Every explainer settles carbon removal vs carbon capture the same way, and the answer is correct. Carbon capture takes CO2 at the point of emission, in the stack of a cement plant or a power station, before it reaches the air. Carbon removal takes CO2 that is already in the atmosphere, which makes capture a form of emissions reduction and removal a separate thing entirely.
That definition will get you through a dinner party. It will not get you through a term sheet, because it does not tell you the thing that decides whether a project is an asset or a liability: who pays for it.
The molecule is identical. The separation equipment is often identical. The business is not. One of these categories produces a certificate that corporations queue up to buy at several hundred dollars a tonne. The other mostly cannot issue that certificate at all, and earns its money from tax credits, compliance obligations and, frequently, from selling the CO2 to an oil producer. Everything else in this comparison follows from that split.
Where the molecule came from sets the cost floor
Start with the physics, because it explains the prices later.
A point source is a rich stream. CO2 makes up 4 to 30 percent of the flue gas leaving an industrial plant, against 0.04 percent of the ambient atmosphere. You are separating the same molecule in both cases, but in one you are sorting a crowded room and in the other you are searching an empty stadium. No amount of engineering repeals that ratio.
The cost consequence is direct. The IEA range for capture runs from roughly 15 to 25 dollars per tonne on concentrated industrial streams to 40 to 120 dollars on dilute ones, before transport and storage. Removal spans a much wider band: natural climate solutions can start near zero dollars per tonne, BECCS runs 30 to 400 dollars and direct air capture 100 to 1000 dollars, while post-combustion capture rates sit at 85 to 90 percent.
Hold on to one caveat. Those are engineering costs, what it takes to do the job. They are not prices, which is what somebody agrees to pay. The gap between the two is where the actual businesses live, and the two categories sit on opposite sides of it.
The confusing middle: capture hardware that produces removal
This is where most readers get lost, and it is fair to get lost here, because the naming is genuinely bad.
Direct air capture has the word capture in it and is a removal technology. BECCS, bioenergy with carbon capture and storage, has the word capture in it twice and is a removal technology. Both use the same amine-and-sorbent equipment a coal plant would use. As the Grantham Institute puts it, BECCS and DACCS use carbon capture and storage technology to deliver net negative emissions, yet applying that same technology to fossil-fuel emissions can never result in removal.
So do not classify by equipment. Classify by carbon provenance, and follow the carbon backwards.
If it was in the atmosphere twelve months ago and a plant pulled it down, capturing it now is removal. If it has been underground for two hundred million years and you are intercepting it on its way up, that is capture, no matter how good the separator is. The US Department of Energy makes the same cut and adds two conditions worth memorising: removal must be an intentional intervention rather than the natural cycle carrying on, and CO2 conversion into a product only counts as removal if the carbon came from the air and the product stores it for a long duration. CO2 turned into fuel and burned within a year is not removal. Existing trees doing what trees do are not removal either.
Different business: who actually pays for each one
Here is the part the ranking pages leave out, and it is the whole commercial answer.
Capture is paid by tax credits, compliance rules and the oil barrel
A capture unit on an emitter's site is a cost centre. It makes the plant more expensive to run and produces nothing the plant can sell, so its revenue has to come from somewhere other than the CO2 itself.
Usually it cannot come from carbon credits. Double counting is prohibited across every serious registry, which means that when the emitter or the host government has already claimed the emissions reduction, nobody can issue a credit for it. The practical result is that most CCUS projects are ineligible for voluntary market credit issuance, and their financial viability comes instead from CO2 utilisation, usually enhanced oil recovery, from government subsidies, or from compliance carbon pricing.
Even the subsidy is written with the provenance rule baked in. Under the US 45Q regime, the tax credit pays different rates depending on whether the CO2 came from a facility or from the atmosphere, and eligibility for California's Low Carbon Fuel Standard requires the capture to be embedded in a fuel production pathway, so a cement plant with capture does not qualify while any direct air capture project with storage does. Two projects, identical vessels, different tax treatment, because the accountants are asking where the carbon came from.
Removal is paid by companies buying a product
Removal inverts the model. The tonne is the product. Nobody obliges a corporate buyer to purchase it, and that is precisely why it commands a premium: buyers are paying voluntarily for a claim they cannot make any other way. Roughly 6.4 billion dollars of carbon dioxide removal has already been contracted globally, representing about 28 million tonnes once executed, most of it still in development.
Stacking is where the two models finally meet. A BECCS project sits in both worlds: biogenic carbon, so it qualifies as removal, plus capture equipment eligible for capture subsidy. Developers stack them, running the 85 dollar per tonne 45Q credit alongside renewable energy credits and removal credits, with BECCS credits averaging 387.49 dollars per tonne and 60 percent of all CDR volume transacted to date, while direct air capture trades at around three times the price in deals averaging just 9,318 tonnes. Microsoft alone accounts for 95 percent of BECCS sales.
Read that last clause twice before building a model on it. This is not a market. It is one customer with a procurement budget.
The price gap, and why it is not closing
Put the two revenue models side by side and the spread is startling.
Compliance allowances, the price an emitter pays for the right to emit, sit at around 84 dollars per tonne in the EU Emissions Trading System and 66 dollars in the UK system, against mean UK credit prices of roughly 317 dollars per tonne for BECCS and 908 dollars for DACCS, with only 1.2 percent of contracted durable removal actually delivered by the end of 2025. The same analysis finds learning rates for most removal methods to be low or even negative, which points at structural bottlenecks rather than an immature technology working its way down a cost curve. Convergence is not scheduled.
California tells the same story from the compliance side, where allowances trade around 30 dollars per tonne and offsets are capped at 4 percent of a polluter's obligation, rising to 6 percent from 2026. A buyer paying ten times the allowance price for a removal tonne is not being rational about compliance. They are buying something compliance does not sell. The words credit and offset are not synonyms either, and the difference matters as much as this one, which our sister company Energaia unpacks in carbon credits vs carbon offsets.
For an investor that reframes the question. The removal premium is not the market pricing superior technology. It is a small pool of voluntary buyers paying for permanence and reputation in the absence of any regulation requiring it. Which pathways are converting that premium into delivered tonnes rather than signed contracts is a separate exercise, and we run it in which removal pathways are actually delivering tonnes.
Both categories are still rounding errors
Scale keeps the enthusiasm honest on both sides.
CCUS today captures around 0.1 percent of global emissions, roughly 50 million metric tons of CO2, while IPCC and IEA scenarios consistent with 1.5 degrees call for about 1 billion tons by 2030 and several billion by 2050. Set that against human CO2 emissions exceeding 40 billion tonnes in 2022, with engineered removal running at roughly 2 million tonnes a year.
Neither category is displacing an incumbent. Both are bets on policy and on procurement, which is a specific kind of risk and should be underwritten as one.
What to ask before funding either one
Five questions, in the order that resolves deals fastest.
Where was this carbon twelve months ago? That single answer determines the category, the eligible subsidy and whether a credit can exist at all.
Who is the counterparty, and is the revenue a tax credit, an allowance, a fuel-standard credit or a voluntary purchase? Each has a different duration and a different political half-life.
Is anyone else claiming this tonne? If the host government counts it in a national inventory, the voluntary credit is not there to sell, however good the project is.
How much has been delivered, not contracted? Signed volume is a forecast. Retired tonnes in a registry are a fact, and the ratio between them is the most honest single number about any supplier.
And who owns the process? A removal method that anyone can copy from a permit filing is a service business with a climate story attached, not an asset. That distinction is the same one that separates a good deep-tech deal from a good deep-tech press release, and it is worth applying with the same rigour as how to evaluate a deep tech startup.
We write this from the deployment side rather than the commentary side. EX EPIC finances, patents and deploys deep tech across four continents on a EUR 160M+ capital track record, with Zero-X having designed, financed and installed 200+ waste-to-energy units across 11 countries, EX Carbon as the portfolio's biochar carbon removal venture, and EX IX holding 24 patent families filed with roughly 100 more validated in the pipeline. What that experience teaches is unglamorous: the projects that survive are the ones where the CO2 economics ride on top of a machine that was already worth building for another reason, which is the logic behind the waste-to-energy business model. Getting a validated mechanism from a paper to a patented, financed, operating asset is a sequence, and running that sequence is what a deep-tech venture builder actually does.
FAQ
Is direct air capture carbon capture or carbon removal? Both, depending on which question you are asking. Mechanically it is capture equipment, the same class of sorbent and separation hardware used on a smokestack. By accounting it is removal, because the CO2 came from the atmosphere and is durably stored. The word capture sitting inside the name is the single largest source of confusion in this whole category.
Does carbon capture count as negative emissions? No. Capture on a fossil source prevents an emission that would otherwise have occurred, which is emissions reduction. Negative emissions require taking CO2 that is already in the air and storing it durably, so a well-run capture project moves the total towards zero and never past it.
Can a carbon capture project sell carbon credits? Usually not on the voluntary market, because double counting is prohibited and the emitter or the national government has typically already claimed the reduction. The credit cannot be issued twice for the same tonne. Capture projects monetise through tax credits, compliance schemes and CO2 utilisation instead.
Is CCUS the same as CCS? CCUS adds utilisation: the CO2 goes into a product rather than only into geology. Whether that has climate value depends entirely on how long the product holds the carbon, which is why CO2 made into fuels counts for very little and CO2 mineralised into building materials counts for considerably more.
Does planting trees count as carbon removal? Only where it is an intentional intervention that adds removal beyond what was already happening, with durable storage behind it. Forests continuing to do what they have always done are part of the natural carbon cycle rather than a credited removal, which is also why reversal risk from fire or land-use change is priced into forestry credits.