Preventing a tonne of carbon dioxide from entering the atmosphere and removing a tonne that is already in the atmosphere can produce similar arithmetic in a spreadsheet while representing very different physical acts.
One stops an addition. The other lowers the atmospheric stock.
Carbon dioxide removal, or CDR, refers to deliberate human activities that remove carbon dioxide from the atmosphere and durably store it in geological, terrestrial or ocean reservoirs, or in durable products.
That definition has two inseparable halves:
- carbon dioxide must be removed from the atmosphere;
- the removed carbon must be stored durably enough to matter to climate outcomes.
If atmospheric carbon is captured and immediately returned, removal has not been durable. If fossil carbon is captured before emission, an emission may have been avoided, but atmospheric carbon has not necessarily been removed. If a forest stores carbon for twenty years and then burns, some or all of the earlier removal can reverse.
Carbon removal is therefore a problem of source, flow, storage and time.
Wait, what? Carbon capture and carbon removal are not synonyms
A fossil power plant that captures CO₂ before it reaches the atmosphere reduces an emission relative to unabated operation.
A direct-air-capture plant that removes CO₂ from ambient air and stores it geologically removes atmospheric carbon.
A biomass power plant takes carbon that plants recently removed from air, generates energy, captures the resulting CO₂ and stores it. Under suitable lifecycle conditions, that can create net removal.
The hardware can overlap. The climate accounting does not.
The direct answer
A carbon dioxide removal system works through four accounting questions:
- Removal: how did carbon dioxide leave the atmosphere?
- Storage: where is the carbon now?
- Durability: how long is it expected to remain there?
- Net balance: after energy, materials, land-use change, leakage and reversal risks, how much atmospheric CO₂ was actually removed?
A system that cannot answer all four questions has not yet established a robust removal claim.
The IPCC definition
The Intergovernmental Panel on Climate Change defines CDR as anthropogenic activities that remove CO₂ from the atmosphere and durably store it in geological, terrestrial or ocean reservoirs, or in products.
The definition includes human enhancement of biological and geochemical sinks and direct air carbon dioxide capture and storage. It excludes natural CO₂ uptake that is not caused by deliberate human activity.
This distinction matters because forests, oceans and soils already absorb enormous amounts of carbon naturally. CDR accounting asks what additional removal was caused by the intervention.
Removal versus avoided emissions
Avoided emissions prevent additional carbon from entering the atmosphere.
Examples include:
- replacing coal power with renewable electricity;
- improving industrial efficiency;
- capturing fossil process emissions before release;
- switching from fossil fuel to low-carbon hydrogen.
Removal lowers atmospheric CO₂ relative to a world without the removal intervention.
Both are important. They are not interchangeable physically.
Why removal cannot substitute for ordinary emissions reduction
It is usually more efficient to avoid emitting a tonne in the first place than to emit it and later spend energy, land or materials removing it.
Removal also faces scale limits, durability constraints, land competition, energy demand and uncertainty.
IPCC assessments therefore treat CDR as complementary to deep emissions reduction, especially for counterbalancing difficult residual emissions and potentially achieving net-negative emissions—not as permission to ignore avoidable emissions.
Gross removal versus net removal
A project can physically remove 100 tonnes of CO₂ from air and still deliver less than 100 tonnes of net removal.
Suppose the project causes 20 tonnes of lifecycle emissions from electricity, construction, transport and materials.
Gross removal: 100 tonnes.
Lifecycle emissions: 20 tonnes.
Net removal: 80 tonnes, before considering any later reversal.
The difference between gross and net is fundamental.
Durability
Durability describes how long removed carbon is expected to remain stored.
Different pathways have very different storage timescales.
- trees can store carbon for decades to centuries, with fire and harvest risks;
- soils can gain or lose carbon as management and climate change;
- biochar can persist much longer depending on feedstock and conditions;
- geological storage can target very long containment;
- mineral carbonates can be extremely durable.
A tonne stored for 30 years does not have the same climate durability as a tonne mineralised or geologically stored for millennia.
Reversal
A reversal occurs when previously stored carbon returns to the atmosphere earlier than intended.
Examples include:
- forest fire;
- deforestation;
- soil disturbance;
- biochar decomposition beyond assumptions;
- storage-system leakage.
Credible removal systems therefore need reversal management as well as initial measurement.
Permanence is not binary
Carbon storage exists on a continuum of durability.
Some methods store carbon in living systems that require continued stewardship. Others convert carbon into minerals or place it in deep geological formations designed for very long containment.
Good accounting does not pretend these are identical. It records expected storage time, reversal probability and monitoring obligation.
Additionality
A removal is additional when it would not have occurred without the intervention or incentive being claimed.
Suppose a forest was already legally protected and expected to grow normally.
Claiming all future natural growth as newly caused removal can exaggerate the intervention’s impact.
Additionality asks: what changed because of this project?
Baseline
A baseline describes what would have happened without the project.
Removal is measured relative to that counterfactual.
Baselines are difficult because the counterfactual world cannot be directly observed.
Strong projects therefore use conservative, transparent assumptions and update them when evidence changes.
Leakage in carbon accounting
In CDR accounting, leakage can mean activity shifts rather than physical CO₂ leakage.
Example: a project protects one forest, but logging moves to another nearby forest.
The project area looks successful while broader atmospheric benefit is reduced.
Accounting therefore needs a system boundary large enough to detect displacement.
Direct air capture and storage
Direct air capture uses engineered chemical systems to separate carbon dioxide from ambient air.
The captured CO₂ can then be compressed and stored geologically.
The attraction is location flexibility: air mixes globally, so capture does not need to sit beside the original emitter.
The difficulty is dilution. Atmospheric CO₂ is far less concentrated than most industrial process streams, so moving air and regenerating sorbents requires energy and equipment.
eduKateSingapore already has a dedicated Science ROUTE Learning Manual owning the molecular DAC mechanism: One CO₂ Molecule in Direct Air Capture. This article does not duplicate that mechanism; it positions DACCS inside the wider removal portfolio.
DACCS
Direct Air Carbon Dioxide Capture and Storage, often abbreviated DACCS, combines atmospheric capture with durable storage.
Net removal depends on:
- capture efficiency;
- electricity and heat source;
- construction emissions;
- CO₂ transport;
- storage integrity;
- system lifetime.
A DAC plant powered by carbon-intensive energy can remove gross CO₂ while creating large counter-emissions.
Afforestation and reforestation
Trees remove carbon dioxide through photosynthesis and store carbon in wood, roots, litter and soils.
Afforestation establishes forest where forest has not existed recently under the relevant definition. Reforestation restores forest where it was lost or degraded.
These approaches can provide biodiversity, water and social co-benefits.
They also face limits:
- land competition;
- fire;
- drought;
- pests;
- harvest;
- slow growth;
- saturation as forests mature.
A forest is therefore both a carbon store and a living ecosystem exposed to changing climate.
Improved forest management
Existing forests can sometimes store more carbon through changes in harvest cycles, restoration, species management, fire management or reduced degradation.
The removal claim depends strongly on baseline: what would the forest have done without the management change?
Measurement and additionality are therefore as important as tree growth.
Soil carbon
Soils contain enormous carbon stocks.
Changes in tillage, cover crops, grazing, organic inputs and land management can increase soil organic carbon in some contexts.
The challenge is measurement and durability.
Soil carbon varies spatially. Gains can saturate. Later changes in land management can reverse them.
Credible soil CDR therefore requires repeated measurement and long-term management commitments.
Biochar
Biochar is carbon-rich material produced by heating biomass under limited-oxygen conditions.
Part of the biomass carbon becomes a relatively stable solid that can be incorporated into soils or other applications.
The removal chain is:
atmospheric CO₂ → plant growth → biomass → pyrolysis → stable carbon storage.
Net removal depends on feedstock sourcing, transport, pyrolysis energy, avoided alternative uses and long-term biochar stability.
BECCS
Bioenergy with carbon dioxide capture and storage, or BECCS, combines biological uptake with engineered capture and geological storage.
Plants remove CO₂ while growing. Biomass is converted into electricity, heat, fuels or chemicals. CO₂ from conversion is captured and stored geologically.
When the biomass is genuinely sustainable and lifecycle emissions are sufficiently low, the system can create net atmospheric removal while producing energy.
The difficulty is biomass.
- Where did it grow?
- What land did it use?
- Would the biomass have stored carbon otherwise?
- How much fertiliser was used?
- How far was it transported?
- What happened to soil carbon?
BECCS is not automatically carbon-negative simply because biomass and capture appear in the same project.
Biogenic industrial CO₂
Some processes create concentrated biogenic carbon dioxide streams without requiring a dedicated bioenergy power plant.
Fermentation is one example.
Capturing and geologically storing a high-purity biogenic stream can potentially create removal with relatively modest separation energy.
This can be one of the simpler engineered CDR opportunities where sustainable biomass feedstock already exists for another purpose.
Enhanced weathering
Natural rock weathering consumes carbon dioxide over geological time.
Enhanced weathering attempts to accelerate that process by increasing reactive mineral surface area or placing suitable minerals in environments where they react more quickly.
Potential applications include finely ground silicate minerals applied to soils or other engineered contact systems.
The major questions include mining, grinding energy, transport, reaction rate, trace metals, soil effects and how accurately removal can be measured.
Mineral carbonation
Carbon dioxide can react with calcium- or magnesium-bearing minerals to form solid carbonates.
Mineral carbonates can be highly durable.
The challenge is reaction speed, material handling and energy.
Some pathways use mined rock. Others use industrial wastes, mine tailings or alkaline materials already produced by industry.
Using reactive waste can reduce mining demand and solve two problems at once, but feedstock quantity and contaminants matter.
Carbon mineralisation in concrete
Carbon dioxide can be mineralised into some concrete, aggregate or curing processes.
The carbon becomes part of a solid mineral phase.
The climate benefit depends on whether the CO₂ was atmospheric or biogenic, whether ordinary cement emissions are reduced, and how much carbon the product actually stores.
Carbon utilisation can create durable storage in products—but not every CO₂-derived product qualifies as removal.
Ocean alkalinity enhancement
The ocean naturally absorbs carbon dioxide.
Increasing ocean alkalinity can, in principle, shift seawater chemistry so the ocean can take up additional atmospheric CO₂ and store more carbon in dissolved inorganic forms.
The concept is potentially large-scale but scientifically and operationally complex.
Questions include mineral sourcing, ecological effects, reaction rates, measurement, spatial attribution and international governance.
Ocean pathways therefore require careful research and verification before claims can be treated like well-characterised geological storage.
Blue carbon
Coastal ecosystems such as mangroves, salt marshes and seagrass meadows can store substantial carbon in biomass and sediments.
Protecting or restoring these ecosystems can preserve and increase carbon stocks while supporting biodiversity and coastal resilience.
Removal claims need to distinguish avoided loss of existing carbon from additional new atmospheric uptake.
Wood products
Carbon taken up by trees can remain stored in long-lived wood products after harvest.
The duration depends on product lifetime, disposal and recycling.
Wood used in a long-lived building stores carbon differently from paper burned after months.
Product pools therefore need survival curves rather than a binary “stored/not stored” label.
The durability ladder
CDR methods can be arranged loosely by expected durability, although every project needs its own assessment.
- shorter and more reversible biological storage;
- longer-lived soils and products;
- biochar and some mineralised products;
- deep geological storage and stable mineral carbonates.
A tonne stored more durably can carry a different risk profile from a tonne requiring continual management.
Storage duration should match the climate job
If removal is used to counterbalance a fossil CO₂ emission that would affect climate for centuries, compensating storage should also be durable.
Short-lived storage can delay atmospheric carbon but may not provide equivalent long-term compensation.
This is why durability cannot be ignored when tonnes from different methods are compared.
Reversal buffers
Some carbon-credit systems withhold a fraction of credits in a shared buffer pool.
If a forest burns, buffer credits can be cancelled to cover the reversal.
This is a financial risk-management mechanism. It does not physically prevent fire.
Strong CDR combines physical risk reduction with accounting mechanisms for residual risk.
Monitoring, reporting and verification
Every removal pathway needs an MRV system appropriate to its physics.
- DACCS can meter captured and injected CO₂ directly while monitoring lifecycle energy.
- Forests use inventories, remote sensing and growth models.
- Soils require sampling and statistical spatial design.
- Biochar requires feedstock, production and stability accounting.
- Mineralisation needs chemical analysis of reacted carbon.
- Geological storage tracks injection, plume behaviour and containment.
There is no single universal CDR meter.
Measurement uncertainty
Removal estimates have uncertainty.
A pipeline flow meter can be highly precise. Soil-carbon estimates across a large landscape can be much more uncertain.
Good accounting reports uncertainty rather than hiding it behind a single exact tonne number.
Lifecycle emissions
CDR equipment and land management consume resources.
- DAC uses heat and electricity.
- Enhanced weathering uses mining and grinding.
- BECCS uses cultivation, harvest and transport.
- Biochar uses pyrolysis equipment.
- Geological storage uses compression and drilling.
Net removal subtracts these counter-emissions from gross atmospheric uptake.
Energy source can decide whether engineered removal is actually negative
An energy-intensive removal plant powered by high-carbon energy can undermine its own purpose.
Low-carbon electricity and heat therefore become critical inputs to DACCS, mineral processing and other engineered methods.
This creates an opportunity cost: clean electricity used for removal cannot simultaneously displace fossil electricity elsewhere.
Energy planning must therefore compare marginal system value.
Land competition
Land-based CDR can compete with:
- food production;
- biodiversity;
- housing;
- renewable generation;
- water;
- local livelihoods.
A removal pathway that stores carbon while causing deforestation or food-system pressure elsewhere can lose much of its benefit.
Water
Some CDR methods increase water demand.
Forests can change evapotranspiration. Biomass requires water. DAC technologies differ in water use. Enhanced weathering can affect soil chemistry and hydrology.
Water-constrained regions therefore need to evaluate removal per unit water as well as removal per unit land or money.
Biodiversity
Carbon and biodiversity can align—or conflict.
Restoring native ecosystems can support both. Planting a monoculture plantation on biodiverse grassland can increase tree carbon while harming ecosystems.
Carbon tonnes should therefore not be treated as the only environmental metric.
Food and biomass
Large BECCS or afforestation scenarios can demand substantial land.
If dedicated biomass crops displace food production, prices and land-use change can follow.
Residues and wastes can reduce direct competition but are finite and often already have useful roles.
Residual emissions
Even a deeply decarbonised economy can retain difficult residual emissions.
Examples can include some agricultural methane and nitrous oxide, industrial process emissions, aviation effects and dispersed sources that are very costly to eliminate completely.
CDR can counterbalance part of that residual climate burden, but the amount needed depends on how aggressively direct emissions are reduced first.
Net zero
Net-zero CO₂ is reached when anthropogenic CO₂ emissions are balanced by anthropogenic CO₂ removals over a specified period and boundary.
This does not mean emissions become harmless.
It means remaining emissions are counterbalanced by removals.
The stronger strategy is usually:
- reduce avoidable emissions deeply;
- reserve removal for residual emissions;
- use durable storage for durable climate compensation;
- continue improving both sides.
Net-negative emissions
A system becomes net-negative when removals exceed emissions across the defined boundary and period.
Example:
Annual residual emissions = 10 million tonnes CO₂.
Verified net durable removals = 14 million tonnes CO₂.
Net balance = negative 4 million tonnes.
The credibility depends on whether both emissions and removals use compatible boundaries and whether the removals are genuinely additional and durable.
Carbon debt
Some removal pathways incur emissions upfront before later uptake repays them.
Clearing land, constructing equipment or transporting biomass can create an initial carbon debt.
The time required to repay that debt matters to near-term climate outcomes.
Saturation
Biological reservoirs cannot grow indefinitely.
A forest eventually approaches a mature carbon stock. Soil-carbon gains can slow as a new equilibrium is approached.
Once saturation is reached, continued stewardship preserves the stock but may no longer create the same annual removal flow.
Removal rate and stored stock are different metrics.
Stock versus flow
A forest stores a carbon stock.
Annual tree growth creates a removal flow.
If growth stops but the forest remains intact, the flow falls toward zero while the stock remains valuable.
Destroying the forest can release part of the stock, reversing past removal.
Good CDR accounting tracks both.
Atmospheric rebound
The carbon cycle responds when humans remove CO₂ from the atmosphere.
Some carbon can move back toward the atmosphere from oceans and land as the system re-equilibrates.
IPCC assessments note that the climate response to removals is not perfectly symmetric with emissions.
This is another reason preventing emissions first is generally preferable to emitting and attempting to remove later.
Removal portfolios
No single CDR method is likely to dominate every criterion.
- forests can be low-cost and provide ecosystems but face reversal;
- DACCS can be highly measurable and durable but energy-intensive and expensive;
- biochar can store carbon and improve some soils but feedstock is limited;
- BECCS can combine energy and removal but competes for biomass and land;
- mineralisation can be highly durable but material handling can be large;
- soil carbon can support agriculture but is difficult to measure precisely and can reverse.
A diversified portfolio can reduce dependence on one failure mode.
The Pareto frontier of removals
Removal methods trade off:
- cost;
- durability;
- measurement precision;
- land;
- water;
- energy;
- biodiversity;
- scale;
- social acceptance.
The cheapest tonne may be less durable. The most durable tonne may be more expensive. The most scalable method may use more energy.
The decision logic belongs to How Energy Trade-Offs Work.
Removal credits and tonne equivalence
Markets often want to express removals in one common unit: tonnes of CO₂.
But one tonne stored in a forest, one tonne stored in a mineral and one tonne stored geologically do not necessarily carry identical reversal and durability risks.
High-integrity systems therefore need either differentiated products or methods that adjust for durability and risk rather than pretending every tonne is perfectly fungible.
Double counting
One removal should not be claimed by two actors.
If a project sells a removal credit to an airline, the host country, project developer and airline cannot all independently claim the same atmospheric benefit under incompatible accounting systems.
Registries and chain-of-custody systems exist partly to prevent this.
Temporal accounting
Removal happens through time.
A tree plantation may take decades to accumulate carbon. A DAC plant can meter captured CO₂ continuously. Enhanced weathering can continue reacting after material application.
Crediting should reflect when removal actually occurs rather than granting the entire future estimate on day one without risk adjustment.
Ex ante versus ex post crediting
Ex ante crediting issues value based on expected future removal.
Ex post crediting waits until removal is measured or verified.
Ex ante systems can finance projects earlier but carry delivery risk. Ex post systems provide stronger evidence but make financing harder.
This is a financial trade-off layered onto a physical carbon process.
Singapore and carbon dioxide removal
Singapore has limited land, dense urban infrastructure and high energy demand.
That constrains large land-intensive removal pathways within national borders.
Singapore’s removal portfolio is therefore likely to place unusual value on:
- high-quality international removals with robust accounting;
- regional geological storage;
- durable mineralisation in construction materials where practical;
- biogenic industrial CO₂ capture where appropriate;
- future engineered removals that can operate with limited land.
The strategic challenge is verification. When removal happens outside the country, Singapore must still know that the tonne was additional, durable, not double-counted and correctly attributed.
This makes CDR partly an infrastructure problem and partly an evidence problem.
Worked example 1: DACCS
A direct-air-capture facility removes 100,000 tonnes of CO₂ from ambient air.
Low-carbon electricity and heat run the system. Compression, transport and construction create 12,000 tonnes of lifecycle emissions. Geological storage receives and verifies 100,000 tonnes.
Approximate net removal before other uncertainties: 88,000 tonnes.
The accounting subtracts system emissions rather than crediting gross capture.
Worked example 2: forest restoration
A degraded forest is restored and accumulates additional carbon relative to a credible baseline.
Twenty years later, a severe fire releases part of the stored carbon.
The project delivered real removal for twenty years, but some durability was lost.
A credible system monitors the stock and accounts for the reversal rather than pretending the original credit is permanent.
Worked example 3: BECCS with bad biomass
A BECCS plant captures and stores large volumes of biogenic CO₂.
However, dedicated biomass plantations displaced natural forest and increased fertiliser use.
The stack measurement looks excellent while lifecycle land-use emissions undermine the net removal.
The lesson: capture technology cannot repair a poor feedstock boundary.
Worked example 4: biochar residue pathway
A farm region has crop residues that would otherwise decompose rapidly.
Part of the residue is converted to biochar using a controlled pyrolysis system. Process heat is recovered. The biochar is applied to suitable soils and tracked.
Net removal depends on actual biochar stability, alternative residue use, transport and pyrolysis emissions.
Worked example 5: mineralisation in waste material
An industrial waste stream contains reactive alkaline minerals.
Captured atmospheric or biogenic CO₂ reacts with the material and forms stable carbonate minerals.
The method avoids new mining because the reactive material already exists.
The project must still measure how much carbon actually mineralised and whether the product remains stable in its final use.
Failure mode: avoided emissions sold as removal
A fossil plant installs capture and calls every captured tonne atmospheric removal.
Repair: distinguish prevented emissions from atmospheric removal.
Failure mode: gross removal sold as net removal
A DAC plant credits every tonne captured without subtracting electricity and construction emissions.
Repair: use lifecycle net removal.
Failure mode: no durability distinction
A twenty-year forest commitment is treated as identical to verified geological storage.
Repair: report storage duration, reversal risk and monitoring obligation explicitly.
Failure mode: weak baseline
A project claims carbon that would have accumulated without intervention.
Repair: use conservative additionality and transparent counterfactual assumptions.
Failure mode: displacement ignored
One forest is protected but logging moves elsewhere.
Repair: expand system boundaries and account for market leakage where relevant.
Failure mode: future removals credited immediately
A 50-year tree-growth forecast is sold today as though all carbon is already stored.
Repair: distinguish ex ante finance from ex post verified removal and manage delivery risk.
Failure mode: carbon removal used to justify avoidable emissions
A company continues easily avoidable fossil use because cheap short-duration offsets appear available.
Repair: prioritise direct emissions reduction and reserve CDR for residual emissions and net-negative goals.
Failure mode: one removal technology becomes the whole plan
A national strategy assumes one method will scale indefinitely.
Land, energy, feedstock or public acceptance later constrain it.
Repair: build a diversified portfolio and update it as real-world performance becomes visible.
Common misconceptions
- Carbon capture is not automatically carbon dioxide removal.
- CDR requires both atmospheric removal and durable storage.
- Natural uptake without deliberate human intervention is not automatically anthropogenic CDR.
- Gross removal is larger than or equal to net removal.
- Durability differs strongly across methods.
- Forests and soils can reverse.
- Geological and mineral storage can offer much longer durability.
- DACCS is only net-negative if lifecycle emissions are sufficiently low.
- BECCS is not automatically negative; biomass sourcing decides much of the result.
- Carbon utilisation is not automatically removal.
- Recycled fossil carbon is not the same as atmospheric carbon removal.
- Additionality, baseline and leakage can matter as much as the physical capture process.
- One tonne from every CDR method is not necessarily equivalent in durability and risk.
- Removal complements deep emissions reductions; it does not make avoidable emissions irrelevant.
A universal carbon-removal audit
- Define how CO₂ leaves the atmosphere.
- Define the counterfactual baseline.
- Test additionality.
- Measure gross atmospheric removal.
- Define where carbon is stored.
- Estimate storage duration.
- Identify reversal pathways.
- Quantify lifecycle emissions.
- Calculate net removal.
- Assess land use.
- Assess water use.
- Assess biodiversity effects.
- Assess energy opportunity cost.
- Assess feedstock sustainability.
- Check displacement or market leakage.
- Define MRV appropriate to the method.
- Report measurement uncertainty.
- Prevent double counting.
- Define chain of custody for credits or claims.
- Define reversal responsibility.
- Match durability to the climate job being claimed.
- Review the project as baselines, science and operating conditions change.
The deepest carbon-removal principle
Carbon removal is easy to describe as a negative emission.
But a negative sign in an inventory is the end of the reasoning, not the beginning.
Behind that sign must be a physical story:
atmospheric carbon moved into a measured reservoir, because of a deliberate intervention, with lifecycle emissions deducted, with reversal risk managed, for a storage duration appropriate to the claim.
When that story is strong, CDR can help balance residual emissions and eventually support net-negative systems.
When the story is weak, the tonne can become an accounting fiction.
How Carbon Dioxide Removal fits the Energy series
How Carbon Capture Works owns source-side CO₂ separation. How Carbon Dioxide Transport Works owns movement. How Geological Carbon Storage Works owns subsurface injection and permanence. The existing eduKateSingapore Direct Air Capture Learning Manual owns the detailed DAC molecule mechanism. This article owns the umbrella CDR framework: what qualifies as removal, how methods differ, and how durability, lifecycle emissions, additionality, reversal and net-negative accounting should be compared.
Current evidence and further reading
- Intergovernmental Panel on Climate Change — AR6 Synthesis Report Glossary: Carbon Dioxide Removal
- Intergovernmental Panel on Climate Change — Carbon Dioxide Removal Factsheet
- Intergovernmental Panel on Climate Change — 2027 Methodology Report on CDR and CCUS
- International Energy Agency — Carbon Capture, Utilisation and Storage
The final lesson is simple: real removal is not a promise to take carbon out later; it is an evidenced transfer of atmospheric carbon into storage durable enough to carry the climate claim being made.