Carbon Removal: what are the biggest unsolved problems?

Why is carbon removal still such a big problem if we already remove billions of tonnes of CO₂?

Carbon removal is already happening at huge scale, but almost all of today's volume comes from conventional land-based methods; the newer technologies expected to provide much more durable removal are still tiny.

The latest State of Carbon Dioxide Removal assessment estimates that humanity removes around 2 billion tonnes of CO₂ a year through activities such as afforestation, reforestation and forest management. Novel carbon removal, which includes approaches such as direct air capture, biochar, enhanced weathering and biomass with geological storage, is only around 2 million tonnes a year.

So two very different industries sit under the same "carbon removal" label.

Conventional land-based removal already operates at gigaton scale, but forests and soils face limits around land, saturation, measurement and reversal. Novel removal can often store carbon for much longer, yet today's annual volume is roughly one-thousandth as large.

The encouraging part is that novel CDR is currently growing by around 40% a year according to the latest State of CDR assessment. That is fast, but the starting point is tiny. Even many years of 40% growth are required before a 2-million-tonne industry becomes a billion-tonne industry.

Paris-consistent climate pathways generally require several additional gigatonnes of annual carbon removal later this century.

We already know numerous ways to take CO₂ out of the atmosphere. The unsolved part is doing it billions of times a year, keeping the carbon stored and making the economics work.

Carbon removal today	Approximate scale
Conventional land-based CDR	~2 GtCO₂/year
Novel CDR	~0.002 GtCO₂/year
Durable CDR delivered commercially to date	~0.0017 GtCO₂
Novel CDR recent annual growth	~40%

Are companies actually removing the carbon they sold, and can they build projects fast enough?

Carbon removal sales are running far ahead of physical delivery today, and the industry's ability to turn those contracts into operating projects is still unproven.

CDR.fyi currently tracks roughly 49.5 million tonnes of durable carbon removal sold and about 1.67 million tonnes delivered. The gap is enormous, although it does not mean suppliers have defaulted on 96% of their contracts. Many of the industry's biggest agreements deliberately cover deliveries extending for ten years or more.

Recent purchasing has made that especially visible. Biomass-based carbon removal continued to dominate new durable CDR sales in early 2026. Liferaft, for example, signed a one-million-tonne deal with Microsoft covering ten years. Exomad Green added more than 600,000 tonnes of sales through several agreements and passed 2.3 million tonnes sold overall.

Those are large contracts for an emerging industry, yet operational deliveries remain concentrated in smaller and more mature pathways such as biochar. Much of today's apparent scale therefore consists of projects that still need financing, permits, equipment, construction and commissioning.

At one gigaton, even million-tonne facilities would need to be replicated roughly a thousand times. Biochar can spread that burden across many smaller plants, while DAC and BioCCS require much larger facilities, energy systems and CO₂ transport infrastructure.

Climeworks' Mammoth experience shows the challenge at facility level. New hardware had to be installed, tested, improved and gradually pushed toward better capacity factors. Its net output is rising sharply, but industrial ramp-up takes repeated engineering work.

What deserves much more attention now is the conversion rate: how many contracted tonnes arrive on schedule, how many projects actually reach operation, and whether suppliers can repeat delivery year after year.

Can carbon removal ever get cheap enough?

Some carbon removal methods should become much cheaper, but today's prices remain far above what we would want for a multi-gigaton industry.

Frontier's current portfolio shows how wide the range still is. Enhanced weathering contracts average about $362 per tonne, with its portfolio ranging from roughly $270 to $500. Marine carbon removal projects span an even wider range, from around $238 to nearly $2,000 per tonne. Early direct air capture purchases have often landed in the several-hundred-dollar range or higher.

These are early-market prices. Frontier deliberately funds technologies before they reach large scale, so the contracts carry technical, financing and construction risk that should decline if the technologies work.

We can already see some cost movement. Climeworks says the operating cost per tonne at its Mammoth direct air capture plant has fallen by more than 50% over the past year as it improved its sorbents and equipment. That is meaningful, although Mammoth is still producing carbon removal on a very small scale.

The arithmetic is harsh. Removing 1 billion tonnes at $500 per tonne costs $500 billion. Five billion tonnes would cost $2.5 trillion every year. Even at $100 per tonne, five billion tonnes still means $500 billion annually.

That is why $100 has become such a powerful target across the industry. Direct air capture companies are trying to extend sorbent life, cut energy use and mass-produce equipment. Biomass projects can earn revenue from electricity, heat or waste disposal. Enhanced-weathering companies can piggyback on existing mining and agricultural systems.

We still have no proof that any highly durable pathway can deliver several gigatonnes at around $100 per tonne.

Selected early-market pathway	Current Frontier pricing
Enhanced weathering	~$270–500/t
Marine carbon removal	~$238–1,961/t
Typical early DAC purchases	Several hundred dollars/t or more

Can direct air capture ever use little enough energy?

Direct air capture is getting more efficient, but energy will remain a serious constraint because capturing CO₂ from ordinary air means processing a gas where CO₂ makes up only around 0.04%.

DAC still has some unusual advantages. A plant can theoretically be located near cheap clean energy and suitable geological storage. Operators can measure captured CO₂ relatively directly, and geological storage can keep it underground for thousands of years.

Climeworks' Mammoth plant produced 675 tonnes of net carbon removal during the first half of 2026, up from 119 tonnes over the same period a year earlier. Its upgraded collector units are currently reaching capacity factors of roughly 40% to 50%. Climeworks also says laboratory work has increased sorbent lifetime tenfold and packed roughly four times more process capacity into the same module size.

Those are real engineering gains. Mammoth was designed around a nameplate capacity of up to 36,000 tonnes annually, while actual net removal during its ramp-up remains far below that figure.

A DAC plant also consumes electricity and heat, uses materials, requires construction and eventually compresses and transports CO₂. The useful climate number is what remains after those emissions are deducted.

Energy sourcing adds another constraint. A DAC plant supplied by abundant geothermal power in Iceland has a different footprint from one competing for scarce clean electricity in a grid that still burns fossil fuels.

Companies such as Phlair are therefore developing electrochemical systems designed to operate flexibly when renewable electricity is cheap. Others are testing passive contactors, new sorbents and lower-temperature processes.

Whether DAC becomes efficient enough to remove hundreds of millions or billions of tonnes every year remains wide open.

Can we actually measure carbon removal well enough to know what one tonne is worth?

Carbon removal measurement is getting stricter, but one tonne of removal can still represent very different climate outcomes depending on how it is measured, stored and protected from reversal.

With direct air capture and geological storage, operators can measure concentrated CO₂ leaving a plant, meter what enters a storage site and subtract lifecycle emissions. The physical carbon flow is relatively visible.

Enhanced rock weathering is harder. A company can weigh every tonne of basalt spread on a farm, yet the amount of CO₂ ultimately removed depends on mineral composition, particle size, rainfall, temperature, soil chemistry and water flows.

Ocean removal pushes the problem further. Changing ocean alkalinity can cause additional atmospheric CO₂ to enter seawater, but that exchange happens across moving water masses. Sensors can measure local chemistry, while models are still needed to estimate how much atmospheric CO₂ ultimately entered and remained stored.

The more uncertain a pathway is, the more monitoring we want. Field sampling, laboratory analysis, sensors and scientific modeling can then become expensive enough to damage the economics of the removal itself.

Durability adds another layer. A tonne injected into a well-characterized geological reservoir can plausibly remain underground for thousands of years. Carbon held in a forest may return to the atmosphere after a fire, drought, disease outbreak or future harvest. Biochar usually sits somewhere between those extremes.

Standards are becoming more explicit about these differences. Puro.earth's latest biochar methodology moved eligible biochar into a 200-plus-year durability category after reviewing newer persistence research. The EU has also created dedicated permanent-removal methodologies for DACCS, BioCCS and biochar covering net removal, lifecycle emissions, monitoring, additionality, sustainability and long-term liability.

That makes the simple "one tonne equals one tonne" model increasingly difficult to defend. If two projects each claim 100,000 tonnes but one has direct measurement and millennial storage while the other relies heavily on modeling and stores carbon for perhaps a century, buyers are not receiving the same thing.

We still have not found a universal way to price those differences without making verification too costly.

What changes the value of a removal	Why it matters
Measurement	Determines how sure we are the tonne was actually removed
Additionality	Tests whether the removal would have happened anyway
Durability	Storage can range from decades to millennia
Reversal risk	Some storage can burn, decay or otherwise return CO₂
Lifecycle emissions	Energy, transport and materials reduce net removal

Can forests, soils and biomass provide most of the carbon removal we need?

Forests, soils and biomass can provide enormous carbon removal, but relying on them for most additional future removal would put too much pressure on land, feedstocks and storage that can reverse.

Their existing contribution is already huge. Conventional land-based removal accounts for roughly 2 GtCO₂ a year, vastly more than novel carbon removal.

Biological methods also bring benefits that engineered technologies often cannot match. Reforestation can improve habitat, reduce erosion and restore watersheds. Better soil management can improve soil health and agricultural resilience.

The limitations appear when we try to multiply these methods indefinitely. New forests compete with agriculture, settlements, ecosystem restoration and biomass production. Forest carbon remains exposed to fire, drought, pests, logging and future land-use changes, while accumulation slows as forests mature.

Soils face similar limits. Farmers can build soil organic carbon through better management, but gains vary with climate, soil and practice, are difficult to measure across huge areas and can disappear when land management changes.

Biomass-based engineered removal solves some of the permanence problem by moving biological carbon into biochar or geological storage, and it is currently one of the strongest commercial parts of durable CDR. CDR.fyi's latest market data show biomass-based methods taking seven of the ten largest supplier positions by sales during the first quarter of 2026.

Projects can also exploit infrastructure that already exists. Pulp mills, biomass boilers and waste facilities already gather biomass and often produce concentrated biogenic CO₂.

The difficulty returns at larger scale because residues are not automatically free resources. Crop residues can maintain soil nutrients. Forest residues may already have industrial uses. Waste biomass can generate heat or electricity. Rising carbon-removal demand can eventually change harvesting behavior and feedstock prices.

The EU's latest BioCCS methodology reflects this concern by requiring biogenic CO₂ to arise as a by-product of producing goods, energy or services rather than from facilities burning biomass solely to generate CO₂ for storage.

Biological carbon removal will remain essential, but sustainable land and biomass supply will put a ceiling on how far we can push it.

Is biochar actually the carbon removal technology working best today?

Biochar is currently the strongest proof that durable carbon removal can move beyond pilots and deliver meaningful commercial volumes.

It has several advantages at once. Pyrolysis equipment is modular. Biomass wastes exist across many regions. Carbon content can be measured. The resulting biochar can be used in soils, construction materials and other applications.

The delivery numbers show the difference. CDR.fyi's rankings currently place biochar suppliers Exomad Green, Varaha and Carboneers among the largest durable CDR suppliers by tonnes physically delivered. Exomad alone has delivered more than 400,000 tonnes according to the platform.

Puro.earth has also certified more than 70 biochar-based removal projects, and its newer methodology now assigns qualifying biochar storage a 200-plus-year durability level.

Biochar benefits from another practical advantage: buyers do not always have to wait a decade. Operating pyrolysis projects can issue credits from production happening today.

Hundreds of thousands of delivered tonnes is impressive inside the current durable CDR market, but global climate needs are measured in billions. Feedstock availability will eventually tighten, production has to control methane and other emissions, and end uses must keep the biochar in stable storage.

Biochar has crossed a threshold that many competing technologies have not: we can increasingly judge it on delivered carbon rather than mainly on engineering projections.

If you want more recent data on this point, please see our latest CCUS market report.

Can enhanced rock weathering really reach billions of tonnes?

Enhanced rock weathering has a believable route to enormous scale, but today we are still proving how accurately and cheaply that weathering can be turned into verified carbon removal.

The physical idea fits existing industries unusually well. Mines already extract and crush huge quantities of rock. Farmers already spread minerals such as lime across enormous areas. Enhanced weathering adds finely ground alkaline minerals, often basalt, to soils so natural chemical reactions consume CO₂ more quickly.

Frontier currently estimates potential capacity of around 1–2 GtCO₂ a year for the pathway. It has contracted roughly 333,000 tonnes from eight companies, with an average offtake price around $362 per tonne.

The future capacity looks huge, yet contracted carbon removal is still a tiny fraction of one gigaton.

Weathering rates change with mineralogy, grain size, soil acidity, rainfall, temperature and agricultural conditions. Grinding the rock consumes energy. Transporting heavy low-value material over long distances can ruin project economics. Some rocks also contain metals that require careful monitoring.

Verification is especially difficult because researchers need to separate enhanced weathering from natural changes already happening in soils and water.

Farmers will also care about crop performance, soil effects, logistics and cost before atmospheric carbon accounting. The strongest projects may therefore be those where the rock provides a useful agronomic service too.

Enhanced weathering has enough raw material and existing infrastructure to become much larger than it is today. Cheap, reliable measurement remains the main hurdle.

Could ocean carbon removal become one of the biggest carbon removal methods?

Ocean carbon removal could eventually become enormous, but today its verification and ecological questions remain too unsettled for us to assume that scale will happen.

Oceans already hold vastly more carbon than the atmosphere and continuously exchange CO₂ with the air. Increasing seawater alkalinity can let the ocean absorb additional atmospheric CO₂ and keep much of that carbon in bicarbonate form for very long periods.

Frontier describes the potential capacity of the marine pathways it evaluates as functionally unlimited at the scale relevant to CDR. Its portfolio now covers around 249,000 contracted tonnes across 16 companies.

Current prices show how experimental the field still is. Frontier's marine contracts range from roughly $238 to $1,961 per tonne.

Companies are trying very different deployment models. Some add alkaline minerals to seawater. Others use electrochemical systems. CREW Carbon works inside wastewater-treatment infrastructure, where water chemistry can be controlled and measured before discharge. Planetary is developing ocean alkalinity projects linked to existing coastal infrastructure.

The open ocean is where verification gets difficult. Altered water moves, air-sea CO₂ exchange can take time, and biological activity changes local chemistry. A reading beside the discharge point cannot by itself tell us exactly how much atmospheric carbon entered the ocean later.

Environmental confidence also has to survive a huge increase in volume. A successful pilot says much less about what happens if alkalinity enhancement becomes a billion-tonne industry.

Research these days is therefore heavily focused on monitoring, models, dissolution rates and ecological effects. Ocean removal has enormous upside, but its safe practical ceiling is still unknown.

If you want more recent data on this point, please see our latest CCUS market report.

Do we have enough underground space to store billions of tonnes of CO₂?

The world probably has plenty of geological storage capacity, while the infrastructure needed to access that storage remains nowhere near climate scale.

Deep saline formations and other suitable geological reservoirs can collectively hold enormous quantities of CO₂. Physical pore space is therefore unlikely to stop carbon removal first.

A usable storage site needs much more than rock. Developers have to characterize the reservoir, prove containment, obtain permits, drill injection wells, build compression equipment, connect pipelines or shipping terminals and monitor the site for years.

Carbon removal will also share this infrastructure with conventional carbon capture. Cement plants, steel facilities, hydrogen projects, waste incinerators, refineries and power projects may all want access to the same transport networks and reservoirs.

The latest IEA data show that CCUS investment reached more than $5 billion in 2025, more than fifteen times its 2020 level. More than 30 projects reached final investment decisions over the previous two years, and projects already under construction could nearly double operational capture capacity by 2030.

That is one of the stronger infrastructure trends in carbon management right now, but a doubling from today's relatively small base still falls far short of a gigaton-scale CO₂ network.

Shared hubs may help. Several European projects are building transport and storage networks designed to accept CO₂ from multiple industrial sources rather than forcing every capture project to construct its own reservoir.

The bottleneck is increasingly development speed: permits, wells, pipelines, financing and long-term responsibility for stored CO₂.

Who will pay for carbon removal when Microsoft stops being enough?

Carbon removal still depends far too heavily on a small group of voluntary corporate buyers, and gigaton scale will require governments or regulated emitters to become much larger customers.

Buyers including Microsoft, Google, Stripe, Shopify and Meta have funded technologies years before normal customers would have touched them.

Microsoft in particular has become extraordinarily important. The company disclosed almost 22 million tonnes of carbon-removal contracts in its FY2024 sustainability reporting alone, and it has kept signing large agreements since then.

That concentration creates obvious fragility. A climate strategy requiring billions of tonnes annually cannot depend on a few technology companies voluntarily spending increasingly large amounts of money.

Permanently putting a tonne of CO₂ underground also does not usually create a consumer product worth hundreds of dollars. Some pathways produce electricity, biochar, construction materials or waste-management value, but those side revenues rarely cover the entire cost of removal.

Policy is beginning to move closer to the market. The EU has now established official certification methodologies for DACCS, BioCCS and biochar under its Carbon Removals and Carbon Farming framework. Those rules could eventually support procurement or compliance markets.

A common certification system gives governments a basis for buying removals, setting obligations or allowing certified CDR into regulated climate policy.

Voluntary buyers have kept the industry alive. They cannot carry it to several gigatonnes.

If you want more recent data on this point, please see our latest CCUS market report.

Could carbon removal cause more environmental damage than it fixes?

Some carbon removal projects can create serious local damage, and those risks become much harder to dismiss once deployment reaches millions or billions of tonnes.

DAC uses energy, equipment, sorbents and sometimes water. Enhanced weathering mines, crushes and transports rock. Biochar and BioCCS consume biomass. Afforestation changes land use. Ocean alkalinity changes seawater chemistry and requires minerals or chemicals.

At small pilot scale, many of these pressures look manageable. Gigaton scale changes the arithmetic.

Depending on the mineral and conditions, enhanced weathering can require several tonnes of rock for each tonne of CO₂ removed. A pathway removing hundreds of millions of tonnes of CO₂ could therefore create material flows measured in billions of tonnes, putting it in the same physical category as a major mining industry.

Biomass has the same issue in a different form. Large carbon-removal demand could compete for residues, timber, agricultural land and transport capacity.

Regulators are already responding. The EU's permanent-removal methodology limits certified biochar application on agricultural and forest soils to a cumulative 50 tonnes per hectare and requires projects to consider effects on productivity, soil health, biodiversity and downstream ecosystems.

Ocean removal has an even higher evidence burden because deployment changes shared marine environments. Large-scale intervention will need much stronger evidence than a successful local experiment.

Theoretical capacity tells us how far a technology could go chemically or geologically. Its sustainable capacity may be much lower.

Could carbon removal become an excuse to keep emitting CO₂?

Yes. Weak net-zero rules can easily turn carbon removal into a way of delaying emissions cuts that are cheaper and easier to make directly.

Climate models still need carbon removal. Agriculture, aviation, shipping and some industrial processes are unlikely to eliminate every last tonne of emissions, and net zero requires balancing whatever genuinely remains.

The problem lies in deciding what counts as "genuinely remains."

Almost every industry would prefer its own emissions to be classified as difficult to eliminate. If companies can buy removal instead of making uncomfortable changes, the residual category can quietly expand until it includes emissions that were perfectly avoidable.

Spending several hundred dollars on DAC to cancel a tonne of CO₂ that could have been prevented much more cheaply makes little sense. Future removal also carries execution risk: a company can emit fossil CO₂ today and buy a promise for removal years from now, while the atmosphere receives the emission immediately.

Climate overshoot makes the stakes larger. If global temperatures rise beyond the intended limit, net-negative emissions may eventually be needed to bring atmospheric CO₂ back down.

Carbon removal works best when direct emissions cuts come first and durable removal is reserved for genuinely stubborn emissions and eventual net-negative needs.

The unresolved part is governance. We still lack a universally credible line between residual emissions and emissions that companies simply prefer not to eliminate.

If you want more recent data on this point, please see our latest CCUS market report.

Which carbon removal technologies are actually ahead right now?

Biochar and other biomass-based approaches are ahead in commercial delivery today, while enhanced weathering, DAC and ocean removal are still competing for much larger future roles.

Biochar is ahead on delivered durable carbon. It already has many operating suppliers, modular production and a relatively broad buyer base.

Biomass with geological storage looks strongest for very large individual contracts. Existing pulp mills, waste plants and biomass facilities can provide concentrated streams of biogenic CO₂, letting projects move toward hundreds of thousands or millions of tonnes.

Enhanced weathering has one of the more believable routes into existing physical infrastructure. Its biggest remaining weakness sits in measurement rather than access to raw materials.

DAC offers cleaner accounting and highly durable storage, but today's operating plants remain expensive and small. The latest Mammoth results show rapid improvement alongside the huge distance still left to travel.

Ocean carbon removal may ultimately have the largest physical storage opportunity of the group, yet it carries the greatest unanswered questions around monitoring and ecosystem effects.

Near-term volumes will probably keep leaning toward biochar and well-sited biomass projects. Enhanced weathering has a credible route into much larger regional deployment if verification improves. DAC could become strategically valuable where clean energy and geological storage are abundant. Ocean approaches remain the highest-upside scientific wildcard.

So what are the biggest unsolved problems in carbon removal?

Carbon removal looks technically credible today, but we still have no proof that the world can deliver several gigatonnes of high-quality removal every year at a price people will actually pay.

The biggest uncertainty is scale. Novel CDR has reached roughly 2 million tonnes annually and is growing fast, yet future climate needs are measured in billions.

Cost comes next. Several hundred dollars per tonne can support an early market funded by wealthy corporate buyers. It becomes painful when multiplied by billions of tonnes every year.

Measurement remains unresolved for pathways such as enhanced weathering and ocean removal, while permanence creates very different levels of climate value across forests, biochar, minerals and geological storage.

Resources become harder too. Biomass, clean electricity, rock, land, pipelines, storage wells and suitable project sites all look abundant when the industry is small. Carbon removal needs to prove that they remain available once thousands of projects compete for them.

Demand may be the most underrated constraint. A handful of companies has successfully created the first market, but voluntary purchasing will never be enough. Governments and regulated emitters eventually need to finance removal at a completely different scale.

There is still plenty of reason to build. Novel CDR is growing quickly. Biochar has moved into real commercial delivery. Biomass-based projects are signing million-tonne contracts. DAC performance is improving. Standards are getting stricter. Governments are beginning to create common rules.

Our judgment is that carbon removal has moved beyond the stage where its main question is whether the technologies work. Several clearly do. The harder test is whether they still work economically, environmentally and politically when the industry becomes a thousand times larger.

That is the problem carbon removal has to solve now.

If you want more recent data on this point, please see our latest CCUS market report.

Who is the author of this content?

NEW MARKET PITCH TEAM

We track new markets so founders and investors can move faster

We build living "market pitch" documents for emerging markets: AI, synthetic biology, new proteins, and more. Instead of outdated PDFs or hallucinated LLM answers, our clients get a clean, visual, always-updated view of what's really happening: key players, deals, regulations, and signals that matter. Learn more about us.