Unbiased

Carbon Capture's Real Fight Is Cement's Chemistry Against the Oil Field

Climate models cannot close their books without it, most of the CO2 captured today is sold to push more oil out of the ground, and named engineers stake careers on both facts.

9 min read2026-07-14

Capture is one process pointed at two different jobs

Make a tonne of Portland cement and most of the carbon dioxide never touches the fuel. About 40% of it comes from burning coal or gas to heat the kiln; the remaining roughly 60% is released when limestone (calcium carbonate) is cooked into lime, a reaction that breaks CO2 out of the rock itself.¹ Switch the kiln to clean electricity and that second fraction does not move. It is the clearest instance of what engineers call a process emission, and it is the reason carbon capture is a serious climate proposal rather than a curiosity: some carbon dioxide has no combustion to electrify away.

The apparatus itself is mundane. A chemical solvent or a solid sorbent binds CO2, either from a smokestack's flue gas or, in direct air capture, from the open atmosphere; heat then drives the gas back off the medium in concentrated form; the stream is compressed and sent somewhere. Where it is sent is the whole argument. One destination is a saline formation or a depleted reservoir a mile underground, where the gas is meant to stay for good. The other is a working oil field, where the CO2 is injected to push additional crude to the surface, a practice called enhanced oil recovery. The chemistry of capture is the same in both cases. The climate accounting is not.

Two questions travel together here and are constantly mistaken for one another. The first is whether capture works: whether a plant removes the tonnes it promises at a price someone will pay. The second is what it is pointed at: whether those tonnes come from emissions that truly have no cheaper route to zero, or from fossil operations that would be better shut than scrubbed. Almost every live dispute lives in the second question. The IPCC's Sixth Assessment finds carbon capture and carbon dioxide removal inside essentially all scenarios that hold warming near 1.5°C.² The IEA, in its 2020 assessment, had carbon capture delivering close to 15% of the cumulative emissions cuts its central decarbonization scenario needed, and singled out cement's process emissions as "hard or, in the case of process emissions in cement, currently impossible to abate in other ways."³ Against that institutional confidence sits an operating record of built plants that is patchy at best, and a commercial reality in which much of what gets captured is routed to producing more oil. Both pictures are real.

Net-zero modelers and hard-to-abate engineers

Some emissions, cement's above all, have no cheaper route to zero, so the pathways that actually reach net zero all keep carbon capture; removing it makes the target unreachable, not merely dearer.

The case begins with a claim about cost, not virtue. Julio Friedmann, a carbon-management scientist who ran the U.S. Department of Energy's fossil-energy office, put it to the Senate plainly: "Our research shows these technologies often present the lowest-cost, most actionable pathway for profound emissions reduction and that in some sectors they are the only pathway."⁴ The logic runs through the arithmetic of net zero. If any emissions remain once every cheaper cut is made, balancing them requires an equal mass of removal, and those leftover emissions are "hard-to-abate," meaning "either the cost is extremely high (e.g., for aviation) or the technology does not exist."⁴

Cement is the worked example. Because roughly 60% of its CO2 comes from calcining limestone rather than from fuel, "eliminating the process emissions requires fundamental changes to the cement-making process and cement materials and/or installation of carbon-capture technology," as the 29-author Science review of net-zero energy systems states it.¹ The Energy Transitions Commission, chaired by Adair Turner, estimates that the harder-to-abate sectors, about 30% of energy CO2, can be fully decarbonized, with capture one of the necessary routes and cement the costliest because of those process emissions, at a cost it puts at US$130 or less per tonne for cement and US$60 or less for steel.⁵ This is why the IPCC calls capture "a critical mitigation option" precisely in cement and chemicals, where it is least mature.² The argument is not that capture should power the grid; it is that a handful of industrial reactions cannot be run any other way, and a pathway that pretends otherwise does not add up.

Julio Friedmann, Columbia University, standing here as a Senior Research Scholar at the Center on Global Energy Policy who served as the DOE's Principal Deputy Assistant Secretary for Fossil Energy and spent fifteen years at Lawrence Livermore, testifying under his own name.⁴

Energy-system researchers and project post-mortems

The captured tonne is mostly a subsidy for the fossil economy: real plants under-deliver, most of their CO2 is pumped into oil fields, and the same dollar would cut more carbon spent on wind and solar instead.

The sharpest quantitative form of the critique is Mark Z. Jacobson's. His peer-reviewed lifecycle model finds that a gas-powered capture plant nets only 10.8% of its CO2-equivalent emissions captured over twenty years, and a synthetic direct-air-capture plant 10.5%, once the uncaptured emissions from the natural gas that runs the equipment and from upstream leakage are counted, even though the smokestack device itself captures 55.4%.⁶ His conclusion is categorical: "No improvement in CCU or SDACCU equipment can change this conclusion while fossil fuel emissions exist, since carbon capture always incurs an equipment cost never incurred by wind."⁶

The second charge is about what capture is for. Charles Harvey, an MIT environmental engineer who co-founded the first privately funded U.S. carbon-capture company, wrote with Kurt House that of the twelve commercial capture projects operating in 2021, more than 90% fed enhanced oil recovery, and that "every dollar invested in renewable energy, instead of C.C.S. power, will eliminate far more carbon emissions." By promoting capture, they argue, "the fossil fuel industry is slowing the transition away from fossil fuels."⁷ Chevron's Gorgon project, required to capture 80% of its reservoir CO2, has managed 44% since 2019 and just 30% in its latest year, at a cost that rose from an estimated $70 a tonne toward $222.⁸ Saskatchewan's Boundary Dam 3, a billion-dollar-plus coal retrofit promised at 90% capture, has run closer to 57% over its life, with every captured tonne sold for oil recovery.⁹ Petra Nova captured 662,000 tonnes below target in its first three years and was mothballed in 2020.¹⁰ Reviewing thirteen flagship projects that hold about 55% of the world's nominal capture capacity, IEEFA's Bruce Robertson found failures outnumbering successes and captured carbon "mostly used for enhanced oil recovery: enhancing oil production is not a climate solution."¹¹

Mark Z. Jacobson, Stanford University, standing here as a professor of civil and environmental engineering whose peer-reviewed lifecycle accounting quantifies net capture and its opportunity cost against wind and solar.⁶

It narrows to cement's price and the dollar's best use

The two camps are largely indicting different machines. The skeptics' hardest evidence, Gorgon, Boundary Dam, and Petra Nova, is drawn from gas-processing and coal-power retrofits whose CO2 goes to enhanced oil recovery.⁸⁹¹⁰ The necessity case rests almost entirely on cement and steel process emissions.¹⁵ The two barely overlap, and where they meet, they converge: Robertson, a pillar of the distraction case, allows that capture in a hard-to-abate industry "such as cement could be studied as an interim partial solution with careful consideration."¹¹ Both sides mostly agree that scrubbing a coal plant to pipe its CO2 into an oil well is a poor climate trade, and mostly agree that cement's calcined CO2 has no easy substitute.

The one fact that most unsettles the necessity framing comes from its own marquee authority. The IEA that put capture near 15% of cumulative cuts in 2020 has, in its World Energy Outlook 2025, cut the technology to at most 4.9% of modelled 2050 emissions reductions, 1.2% from power and 3.7% from industry, down from 13% in its 2021 outlook, as cheaper renewables, electrification, and efficiency absorbed the work.¹² That is the necessity side's own institution scaling the role down by roughly two-thirds. But read against the different-targets point, the cut lands mostly where the skeptics were already right. What collapsed is the broad and power-sector deployment; the industrial share the cement case depends on is what remains standing. The downgrade is real, and it retires the applications both camps doubt more than the process-emission core the necessity case actually defends.

So the split is not about whether capture physically works, nor whether the planet can hold the gas: the IPCC puts technical geological storage near 1,000 gigatonnes of CO2, more than this century's 1.5°C requirement, which makes cost and deployment, not geology, the binding limit.² What remains is a forecast bound to a value judgment. The forecast is whether cement and steel process capture actually gets built at anything like the Energy Transitions Commission's projected US$130 a tonne,⁵ or whether, as the IEA's own trend now implies, cheaper cuts keep winning the capital first. The value judgment is whether scarce capture capacity and public money belong on the hardest tenth of emissions or, as Jacobson argues, on the wind and solar that abate more per dollar today.⁶ The evidence that would move an undecided reader is specific and near-term: a full-scale cement or steel plant capturing its process CO2 for permanent storage rather than oil recovery, at a verified cost per tonne, and the direction of the IEA's next several outlooks. The Science review notes that a large fraction of cement's process CO2 is reabsorbed through natural carbonation over about fifty years, so capturing it may run net-negative and the lifetime accounting is not the clean subtraction either side reaches for.¹ Until a process-emission plant runs at scale and at a known price, the disagreement stays exactly where the chemistry leaves it: not over whether capture works, but over what it is worth pointing at.

Capture is one process pointed at two different jobs

Net-zero modelers and hard-to-abate engineers

Energy-system researchers and project post-mortems

It narrows to cement's price and the dollar's best use

Sources