Carbon capture: the honest assessment

Carbon capture covers several distinct technologies used for different purposes, and conflating them produces most of the confusion in the argument.

The categories

Point-source capture at a flue stack, where carbon dioxide concentration is relatively high — a few percent for power generation, higher for some industrial processes.

Direct air capture, extracting carbon dioxide from ambient air at around four hundred parts per million.

The concentration difference is enormous and it drives the energy requirement. Separating a dilute gas requires substantially more work than separating a concentrated one, which is thermodynamics rather than engineering deficiency.

Which is why direct air capture costs far more per tonne and always will, relative to point-source capture.

Utilisation versus storage. Captured carbon dioxide can be stored geologically or used industrially.

The distinction matters enormously. Storage removes carbon from the atmosphere; most utilisation returns it promptly. Carbon dioxide used in beverages or in enhanced oil recovery is not sequestration in any meaningful sense.

Where it is genuinely necessary

Cement. The strongest case.

Roughly half to two-thirds of cement's emissions come from the chemistry itself — calcining limestone releases carbon dioxide regardless of the heat source. Electrifying the kiln does not address it.

Cement is essential, alternatives are limited, and capture is the main available route.

Steel, where blast furnace chemistry produces carbon dioxide directly. Hydrogen direct reduction is the competing route and both are being pursued.

Chemicals and refining, where some processes produce high-purity carbon dioxide streams that are unusually cheap to capture.

Waste incineration, where the alternative is emitting.

The pattern: capture is most defensible where emissions are inherent to a necessary chemical process rather than to a choice of fuel.

Where it fits poorly

Power generation. The most contested case.

Capture on a gas or coal plant imposes an energy penalty — a substantial fraction of the plant's output is consumed running the capture process — which raises the cost per unit of electricity considerably.

Against renewables plus storage, the economics have deteriorated steadily as renewable costs fell.

Several high-profile power sector projects have been cancelled, underperformed against capture rate targets, or closed. That record is part of the evidence base and should be counted.

Direct air capture

Worth assessing separately because the argument is different.

Costs are currently very high per tonne, with projections of substantial reduction that depend on scaling and on cheap energy.

The strongest argument for it is that some emissions are genuinely unavoidable, and that reaching net zero requires removing them. It is also the only route to net negative emissions if that becomes necessary.

The strongest argument against is scale: current global capacity is a rounding error against annual emissions, and building capacity at the required scale would consume enormous quantities of low-carbon energy that could otherwise displace fossil generation directly.

The reasonable position is that it is worth developing for the residual and is not a substitute for reduction — and that using it to justify continued emissions is the failure mode to guard against.

The storage question

Geological storage in deep saline formations and depleted hydrocarbon reservoirs is technically established, with projects operating for decades and monitoring showing retention.

The concerns are induced seismicity from injection, leakage through legacy wells, long-term liability, and the cost of monitoring in perpetuity.

Capacity estimates are large and are estimates. Actual injectable capacity at acceptable cost in specific locations is more limited than headline figures suggest.

The measurement problem

Underdiscussed and important.

Claimed capture rates and actual performance have diverged at several projects. Reporting standards vary, and the difference between capture capacity and captured tonnes is frequently blurred.

For carbon credits based on removal, verification is a genuine unresolved issue: demonstrating that a tonne was removed and will stay removed requires monitoring commitments that outlast most companies.

The reasonable summary

Necessary for a specific set of industrial processes where no alternative exists.

Expensive, energy-intensive and slower to deploy than proponents claim.

Frequently invoked to justify continued emissions where alternatives do exist, which is the use that deserves the scepticism it receives.

And a much smaller part of any credible decarbonisation pathway than the attention it attracts would suggest.