Flue gas or ambient air is modelled as a stream of individual molecules — grey N₂/O₂, teal CO₂ — rising through a capture column. A horizontal sorbent zone (liquid amine solvent, a solid sorbent bed, or an aqueous mineral slurry, depending on the technology) reacts selectively with CO₂ as it passes through. Every simulated timestep, each CO₂ molecule inside the zone has a chance to be captured that follows a first-order removal law:
P(capture) = 1 − e^(−k·τ)
τ = bed_height / flow_speed (residence time)
- k (rate constant) — set by the technology: amine solvents react fastest (k≈2.4 s⁻¹) because MEA binds CO₂ directly at flue-gas concentrations; solid DAC sorbents are slower (k≈0.9 s⁻¹) since they must pull CO₂ from dilute ambient air (~0.04%); mineralization is slowest (k≈0.5 s⁻¹) but locks carbon away permanently as solid carbonate rock.
- CO2 concentration — how much of the incoming gas is CO₂ molecules vs. inert N₂/O₂. Flue gas from a power plant runs ~10–15%; open air is ~0.04%, which is exactly why DAC needs huge sorbent surface area to catch enough molecules.
- Gas flow speed — faster flow means less time in the sorbent zone (shorter τ), so efficiency drops even though more total gas passes through per second.
- Sorbent bed height — a taller bed gives each molecule more residence time inside the reactive zone, raising τ and therefore capture efficiency, at the cost of more pressure drop in a real plant.
Captured CO₂ leaves the gas stream and routes to the side: amine and DAC systems compress and pipe it toward geological storage or utilization (CCUS), while mineralization solidifies it on the spot as carbonate mineral — the same basic chemistry used in BECCS and enhanced-weathering pilots today.