Troubleshooting

Carbon Capture and CO2 Compression: Costly Mistakes and How to Catch Them

The recurring errors that throw off capture efficiency, compressor power, and regen heat numbers, each with a symptom, a root cause, and a numeric fix.

Symptom: your reported capture rate reads 94 percent but the stack analyzer says CO2 slip is climbing. Root cause is almost always a mass balance closed on volumetric flow instead of molar flow, so wet and dry basis get mixed. A flue gas at 12 percent CO2 wet versus 13.6 percent dry is a 1.6 point gap that propagates straight into the Capture Efficiency calculation. Fix: convert both inlet and treated streams to a dry molar basis before subtracting, and confirm your CEMS moisture correction. A 2 percent moisture error on a 500,000 Nm3/h train hides roughly 3 to 4 tonnes per day of uncounted CO2.

Symptom: solvent makeup invoices run 20 to 40 percent over budget within the first quarter. The root cause is usually treating amine degradation and mechanical carryover as one lumped rate. Thermal and oxidative degradation of MEA typically runs 0.3 to 1.5 kg per tonne CO2, while entrainment losses depend entirely on demister condition. If you feed a single blended loss figure into Solvent Makeup Cost, you cannot tell whether to reclaim solvent or fix a mist eliminator. Fix: split the two streams, then check reclaimer duty. Recovering degraded amine below 0.5 kg per tonne CO2 often pays back the reclaimer inside 14 months.

Symptom: the regeneration heat load in your model sits at 3.2 GJ per tonne CO2 but the reboiler steam meter shows 4.1. The missed variable is lean loading drift. Operators tune for capture rate and let lean loading creep from 0.20 to 0.28 mol CO2 per mol amine, which raises specific reboiler duty by 15 to 25 percent. The Regeneration Heat Load calculator only tells the truth if lean and rich loadings are measured, not assumed. Fix: pull a weekly loading titration and hold lean loading within 0.02 of design. Each 0.05 mol per mol of excess lean loading is roughly 0.4 GJ per tonne of wasted steam.

Symptom: compressor shaft power comes in 12 percent above the datasheet at the same throughput. The classic error is applying a single polytropic efficiency across all stages while ignoring that suction temperature drifted up. CO2 Compressor Power scales with absolute suction temperature, so a 15 degree C rise from 40 to 55 at the first stage inlet adds about 5 percent to head per stage. Across a 5 stage machine reaching 150 bar, that compounds. Fix: verify interstage cooler approach temperatures and log suction conditions per stage. Restoring a 40 degree C suction typically recovers 300 to 600 kW on a 100 tonne per hour unit.

Symptom: blower energy is higher than the model predicts and the absorber floods intermittently. Root cause is an outdated pressure drop assumption after packing or trays foul. People size the fan on clean dP, then never revisit it. Absorber Pressure Drop above 6 to 8 mbar per meter of packing signals liquid maldistribution or fouling, and every extra 10 mbar of column dP pushes the Blower Energy Cost up measurably. Fix: trend dP against gas load. A jump from 30 to 45 mbar at constant flow means schedule a wash, not a bigger blower, saving 40 to 80 kW.

Symptom: skid nameplate says 50 tonnes per day but sustained output is 38. The mistake is quoting instantaneous peak instead of availability adjusted throughput. Skid Throughput must be derated by real uptime, and small modular capture skids commonly run 82 to 90 percent availability in year one, not the 95 percent on the brochure. Feeding 100 percent uptime into the number overstates annual capture by thousands of tonnes. Fix: multiply design rate by measured availability and by turndown time at part load. A skid at 86 percent availability and 8 percent part load operation delivers closer to 40 tonnes per day.

Symptom: your CO2 balance shows a persistent 2 to 5 percent shortfall between captured and delivered product. Root cause is unaccounted fugitive and vent losses treated as measurement noise. Flash gas from the stripper, dehydration vents, and flange leaks add up, and the CO2 Leak Loss calculator exposes them only if you meter vent streams rather than back-calculate. Fix: install vent metering and pressure decay tests on the compression train. A single 150 bar flange weeping at 2 kg per hour is 17.5 tonnes per year, and a poorly set stripper reflux can vent 1 to 2 percent of product outright.

Symptom: overall energy per tonne slowly climbs with no operating change, and cooling water outlet runs warm. The overlooked driver is heat exchanger fouling on the lean rich cross exchanger, where a 5 degree C rise in cold end approach forces the reboiler and trim coolers to compensate. Heat Exchanger Fouling Loss quantifies this, but only if you log both approach temperatures rather than a single duty number. Fix: trend the cross exchanger approach weekly. When approach widens from 10 to 15 degrees C, a clean adds back roughly 0.2 to 0.3 GJ per tonne CO2 and cuts trim cooling load 8 to 12 percent.

Published 2026-07-02.