Carbon Capture Costs: FEED & pre-FEED Cost Reports

Electric Power Research Institute (EPRI), with Fluor Corporation and California Resources Corporation (CRC), conducted a FEED study to assess the feasibility of retrofitting Fluor’s solvent for post-combustion CO₂ capture at the 550 MW Elk Hills NGCC power plant. The system aims to capture approximately 4,000 tonnes of CO₂ per day (75% of total emissions, or 90% of an 83% slipstream), with the captured CO₂ intended for enhanced oil recovery. Deliverables include the full engineering design package—such as process flow diagrams, equipment datasheets, and capital cost estimates—optimized for site-specific performance, operations, and construction practices.

Analysis sets the design and cost-estimating basis for evaluating pulp and board mills with and without CCS. Two base cases are considered: a market pulp mill and an integrated pulp and board mill. Six CCS cases are evaluated, capturing CO₂ from the recovery boiler, multi-fuel boiler, lime kiln, or their combinations. The mills are assumed to be energy independent, with black liquor and bark burned to produce steam and electricity, and excess electricity exported to the grid. The CO₂ capture system uses post-combustion MEA technology with a 90% capture rate, and if on-site electricity is insufficient, an auxiliary boiler firing forest residues will supply the additional energy. Capture of CO2 in both Kraft & Multi-boilers

The University of Texas at Austin, with AECOM Technical Services and Trimeric Corporation, is conducting a FEED study for the Piperazine Advanced Stripper (PZAS) CO₂ capture process at Golden Spread Electric Cooperative’s Mustang Station in Denver City, Texas. Designed for two GE gas turbines with HRSGs and a steam turbine totaling 464 MWe, the PZAS process uses 30 wt% piperazine for higher efficiency, solvent stability, smaller absorber size, and cost savings compared to conventional amine systems. The project will deliver a 30–60% complete design package and a capital cost estimate with ±15% accuracy.