What Is Carbon Capture? Separating One Gas From a Mixture

The concentration of the gas at the point of capture dominates the cost, because separating something dilute is far harder than separating something concentrated:

• •Post-combustion capture, taking carbon dioxide from flue gas after burning, where it is typically four to fifteen percent of the stream, which can be retrofitted to existing plants and is the most studied route
• •Pre-combustion capture, converting fuel to hydrogen and carbon dioxide before burning, which gives a more concentrated stream at higher pressure and is easier to separate but requires a plant designed for it
• •Oxy-fuel combustion, burning in nearly pure oxygen so the exhaust is mostly carbon dioxide and water, which simplifies capture and adds the cost of producing the oxygen
• •Industrial process capture, from cement, steel, ammonia and ethanol production, where some streams are nearly pure and capture is comparatively cheap, and where the emissions come from chemistry rather than from fuel and so cannot be removed by switching energy source
• •Direct air capture, taking it from the atmosphere at about 0.04 percent, which is thermodynamically the hardest and correspondingly the most expensive per tonne
• •Bioenergy with capture, burning biomass that absorbed carbon while growing and capturing the emissions, which in principle removes carbon from the atmosphere

Several chemistries compete and each trades energy against equipment. Chemical absorption using amine solutions is the mature approach: flue gas is bubbled through a solvent that binds carbon dioxide at low temperature, and the loaded solvent is then heated to release it and regenerated. The release step is where the energy goes, since the same strong bond that makes capture efficient makes reversal expensive, and it typically consumes a substantial share of the host plant's output, which is the energy penalty that dominates the economics. Physical solvents dissolve the gas rather than reacting with it, which suits high-pressure streams and needs less heat. Solid sorbents adsorb the gas onto engineered surfaces and release it with heat or a pressure swing, avoiding the large water inventory of a liquid system. Membranes separate by allowing the gas to pass preferentially, which is simple and struggles with dilute streams. Calcium looping cycles lime between carbonate and oxide at high temperature, and is attractive for cement because the industry already handles those materials.

Captured gas is compressed to a dense supercritical state and transported by pipeline or ship. Geological storage injects it into porous rock under an impermeable caprock, typically depleted oil and gas fields, whose sealing has already been demonstrated over geological time, or deep saline aquifers, which offer far more capacity. Several mechanisms trap it progressively: the caprock physically, then dissolution into the formation water, then residual trapping in pore spaces, and finally mineralisation into carbonate rock, with security increasing over time. Basalt formations mineralise unusually fast, with a project in Iceland reporting conversion of most injected carbon to carbonate minerals within about two years, which removes the leakage question entirely. Monitoring uses seismic surveys and pressure measurement. The awkward exception is enhanced oil recovery, where the gas is injected to push out more oil, which has historically been where most captured carbon went and which produces additional hydrocarbons in the process, making its net effect contested.

The case for capture is strongest where emissions are hard to avoid any other way. Cement releases carbon dioxide from limestone as an unavoidable part of the chemistry, and steel, some chemicals and aviation fuels have no straightforward electrification path, so capture is one of few options for those sectors. The case is weakest for power generation, where deployment has been slow, several flagship projects have underperformed or closed, and building renewable generation instead is now cheaper per tonne avoided. The structural criticisms deserve stating plainly: capture rates in practice have frequently fallen short of design, the energy penalty means more fuel burned for the same output, the technology has been used rhetorically to defend continued fossil fuel investment, and direct air capture at current costs is far more expensive per tonne than almost any alternative mitigation. The defensible position is that it is necessary for specific industrial processes and for eventual net removal, and is not a substitute for reducing emissions where reduction is available.

Capture separates carbon dioxide from a gas stream, and cost tracks concentration, so industrial streams are cheap, flue gas is moderate and pulling it from air at 0.04 percent is the hardest. Amine solvents dominate, and regenerating them consumes a large share of the host plant's energy. Storage injects the compressed gas into porous rock under a seal, where it dissolves, becomes trapped and eventually mineralises. It matters most for cement and steel, where the emissions are chemical rather than from fuel.