Skip to content
ContentLora

    Tip: press / anywhere to search.

    technology

    Direct air capture (DAC)

    Also known as DAC, DACCS, direct air carbon capture and storage

    Direct air capture with storage (DACCS) uses machines with sorbents or solvents to pull CO2 out of ordinary air and store it underground; it is energy intensive and currently costs roughly $400–2,500 per tonne.[1][2] In 2025 it removed only about 1,500 tonnes worldwide, but upgrades at Climeworks' Mammoth and the planned 500,000-tonne STRATOS plant make 2026 a test of whether it can scale.[3][4][5]

    Editor reviewedUpdated WikiClimate science and carbon removal, Energy and climate, Science
    Key facts

    What it is

    Direct air capture (DAC) separates carbon dioxide from ambient air; paired with permanent storage it is called DACCS, one of the “novel” carbon dioxide removal methods.[6] Because CO2 makes up only about 425 parts per million of air, large volumes must be processed.[7] The State of CDR report notes that DACCS is energy intensive and its performance depends on the carbon intensity of the energy it uses.[1]

    How it works

    Capture processes vary. Solid-sorbent systems, which capture CO2 on a filter material, and mineral-looping systems are commercially deployed; liquid-solvent systems are at pilot scale and electrochemical regeneration is earlier still.[1][8] The storage step builds on established geological storage practice.[1]

    Because the process needs a lot of energy, scale depends on clean power. Frontier, a carbon removal buyers’ group, estimates that 1 billion tonnes a year at 2 MWh per tonne would need about 2,000 TWh of clean energy, around 6% of today’s global electricity.[9] For how low-carbon power is balanced on the grid, see how grid-scale storage works.

    Operators also distinguish gross capture from net removal. Climeworks counts as net removal only the CO2 permanently stored after allowing for maintenance, weather, plant availability and the “grey” emissions from running the plant and storing the CO2.[10]

    Scale and cost

    DACCS removed about 0.0015 million tonnes of CO2 in 2025, against total global removal of about 2.2 billion tonnes, nearly all from forests and soils.[3][11] A 2025 review cited by the State of CDR report found levelized costs of roughly $400–2,500 per tonne, possibly falling to $100–600 at gigatonne scale through economies of scale and learning.[2] Constrained potential estimates range widely, and the highest-ambition scenarios use under 0.1 to 1.8 billion tonnes a year by 2050.[12]

    The buyers’ coalition Frontier gives a higher ceiling at a high price: it puts DAC’s potential above 10 billion tonnes a year at $200–300 per tonne, well above its estimates for biomass-based methods and enhanced weathering.[9][13][14]

    Landmark plants

    • Mammoth (Iceland). climeworks switched on Mammoth on 8 May 2024 with a nameplate capacity of up to 36,000 tonnes a year.[15] In September 2026 the company said upgrades had more than doubled capture in upgraded containers, and that net removal reached 675 tonnes in the first half of 2026 against 119 tonnes a year earlier.[4][16] It plans to extend the upgrades to all 12 containers of one plant module by the end of 2026.[17]
    • STRATOS (Texas). Occidental’s 1PointFive plant is designed for up to 500,000 tonnes a year, with permits to store CO2 in saline formations.[5] Originally due by the end of 2024, it hit a further delay in May 2026 after phase-one commissioning revealed what Occidental called a “non-process component issue”.[18][19] In August 2026 Occidental said full plant commissioning should begin around the end of 2026, with operations in 2027.[20]
    • US DAC hubs. In March 2026 E&E News reported that early work on Project Cypress in Louisiana, designed for 1 million tonnes a year, had stalled amid uncertainty over its federal award.[21] In April federal funding for Cypress and the South Texas DAC Hub was reported preserved, but only $50 million of each award had been paid out.[22][23]

    Open questions

    The main questions are whether plants can reach their nameplate capacity, whether costs fall as projected, and whether buyers keep paying. The State of CDR report flags lifecycle emissions of energy-intensive DACCS as an open measurement problem.[24] Demand is concentrated: Microsoft held 78.5% of disclosed durable removal purchases as of April 2026 and was reported that month to be pausing new purchases.[25][26] Its later deals, with BioCirc, Alt Carbon and Crew Carbon, went to biogas capture, enhanced weathering and wastewater alkalinity rather than DAC.[27][28]

    Next in the crash course

    More in Climate & carbon

    Developing in Science & Energy

    Questions readers ask

    How much CO2 does direct air capture remove today?

    Very little. The State of CDR report estimates about 0.0015 million tonnes in 2025, from Climeworks' Orca and Mammoth plants running below capacity.[3]

    How much does direct air capture cost?

    A 2025 review found levelized costs of roughly $400–2,500 per tonne, possibly falling to $100–600 at gigatonne scale.[2]

    When will the STRATOS plant open?

    Occidental first planned operation by the end of 2024 and reported a new delay in May 2026. In August 2026 it said full plant commissioning should begin around the end of 2026, with operations in 2027. It is designed for up to 500,000 tonnes a year.[18][20][5]

    Could direct air capture remove billions of tonnes?

    Constrained estimates range from 0.5–5 billion to 5–40 billion tonnes a year, limited by clean energy and storage, and ambitious scenarios use under 0.1 to 1.8 billion tonnes by 2050.[12]

    How much energy would direct air capture need at scale?

    Frontier estimates that removing 1 billion tonnes a year at 2 MWh per tonne would need about 2,000 TWh of clean energy, around 6% of today's global electricity.[9]

    Sources

    · 31 claims from 14 publishers

    Each numbered claim is a statement we checked against the sources listed with it. Tap a number in the text to see its sources in place.

    1. [1]

      Solid-sorbent and mineral-looping direct air capture are commercially deployed (TRL 9), liquid-solvent systems are at pilot scale (TRL 6–7), and electrochemical regeneration is earlier (TRL 4–6); all are energy intensive. confirmedas of 2026-06-01

    2. [2]

      A 2025 review found levelized costs of direct air capture gross removal of roughly $400–2,500 per tonne, possibly falling to $100–600 per tonne at gigatonne scale. confirmedas of 2026-06-01

    3. [3]

      Direct air capture with storage removed about 0.0015 million tonnes (roughly 1,500 tonnes) of CO2 in 2025, with Climeworks' Orca and Mammoth both still running below full capacity. confirmedas of 2026-06-01

    4. [4]

      On 10 September 2026 Climeworks said sorbent and plant upgrades had more than doubled CO2 capture in upgraded collector containers at Mammoth and cut operating costs per tonne by more than 50%. confirmedas of 2026-09-10

    5. [5]

      Occidental's 1PointFive STRATOS plant in the Texas Permian Basin is designed to capture up to 500,000 tonnes of CO2 a year, with Class VI permits for three wells to store about 722,000 tonnes a year in saline formations. confirmedas of 2026-03-03

    6. [6]

      The State of CDR report splits removal into "conventional" methods, largely forests and soils, and "novel" methods such as biochar, enhanced weathering and direct air carbon capture and storage. confirmedas of 2026-06-01

    7. [7]

      Global mean surface concentrations in 2025 were 425.6 ppm for carbon dioxide, 1936.3 ppb for methane and 339.4 ppb for nitrous oxide. confirmedas of 2026-06-11

    8. [8]

      Climeworks, headquartered in Zurich, began as a laboratory project at ETH Zurich, built the first commercial direct air capture plant in Switzerland, and then built Orca and Mammoth in Iceland. confirmedas of 2026-09-10

    9. [9]

      Frontier estimates direct air capture could scale beyond 10 billion tonnes a year at $200–300 per tonne, but says 1 billion tonnes a year at 2 MWh per tonne would need about 2,000 TWh of clean energy, around 6% of today's global electricity. confirmedas of 2026-06-17· interpretation

    10. [10]

      Climeworks defines net CDR production as CO2 permanently removed and stored after accounting for maintenance, weather, plant availability and grey emissions from capture and storage; these are the tonnes delivered to customers. confirmedas of 2026-09-10

    11. [11]

      Total carbon removal is about 2.2 billion tonnes of CO2 a year, around 5% of gross CO2 emissions; conventional methods provide 99.9% and novel methods about 2 million tonnes, growing about 40% a year. confirmedas of 2026-06-01

    12. [12]

      Constrained estimates of direct air capture potential range from 5–40 billion tonnes a year (IPCC AR6) down to 0.5–5 billion tonnes by 2050, and the highest-ambition scenarios use less than 0.1 to 1.8 billion tonnes by 2050. confirmedas of 2026-06-01

    13. [13]

      Frontier estimates biomass carbon removal and storage, including BECCS and bio-oil injection, could reach 1–5 billion tonnes a year at $60–200 per tonne if feedstock meets sustainable sourcing criteria; 1 billion tonnes would take about 1,250 plants the size of Stockholm Exergi's. confirmedas of 2026-06-17· interpretation

    14. [14]

      Frontier estimates enhanced rock weathering could reach about 1–2 billion tonnes a year at $75–200 per tonne; 1 billion tonnes would mean crushing about four billion tonnes of rock and spreading it on roughly 28% of global cropland, and would require moving from direct soil measurements to modelled quantification. confirmedas of 2026-06-17· interpretation

    15. [15]

      Climeworks' Mammoth plant at Hellisheiði, Iceland started operations on 8 May 2024, with 12 of its 72 collector containers installed, and has a nameplate capacity of up to 36,000 tonnes of CO2 a year, about ten times its predecessor Orca. confirmedas of 2024-05-08

    16. [16]

      Mammoth produced 675 tonnes of net CO2 removal in the first half of 2026, compared with 119 tonnes in the first half of 2025. confirmedas of 2026-09-10

    17. [17]

      Climeworks says it will extend the new sorbent and mechanical upgrades to all 12 collector containers in one Mammoth plant module by the end of 2026 and begin testing its next-generation technology at Mammoth in early 2027. reportedas of 2026-09-10· forecast

    18. [18]

      In May 2026 Occidental said STRATOS, first planned to operate by the end of 2024, had hit an unexpected delay from component issues after its core process units performed as expected in testing. E&E News noted the plant is designed for 500,000 tonnes a year, against 36,000 tonnes for the largest DAC plant then operating. confirmedas of 2026-05-07

    19. [19]

      Occidental said on its first-quarter 2026 call that STRATOS's second construction phase was complete but that phase-one commissioning had revealed a "non-process component issue". confirmedas of 2026-05-12

    20. [20]

      On 6 August 2026 Occidental CEO Richard Jackson said repair and commissioning of STRATOS trains 3 and 4 was progressing and that full plant commissioning was expected to begin around the end of 2026, with the plant transitioning to operations in 2027 (company statement). confirmedas of 2026-08-06

    21. [21]

      In March 2026 E&E News reported that preliminary work on Project Cypress, designed to remove 1 million tonnes of CO2 a year in Louisiana, had stalled amid uncertainty over up to $550 million in DOE funding. reportedas of 2026-03-20

    22. [22]

      In April 2026 Carbon Herald reported that the US Department of Energy had kept funding for the two large DAC hubs, Project Cypress in Louisiana (a Heirloom, Climeworks and Battelle partnership) and the South Texas DAC Hub (1PointFive/Occidental), after earlier signals they could be cancelled. reportedas of 2026-04-17

    23. [23]

      Only an initial $50 million tranche had been disbursed to each of the two DAC hubs as of April 2026, against awards of $500 million for the South Texas DAC Hub and $550 million for Project Cypress. reportedas of 2026-04-20

    24. [24]

      The State of CDR report lists open problems for novel CDR quality, including sustainable biomass sourcing for BECCS and biochar, permanence of storage, measurement of enhanced weathering and ocean alkalinity enhancement, and lifecycle emissions of energy-intensive DACCS. confirmedas of 2026-06-01

    25. [25]

      Microsoft accounted for 78.5% of all disclosed durable carbon removal tonnes purchased as of April 2026, according to CDR.fyi. confirmedas of 2026-06-12

    26. [26]

      In April 2026 Heatmap News and Bloomberg reported that Microsoft was pausing future carbon removal purchases; Microsoft said it may "adjust the pace or volume" of its procurement. confirmedas of 2026-04-16

    27. [27]

      In May 2026, weeks after the reported pause, Microsoft agreed to buy 650,000 tonnes of carbon removal credits from BioCirc, which captures CO2 at Danish biogas plants and stores it offshore; Microsoft's chief sustainability officer said its carbon removal programme had not ended. confirmedas of 2026-05-20

    28. [28]

      In August 2026 Microsoft signed a deal with Crew Carbon for up to 23,602 removal units from wastewater alkalinity enhancement, its third removal deal since reports that it was adjusting the pace of purchases, after BioCirc (650,000 tonnes) and enhanced weathering firm Alt Carbon (36,920 tonnes); DatacenterDynamics said its purchasing cadence had dropped significantly in 2026. reportedas of 2026-08-05

    29. [29]

      Frontier is an advance market commitment that aims to accelerate carbon removal technologies by guaranteeing future demand for them. confirmedas of 2026-10-10

    30. [30]

      BECCS is an umbrella term for technologies that turn biomass into energy and capture and geologically store the CO2; bioethanol with CCS is commercial (TRL 9) while biomass combustion with CCS is at demonstration stage (TRL 7). confirmedas of 2026-06-01

    31. [31]

      Direct ocean capture pumps seawater through electrochemical systems to extract dissolved CO2, after which the CO2-depleted seawater gradually takes up atmospheric CO2; it is at lab-to-pilot scale (TRL 5) with system-level costs estimated at $1,000–2,000 per tonne. confirmedas of 2026-06-01

    Revision history (2)
    1. Page created.
    2. Refresh: STRATOS now expected to operate in 2027; added DAC hub funding status, Frontier's energy and cost outlook, and Mammoth's roll-out plan.

    Created Oct 10, 2026. Last reviewed by an editor on Oct 10, 2026. Next scheduled review: Jan 10, 2027.

    Cite this page

    "Direct air capture (DAC)." ContentLora, updated Oct 10, 2026. https://contentlora.com/wiki/direct-air-capture

    Spotted an error? Suggest a correction or emailcorrections@contentlora.com.

    Connections
    Climeworks Leading developerFrontier Buyer coalitionBECCS Alternative methodOcean-based carbon … Ocean analogue
    ☰ Contents