Mineral storage

Mineral storage converts CO2 into carbonate minerals. Calcium and magnesium silicates react with dissolved CO2 to precipitate calcite, magnesite and related phases, and because that reaction is thermodynamically downhill the product is stable indefinitely without a seal, a caprock or a monitoring programme to detect leakage. This is the same mineral trapping that terminates the geological storage hierarchy, treated here as a route in its own right rather than as the last stage of a slower process [1].

The engineering question is entirely one of rate. Carbonation happens in nature continuously and far too slowly to matter: natural carbonation of the Samail ophiolite in Oman proceeds at roughly a thousand tonnes of CO2 per cubic kilometre per year [2]. Everything in the field is an attempt to accelerate that by orders of magnitude, and the routes divide by where the reaction is made to happen: in-situ, in reactive rock underground; surficial, in mine tailings and other exposed reactive material; or ex-situ, in an engineered reactor [2].

In-situ carbonation

Basalt and peridotite contain the divalent cations that sedimentary reservoirs lack, so injecting CO2 into them converts storage from a containment problem into a chemical one. Basaltic formations alone offer global storage capacity exceeding anthropogenic emissions, onshore and offshore [1].

The CarbFix pilot in Iceland is the result that established the field. CO2 was dissolved in water before injection rather than injected as a supercritical phase, which removes buoyancy as a failure mode, and over 95 percent of the injected CO2 was mineralised to carbonate in under two years [3]. A second field programme at Wallula validated the reactivity of supercritical CO2 with basalt independently [5].

The scaled operation is the more informative record. CarbFix2 injected 23,200 tonnes of CO2 alongside 11,800 tonnes of H2S over three and a half years at 750 metres depth into rock hotter than 250 degrees C. Over half the injected carbon mineralised within four to nine months, and after the injection rate was doubled, over 60 percent mineralised within four months [4]. Those are strong numbers, but they are not the 95 percent of the pilot, and the difference matters when the pilot figure is quoted as though it were the operating expectation. Mineralised fraction is site-, temperature- and time-specific, not a property of the method.

The other constraint is water. Dissolving CO2 before injection is what makes the CarbFix approach behave so well, and it is also why the approach is water-intensive [1], which bears directly on where it can be sited.

Ex-situ and surficial carbonation

Bringing the rock to the CO2 instead removes the geological uncertainty and replaces it with an energy bill. Silicate feedstock must be mined, ground to expose surface area, and reacted at elevated temperature and pressure. Compiled kinetics for olivine, the most studied feedstock, put the carbonation optimum at 185 to 200 degrees C, with a change in reaction behaviour above about 90 degrees C [6]. Heat and comminution are not incidental costs; they are the reason ex-situ carbon mineralisation runs roughly an order of magnitude more expensive than injecting CO2 into a subsurface reservoir [2].

The exception is material that has already been mined and ground for some other purpose. Steel slag, cement kiln dust, coal ash, mine tailings and similar alkaline residues are produced at something like seven billion tonnes a year globally, and carbonating them has been estimated to offer 2.9 to 8.5 billion tonnes of CO2 removal per year by 2100 [7]. The feedstock is free, fine-grained and often located at an emissions source. Where the carbonated product is used as aggregate or supplementary cementitious material, the boundary with utilisation blurs, and that end-use case belongs with building materials rather than here.

Open challenges

• Passivation. Carbonate and silica-rich layers form on reacting grain surfaces and cut the reaction off before the mineral is spent. Secondary surface coatings remain one of the named gaps in olivine carbonation kinetics [6].
• Clogging versus cracking. In-situ, precipitating carbonate occupies more volume than the mineral it replaces. That can crack the rock and expose fresh surface, or it can seal the pore space and destroy the permeability the injection depends on. Which feedback dominates, and under what conditions, is not yet well understood [2].
• Energy and cost. The ex-situ energy penalty is intrinsic to the kinetics, not an engineering detail awaiting optimisation [6][2].
• Deployment record. Despite two decades of work, large-scale implementation remains largely confined to laboratory and field experiments [1]. The capacity argument is much further advanced than the delivery record.