BR-4

A branched amine grafted onto mesoporous silica, 82 kg to a column. It is not exotic and we did not invent it. What we do with it is run it at a duty we publish, and change it out on a number rather than on a calendar.

A drawn field of 1746 pores with every pore inked, standing for a bed at equilibrium loading.
The pore field at equilibrium, every pore inked. This is a drawing and not a micrograph: one seeded field of 1,746 pores in a 1176 by 160 box, 70.33 per cent of it by drawn area and 66.92 per cent as the shipped file rasterises. It is re-inked at each of the 11 loadings this site shows, and the geometry is byte-identical in all of them (sha256 93219fe421cbb8fd).

What it holds

Equilibrium capacity at 420 ppm, 10 °C1.250 mmol/g
Loading at breakthrough, 120 min1.188 mmol/g
Residual after regeneration, the heel0.062 mmol/g
Working capacity a cycle1.126 mmol/g
Adsorption time constant at design face velocity40 min
Desorption time constant at 95 °C, 80 mbar12 min
Bed mass a column82 kg
Change-out trigger, measured heel0.090 mmol/g

How it fails

Amines oxidise. The bond that holds carbon dioxide at ambient temperature is the same chemistry that oxygen attacks over thousands of thermal cycles, and the first symptom is not a fall in capacity. It is a rise in the heel: the bed keeps more of what it caught, so the working capacity drops while the equilibrium number looks healthy. We take a column out at a measured heel of 0.090 mmol/g, which on the four production columns has come out between 5,900 and 7,400 cycles.

Spent sorbent goes back to the supplier for silica recovery. It is 82 kg a column, and we count the embodied carbon of the replacement in the deductions on measurement.

Why the type on this site changes weight

A bed gains mass as it adsorbs. That is the whole of what a capture plant does, and it is the one physical fact everything here is drawn from: the weight axis of the typeface is the loading on the bed, at 640 weight units to the millimole per gram. An empty bed is drawn at weight 100 and an equilibrium bed at weight 900. Nothing else about the headings changes: one family, one size, one colour, one tracking.

You can check it on two pages that are deliberately dull. The blank column has no sorbent in it, so its loading is zero at every station and every heading on that page is at weight 100. The saturated column is already at equilibrium, so every heading is at 900. Both pages are flat, and that is the point of them.

Reserve a tonne