Two Clocks In Carbon Capture

A cement kiln vents its flue gas at 12 to 30 per cent carbon dioxide by volume. Four hundred metres away, the ambient air holds 420 parts per million, which is 0.042 per cent. Same molecule, same atmosphere, a concentration 300 to 700 times lower. That gap is not a detail. It is the business. Thermodynamics puts a floor under each job. Lifting carbon dioxide out of a rich stream costs a minimum of about 0.25 megajoules per kilogram. Pulling it out of the sky costs about 0.45. Those are floors, not prices, and real plants run 10 to 20 times above them. Point-source capture began commercial life more than 40 years ago and sits at technology readiness level 9. Direct air capture sits at level 7 and has been shipping hardware for about five years. I have deployed 210 energy systems across Africa and Asia, and every one of them was priced on the same rule. The richer the input, the cheaper the output. The probability that one cost curve describes both technologies is not zero. Most of the market is pricing it at one.

Amine scrubbing is the incumbent and it works. Flue gas is bubbled through a solvent such as MEA, MDEA or KS-1 in an absorber column, and the loaded solvent is regenerated in a stripper with steam at about 120 degrees Celsius. Capture efficiency runs 85 to 95 per cent. Cost runs 50 to 74 dollars per ton of carbon dioxide at a coal plant and 70 to 100 dollars at a gas plant. Capital runs about 400 dollars per ton of annual capacity. Vendors in the market include Shell CANSOLV, Mitsubishi Heavy Industries, BASF with OASE blue and Aker Carbon Capture. The penalty is the part nobody prints on the front page. Regeneration eats 3.0 to 3.7 gigajoules per ton, which takes 15 to 45 per cent off the efficiency of the plant it is bolted to. Solvent is lost at 0.5 to 3.0 kilograms of MEA per ton captured and has to be replaced. In 2024 Thunder Said Energy revised the honest number upward to 80 to 130 dollars per ton, against the 40 dollars that 2019 studies promised. Project budgets did the revising, not models. An efficiency penalty is charged against the output you are trying to sell.

Direct air capture has two working families and both of them are energy businesses wearing a chemistry coat. Liquid solvent, the route Carbon Engineering took before Occidental bought it, contacts air with potassium or sodium hydroxide, captures the carbon as a carbonate, then calcines it at 800 to 900 degrees Celsius. Heat demand is 6 to 9 gigajoules per ton, electricity is 0.5 to 1.0 megawatt hours per ton, and a first-of-a-kind plant at 10,000 tons a year runs 700 to 1,000 dollars per ton. The Stratos project in Texas, 500,000 tons a year for Occidental, is quoted at 180 dollars per ton. Solid sorbent, the route Climeworks took, uses amine-functionalised sorbents on contactors and regenerates at 80 to 120 degrees Celsius. Heat demand is 3 to 5 gigajoules per ton, electricity is 0.6 to 1.0 megawatt hours, and current credits clear between 600 and 1,000 dollars per ton. Orca in Iceland captures 4,000 tons a year, and Mammoth beside it is designed for 36,000. Add the thermal and the electric demand together and direct air capture spends 1.5 to 2.5 megawatt hours on every ton. Removing 1 gigaton a year would take 1,500 to 2,500 terawatt hours a year, which is 5 to 8 per cent of global electricity, and the wind, solar and battery build for it has been estimated at 4 trillion dollars. A stack is a river. The atmosphere is fog, and the river is the only place where the water is already in a channel.

Policy has already priced the two jobs differently, and the spread is the clearest number in the file. In the United States, section 45Q pays 180 dollars per ton for carbon removed by direct air capture and 85 dollars per ton for carbon avoided at a point source. A factor of 2.1, written into a tax code. The Stratos plant is quoted at 180 dollars per ton. That is the credit, not the cost curve. A plant sized to its subsidy has already decided who its customer is. Europe writes the same distinction through different doors. The EU Innovation Fund and the German BMWK programmes fund both, but the eligibility line moves with the readiness level. Amine scrubbing at level 9 is bankable and boring, which is why it is crowded. Solid sorbents at level 7 or 8 are the innovation sweet spot for industrial grant money. Direct air capture at level 7 is fundable, but only with a subsidy structure underneath it, and the project economics only work on paper at a size nobody has built yet. When the credit is 180 dollars and the cost is 180 dollars, you have not built a business. You have built a mirror.

That gap is not a detail. It is the business.

Grant money in this sector is not a reward for being early. It is a payment for carrying risk that the market will not carry at the price the market wants. A technology at readiness level 9 with 40 years of operating history does not need a grant to exist. It needs a customer. A technology at level 7 with two commercial plants and a cost between 500 and 1,000 dollars per ton needs somebody to underwrite the gap between the first plant and the hundredth, and that is the job a programme is for. The bind is arithmetic. A biochar composite sorbent at readiness level 3 to 5, with a capacity of 0.5 to 2.5 millimoles per gram, is not competitive with an amine at 50 to 74 dollars per ton, and no grant changes that. Membranes at level 6 to 7 with a selectivity of 30 to 60 sit in the same place. What a grant buys is time to close a cost gap, and the only honest question left is how many doublings it takes. Sixteen years of field work, from rural Nigeria to post-conflict Serbia to Fukushima, taught me the same rule every time. The money follows the risk that nobody else will hold.

There are two clocks in this problem and they do not tick at the same rate. The first clock is the tap. Every ton of carbon dioxide leaving a stack today is a new ton, and point-source capture is the only instrument that stops it at the flange. That is the brake. The second clock is the bathtub. The carbon already in the atmosphere does not care when it was emitted, and direct air capture is the only instrument that takes it back out. That is the mop. A brake and a mop are not substitutes, and the market keeps pricing them as one product with one cost curve. Point-source capture is 5 to 20 times cheaper per ton and it does a different job at the same time. Confusing the two is how a project ends up applying a 180 dollar credit to a 60 dollar problem, or pointing a 60 dollar budget at a 900 dollar one. You cannot mop a floor while the tap is still running. Both instruments are needed, both are priced, and neither one removes the need to stop emitting in the first place. The probability that a portfolio of the two resolves into a single number a board can underwrite is not zero. Every operator I have met is pricing it at zero.