The standard way to capture carbon dioxide from flue gas is with an amine tower. In this system, amines are dissolved in a column of water, and flue gas is bubbled through the column to allow the CO2 to react with amines. The reaction is selective, and the binding between the amines and the CO2 is tight. The CO2 is released by heating the water, to near boiling, and the amines are returned to the column to react again.
This costs about $50/ton today for a flue gas with high concentrations of CO2, such as those from a concrete or steel plant. For a leaner gas stream, such as the 3.5% CO2 flue gas from a natural gas turbine, the cost goes up, as more energy must be spent bubbling through the column, and more entropy must be overcome to achieve the baseline 90% capture rates typically specified for such a plant. These costs might come down 20% with better amines that are already demonstrated at lab scale, but it's unlikely to drop further.
The Next Big Thing in carbon capture has proven to be solid sorbents - materials with pores almost exactly the size of CO2, which can bind to the CO2 selectively over N2 in the flue gas stream. The binding of CO2 to solid sorbents isn't as strong as with amines - the bonds between CO2 and the solid are more like the attraction of water to a sponge than a chemical reaction. But with no bubbling required, the throughput of solid sorbent systems is much higher, and the required energy costs and capital expenditures are much lower. The great hope of many companies - Svante is the leader here, but there are others - is that these new materials will lead to substantial cost improvements. To get the the ultimate low cost requires some fancy chemistry, fine-tuning the sorbents to get just the right pore size, with just the right interactions with CO2 to favor its binding over nitrogen.
How cheap can such a system be? This was the topic of a recent paper with a super-long title, How much can novel solid sorbents reduce the cost of post-combustion capture? A techno-economic investigation on the cost limits of pressure–vacuum swing adsorption. The paper didn't model specific chemical sorbents, but instead asked what the ideal properties of a sorbent might be, without overly constraining the physical system.
The first thing that fell out of this analysis is that the concentration of CO2 matters a lot more for solid sorbents than for liquid-phase amines. A flue gas that is 30% CO2 costs just a little over $10/ton to abate if the sorbent is free (remember, these are ideal materials), while a flue gas at 3.5% costs $85/ton. It's a much bigger difference than with amines, consistent with the intuition that the lower binding strength of solid sorbents makes them less effective at handling dilute flue gas.
What about a real system? Adding in the cost of materials does not alter the situation much if the materials are robust - when the sorbent costs $5/ton, the cost of capture only rises by 15% or so. This is consistent with the cost profile at Svante, though other more esoteric materials may be more expensive still.
The net is that a system that captures a 20% CO2 flue gas stream, as from concrete or steel, might plausibly designed to cost only $20/ton. Real implementations will almost certainly require more fine-tuning than this, including compromises that are beyond the scope of this paper. But certainly this analysis lends credence to Svante's approach in particular, and to the notion that the $30/ton target of the DOE is entirely within reach.
Kommentarer