Amines, Membranes, and the Energy Penalty

7 min read

What you'll take away Compare amine absorption with membrane separation, and explain why the capture rate on the datasheet is not the capture rate the climate sees.

Strip away the vendor slides and onboard capture is two competing ways of pulling a few per cent of CO2 out of a hot, wet, dirty gas stream, one chemical and mature, one physical and younger. Both obey the same tax: the energy to run them comes from the ship, and the ship makes energy by burning fuel.

Amine absorption, the incumbent

The dominant design is chemical absorption with an amine solvent, the same process that has scrubbed onshore flue gas for decades. The exhaust is first cooled and cleaned, sulphur oxides must come out, because they react with the solvent and destroy it, then enters the bottom of an absorber column. Lean solvent trickles down through packing as the gas rises; the amine binds CO2 chemically and lets the nitrogen pass to the funnel. The CO2-rich solvent is pumped to a second column, the regenerator, and heated. Heat reverses the reaction: the solvent releases a near-pure stream of CO2 and returns, lean again, to the absorber. The CO2 is then dried, compressed and liquefied for storage: lesson 3's subject.

The amine loop: the solvent goes round, the heat goes in, the CO2 comes out Absorber amine binds CO2 Regenerator heat releases it exhaust, cooled and cleaned nitrogen and the rest → funnel rich solvent, CO2 loaded lean solvent returns near-pure CO2 → liquefaction reboiler heat, steam-grade: the plant's largest energy demand Heat closes the loop, and the ship makes heat by burning fuel
The amine route in one picture: the gas gives up its CO2 in the absorber, heat takes it back out in the regenerator, and the solvent circulates between them. The amber arrow is where the energy penalty physically enters the plant.

The strengths are real: proven chemistry, high single-pass capture, and decades of operating experience ashore. The costs are equally real: a tall absorber that must live near the casing, a solvent inventory that degrades and needs make-up, and above all heat. Regenerating the solvent takes steam-grade heat in quantities that dominate the plant's energy budget.

Membranes, the challenger

Membrane separation does away with the solvent entirely. The exhaust is pushed against a polymer membrane that CO2 permeates faster than nitrogen; a pressure difference across the membrane drives the separation. No regeneration column, no steam demand, no chemical inventory: a compact, modular plant that suits a ship well on paper.

The catch is the same dilute exhaust from lesson 1. With CO2 at only a few per cent, its partial pressure, the driving force through the membrane, is low, so the system needs compression or vacuum on the gas side and usually several stages to reach a useful purity. The steam bill becomes an electricity bill. The route is credible enough that ClassNK added membrane separation to its guidelines for onboard CO2 capture and storage in the October 2025 revision; the open Technical Journal walkthrough of the guidelines explains what that revision covers. When class writes rules for a technology, it expects to survey it.

How much can these plants capture?

Bureau Veritas studied shipboard capture across technologies in its whitepaper Onboard Carbon Capture (free after registration) and found capture rates of 82–90% technically feasible. Treat that as the gross figure: what the plant removes from the gas that passes through it, in the conditions it was designed for.

The energy penalty, honestly counted

Now the tax. Solvent regeneration wants heat; compression, liquefaction and fans want power. Some of the heat can be recovered from the exhaust itself, but on most ships recovery does not cover the reboiler, so an auxiliary boiler burns fuel to make up the difference, and every extra tonne burned makes roughly three more tonnes of CO2 for the plant to deal with.

This is why the Maersk Mc-Kinney Moller Center's study of onboard capture distinguishes gross from effective capture: once the energy penalty is counted, the effective figure lands around 74–78%: several points below the 82–90% the hardware achieves on the gas stream. The distinction matters commercially, not just intellectually. If a regulation ever pays per tonne captured, it will pay on a metered, net basis; and if a vendor quotes a capture rate, the first question to ask is whether it is gross or effective, and what fuel consumption it assumes.

Gross is what the hardware does; effective is what the atmosphere gets Gross capture: what the plant removes from the gas 82–90% Effective capture: net of the fuel burned to run it 74–78% 60% 70% 80% 90% 100% share of the CO2 in the exhaust: scale starts at 60% the distance between the ranges is the CO2 emitted to make the plant's own heat and power
Two honest numbers for the same plant: 82–90% of the CO2 comes out of the gas stream (Bureau Veritas), but only about 74–78% is a net removal once the energy penalty is counted (MMMCZCS). The scale starts at 60% so the shift is visible.

One design consequence follows and is worth carrying into the next lesson: chasing the last few points of capture rate is expensive. Pushing a plant from 80% toward the high nineties drives up the columns, the heat, and the power faster than the captured tonnes grow. Every OCC design is therefore a chosen compromise between capture rate, energy penalty, and the space and money the ship can afford, which is exactly where ship integration begins.

Check yourself

1. In an amine capture plant, where is the CO2 released from the solvent?
2. What drives the separation in a membrane system?
3. Which technology was added to the ClassNK guidelines for onboard CO2 capture in the October 2025 revision?
4. What is the largest single energy demand of an amine plant?
5. Why does effective capture fall below the plant's rated capture rate?
6. Which of these statements about the two technologies are true?

Select all that apply.

7. A capture plant rated at 90% removes 90% of the ship's total climate impact.
8. The MMMCZCS study put effective capture, net of the energy penalty, in a band of roughly 74–78%. Enter a value in that band.
%