the stored carbon returns to the atmosphere. The net removal was temporary. A tonne of biochar produced at 600°C resists decomposition for over 500 years. CO2 injected into basalt formations (as in Climeworks' Carbfix process in Iceland) mineralizes into carbonate rock within 2 years and remains stable for geological timescales: tens of thousands of years.
Permanence determines whether a CDR method is a climate solution or a climate loan. Short-permanence removal buys time. Long-permanence removal solves the problem. Both have value, but they are not interchangeable, and the market is learning to price the difference.
The carbon credit market reflects this hierarchy. Puro.earth, the leading removal credit registry, requires projects to demonstrate permanence of at least 100 years for certification. Biochar credits certified under Verra's VM0044 methodology must demonstrate production temperatures above 550°C. Climeworks sells removal certificates with permanence guarantees backed by the geological science of basalt mineralization.
Nature-based removal (forests, soil, wetlands) operates on shorter timescales but delivers co-benefits that engineered methods do not: biodiversity, water cycling, food production, and ecosystem resilience. The strongest CDR strategies deploy both: nature-based removal for near-term benefit and ecosystem value, and engineered removal for durable, verifiable atmospheric CO2 reduction. This is not a competition. It is a portfolio.
CDR costs span a wide range. The cheapest methods are the least permanent. The most permanent methods are the most expensive. This is not a coincidence: durably locking carbon out of the atmospheric cycle requires either geological processes (slow and capital-intensive to accelerate) or high-temperature chemistry (energy-intensive). Nature does the work cheaply but on her own timeline and with no guarantees.
Cost curves are falling. DAC has declined from $600-1,000/tCO2 in 2020 to $400-600+ in 2025, driven by Climeworks' and Carbon Engineering's scaling efforts, improved sorbent chemistry, and the US Inflation Reduction Act's 45Q tax credit ($180/tCO2 for DAC with geological storage). The Department of Energy's Carbon Negative Shot targets $100/tCO2 for DAC by 2032.
Biochar costs have compressed as the supply chain matures. Continuous pyrolysis systems from companies like Pacific Biochar and Carbon Gold achieve higher throughput and lower per-tonne costs than batch systems. Feedstock flexibility (agricultural waste, food waste, invasive species removal) keeps input costs low. The credit revenue ($131-164/tCO2e) combined with the product value of the biochar itself (sold as a soil amendment for $200-600/tonne) creates a dual revenue stream that makes many biochar operations commercially viable without subsidy.
Enhanced rock weathering costs are dominated by grinding and transport. Basalt is abundant and cheap, but crushing it to particle sizes that react on agricultural timescales requires energy. UNDO, the UK-based ERW company, reported costs below $100/tCO2 at scale in 2024. The co-benefit of replacing agricultural lime (which farmers already buy) creates a partial cost offset.
The economic trajectory mirrors early-stage solar and batteries: high initial costs, learning rates of 10-20% per doubling of cumulative deployment, and a policy environment (45Q, EU Innovation Fund, Frontier commitments) that underwrites the first commercial scale-ups.
The carbon removal market is driven by a relatively small number of high-conviction corporate buyers. These are not greenwashing purchases. They are advance procurement commitments designed to build supply at scale, published transparently, and often at prices well above the market because the buyers recognize that current prices reflect a nascent industry.
Microsoft committed to becoming carbon negative by 2030. Its 2024 removal portfolio purchased removal credits averaging over $120/tCO2e across biochar, DAC, enhanced weathering, and ocean-based methods. Microsoft publishes its full portfolio data annually. In 2024, Microsoft procured approximately 1.4 million tonnes of carbon removal, making it the largest single buyer.
Stripe Frontier is an advance market commitment (AMC) that pools funds from Stripe, Alphabet, Meta, Shopify, McKinsey, and others to purchase carbon removal at any price, aiming to build the market from the demand side. Total commitments exceed $1 billion. Frontier has contracted over 250,000 tonnes of removal across 30+ companies. Purchase prices range from $50/tCO2 (enhanced weathering at scale) to $600+/tCO2 (early DAC).
Shopify Sustainability Fund allocates a portion of revenue to climate action, including removal purchases. Shopify has been an early buyer of biochar credits, enhanced weathering, and ocean CDR pilots.
The US government, through the Department of Energy's Regional DAC Hubs program, allocated $3.5 billion to build four large-scale DAC facilities. The first two hubs (Project Cypress in Louisiana and South Texas DAC Hub) are in development. The 45Q tax credit provides $180/tCO2 for DAC with geological storage, the most generous CDR subsidy globally.
CDR.fyi tracks all disclosed removal purchases. As of early 2026, total disclosed carbon removal purchases exceed 4 million tonnes, with market value approaching $1 billion. Volumes grew 166% year-over-year in 2024. The buyer pool is small but growing, and their willingness to pay premium prices is funding the infrastructure buildout that will eventually bring costs down for everyone.
The gap between current CDR capacity and what climate scenarios require is the defining challenge of the carbon removal industry. It is not a gap measured in percentages. It is measured in orders of magnitude.
Current engineered CDR capacity is approximately 2 million tonnes (0.002 Gt) per year. The IPCC's median 1.5°C pathway requires 6-10 Gt per year by 2050. That is a 3,000-5,000x scale-up in 25 years. For context, solar PV achieved a roughly 5,000x scale-up in generation capacity between 2000 and 2025. The precedent exists, but it required sustained policy support, massive capital deployment, and continuous cost reduction over two decades.
Nature-based removal adds significant volume. The State of Carbon Dioxide Removal 2024 report estimates that land-based biological CDR (forests, soil, wetlands) currently removes approximately 2 GtCO2/year from the atmosphere. But much of this is difficult to attribute to deliberate CDR activity versus natural ecosystem function, and the permanence limitations mean it cannot substitute for durable engineered removal in long-term climate accounting.
The scale gap is closing faster than most observers expected. Biochar production is doubling roughly every 18 months. DAC capacity increased 10x between 2022 and 2025. Enhanced weathering deployments expanded from pilot plots to thousands of hectares. But "faster than expected" from a near-zero base is still far from sufficient. The next decade determines whether CDR becomes a meaningful climate tool or remains a rounding error.
The path from 0.002 Gt to 6 Gt is not a single technology's job. It requires a portfolio: biochar scaling through distributed pyrolysis, DAC scaling through industrial standardization, ERW scaling through agricultural deployment, ocean methods proving out at pilot scale, and policy frameworks (carbon prices, tax credits, procurement mandates) that make the economics work at each stage of the curve. No single method can close the gap alone. All of them together, deployed in parallel, are the thesis.
Carbon removal, also called carbon dioxide removal (CDR), is any process that actively extracts CO2 from the atmosphere and stores it in a durable form. Unlike emission reduction (which prevents new CO2 from entering the atmosphere), removal reverses existing atmospheric concentrations. Methods include biochar production, direct air capture with geological storage, enhanced rock weathering, ocean alkalinity enhancement, BECCS, and reforestation. The IPCC states that large-scale carbon removal is necessary to limit warming to 1.5 degrees Celsius.
Carbon removal costs vary by method. Biochar costs $131-164 per tonne of CO2 removed (Puro.earth 2024 data). Enhanced rock weathering costs $80-200 per tonne. Direct air capture with geological storage costs $400-600+ per tonne (Climeworks pricing). Reforestation is cheaper at $10-50 per tonne but has lower permanence and reversal risk from fire or drought. Ocean alkalinity enhancement is $50-150 per tonne but is still in early pilot stages. Costs are falling as the industry scales.
Carbon offsets typically represent emissions avoided or reduced, such as protecting a forest from logging or replacing a coal plant with wind power. The climate benefit depends on a counterfactual scenario (what would have happened without the project). Carbon removal represents CO2 actively pulled from the atmosphere and stored. The distinction matters: offsets prevent a hypothetical tonne of future emissions, while removal physically extracts an existing tonne. Major corporate buyers like Microsoft, Stripe, and Shopify have shifted procurement from offsets to removal credits because the climate benefit is more measurable and verifiable.
The IPCC estimates that limiting warming to 1.5 degrees Celsius requires 6-10 billion tonnes (gigatonnes) of CO2 removal per year by 2050. Current engineered CDR capacity is approximately 0.002 gigatonnes per year, a gap of more than three orders of magnitude. Tree planting and soil carbon contribute additional removal but face permanence challenges. Bridging this gap requires simultaneous scale-up across multiple CDR pathways, massive cost reductions, and coordinated policy support including carbon pricing and procurement commitments.
No single CDR method is best for all contexts. Biochar offers the best current combination of cost ($131-164/tCO2e), permanence (500+ years at production temperatures above 550 degrees Celsius), and co-benefits (soil improvement, water retention). Direct air capture has the highest permanence (geological storage for 10,000+ years) but the highest cost ($400-600+/tCO2e). Enhanced rock weathering is promising for agricultural regions with lower costs ($80-200/tCO2e) and soil benefits. The IPCC and major CDR researchers agree that a portfolio approach, scaling multiple methods simultaneously, is necessary because no single pathway can deliver the gigatonne-scale removal needed by 2050.
The Grove Briefing: a weekly digest of which green technologies are winning on cost curves, capital flows, and market signals.
Free via Substack. Unsubscribe anytime.