CO2 captured at the surface is dissolved into water at the injection well, then pumped down into a porous basaltic formation — the approach pioneered by Iceland's Carbfix project. Unlike injecting dry supercritical CO2 (which can migrate and needs a caprock to stay trapped), carbonated water sinks and diffuses through the pore network as a dense, mildly acidic plume that cannot leak back to the surface as gas.
Each cube in the grid is a block of basalt. As the plume front (blue) reaches a block, the acidic dissolved CO2 attacks calcium, magnesium and iron silicate minerals in the rock:
CaSiO3 + CO2 → CaCO3 + SiO2
(Mg,Fe)2SiO4 + 2CO2 → 2(Mg,Fe)CO3 + SiO2
The reaction converts CO2-charged fluid into solid carbonate minerals (calcite, magnesite, siderite) that fill the pore space permanently — blocks turn white as they mineralize. Warmer reservoir rock reacts faster (an Arrhenius-type rate law), but the model caps it: past roughly 60 °C the carbonate minerals themselves start dissolving faster than they form, so extremely hot basalt is not necessarily better.
- Injection rate — how fast carbonated water is pumped in; higher rate pushes the plume outward faster but dilutes local CO2 concentration.
- Water : CO2 ratio — how much water carries each unit of CO2; a higher ratio (more water) spreads CO2 thinner but keeps it fully dissolved, avoiding a free gas phase.
- Reservoir temperature — controls the silicate→carbonate reaction rate constant.
- Simulation speed — time compression only; real Carbfix wells mineralize roughly 95% of injected CO2 within about two years.
Real-world relevance: basalt is the single most abundant volcanic rock on Earth's surface and ocean floor, giving mineral carbonation enormous long-term storage potential — once mineralized, the carbon cannot leak, corrode a well, or escape a caprock failure the way gas-phase CO2 storage can.