Lessons Learned: Comparing Two Detailed Capital Cost Estimates for Carbon Capture by Amine Scrubbing

Two front-end engineering designs (FEED) for commercial -scale carbon capture from natural gas combined-cycle (NGCC) power plants have been published in full, including detailed capital cost estimates. The two studies are the Piperazine with the Advanced Stripper (PZAS) FEED at Mustang Station in Denver City, TX, U.S.A. ("Mustang FEED") and the FEED at Panda Sherman Power Station in Sherman, TX, U.S.A. ("Panda FEED"). The design capacities are 197 tonne/h of CO2 captured in the Mustang FEED and 129 tonne/h in the Panda FEED. This work compares both designs with the aim of gaining insight about the cost of carbon capture processes. Direct field costs for each process are adjusted to account for differences in the estimating methodology in each FEED. This adjustment results in estimated direct field costs of $574MM for Mustang and $411MM for Panda. The breakdowns of capital costs for each process area are compared, to highlight differences in costs due to the respective processes or particular requirements of each host power plant. A key finding is that site-specific factors not typically represented in less rigorous cost estimates are observed to have a significant impact on cost. To compare both estimates on a similar design basis, the Mustang unit was scaled to the Panda FEED design capacity, resulting in direct field costs of $489MM for Mustang and $411MM for Panda. Additionally, the cost estimates of the CO2 absorbers are analyzed in greater detail to directly compare three equivalent designs that include rectangular and cylindrical absorber columns. Adjusting the absorber estimates to the same scope and scaling to the same size results in estimated costs of $19.7MM, $18.2MM, and $14.9MM, with the rectangular absorber being approximately 10% to 30% more expensive than the cylindrical designs.