Partial Repeal of the Carbon Pollution Standards for Fossil Fuel-Fired Electric Generating Units

CCS with 90 percent capture involves the capture of 90 percent of the CO2 emissions from the EGU, compression and transport of the captured CO2 via pipeline, and sequestration in geologic storage. Due to the relatively low concentration of CO2 in the flue gas, an amine solvent-based capture system is better suited for application to the post-combustion flue gas of fossil fuel-fired EGUs than other CO2 removal technologies (e.g., pressure- swing adsorption). CO2 removal occurs by reactive absorption of the CO2 from the flue gas into the amine solution in an absorption column. The amine reacts with CO2 but will also react with impurities in the flue gas, including sulfur dioxide (SO2). Particulate matter (PM) will also affect the capture system. Adequate removal of SO2 and PM prior to the CO2 capture system is therefore necessary. After pretreatment of the flue gas with conventional SO2 and PM controls, the flue gas goes through a quencher to cool the flue gas and remove further impurities before the CO2 absorption column. After absorption, the CO2-rich amine solution passes to the solvent regeneration column, while the treated gas passes through a water and/or acid wash column to limit emission of amines or other byproducts. In the solvent regeneration column, the solution is heated (using steam) to release the absorbed CO2. The released CO2 is then compressed and transported to a sequestration site. In an integrated CO2 capture system, steam and electricity for the capture process are provided by the host-EGU; this avoids the need to capture additional emissions from any auxiliary boilers or cogeneration units.

In the 2024 CPS, the CO2 capture component of the 90 percent CCS BSER was premised on an integrated amine solvent-based CO2 capture system with 90 percent removal of CO2 from the post-combustion flue gas of the host EGU. The EPA previously argued that such a system was adequately demonstrated.

The EPA has reevaluated the record and is determining in this final rulemaking that, critically, 90 percent capture of the CO2 from flue gas of an EGU has not been adequately demonstrated. As a result, 90 percent CCS has not been adequately demonstrated and cannot be the BSER for long-term coal-fired steam generating units. The EPA is basing this conclusion primarily on a revised evaluation of the record in the 2024 CPS, as detailed in this section of the preamble.\141\ The EPA is additionally considering several developments since the EPA promulgated the 2024 CPS (e.g., changes in the plans of certain CCS projects).

\141\ See Nat'l Ass'n of Home Builders v. EPA, 682 F.3d 1032, 1038 (D.C. Cir. 2012) (“EPA did not rely on new facts, but rather on a reevaluation of which policy would be better in light of the facts . . . . Fox makes clear that this kind of reevaluation is well within an agency's discretion.” (citing 556 U.S. at 514-15)).

In the 2024 CPS, the emission guidelines required States to establish plans that would require long-term existing coal-fired steam generating units to achieve an annual standard of performance based on capturing 90 percent of the unit's total CO2 emissions. However, the record for 90 percent capture as adequately demonstrated did not include an example of a commercial scale coal-fired steam generating unit that was already capturing 90 percent of its annual CO2 emissions. Instead, the EPA attempted to argue that 90 percent capture has been adequately demonstrated based on other evidence for the technology at that time. Specifically, the EPA relied primarily on evidence that consisted of the operation of the CO2 capture system at Boundary Dam Unit 3, fixes applied at Boundary Dam Unit 3, and testing on new solvents from different vendors.\142\ Consequently, the EPA extrapolated from that combination of primary evidence to determine that 90 percent capture would perform as anticipated for the affected sources. On that basis, the EPA concluded that 90 percent capture was adequately demonstrated for existing coal-fired steam generating units.

\142\ The EPA also included other, secondary observations (e.g., projects in development) that would be insufficient on their own to conclude 90 percent capture is adequately demonstrated.

The only datapoint for commercial scale post-combustion CCS on a fossil fuel-fired EGU, with integrated steam and power, is Boundary Dam Unit 3. However, between 2014 and 2022, the capture system at Boundary Dam achieved a total capture efficiency of not more than 63 percent over the course of a calendar year.143 144 This total annual capture efficiency is substantially below the 90 percent capture level specified by the BSER. While the EPA had acknowledged the challenges and underperformance of the capture system at Boundary Dam in the 2024 CPS, the Agency asserted that fixes were available or could be made to address those issues. However, many of those fixes were already made, and performance remained below the design capture efficiency.\145\ The EPA also previously argued that new solvents were available that could capture CO2 at higher rates to address these gaps.\146\ However, in the 2024 CPS, the EPA failed to reasonably account for potential underperformance of capture systems using new solvents, and the experience at Boundary Dam shows it would be reasonable to anticipate that such capture systems would similarly underperform.\147\ Furthermore, the EPA also failed to account for any decrease in operating-availability of capture, even though the annual operating- availability of the capture system at Boundary Dam has been less than 100 percent.\148\ In combination, a capture system would achieve much less than 90 percent total capture efficiency. On review, the EPA's

prior extrapolation fails to support CCS with 90 percent capture as adequately demonstrated. The EPA's other tangential arguments in the 2024 CPS similarly fail to show that 90 percent capture has been adequately demonstrated (including projects in development, as discussed in section IV.A.1.a.iv of this final rule preamble). Considering these factors, the EPA is finalizing the determination that CCS with 90 percent capture and, consequently, 90 percent CCS are not adequately demonstrated for existing coal-fired steam generating units. The following subsections provide further explanation.

\143\ Here, total capture efficiency is equivalent to the mass of CO2 captured relative to (i.e., divided by) the mass of CO2 that the EGU would otherwise emit (including the mass of CO2 produced in the combustion chamber of the EGU plus the mass of CO2 produced by any auxiliary equipment that supports the capture process) over a given period (e.g., annual).

\144\ Jacobs, B., et al. Reducing the CO2 Emission Intensity of Boundary Dam Unit 3 Through Optimization of Operating Parameters of the Power Plant and Carbon Capture Facilities. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies (2022). Available at: https://dx.doi.org/10.2139/ssrn.4286430.

\145\ The most recently reported total capture efficiency when operating was 83 percent on an instantaneous basis. Accounting for periods when the capture system is offline, the total annual capture efficiency would be lower.

\146\ 95 percent total capture on an instantaneous basis, not accounting for uncertainty in real-world operation or availability.

\147\ As the only commercial scale and long-term application of CO2 capture on a fossil fuel-fired EGU, Boundary Dam's underperformance is a reasonable quantification of potential underperformance.

\148\ Between 2015-2022, the availability of the capture system relative to the EGU was, at best, 94 percent.

In the 2024 CPS, the EPA based the determination that 90 percent CO2 capture was adequately demonstrated on the record for amine-solvent CO2 capture.\149\ Thus, the EPA relied heavily on the operation of carbon capture at the commercial scale 110 megawatt (MW) coal-fired Boundary Dam Unit 3 (Saskatchewan, Canada) to demonstrate 90 percent capture. Boundary Dam has operated CCS since 2014. The unit uses Shell's amine-based CANSOLV[supreg] solvent technology to capture CO2 from the post-combustion flue gas of the coal-fired boiler.\150\ Captured CO2 is then compressed, transported by pipeline, and used for enhanced oil recovery (EOR) or stored in a saline aquifer at the Aquistore site.\151\ While Boundary Dam Unit 3 achieved 89.7 percent capture over a 3-day test early in its operation, longer-term capture levels have been lower.\152\ Between 2015 and 2022, Boundary Dam achieved a total capture efficiency of not more than 63 percent in a calendar year.\153\ This total long term capture efficiency is substantially below the 90 percent capture efficiency of the BSER.

\149\ 89 FR 39798, 39848 (May 9, 2024).

\150\ Giannaris, S., et al. SaskPower's Boundary Dam Unit 3 Carbon Capture Facility--The Journey to Achieving Reliability. Proceedings of the 15th International Conference on Greenhouse Gas Control Technologies (2021). Available at: http://dx.doi.org/10.2139/ssrn.3820191.

\151\ Aquistore. Available at: https://ptrc.ca/aquistore.

\152\ SaskPower Annual Report (2015-16). Available at: https://www.saskpower.com/-/media/saskpower/about-us/reports/past-reports/report-annualreport-2015-16.pdf.

\153\ Jacobs, B., et al. Reducing the CO2 Emission Intensity of Boundary Dam Unit 3 Through Optimization of Operating Parameters of the Power Plant and Carbon Capture Facilities. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies. (2022). Available at: https://dx.doi.org/10.2139/ssrn.4286430.

This lower total capture efficiency is due to, among other things, the capture system at Boundary Dam Unit 3 typically processing less than all of the flue gas, in part to “maintain long-term reliable operation.” \154\ Prior to 2023, the CO2 capture system at Boundary Dam Unit 3, when operating, processed up to approximately 75 percent of the flue gas with 90 percent CO2 capture from the processed flue gas.\155\ The EPA argued in the 2024 CPS that such capture from the majority of the flue gas supported the determination of 90 percent capture from all of the flue gas as adequately demonstrated; however, this ignores that the total capture efficiency was substantially less than the 90 percent design capture efficiency.

\154\ SaskPower. “Docket ID No. EPA-HQ-OAR-2023-0072: SaskPower Correction of Reference to Boundary Dam Unit 3 Emissions Performance in Proposed Rule” (August 4, 2023). Document ID No. EPA-HQ-OAR- 2023-0072-0687.

\155\ Jacobs, B., et al. Reducing the CO2 Emission Intensity of Boundary Dam Unit 3 Through Optimization of Operating Parameters of the Power Plant and Carbon Capture Facilities. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies. (2022). Available at: https://dx.doi.org/10.2139/ssrn.4286430.

Additionally, Boundary Dam Unit 3 has experienced various technical challenges that have reduced its performance.\156\ These include fouling of the CO2 absorber due to PM (fly ash), buildup of scale on heat exchangers, biological fouling in the wash-water section of the CO2 absorber, foaming of the amine solvent in the CO2 absorber, and damage to the CO2 compressor. Fouling in the CO2 absorber affects the throughput of the process by increasing the pressure drop (i.e., difference in pressure between the bottom and top of the absorber) such that it cannot be overcome by the draft fans (i.e., more energy is required to move the same volume of flue gas than the fans are designed to provide). Boundary Dam Unit 3 has implemented improvements to its particulate controls (electrostatic precipitator) and added more spray-wash systems to mitigate fouling due to fly ash. Caustic shocking of the wash-water section can reduce the buildup of biological material. Buildup of scale on heat exchangers reduces throughput by increasing pressure drop, while the layer of scale reduces the rate of heat transfer. To address this issue, redundant heat exchangers with isolations were installed in 2017 to allow for removal of scale without shutting down the CO2 capture system. Finally, damage to the CO2 compressor caused by a loose bolt, and issues with a leaking intercooler in the compressor, forced the CO2 capture system to be offline for several months in 2021 and the start of 2022. While the compressor was repaired, the unit lacks a redundant compressor in the event of a similar outage. Importantly, despite these attempts to improve operation, the unit continues to underperform. Furthermore, outages to address these issues have contributed to a lower total capture efficiency.

\156\ Jacobs, B., et al. Reducing the CO2 Emission Intensity of Boundary Dam Unit 3 Through Optimization of Operating Parameters of the Power Plant and Carbon Capture Facilities. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies. (2022). Available at: https://dx.doi.org/10.2139/ssrn.4286430.

Comments: Some commenters agreed that Boundary Dam Unit 3 does not support 90 percent capture as adequately demonstrated. Commenters noted the low total capture efficiency achieved by the unit over time and the challenges faced by the unit. Other commenters argued Boundary Dam Unit 3 does support 90 percent capture as adequately demonstrated because of the capture rate from the processed flue gas and the fixes to the unit.

EPA Response: In approximately 2024, SaskPower made additional improvements at Boundary Dam Unit 3 to increase throughputs, and SaskPower noted that the capture system was processing a greater portion of the flue gas (up to 95 percent of the flue gas, with 87 percent capture from the processed flue gas, resulting in 83 percent total capture when operating).\157\ SaskPower has not reported whether Boundary Dam Unit 3 has maintained that performance in the long term.\158\ Notably, at those higher throughputs, the capture efficiency from the processed flue gas is lower. Moreover, even with those improvements, Boundary Dam continues to operate with capture efficiencies below design specification. Therefore, the fixes applied at Boundary Dam Unit 3 do not support 90 percent capture as adequately demonstrated.

\157\ U.S. EPA, “Meeting with SaskPower to Discuss CCS at Boundary Dam Unit 3” (January 18, 2024). Document ID No. EPA-HQ- OAR-2023-0072-8906.

\158\ Status updates from Boundary Dam from the second quarter of 2022 onward report average daily capture rates on a metric tons per day basis, capture system availability, and emissions intensity. Capture efficiency is not reported. See SaskPower. BD3 Status Update: Q4 2024. Available at: https://saskpower.com/about-us/our-company/blog/2025/bd3-status-update-q4-2024.

Additionally, the operating-availability of the capture system at Boundary Dam Unit 3 has been less than 100 percent.\159\ Between 2015 and 2022, annual operating-availability of the capture plant relative to the EGU varied

between 58 and 94 percent.\160\ In 2023, the average quarterly operating-availability of the capture plant was approximately 85 percent.\161\ Operating-availability remained at this level in 2024.\162\ Lower operating-availabilities further contribute to lower total capture efficiencies.

\159\ Here, operating-availability is the percent of time that the capture system is operating relative to the time that the EGU is operating.

\160\ Jacobs, B., et al. Reducing the CO2 Emission Intensity of Boundary Dam Unit 3 Through Optimization of Operating Parameters of the Power Plant and Carbon Capture Facilities. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies. (2022). Available at: https://dx.doi.org/10.2139/ssrn.4286430.

\161\ SaskPower. BD3 Status Update: Q4 2023. Available at: https://www.saskpower.com/about-us/our-company/blog/2024/bd3-status-update-q4-2023.

\162\ SaskPower. BD3 Status Update: Q4 2024. Available at: https://saskpower.com/about-us/our-company/blog/2025/bd3-status-update-q4-2024.

The total capture efficiency at Boundary Dam Unit 3 has been less than 90 percent because the capture system has not processed all the flue gas. Also, the capture efficiency is still less than 90 percent when the capture system is operating even after applying fixes. Additionally, the operating-availability of the capture system is less than 100 percent. Considering this, the EPA concludes that the experience at Boundary Dam Unit 3 does not support 90 percent CCS as adequately demonstrated. Furthermore, the capacity of Boundary Dam Unit 3 is less than the capacity of the average U.S. coal-fired steam generating unit of approximately 430 MW. Because CCS at Boundary Dam Unit 3 underperformed at 110 MW, the EPA concludes that 90 percent capture would similarly underperform at any units greater than 25 MW, including larger units.

In the 2024 CPS, the EPA argued that new solvents achieving 95 percent capture efficiency were evidence that 90 percent capture was adequately demonstrated. However, the EPA failed to reasonably account for the performance that could be achieved in practice. The only datapoint for commercial scale post-combustion CCS on a fossil fuel- fired EGU, with integrated steam and power, is Boundary Dam Unit 3. As noted in the June 2025 NPRM, it would be reasonable to anticipate that a capture system using a new solvent would underperform to a similar degree as Boundary Dam.\163\ A capture system using a new solvent, even in a process designed to achieve 95 percent capture on an instantaneous basis, would achieve just 66 percent total capture efficiency and still fail to achieve 90 percent capture if the capture system using a new solvent performed proportionately to Boundary Dam's best annual performance.\164\ Even under more optimistic circumstances, assuming Boundary Dam's best annual operating-availability and that a new solvent capture system performs proportionally to Boundary Dam's recent performance, the resulting total annual capture efficiency would be only 82 percent.\165\ In combination, a capture system would achieve much less than 90 percent total capture efficiency. On review, the EPA's prior extrapolation fails to support CCS with 90 percent capture as adequately demonstrated. Considering these factors, the EPA is finalizing the determination that CCS with 90 percent capture is not adequately demonstrated for existing coal-fired steam generating units.

\163\ 90 FR 25752, 25769 (June 17, 2025).

\164\ 95 percent design capture x (63 percent total annual capture/90 percent design capture) = 66 percent total annual capture efficiency. Between 2015 and 2022, Boundary Dam Unit 3 achieved a total annual capture efficiency of not more than 63 percent. See Jacobs, B., et al. Reducing the CO2 Emission Intensity of Boundary Dam Unit 3 Through Optimization of Operating Parameters of the Power Plant and Carbon Capture Facilities. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies. (2022). Available at: https://dx.doi.org/10.2139/ssrn.4286430.

\165\ 94 percent availability x 95 percent design capture x (83 percent total capture/90 percent design capture) = 82 percent total annual capture efficiency.

ii. CO2 Capture at Other Coal-Fired Steam Generating Units

In the 2024 CPS, to support the determination of 90 percent capture as adequately demonstrated, the EPA cited other applications of CCS at coal-fired steam generating units. These included CO2 capture at the Argus Cogeneration Plant (Trona, California), at AES's Warrior Run (Cumberland, Maryland) and Shady Point (Panama, Oklahoma) plants, and at Plant Barry's (Mobile, Alabama) 25 MWe (megawatt- equivalent) project.\166\ These projects were not of an equivalent size to commercial scale or, in the case of the Argus Cogeneration Plant, captured far less than 90 percent of CO2. These earlier examples would have informed the design of the capture system at Boundary Dam Unit 3. However, the lessons learned from such projects failed to limit the underperformance of the CO2 capture system at Boundary Dam. Consequently, they do not mitigate the anticipated underperformance in the extrapolation of 90 percent CCS and cannot support 90 percent capture as adequately demonstrated.

\166\ Dooley, J.J., et al. “An Assessment of the Commercial Availability of Carbon Dioxide Capture and Storage Technologies as of June 2009.” U.S. DOE, Pacific Northwest National Laboratory, under Contract DE-AC05-76RL01830. (2009). Available at: https://doi.org/10.2172/967229.

In the 2024 CPS, the EPA also cited the Petra Nova project at W.A. Parish Unit 8 (Thompsons, Texas). The Petra Nova project began operation in 2017, and the owner put the facility into reserve shutdown (i.e., idled) in May 2020, citing the poor economics of utilizing captured CO2 for EOR at that time. On September 13, 2023, the carbon capture facility at Petra Nova restarted.\167\ A final report from the National Energy Technology Laboratory (NETL) details the challenges that the project faced over an initial 3-year period, including leaks from heat exchangers, build-up of slurry and solids on the flue gas blower, and build-up of scale on various components.\168\ Petra Nova captured on average 92.4 percent of the CO2 from the 240 MWe flue gas processed over a 3-year period while operating. However, that does not account for emissions during outages of the CO2 capture system. Maintenance to address outages directly attributable to the CO2 capture facility was approximately 10 percent of the year on average over that timeframe. Accounting for those outages alone would, approximately, result in a total capture efficiency of 83.2 percent. Furthermore, Petra Nova processes a 240 MWe portion of the flue gas from the 610 MW W.A. Parish Unit 8. At full load, that would equate to a capture efficiency of approximately 36 percent of the emissions from the coal-fired steam generating unit.\169\ With 10 percent outages, this would be reduced further to 32.4 percent. Additionally, the 90 percent CCS BSER in the 2024 CPS was premised on the CO2 capture plant using integrated steam and electricity from the host EGU. However, Petra Nova uses an auxiliary natural gas-fired combustion turbine cogeneration unit to provide steam and electricity to the CO2 capture process, and the system does not capture the CO2 emissions from the auxiliary cogeneration unit. This design is inconsistent with the premise of the CCS BSER in the 2024 CPS. A system consistent with the premise of the BSER uses integrated steam and power and would need to meet the electricity and steam load requirements of the capture process. Furthermore, accounting for

emissions from the auxiliary cogeneration unit would lower the capture efficiency at Petra Nova further. In the 2024 CPS, by ignoring the emissions from the auxiliary cogeneration unit, the EPA failed to reasonably extrapolate the results at Petra Nova to a system using integrated steam and power. Considering these factors, the experience at Petra Nova does not support 90 percent capture as adequately demonstrated.

\167\ JX Nippon Oil & Gas Exploration Corporation. Restart of the large-scale Petra Nova Carbon Capture Facility in the U.S. (September 2023). Available at: https://www.eneos-xplora.com/english/newsrelease/upload_files/20230913EN.pdf.

\168\ W.A. Parish Post-Combustion CO2 Capture and Sequestration Demonstration Project, Final Scientific/Technical Report (March 2020). Available at: https://www.osti.gov/servlets/purl/1608572.

\169\ 90 percent x 240 MWe / 610 MW = 36 percent.

Comments: Some commenters cited additional examples of CO2 capture on coal-fired steam generating units in China. Commenters cited the 150,000 metric tons of CO2 per year Jinjie demonstration project. Commenters also cited the 500,000 metric tons of CO2 per year Taizhou CCS project, which began operation in June 2023 and captures less than 12.5 percent of the 1,000 MW EGU's total CO2 emissions. Commenters also referenced the 1.5 million metric tons of CO2 per year Longdong CCS project at the coal-fired Zhengning Power Plant. Captured CO2 will be stored in geologic storage and used for EOR. The 270 MWe project has a design capture efficiency of 95 percent from a portion of the flue gas from a 1,000 MW coal-fired EGU.\170\

\170\ State of the Art: CCS Technologies 2025. Global CCS Institute (2025). Available at: https://www.globalccsinstitute.com/wp-content/uploads/2025/08/State-of-the-Art-CCS-Technologies-2025-Global-CCS-Institute.pdf. 27 percent of the flue gas is treated.

EPA Response: The Jinjie demonstration project began operation in June 2021 and processes less than five percent of the flue gas from one of the units at the coal-fired power plant.\171\ By January 2025, the Taizhou CCS project had captured just 300,000 metric tons of CO2, far below its design basis. While a report states that this project achieves a capture rate of 90.86 percent, detailed data (e.g., operating-availability, amount of flue gas processed) is limited.\172\ According to a press release, the Longdon CCS project completed a 72-hour test on September 25, 2025, but the press release did not contain detailed performance information.\173\ Beyond that press release, no reports on the project are available. Because of the limited data available, the EPA has concluded that these projects do not support 90 percent capture as adequately demonstrated.

\171\ Bongers, N. China's impressive strides towards CCUS. Low Emission Technology Australia. (2025). Available at: https://letaustralia.com.au/wp-content/uploads/Executative-Summary-Chinas-Impressive-Strides-Towards-Carbon-Capture-Utilisation-and-Storage-CCUS.pdf.

\172\ Gong, H., et al. Taizhou 500kt per year post-combustion carbon capture demonstration project. Clean Energy, 9 (4). (2025). Available at: https://doi.org/10.1093/ce/zkaf011.

\173\ China Launches World's Largest Coal-fired Carbon Capture Project (September 29, 2025). Available at: en.sasac.gov.cn/2025/09/ 29/c_19886.htm.

iii. Variations in Performance of CO2 Capture

The determinations in the 2024 CPS assumed that the CO2 capture system is available every hour the EGU is operational and performs at its design capture efficiency (or better) during each of those hours. The EPA finds that the Agency did not adequately account for variations in performance of CO2 capture that would result in a lower capture efficiency. This further supports the conclusion that 90 percent CO2 capture is not adequately demonstrated for existing coal-fired steam generating units.

In the 2024 CPS, the EPA did not account for periodic decreases in the performance of the CO2 capture system due to solvent degradation and fouling of components between maintenance cycles. Boundary Dam Unit 3 experienced challenges with respect to solvent foaming, biological fouling, scaling, and fouling from fly- ash.174, 175 While units could take actions to address those issues, performance and capture efficiency would necessarily decrease in between treatments or maintenance (e.g., fouling would steadily accumulate after cleaning). On average, the capture efficiency would therefore be less than optimal. SaskPower indicated that even after applying such fixes, Boundary Dam Unit 3 achieved, at best, a total capture efficiency of 83 percent when the capture system was operating.\176\

\174\ Giannaris, S., et al. SaskPower's Boundary Dam Unit 3 Carbon Capture Facility--The Journey to Achieving Reliability. Proceedings of the 15th International Conference on Greenhouse Gas Control Technologies. (2021). Available at: http://dx.doi.org/10.2139/ssrn.3820191.

\175\ Pradoo, P., et al. Improving the Operating Availability of the Boundary Dam Unit 3 Carbon Capture Facility. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies. (2022). Available at: http://dx.doi.org/10.2139/ssrn.4286503.

\176\ U.S. EPA, “Meeting with SaskPower to Discuss CCS at Boundary Dam Unit 3” (January 18, 2024). Document ID No. EPA-HQ- OAR-2023-0072-8906.

Furthermore, the EPA did not adequately account for periods of startup on the operation of the capture system.\177\ After absorption, thermal energy (heat) in the form of steam is required to release the CO2 from the CO2-rich solvent and electricity is required to power the compressor to compress the CO2 for transport via pipeline. However, prior to substantial production of steam and electricity, major components of the capture process may be offline. Even assuming the capture system could consistently capture 90 percent CO2 when operating, any CO2 emitted prior to operation of the capture equipment would necessarily result in an average capture efficiency of less than 90 percent.

\177\ 89 FR 39798, 39854, 39929 (May 9, 2024).

To consistently achieve 90 percent capture on average, the source would have to overperform during certain hours. The EPA cited results from Boundary Dam that suggested higher capture efficiencies were achieved at lower throughputs.\178\ However, in its justification of the BSER, the EPA relied on an assumption that sources would operate at high capacity throughout the course of the year. If that were the case, the hypothetical higher capture efficiencies a system could potentially achieve at lower throughputs would not occur in practice. To otherwise achieve an annual average capture efficiency of 90 percent, higher instantaneous capture efficiencies likely would need to be achievable. In the 2024 CPS, the EPA cited vendor statements of pilot tests for different commercial amine solvents where operators observed higher capture efficiencies under specific conditions.\179\ However, the experience at Boundary Dam shows that it would be reasonable to anticipate that the total capture efficiency a system achieves in practice would be less than design specifications.

\178\ Jacobs, B., et al. Reducing the CO2 Emission Intensity of Boundary Dam Unit 3 Through Optimization of Operating Parameters of the Power Plant and Carbon Capture Facilities. Proceedings of the 16th International Conference on Greenhouse Gas Control Technologies. (2022). Available at: https://dx.doi.org/10.2139/ssrn.4286430.

\179\ 89 FR 39798, 39852 (May 9, 2024).

iv. Planned Projects

In the 2024 CPS, the EPA also previously cited planned projects and front-end engineering and design (FEED) studies.180, 181 However, the planned projects are neither operational nor provide measured data. While the equipment for those planned projects may have been designed for 90 or even 95 percent CO2 capture, simply designing a project for a certain percentage capture does not ensure that the project will achieve that percentage capture in practice. Boundary Dam Unit 3 did not achieve its design percentage capture, as detailed in section IV.A.1.a.i of this preamble. Therefore, because those hypothetical projects have not yet produced any data, they do not mitigate

the potential underperformance of CO2 capture, and, therefore, are not sufficient to show that 90 percent CO2 capture is adequately demonstrated.\182\ Moreover, none of the projects (for post-combustion CO2 capture from fossil fuel-fired EGUs) with feasibility or FEED studies previously cited by the EPA have moved forward to construction.

\180\ 89 FR 39798, 39851 (May 9, 2024).

\181\ See Chapter 4.4 and Table 13 of Greenhouse Gas Mitigation Measures for Steam Generating Units. Document ID No. EPA-HQ-OAR- 2023-0072-9095.

\182\ This includes projects for CCS on coal-fired steam generating units in West Virginia and Alaska that were recently selected for funding by DOE. See Project Selections for Broad Agency Announcement DE-FOA-0003605, Restoring Reliability: Coal Recommissioning and Modernization (Topic 1). Available at: https://www.energy.gov/hgeo/project-selections-broad-agency-announcement-de-foa-0003605-restoring-reliability-coal-0.

There are no post-combustion CCS applications on fossil fuel-fired EGUs that have begun operation since the finalization of the 2024 CPS that are sufficient to support 90 percent capture as adequately demonstrated. Rather, some of the planned projects cited in the 2024 CPS either have been abandoned or have faced other challenges. Project Diamond Vault was a planned project to capture up to 95 percent of CO2 emissions from the 600 MW Madison Unit 3 at Brame Energy Center in Lena, Louisiana.\183\ The FEED study and current plans for carbon capture were abandoned in late 2024.\184\ Project Tundra is a carbon capture project in North Dakota at the Milton R. Young Station lignite coal-fired power plant that planned for the capture plant to treat the flue gas from the 455 MW Unit 2 and some additional flue gas from the 250 MW Unit 1 (an equivalent capacity of 530 MW in total).\185\ TC Energy, a primary sponsor of Project Tundra, has since withdrawn from the project, although the project may continue to move forward depending on various factors.\186\ The timeframes for several other CCS projects on coal-fired EGUs are unclear.\187\

\183\ Project Diamond Vault Overview. Document ID No. EPA-HQ- OAR-2025-0124-0027.

\184\ Cleco Corporate Holdings, LLC SEC Form 10Q, at 51 (August 18, 2024). Available at: https://www.sec.gov/Archives/edgar/data/18672/000108981924000026/cnl-20240630.htm.

\185\ “An Overview of Minnkota's Carbon Capture Initiative-- Project Tundra,” 2023 LEC Annual Meeting (October 5, 2023).

\186\ Power Engineering. Key partner withdraws from large-scale CO2 capture project. Available at: https://www.power-eng.com/environmental-emissions/carbon-capture-storage/key-partner-withdraws-from-large-scale-co2-capture-project/.

\187\ D. Gearino, A Carbon Capture Project Faces a New Delay in a Year of Slow Progress for Coal Power Plants Looking for Retrofits, Inside Climate News (December 10, 2024). Available at: https://insideclimatenews.org/news/10122024/north-dakota-coal-plant-carbon-capture-project-faces-new-delay/.

The EPA has re-evaluated the costs and associated assumptions of 90 percent CCS on existing long-term coal-fired steam generating units and is finalizing the determination that the costs are not reasonable based on the rationale detailed in this section of the preamble. i. Capacity Factor, Effective Capture Efficiency, and Other Assumptions

In the 2024 CPS, costs for CCS on existing coal-fired steam generating units were determined assuming a best-case scenario. Specifically, the cost assessment assumed sources operated at high annual capacity factors (80 percent) and that the CO2 capture equipment was available and performing optimally every hour the EGU was operating. However, in 2023, coal-fired EGUs had an average capacity factor of 42 percent.\188\ Lower capacity factors typically result in less revenue from electricity generation. Moreover, as detailed in the preceding section of this preamble, even with a design capture efficiency of 90 percent, the actual total capture efficiency over the course of the year is lower, and under some circumstances significantly lower. Consequently, less CO2 captured (due to lower actual capture efficiency, lower EGU capacity factor, or both) results in higher costs due to reduced revenue from the IRC section 45Q tax credit.\189\

\188\ U.S. Energy Information Administration. Electric Power Annual. Available at: https://www.eia.gov/electricity/annual/.

\189\ These tax credits are currently available for a twelve- year period for facilities that commence construction before January 1, 2033, and can be used to offset tax liability. 26 U.S.C. 45Q (2025).

Furthermore, rather than directly considering the costs for any operation after the expiration of availability of the IRC section 45Q tax credit for existing coal-fired steam generating units in the 2024 CPS, the EPA committed to review the requirements of the emission guidelines pertaining to existing coal-fired steam generating units by January 1, 2041, and posited that other mechanisms for potential valuation of EGUs operating with 90 percent CCS could arise in the future.\190\ However, those assumptions are no longer reasonable because the EPA believes that coal-fired steam generating units are now more likely to operate longer than they will be able to claim the tax credit. As noted in the June 2025 NPRM, the EPA believes that coal- fired steam generating unit capacity and generation will continue to comprise a substantial portion of the nation's electricity supply.\191\ A number of coal-fired steam generating units are delaying or canceling their scheduled retirements in light of increasing electricity demand, among other factors.\192\ The EPA's projections further show a substantial capacity of coal-fired steam generating units operating in the long term.\193\ Based on a lower capacity factor and operation beyond 12 years, the EPA proposed that, even if the IRC section 45Q tax credit is accounted for as a reduction, the costs are unreasonable and solicited comment on the assumptions in evaluation of the reasonableness of the cost of the BSER.

\190\ 89 FR 39798, 39902 (May 9, 2024).

\191\ See 90 FR 25752, 25772, 25774 (June 17, 2025).

\192\ D. Proctor, U.S. Coal Plants Get Reprieve as Market and Policies Change, Power (February 6, 2025). Available at: https://www.powermag.com/u-s-coal-plants-get-reprieve-as-market-and-policies-change/.

\193\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

Comments: Some commenters agreed with the EPA's capacity factor assumptions in the June 2025 NPRM for evaluating the cost of CCS as BSER. Other commenters stated that the 80 percent capacity factor assumed for cost calculations in the 2024 CPS was reasonable and argued that the availability of the IRC section 45Q tax credit would incentivize higher capacity factors. Commenters also argued that the EPA's June 2025 NPRM was internally inconsistent, noting that elsewhere the June 2025 NPRM stated that “coal-fired steam generating unit capacity and generation will continue to comprise a substantial portion of the nation's electricity supply.” \194\ Some commenters disagreed with the EPA's analysis based on a capture system underperforming (i.e., designed for 90 percent capture, but achieving 75 percent total capture efficiency in practice). Commenters argued that the 75 percent total capture efficiency was unjustified.

\194\ 90 FR 25774 (June 17, 2025).

EPA Response: As detailed in section IV.A.1.a of this preamble, it is reasonable to anticipate that a capture system designed to achieve a given capture efficiency would underperform. While CCS may achieve emission reductions, evidence shows the CO2 capture system underperforms. The capture system at Boundary Dam Unit 3 was designed to achieve 90 percent capture but achieved, at best, 63 percent total capture efficiency on an annual basis between 2015 and 2022. Under a set of assumptions that reflect the underperformance of CCS, lower capacity factors, and the limited

availability of the IRC section 45Q tax credit,\195\ the costs are substantially higher ($62/MWh, $124/ton of CO2 reduced) than those determined in the 2024 CPS and more than three times higher on a $/MWh basis than the costs the EPA has previously determined to be reasonable ($18.50/MWh).\196\ Even assuming a higher capacity factor that reflects the average capacity factor of coal-fired steam generating units of approximately 70 percent in the updated baseline projection,\197\ costs remain high ($50/ton, $25/MWh).\198\ Such high costs, particularly on a $/MWh basis, are not reasonable, even considering the potential CO2 emission reductions, and do not support 90 percent CCS as BSER.

\195\ These costs include the costs of capital equipment, etc., consistent with 90 percent design capture rate, 63 percent actual capture rate, a fixed 40 percent capacity factor, and 15-year booklife (12 years of 45Q availability, three years without). Costs are expressed in 2019$. See memorandum entitled Updated Evaluation of Best System of Emission Reduction Costs of Carbon Capture and Sequestration/Storage at Existing Coal-Fired Electric Generating Units in the docket for this rulemaking.

\196\ Costs are expressed in 2019$. In a variety of rulemakings, the EPA has required coal-fired EGUs to install and operate flue gas desulfurization (FGD, or wet scrubbers) to reduce their SO2 emissions. The annualized cost of installing these controls on a representative 700 to 300 MW coal-fired steam generating unit are $14.80 to $18.50/MWh. Hence control costs that are generally consistent with these values should be considered reasonable. See 89 FR 39798, 39882 (May 9, 2024).

\197\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

\198\ These costs include the costs of capital equipment, etc., consistent with 90 percent design capture rate, 63 percent actual capture rate, a fixed 70 percent capacity factor, and 15-year booklife (12 years of 45Q availability, three years without). Costs are expressed in 2019$. See memorandum entitled Updated Evaluation of Best System of Emission Reduction Costs of Carbon Capture and Sequestration/Storage at Existing Coal-Fired Electric Generating Units in the docket for this rulemaking.

ii. The IRC Section 45Q Tax Credit

The costs for 90 percent CCS are even higher if the IRC section 45Q tax credit is not accounted for as a reduction ($77/MWh and $155/ ton).\199\ In the 2024 CPS, the costs of 90 percent CCS for existing coal-fired steam generating units accounted for the IRC 45Q tax credit by reducing the direct costs to the source for every ton of CO2 reduced. The 2024 CPS assessed costs over a period consistent with the 12-year availability of the IRC section 45Q tax credit. The 2024 CPS justified that position on grounds that CAA section 111(a)(1) requires the EPA, in determining the BSER, to account for “the cost of achieving such [emissions] reduction,” and asserted that this provision refers to the cost to the source rather than the societal cost. The Inflation Reduction Act (IRA) extended and expanded the IRC section 45Q tax credit and included CAA section 135(a)(6), requiring the EPA to promulgate regulations under the CAA's authorities to ensure reductions in GHG emissions. In the 2024 CPS, the EPA further stated that the IRA included legislative history stating that Congress intended to authorize the Agency to promulgate regulations under CAA section 111 to reduce GHGs from fossil fuel-fired power plants, including regulations based on CCS that assumed lower cost due to the 45Q tax credit.\200\ In its proposed repeal of the 2024 CPS, the EPA reevaluated this position and proposed that reducing control costs by the amount of the tax credit is an incorrect accounting for the costs of control, and solicited comment on this position.\201\

\199\ These costs include the costs of capital equipment, etc., consistent with 90 percent design capture rate, 63 percent actual capture rate, a fixed 70 percent capacity factor, and 15-year booklife (no reduction in cost from 45Q). Costs are expressed in 2019$. See memorandum entitled Updated Evaluation of Best System of Emission Reduction Costs of Carbon Capture and Sequestration/Storage at Existing Coal-Fired Electric Generating Units in the docket for this rulemaking.

\200\ 89 FR 39798, 39881 (May 9, 2024).

\201\ 90 FR 25752, 25772 (June 17, 2025).

(A) Comments and Responses

Comments: Some commenters opposed the June 2025 NPRM and stated that the EPA must account for the IRC section 45Q tax credit as a reduction when evaluating the costs of 90 percent CCS. Commenters argued that the EPA has long understood the “cost” in CAA section 111(a)(1) to refer to whether the cost to the regulated source would be too great to implement the technology.\202\ Some commenters argued that counting the IRC section 45Q tax credit as a reduction in costs in the evaluation of reasonableness of the costs is the best reading of CAA section 111(a)(1), which directs EPA to “tak[e] into account the cost of achieving such reduction,” when determining the BSER, not costs generally. Commenters stated the phrase “such reduction” refers to the emission reduction “achieve[ed] through the application of the best system of emission reduction.” Because sources are the entities that apply the best system, the phrase “cost of achieving such reduction” is best read as focusing on costs borne by those sources, rather than broader economic impacts. Commenters argued the EPA has followed this approach since the beginning of the regulatory program.\203\

\202\ Commenters cited Portland Cement Ass'n, 513 F.2d at 508 (“The industry has not shown inability to adjust itself in a healthy economic fashion to . . . the standards prescribed.”).

\203\ Commenters cited Portland Cement Ass'n, 486 F.2d at 388, observing that the ruling explained that the “cost” analysis is focused on equipment and operating costs and rejected an argument that a broader cost-benefit analysis evaluating impacts was required because that would “conflict with the specific time constraints imposed on the administrator.”

Other commenters agreed with the June 2025 NPRM and stated that the EPA should not consider the IRC section 45Q tax credit as a reduction when evaluating the cost of CCS as a potential BSER. One commenter stated that the statute requires consideration of total costs and cited the Supreme Court's decision in Michigan.\204\ Commenters claimed that the decision held that the EPA must consider all costs, including system and indirect costs, when evaluating the reasonableness of a standard. Some commenters argued that tax credits do not reduce the cost of 90 percent CCS but simply shift those costs to taxpayers. A few commenters further stated that CAA section 111(a)(1) includes consideration of societal costs beyond those to a source's owner or operator. One commenter argued that the inclusion of “cost” in the parenthetical of CAA section 111(a)(1), along with other factors that account for societal disbenefits, suggests that the best reading of CAA section 111(a)(1) is that the EPA should not focus solely on the cost to the source when evaluating the BSER and, therefore, should not account for the IRC section 45Q tax credit as reducing costs.

\204\ Commenters cited 576 U.S. at 752-53.

EPA Response: The EPA agrees with the commenters who stated that the 2024 CPS erred in excluding the value of the IRC section 45Q tax credit from the cost of CCS by counting the tax credit as a reduction in the cost of CCS. CAA section 111(a)(1) provides that the EPA must determine “the best system of emission reduction . . . (taking into account the cost of achieving such reduction and any nonair quality health and environmental impact and energy requirements).” This provision does not, by its terms, limit the costs to those incurred directly by the source. Under Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), the best interpretation of this provision is that the costs include the full costs of the controls (i.e., without reduction by the amount of the IRC section 45Q tax credit). Under the justification in the 2024 CPS for considering the IRC section 45Q tax credit, if a cost was passed on to the public through a large tax credit or other similar subsidy, and

that transfer was counted as a reduction, then much more expensive controls would be considered reasonable as long as the cost of those controls were passed on to the public. As commenters noted, the tax credit does not eliminate costs, it simply transfers costs to the U.S. taxpayer. Thus, the costs paid by the U.S. taxpayer, in the form of the reduction in tax receipts due to the IRC section 45Q tax credit, are part of “the cost of achieving [the emission] reduction,” under CAA section 111(a)(1). As commenters also noted, this interpretation treats “costs” as consistent with the other factors that CAA section 111(a)(1) directs the EPA to consider because those other factors are not limited to the source. Specifically, CAA section 111(a)(1) directs the EPA to consider “nonair quality health and environmental impact[s]” that affect the public and “energy requirements,” which include effects on the broader energy system.

Comments: Commenters argued that counting the IRC section 45Q tax credit as a reduction in costs is consistent with Congressional intent. Commenters stated that Congress enacted the IRC section 45Q tax credit specifically to encourage CCS deployment because Congress had determined such deployment as sufficiently valuable to justify the cost, and that Congress amended the CAA in the IRA and directed the EPA to regulate with the expanded tax incentives of the IRA in mind.\205\ Commenters stated that if Congress wants to keep a federal funding program from affecting a BSER determination under CAA section 111, Congress would include legislative text stating so, as Congress similarly did for the funding provided as part of the Energy Policy Act of 2005 (EPAct05). Commenters note that no similar provision exists for the IRC section 45Q tax credit. Some commenters took issue with the June 2025 NPRM's observation that there was pending legislation that would have ended the tax credit. Commenters observed that, rather than eliminate the tax credit, Congress in the OBBBA expanded the tax credit by increasing the value for CO2 used for EOR and implementing a more favorable inflation adjustment.

\205\ Commenters cited the Inflation Reduction Act, Public Law 117-169, section 13104, 136 Stat. 1818, 1924-29 (2022).

EPA Response: CAA section 135(a)(6), as adopted by the IRA, provides $18 million to the EPA “to ensure that reductions in [GHG] emissions are achieved through the use of the existing authorities of [the CAA], incorporating [an] assessment” that the EPA is required to conduct of reductions in GHG emissions from changes in domestic electricity generation and use through fiscal year 2031.\206\ However, this provision does not mention CAA section 111 and thus by its terms is not specific enough to indicate Congressional intent to authorize the EPA to promulgate CAA section 111 regulations based on CCS as the BSER and, in doing so, to account for the cost of CCS after reductions by the IRC section 45Q tax credit.

\206\ 42 U.S.C. 7435(a)(6).

The EPA further disagrees with commenters' objection that if Congress had intended for the Agency not to consider the reduction in control costs due to the IRC section 45Q tax credit, Congress would have explicitly said so. Commenters note that in the EPAct05, Congress included provisions explicitly precluding the EPA from considering projects that included emissions controls that had been funded through EPAct05 in determining whether those controls are adequately demonstrated under CAA section 111. Absent such provisions, CAA section 111 would have allowed the EPA to consider those projects. As noted in this section of the preamble, the best interpretation of CAA section 111 is that the EPA must consider the costs of the control device, whether the costs are incurred by the facility or the taxpayer through the IRC section 45Q tax credit, and that if Congress had intended that the EPA not to do so, Congress would have included a specific provision to that effect.

Comments: Some commenters stated that in Michigan v. EPA, the U.S. Supreme Court reiterated that in promulgating rulemakings, Federal agencies “are required to engage in `reasoned decisionmaking,' ” \207\ and that “ordinarily requires paying attention to the advantages and the disadvantages of agency decisions.” \208\ Commenters stated that the failure of the EPA to consider the burden on the average U.S. taxpayer of the IRC section 45Q tax credit constituted a failure of reasoned decisionmaking.

\207\ 576 U.S. at 750 (citing Allentown Mack Sales & Service, Inc. v. NLRB, 522 U.S. 359, 374 (1998)).

\208\ Id. at 753 (emphasis omitted).

EPA Response: The EPA agrees with the commenters, and that considering the burden on U.S. taxpayers tilts this action against adopting CCS as the BSER. The EPA is finalizing that the IRC section 45Q tax credit should not be accounted for as a reduction in costs to the source when evaluating the reasonableness of the costs of the BSER. Without taking into account the IRC section 45Q tax credit as a reduction, the costs of 90 percent CCS as BSER ($77/MWh and $155/ton) are unreasonable for long-term coal-fired steam generating units.\209\

\209\ These costs include the costs of capital equipment, etc., consistent with 90 percent design capture rate, 63 percent actual capture rate, a fixed 70 percent capacity factor, and 15-year booklife (no reduction in cost from 45Q). Costs are expressed in 2019$. See memorandum entitled Updated Evaluation of Best System of Emission Reduction Costs of Carbon Capture and Sequestration/Storage at Existing Coal-Fired Electric Generating Units in the docket for this rulemaking.

The large, widespread, and third-party infrastructure needed to support CCS is unique as compared to other BSER control technologies the EPA has historically analyzed and adopted under CAA section 111. In the 2024 CPS, the EPA determined that the capture, pipeline, and sequestration infrastructure necessary for the affected sources to meet the standards could be deployed by the compliance date of January 1, 2032. However, that position relied on incorrect assumptions in an unrealistic, best-case scenario that experience has already shown to be inaccurate. Therefore, the EPA proposed that the degree of emission limitation is not achievable because it is unlikely that the necessary infrastructure can be deployed by that compliance date.

In general, the capture, pipeline, and sequestration infrastructure necessary for 90 percent CCS for the fleet of existing coal-fired steam generating units does not currently exist. The necessary infrastructure would need to be broadly deployed, and there are challenges that exist for sources (e.g., pipeline permitting and right-of-way) that may not be able to be resolved. It is highly unlikely, if not impossible, that the infrastructure necessary for CCS can be deployed by the January 1, 2032, compliance date, and the EPA is therefore finalizing that the degree of emission limitation in the 2024 CPS for long-term coal-fired steam generating units is not achievable.

With respect to the timeline for implementing infrastructure that does not currently exist, the EPA is revising both the weight the Agency should place on such future projections and its future projections with respect to 90 percent CCS. In the 2024 CPS, the Agency took the general position that an adequately demonstrated control technology could be selected as the BSER, and therefore the basis for a standard, so long as implementation at the scale required for compliance to be feasible could be projected to a date certain. However, upon further consideration and in light of public

comments, the EPA is now clarifying that the uncertainty associated with such projections is a reason to disfavor the selection of such BSERs. The eight-year review cycle laid out in CAA section 111(b)(1)(B) is a signal that Congress thought the BSER generally should be capable of being implemented within eight years such that the EPA's review can meaningfully assess the success of the prior rule.210 211 In the 2024 CPS, the EPA estimated that it would take approximately seven years to implement 90 percent CCS for existing coal-fired units. The 2024 CPS also provided an exemption under which sources were not subject to the rule if they closed by the date that compliance with the standard based on 90 percent CCS would have commenced (January 1, 2032). However, as explained in section IV.A.1.c of this preamble, the EPA is now determining that the January 1, 2032 compliance date for 90 percent CCS for existing coal-fired steam generating units was overly optimistic. That is, implementing 90 percent CCS would take longer than the seven years provided in the 2024 CPS, bringing it close to CAA section 111(b)(1)(B)'s eight-year benchmark, if not exceeding it.

\210\ CAA section 111(b)(1)(B) provides that the Administrator shall review and, if appropriate, revise NSPS at least every eight years. However, the Administrator need not review any such standard if the Administrator determines that “such review is not appropriate in light of readily available information on the efficacy of such standard.” 42 U.S.C. 7411(b)(1)(B). The EPA believes that the eight-year review cycle is relevant to existing sources regulated under CAA section 111(d) because it speaks to the appropriateness of a system of emission reduction as the BSER, and the BSER considerations under CAA section 111(a)(1) apply to both new and existing sources.

\211\ In the event a system of emission reduction would take longer than eight years to implement, the EPA should take a hard look at whether it in fact qualifies as the BSER under CAA section 111(a)(1). The result of this inquiry may depend on whether other, less time-intensive options are available.

With respect to the infrastructure itself, the EPA has, upon further review, determined that the novelty of a BSER predicated on widespread infrastructure operated by third parties warns strongly against its selection. Much of the necessary CCS infrastructure requires the involvement, both for initial development and ongoing performance, of external third parties over whom the owners and operators of regulated facilities have limited control. Each of the three components of CCS--capture, transport, and storage--entail distinct infrastructure projects completed and operated in most instances by different sets of third parties, and all three components would need to be timely completed by those different sets of third parties without delay for owners and operators to be able to implement the CCS requirements by the January 1, 2032, compliance date. Construction of a CO2 capture facility requires years of engineering analysis by third-party experts before ground can be broken and installation of the necessary equipment can commence.\212\ Development of a capture facility also includes entering into necessary agreements and procuring permits through processes that may be governed by multiple jurisdictions, including but not limited to federal and state permitting authorities. And construction itself takes two or more years and involves many different activities that require a range of expert third parties.\213\ In addition to the CO2 capture facility, CCS also requires implementation of CO2 pipeline infrastructure and CO2 injection and storage infrastructure which involves reliance on engineers, geologists, construction firms, and permitting authorities, at least some of whom are likely separate from those involved in design and construction of the capture facility, as well as being outside of the electricity generation sector.

\212\ Sargent & Lundy, CO2 Capture Project Schedule and Operations, Document ID No. EPA-HQ-OAR-2023-0072-9095, at 2-3 (April 2024).

\213\ Id. at 3-4.

A BSER based on 90 percent CCS thus requires involvement, at multiple steps, of a complex web of experts and practitioners across a wide range of disciplines, many of whom may not be part of the regulated source category, or even the broader industry of which it is part, i.e., the electric power industry. The implication is that these third parties are not subject to the same regulatory impetus as the owner or operator of a coal-fired steam generating unit that would be subject to requirements under the 2024 2024 CPS. The scale of the ancillary infrastructure required to support implementation of 90 percent CCS and the multitude of parties outside of an owner or operator's control at multiple steps of the process that are necessary to successfully design, permit, construct, test, and operate a CCS system makes it unreasonable to base federally enforceable requirements on the presumption of a best-case scenario for implementation when tardy- or non-performance by a single party has the potential to cause or exacerbate delays that ripple through the deployment of the CCS system.

The tenuousness of timely compliance given the scope of the requisite infrastructure for 90 percent CCS, coupled with the novelty of CCS and the need for a large number of third parties, is especially apparent when it is contrasted with air pollution control technologies that have been in widespread use in the power sector for decades, such as scrubbers for SO2 emissions and selective catalytic reduction (SCR) for NOX emissions.\214\ Those latter control technologies do not rely on extensive pipeline and sequestration infrastructure. In contrast, the infrastructure to implement CO2 capture in the power sector at a large scale is extensive, relatively new,\215\ and involves a large collection of parties and activities necessary for design, permitting, construction, and implementation. All this compounds the likelihood of delays in implementation as parties work through necessary learning processes, which further supports the unreasonableness of the 2024 CPS's aggressive compliance timeframe. To be sure, CCS has been employed in some industries for certain limited purposes. However, in those instances, CCS has not been required by federal regulation under enforceable timelines or at a similar nationwide scale, meaning that any non-performance by third parties has not resulted in noncompliance with federally enforceable obligations.

\214\ See, e.g., 44 FR 33580, 33580 (June 11, 1979) (promulgating standards of performance for SO2 emitted from new, modified, and reconstructed fossil fuel-fired steam generating EGUs under CAA section 111(b) based on use of scrubbers); 63 FR 49442, 49445 (Sept. 16, 1998) (promulgating standards of performance for NOX emitted from new fossil fuel-fired steam generating units under CAA section 111(b) based on use of SCR).

\215\ While CCS has a longer history of implementation in other industries, such as ethanol production, it has not been broadly deployed in the power sector at the capture rates contemplated in the 2024 CPS. Similarly, while the 2015 NSPS for new coal-fired power plants established a standard of performance based on partial- CCS, 80 FR 64510, 64545 (Oct. 23, 2015), no new coal-fired power plants have been constructed that would be subject to this standard. Therefore, the power sector's experience with CCS has been relatively limited compared to other, similarly complex pollution controls.

In reaching these conclusions, the EPA notes that CAA section 111 anticipates analyses and regulatory requirements that turn on actions by the regulated source. Pursuant to CAA section 111(a)(1), the Agency must promulgate standards that “reflec[t] the degree of emission limitation achievable through the application of the [BSER],” and the resulting emission standards apply, in turn, to the regulated source. Consistent with the EPA's historical understanding, this language is most naturally read as tying the BSER to results that are achievable through the source's application of the selected

control technology. While sources commonly rely on third parties in the normal operation of their business, the extensive infrastructure requirements for CCS, coupled with the novelty of the control technology and the high level of dependence on third parties, undermines the ability of sources to achieve the CCS-based emission standards. If any third party provider responsible for any of the many links in the CCS infrastructure chain fails to develop the necessary infrastructure in a particular area, is delayed in such development for any reason, or ceases operation for any reason, the source would no longer be able to comply. These circumstances are different in kind from other BSERs that entail substantially less novel infrastructure.

The EPA is thus finding that the 2024 CPS erred by not considering the extent of the ancillary infrastructure and reliance on third parties needed for 90 percent CCS to be deployable at scale. This lapse is especially salient given the novelty of CCS at this scale and applied to this particular sector. These considerations provide further support for the conclusion that the January 1, 2032, compliance date and the degree of emission limitation are unachievable.

Comments: Some commenters stated that 90 percent CCS is achievable because the necessary infrastructure can be deployed by the compliance deadline. Commenters reiterated the arguments the EPA previously made in the 2024 CPS, including asserting that the timeline for deployment of capture based on the Sargent and Lundy report was achievable.\216\ Commenters also stated that sequestration potential is broadly available, and that the 2024 CPS was premised on smaller, often intrastate CO2 pipelines from the source to those storage sites. Commenters also stated that the EPA has made significant progress toward granting additional States primacy over Class VI injection wells for geologic storage of CO2.217 218

\216\ Sargent & Lundy, CO2 Capture Project Schedule and Operations, Document ID No. EPA-HQ-OAR-2023-0072-9095, at A-1 (April 2024).

\217\ Class VI wells are used to inject CO2 into deep rock formations and are regulated by the EPA under the Underground Injection Control (UIC) Program as authorized by the Safe Drinking Water Act. Available at: https://www.epa.gov/uic/class-vi-wells-used-geologic-sequestration-carbon-dioxide#authorities.

\218\ Injection wells are overseen by either a state or tribal agency or one of EPA's regional offices. States and tribes may apply for primary enforcement responsibility to implement the UIC program. Primary enforcement responsibility, often called primacy, refers to State, territory, or Tribal responsibilities associated with implementing EPA approved UIC programs. A State, territory, or Tribe with UIC primacy, or primary enforcement responsibility, oversees the UIC program in that State, territory, or Tribe. Available at: https://www.epa.gov/uic/class-vi-wells-used-geologic-sequestration-carbon-dioxide.

Other commenters stated that 90 percent CCS is not achievable because the necessary infrastructure cannot be deployed fast enough to meet the compliance deadline. Commenters argued that the timeline for deployment of the capture equipment would take longer than detailed in 2024 CPS. Commenters asserted that FEED studies can require more than 12 months, and DOE FEED studies can require even longer due to additional reporting requirements. Commenters stated that deployment of a capture facility could take up to eight to 10 years. Some commenters noted that permitting for CO2 pipelines varies by State, and that some States have restrictive policies for CO2 pipelines. Some commenters stated that the Pipeline Hazardous Materials Safety Administration is still working on updated regulations for CO2 pipeline safety.

EPA Response: In general, the capture, pipeline, and sequestration infrastructure necessary for 90 percent CCS for the fleet of existing coal-fired steam generating units does not currently exist.

The equipment for the capture of CO2 takes time to design, permit, and install. In the 2024 CPS, the EPA assumed an aggressive, unrealistic timeline for deployment of capture equipment. The EPA's timeline for installation of capture equipment included a 12- month FEED study in place of an 18-month FEED study, based off the more aggressive project schedule in a report developed by Sargent and Lundy.\219\ The EPA further abbreviated that schedule by two months based on its own assumptions by shortening the duration for commercial arrangements from nine months to seven months, assuming sources immediately begin sitework once permitting is complete, and accounting for 13 months (rather than 14) for startup and testing.\220\ However, those assumptions ignore any potential delays and do not reflect what is actually achievable.\221\ The necessary infrastructure would need to be broadly deployed, and there are significant challenges that could exist for sources that would need to be resolved.

\219\ Sargent & Lundy, CO2 Capture Project Schedule and Operations, Document ID No. EPA-HQ-OAR-2023-0072-9095, Attachment 17 (April 2024).

\220\ 89 FR 39798, 39875 (May 9, 2024).

\221\ See Nat'l Lime Ass'n v. EPA, 627 F.2d 416, 432-33 (D.C. Cir. 1980) (EPA must explain how the standard is “achievable under the range of relevant conditions” which sources may experience).

Regarding transport of CO2, there is no existing network of CO2 pipelines with the capacity capable of meeting the demands in the 2024 CPS. There are approximately 5,000 miles of CO2 pipelines operational in the U.S.\222\ \223\ However, they are largely not located near existing coal-fired sources and additional pipelines would take time to deploy. Planned CO2 pipelines continue to face delays due to several factors, including state permitting and the challenges associated with eminent domain authority and negotiating rights-of-way. For example, after promulgation of the 2024 CPS, Summit Carbon Solutions paused their application for a pipeline in South Dakota after the State banned eminent domain for CO2 pipelines.\224\ \225\ A similar law is progressing through the Iowa legislature.\226\

\222\ Congressional Research Service. Carbon Dioxide Pipelines: Safety Issues, CRS Reports (June 3, 2022). Available at: https://www.congress.gov/crs-product/IN11944.

\223\ U.S. Department of Transportation, Pipeline and Hazardous Materials Safety Administration. Annual Report Mileage for Hazardous Liquid or Carbon Dioxide Systems (May 1, 2026). Available at: https://www.phmsa.dot.gov/data-and-statistics/pipeline/annual-report-mileage-hazardous-liquid-or-carbon-dioxide-systems.

\224\ J. Chilson, Summit pauses CO2 pipeline application in South Dakota, South Dakota Public Broadcasting (March 12, 2025). Available at: https://www.sdpb.org/business-economics/2025-03-12/summit-pauses-co2-pipeline-application-in-south-dakota.

\225\ An Act to prohibit the exercise of eminent domain for a pipeline that carries carbon oxide, South Dakota Legislature House Bill 1052, 100th Session, H.J. 475 (March 6, 2025). Available at: https://sdlegislature.gov/Session/Bill/25581.

\226\ C. Koons, House votes to ban eminent domain for CO2 pipelines, Iowa Capital Dispatch (March 26, 2025). Available at: https://iowacapitaldispatch.com/2025/03/26/house-votes-to-ban-eminent-domain-for-co2-pipelines/.

Furthermore, while the U.S. has broad availability of the geologic formations that are potentially suitable for CO2 sequestration, existing storage infrastructure for sequestration of CO2 is limited. Underground CO2 storage is governed by the Underground Injection Control program as authorized by the Safe Drinking Water Act.\227\ Under that program, Class VI wells are used to inject CO2 thousands of feet underground for geologic storage, and are permitted by the EPA. There are 18 Class VI wells that have been permitted by the EPA and there are six states that have primary enforcement authority (Arizona, Louisiana, North Dakota, Texas, West Virginia, and Wyoming).\228\

In the 2024 CPS, the EPA based assumptions on the availability of “potential” storage sites. Time is required to characterize those sites to ensure the geology in the project area can receive and contain the CO2 within the zone where it will be injected. However, the nearest available “potential” site may, after further investigation, not ultimately be suitable, e.g., if faults or fractures are detected during site characterization. More time would then be required to find and characterize a new storage site, if another suitable site is even available. The timeline in the 2024 CPS did not take into consideration the prospect of project developers having to pivot to a different storage site should the initial site prove unsuitable. Development of planned storage sites may also face delays due to permitting and other issues. These challenges to deployment of CCS provide further support to the conclusion that the January 1, 2032, compliance date and degree of emission limitation are unachievable.

\227\ EPA. Class VI Wells used for Geologic Sequestration of Carbon Dioxide (May 14, 2026). Available at: https://www.epa.gov/uic/class-vi-wells-used-geologic-sequestration-carbon-dioxide#ClassVI_PermittingProcess.

\228\ EPA. Current Class VI Projects under Review at EPA. Accessed June 2, 2026. Available at: https://www.epa.gov/uic/current-class-vi-projects-under-review-epa.

Moreover, a greater number of sources would likely be subject to CCS-based requirements than previously anticipated. Those sources would be competing to build CCS infrastructure, exacerbating any potential schedule delays or supply constraints, further limiting the achievability of the EPA's infrastructure timeline in the 2024 CPS. As the EPA noted in the June 2025 NPRM, the Agency believes that coal- fired steam generating unit capacity and generation will continue to comprise a substantial portion of the nation's electricity supply due to increasing electricity demand.\229\ A number of coal-fired steam generating units are delaying or canceling their scheduled retirements in light of this increasing demand, the changes in tax incentives for various types of electricity-generating resources in the OBBBA, and Administration actions to support the continued operation of coal-fired capacity.230 231 Recent changes in the U.S. electricity market affect the amount of coal-fired steam generating units operating in the long-term, as described in section III.C of this preamble, including increased electricity demand from data centers and scaled back tax credits for renewable generation under the OBBBA. The EPA now projects that more existing coal-fired EGUs will operate in the long term and more base load NGCCs will be built than previously anticipated.\232\ Specifically, at the end of 2024 there were 174 GW of coal-fired EGUs active in the power sector nationwide.\233\ The baseline 2024 analysis for the 2024 CPS projected that, absent requirements, approximately 40 GW of coal capacity would still be active by 2040.\234\ Therefore, the 2024 CPS considered the viability and reasonableness of installing and operating CCS at 40 GW of coal capacity. However, the EPA now projects that, as a baseline, approximately 100 GW of coal capacity will be active in 2040.\235\ Similarly, under the analysis conducted for the 2024 CPS, the EPA projected approximately 26 GW of incremental NGCC capacity additions by 2035.\236\ Under the 2025 analysis, absent the requirements of 2024 CPS, the EPA projects approximately 155 GW of new NGCC builds by 2035. Based on the EPA's updated analytics a much larger number of EGUs would be subject to CCS-based standards than previously estimated in the 2024 CPS. The large amount of CCS infrastructure necessary would likely exacerbate any potential schedule delays or supply chain constraints. Deploying the necessary infrastructure for the affected fleet by the January 1, 2032, compliance date is therefore further unlikely.

\229\ See 90 FR 25752, 25772, 25774 (June 17, 2025).

\230\ D. Proctor, U.S. Coal Plants Get Reprieve as Market and Policies Change, Power (February 6, 2025). Available at: https://www.powermag.com/u-s-coal-plants-get-reprieve-as-market-and-policies-change/.

\231\ See, e.g., U.S. Department of Energy, 2026 DOE 202(c) Orders. Available at: https://www.energy.gov/ceser/2026-doe-202c-orders.

\232\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

\233\ U.S. Energy Information Administration. EIA Power Monthly (December 2024). Available at: https://www.eia.gov/electricity/monthly/archive/december2024.pdf.

\234\ Document ID No. EPA-HQ-OAR-2023-0072-8913. Table 3-14.

\235\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

\236\ Document ID No. EPA-HQ-OAR-2023-0072-8913. Table 3-14.

Considering these factors, it is unlikely the infrastructure necessary for CCS can be deployed by the January 1, 2032 compliance date, and the EPA is therefore finalizing that the degree of emission limitation in the 2024 CPS for long-term coal-fired steam generating units is not achievable. d. Conclusion

Because the EPA is finalizing that 90 percent CCS is not adequately demonstrated as the BSER and that the cost of 90 percent CCS for long- term coal-fired steam generating units is not reasonable, the Agency is finalizing the determination that 90 percent CCS is not the BSER for long-term coal-fired steam generating units. Furthermore, because it is extremely unlikely that the extensive infrastructure necessary for CCS can be deployed by the January 1, 2032 compliance date, the EPA is finalizing a determination that the degree of emission limitation in the 2024 CPS for long-term coal-fired steam generating units is not achievable. Additionally, the challenges posed by the scope of the infrastructure are heighted by the involvement of numerous parties external to the owners and operators of the regulated facility, which complicates deployment and exacerbates delays that thus provides further support for the conclusion that the January 1, 2032 compliance date and the degree of emission limitation are unachievable.

Moreover, the EPA is concluding that in light of current information concerning sources' inability to comply with 90 percent CCS by 2032, retaining 90 percent CCS as the BSER for existing coal-fired steam generating units that did not commit to retire before January 1, 2039, coupled with an exemption from the rule for sources that agreed to retire before January 1, 2032, would be a form of generation shifting prohibited by West Virginia under the circumstances here.\237\ Specifically, because updated projections demonstrate that compliance with 90 percent CCS generally is not feasible by 2032, sources have only one remaining option: retirement by 2032. Failing to revise the standard under these circumstances would therefore amount to forced closures in anticipation of other forms of power generation by facilities not subject to the at-issue 2024 CPS requirements.

\237\ 597 U.S. at 735.

In the immediate aftermath of West Virginia, the EPA initially focused on the Supreme Court's statement that generation shifting could sometimes be a consequence of regulatory requirements under CAA section 111.\238\ The Agency believed at the time that the lesson of West Virginia was that generation shifting was appropriate as a result of regulation, so long as the regulation did not explicitly require it.\239\ The EPA therefore selected 90 percent CCS and 40 percent natural gas co-firing, coupled with an exemption contingent on unit retirement by 2032,

based on the belief that unit closure was an appropriate consequence so long as facilities could also choose to comply with the standards.

\238\ See id. at 731 n.4 (noting that “there is an obvious difference between (1) issuing a rule that may end up causing an incidental loss of coal's market share, and (2) simply announcing what the market share of coal, natural gas, wind, and solar must be”).

\239\ See, e.g., 91 FR 39798, 39899 (May 9, 2024).

However, as detailed in section IV of this preamble, the EPA is now determining that the predictions and analyses underlying 90 percent CCS were overly optimistic. The EPA now concludes that requiring sources either to close by 2032 or comply with 90 percent CCS requirements that have now been determined to be unreasonable and impracticable essentially presents a Hobson's choice. That is, because the EPA's predictions regarding 90 percent CCS have now been determined to be overly optimistic, retaining the 2024 CPS would effectively force generation shifting by leaving regulated sources with no alternative to closure by the compliance deadline in 2032.

Consequently, the EPA is finalizing repeal of the requirements in emission guidelines pertaining to long-term coal-fired steam generating units. As discussed in this section of the preamble, the EPA is addressing only CCS with 90 percent capture and implementation by January 1, 2032, because these were the requirements under the 2024 CPS. This action does not finalize any potential alternative BSERs or implementation timeframes such as CCS with lower rates of capture or a later compliance date. The EPA considered whether to analyze and promulgate potential alternatives and determined it is not necessary to do so under the circumstances presented here. 2. Natural Gas Co-Firing-Based Requirements for Existing Medium-Term Coal-Fired Steam Generating Units

In the 2024 CPS, the EPA determined the BSER for existing medium- term coal-fired steam generating units to be 40 percent natural gas co- firing. Natural gas co-firing can require installation of new gas burners and related boiler modifications. Natural gas co-firing also requires construction of natural gas pipeline infrastructure to supply the necessary amount of natural gas to the unit. The EPA argued in the 2024 CPS that natural gas co-firing qualifies as the BSER for medium- term coal-fired units because, among other things, it does not result in unreasonable adverse consequences related to energy requirements. In the 2024 CPS, the EPA further argued that the degree of emission limitation based on application of the 40 percent natural gas co-firing BSER to the affected sources was achievable by the January 1, 2030, compliance date, considering the time necessary to deploy the necessary natural gas pipeline infrastructure.

The EPA reevaluated the record for and the Agency's determinations in the 2024 CPS and subsequently proposed that 40 percent natural gas co-firing is not the BSER for existing medium-term coal-fired steam generating units based on the potential for significant adverse consequences related to energy requirements. The EPA also proposed that 40 percent natural gas co-firing cannot qualify as the BSER because it constitutes impermissible generation shifting. Finally, the EPA further proposed to determine that the degree of emission limitation based on 40 percent natural gas co-firing is not achievable because it is unlikely that the pipeline infrastructure necessary can be deployed by the compliance date of January 1, 2030. Consequently, the EPA proposed to repeal the requirements for medium-term coal-fired steam generating units.

The EPA is finalizing the determination that 40 percent natural gas co-firing is not the BSER for medium-term coal-fired steam generating units and is repealing the requirements for those sources on that basis. In this section of the preamble, the EPA details the three bases for this determination. The first basis discussed here--that 40 percent natural gas co-firing represents impermissible generation shifting--is sufficient to preclude it from qualifying as the BSER. However, the EPA is also providing two additional bases that provide additional support for this conclusion. a. 40 Percent Natural Gas Co-Firing is Generation Shifting

The EPA is finalizing its proposed interpretation that 40 percent co-firing with natural gas is not the BSER for existing medium-term coal-fired steam generating EGUs because it constitutes generation shifting and is therefore beyond the EPA's authority to require under CAA section 111. In West Virginia, the Supreme Court held that a “system of emission reduction” under that section cannot include a forced shift of nationwide electricity generation from one type of energy source to another.\240\ That is, the BSER cannot be based on generation shifting. In discussing whether the EPA could effect generation shifting through at-the-source measures by, e.g., “simply requiring coal plants to become natural gas plants,” the Court stated that “EPA has never ordered anything remotely like that, and we doubt it could.” \241\

\240\ 594 U.S. at 734-35.

\241\ Id. at 728 n.3.

In the 2024 CPS, the EPA considered whether co-firing natural gas in a coal-fired boiler would constitute generation shifting and concluded that it would not. There, the Agency argued that, in contrast to impermissible generation shifting, 40 percent natural gas co-firing constitutes at-the-source fuel switching, which is a “traditional pollution control measure” as recognized by the Supreme Court in West Virginia.\242\ The EPA further explained in the 2024 CPS that the Agency interpreted the Court's statements in footnote 3 of that opinion as referring to a complete transformation of a coal-fired unit to a 100 percent natural gas-fired unit and contrasted such complete repowering with natural gas co-firing at 40 percent.\243\

\242\ U.S. EPA, Response to Comments Document (April 2024). Chapter 2.7.2, page 101-02. Document ID No. EPA-HQ-OAR-2023-0072- 8914.

\243\ Id.

The EPA has reexamined the question of whether 40 percent natural gas co-firing is impermissible generation shifting and has now determined that requiring a utility to use a completely different fuel type runs counter to the Supreme Court's decision in West Virginia. Critically, in that decision, the Court found that the EPA lacks authority to decide the appropriate share of different types of electricity generation on a nationwide basis.\244\ Requiring a significant portion of the coal-fired fleet to become a different type of electricity generating resource--a hybrid coal and gas-fired fleet that relies on a different set of fuels--has the same effect of dictating the market shares of the different fuel types that comprise the nation's energy supply. The fact that this forced shift would occur within individual sources as opposed to across sources in the electricity generation sector does not rebut this conclusion.

\244\ West Virginia, 594 U.S. at 728.

Moreover, a coal-fired steam generating EGU is a fundamentally different type of plant than a steam generating EGU that fires both coal and natural gas. This is evidenced by the modifications and new infrastructure needed to turn a coal-fired EGU into a hybrid coal and gas-fired EGU, including modifications to or additions of burners to the boiler and potential changes to steam superheaters, reheaters, and economizer heating surfaces. The EPA believes that requiring a coal plant to turn itself into a different type of plant in order to burn a completely different fuel (i.e., natural gas) belies the Agency's earlier assertions that 40 percent natural gas co-firing is fuel switching akin to burning

lower sulfur coal in a coal-fired EGU or ultra-low sulfur diesel in a stationary compression ignition internal combustion engine. In these examples, a source is merely using a particular type of the fuel that the source was always intended to use, as opposed to a different fuel entirely. Therefore, the EPA is finalizing the finding that a BSER based on forcing a coal-fired EGU to become a partially natural gas- fired steam generating units shifts that unit's generation from coal to natural gas and is impermissible under the Court's precedent because it is an attempt to dictate the market share of coal versus natural gas.

The EPA acknowledges that this is a change in position from its earlier interpretation in the 2024 CPS that 40 percent natural gas co- firing is not generation shifting. As explained in the preceding discussion, the Agency believes its updated interpretation of West Virginia and its application to natural gas co-firing is the best one because it recognizes the distinction between traditional fuel switching and requiring a switch to different type of fuel. No party will have relied on the EPA's earlier interpretation in the 2024 CPS, given that that regulatory framework was never implemented.

Comments: Many commenters supported the EPA's proposed interpretation that natural gas co-firing constitutes impermissible generation shifting. Commenters stressed that generation shifting occurs whenever a facility is required to switch from burning one type of fuel to an entirely different type of fuel, regardless of the amount or percentage of co-firing required, because the EPA lacks authority either to require a plant to change to a different fuel type or to require it to become a different type of plant (i.e., a hybrid coal and gas plant). Commenters asserted that 40 percent co-firing would require transformation into a different type of source because co-firing natural gas in a coal-fired boiler entails changes to that boiler that could include modifications to millions of dollars of equipment and additional costs associated with pipeline infrastructure needed to support co-firing.

Other commenters opposed to the EPA's interpretation that 40 percent natural gas co-firing is generation shifting argued that natural gas co-firing is a form of fuel switching, which the EPA has historically relied on under CAA section 111. Commenters contrasted fuel switching, a traditional technology-based control measure focused on improving the performance of individual sources, with the generation shifting described by the Supreme Court in West Virginia, further asserting that co-firing does not entail a transformative expansion in the Agency's regulatory authority. Commenters explained that any new infrastructure and costs associated with natural gas co-firing are accounted for in the EPA's evaluation of the system's cost, which is a separate factor in the BSER analysis. Additionally, commenters asserted that it is common for a CAA section 111 standard to require extensive modifications to a plant, although the modifications associated with co-firing are relatively minor and far less extensive than those required for other control strategies at coal-fired plants, such as flue gas desulfurization or selective catalytic reduction.

Moreover, according to the commenters, co-firing does not involve the type of grid-wide reshuffling that West Virginia prohibits, either in concept or practice. There is no emissions or generation cap on affected sources, and the EPA does not assume any change in generation at co-firing sources. Therefore, commenters argued that the BSER is not premised on a shift of generation from co-firing sources to other sources. In addition, EPA's compliance modeling did not predict such a shift. Commenters also argued that fuel-switching is not generation shifting because a coal plant that is co-firing with 40 percent natural gas does not cease to exist as a coal plant, unlike the generation shifting addressed in West Virginia. Further, the EPA has always treated coal- and gas-fired steam generating EGUs as part of the same source category because the underlying type of source burning the fuel (a steam generating utility boiler) is the same, even though co-firing would require the use of some amount of a different fuel (natural gas instead of coal).

EPA Response: The EPA disagrees that co-firing natural gas in a coal-fired boiler is akin to the traditional pollution control measure under CAA section 111. As noted in this section of the preamble, natural gas co-firing is distinguishable from switching to lower sulfur coal or lower sulfur diesel, or from requiring a source to limit itself to one type of fuel. The EPA takes no position in this rule about these forms of fuel switching vis-[agrave]-vis West Virginia. But the EPA is, in this final rule, finding that it is not permissible for a BSER to be based on the use of an entirely different fuel from what a source was designed and configured to accept. Given that modifications are necessary to accommodate co-firing at 40 percent, this level of co- firing, in particular, is distinct from the traditional forms of fuel switching that the EPA has employed under CAA section 111. The cost of those modifications is not as relevant as the fact that they occur: The EPA considers the need for modifications as part of determining whether 40 percent co-firing is an appropriate system of emission reduction, as opposed to whether it is the best system of emission reduction based on cost and the other factors of CAA section 111(a)(1). That is, because a coal-fired source must transform itself to co-fire natural gas at 40 percent, such co-firing is not fuel switching but is rather shifting generation to a different type of source. While commenters countered that natural gas co-firing is not generation shifting because the source remains a fossil fuel-fired steam generating boiler that continues to fire some amount of coal, the EPA does not believe that it is necessary to shift generation from a boiler to a combustion turbine in order to effectuate impermissible generation shifting. Under the commenters' logic, the EPA's requiring 99 percent natural gas co-firing would still not be considered shifting generation because one percent would still be some amount of coal still being utilized. Rather, the EPA's position is that generation shifting occurs when a system of emission reduction requires a source to use an entirely different type of fuel, particularly when the source was not designed to accept the new type of fuel. The effect of such an approach is that the EPA effectively dictates the fuel mix used in the nationwide energy system, which the Supreme Court held in West Virginia the EPA is not permitted to do. b. Energy Requirements

Even if 40 percent natural gas co-firing could be evaluated as a potential BSER, which it cannot be for the reasons discussed in section IV.A.2.a of this preamble, the EPA further determines that the adverse impacts on the energy system are unreasonable and therefore that 40 percent co-firing cannot be the BSER for medium-term coal-fired EGUs. As part of determining the BSER, the EPA considers energy requirements.\245\ As discussed in section II.C.3 of this preamble, energy requirements may include the impacts, if any, of the air pollution controls on the source's own energy needs.\246\ The EPA may further assess, as part of the energy

requirements consideration, any impacts of a system of emission reduction on the energy system on a sector-wide, regional, or national basis, as appropriate.\247\ As part of such assessment, the EPA has considered potential adverse impacts on the reliability of the bulk power system and its ability to deliver affordable and consistent electricity to end users.\248\ Similarly, the EPA has considered whether potential BSERs might have adverse impacts on the supply or cost of natural gas, which is used in many applications throughout the nation's energy system, including for electricity generation, industrial applications, and transportation.\249\ In this action, the Agency's determination of unreasonable adverse impacts has focused on the sector-wide strain that requirements based on 40 percent natural gas co-firing could place on the availability of natural gas. The demand for natural gas, as explained below, is anticipated to be greater than previously projected. Under these circumstances, diverting natural gas for use in natural gas in steam generating boilers could result in an unreasonable impact on the energy system because it reduces the availability of gas for other, more efficient uses, including for electricity generation in combustion turbines.

\245\ 42 U.S.C. 7411(a)(1).

\246\ See, e.g., 91 FR 1910, 1937 (January 15, 2026) (energy requirements consideration includes auxiliary/parasitic load requirements to run a potential BSER control for fossil fuel-fired combustion turbines).

\247\ See, e.g., Sierra Club, 657 F.2d at 330 (interpreting energy requirements factor as including consideration of effects “on the grand scale”); see also 84 FR 32520, 32534 n.152 (July 8, 2019) (“The EPA may consider energy requirements on both a source- specific basis and a sector-wide, region-wide, or nationwide basis.”).

\248\ See, e.g., 80 FR 64510, 64594 (October 23, 2015); 80 FR 64662, 64721 (October 23, 2015).

\249\ See 84 FR 32520, 32544-46 (July 8, 2019).

The analyses the EPA conducted and relied on to assess the 2024 CPS's projected impacts showed only incremental increases in electricity demand: a 13 percent increase between 2000 and 2022, with demand staying relatively flat over the 2007-2022 period. As discussed in section III.C of this preamble, during this period, the share of coal-fired electricity decreased in both absolute and relative terms, while both natural gas-fired net generation and wind and solar net generation increased. Natural gas surpassed the total net generation from coal on an absolute basis in 2016, while renewables surpassed the total net generation from coal on an absolute basis in 2022.\250\ The information that the EPA analyzed for purposes of the 2024 CPS indicated that the sector trend of moving away from coal-fired generation was likely to continue, that the share of electricity generation from natural gas-fired sources would likely decline, and that the share of generation from non-emitting technologies would likely continue to increase. One important data point for purposes of the 2024 CPS was that the Agency anticipated that the recent trend of retirements of coal-fired capacity (at an average annual rate of 10 GW from 2015 to 2023) would continue due to the economics of coal-fired generation. At that time, more than half of the operating coal-fired steam generating units had announced retirement or plans to convert to natural gas by 2039.\251\ Thus, the EPA predicated the Agency's consideration of energy requirements associated with 40 percent natural gas co-firing in the 2024 CPS on an assumption of a continued and significant decline in the number of coal-fired EGUs.

\250\ Document ID No. EPA-HQ-OAR-2023-0072-8920.

\251\ 89 FR 39798, 39816-18 (May 9, 2024).

In contrast, updated information and analysis of power sector trends indicate a significantly different landscape moving forward. In particular, as outlined in section III.C of this preamble, the projections of electricity markets over the coming decades indicate higher electricity demand due to a number of factors. Recent analyses further predict that this higher electricity demand will, in turn, result in higher utilization of coal and gas fired resources \252\ than projected under the 2024 analysis underpinning the 2024 CPS.\253\ The EPA's updated baseline projections indicate that coal capacity will level off at 100 GW by 2040, in contrast to the 2024 analysis's prediction of continuing to decline from the current level to 52 GW in 2035 and 42 GW in 2040. Similarly, the updated baseline projections indicate that new NGCC additions would be 155 GW in 2035 and 217 GW by 2040, contrasting earlier projections that total just 26 GW in 2035 and 2040. The EPA's updated baseline modeling in the 2025 analysis also projects significantly higher natural gas consumption, even absent the 2024 CPS requirements, with Henry Hub gas prices projected at 18 percent higher by 2030 and 52 percent higher by 2035.

\252\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

\253\ U.S. EPA. RIA for 2024 CPS. Document ID No. EPA-HQ-OAR- 2023-0072-8913.

These significant changes in the power sector complement the EPA's reevaluation of the demands natural gas co-firing puts on both individual coal-fired steam generating units and the energy system more broadly, informing the EPA's analysis and determination here. While coal-fired steam generating units may use small amounts of natural gas for startup purposes, relatively few sources use natural gas in proportions that would have been consistent with the requirements for medium-term coal-fired steam generating units in the 2024 CPS. Therefore, the co-firing-based standards would result in a significant increase in the demand for natural gas. The 2024 CPS failed to adequately address the impacts of this increase, even before the updated projections of coal capacity and natural gas demand that are now available.

Based on the EPA's reexamination of the facts and conclusions in the 2024 CPS and the analysis supporting this rulemaking, the Agency now finds that 40 percent natural gas co-firing is not the BSER because of the potential for unreasonable adverse impacts related to energy requirements.\254\ The EPA's two reasons for this determination-- diverting the volume of natural gas needed to support 40 percent co- firing from other uses in the energy system could have significant impacts and natural gas is more efficiently used in natural gas-fired combustion turbines--are described in further detail below. The EPA is determining that the collective effect of these two phenomena could result in unreasonable adverse impacts on the energy system by forcing natural gas, the availability of which is anticipated to become more constrained, to be used in a relatively inefficient manner. The EPA has also balanced these energy impacts against the relatively small amount of CO2 reductions available from 40 percent natural gas co- firing (16 percent) and the other relevant considerations under CAA section 111(a)(1) (i.e., cost and nonair quality health and environmental impacts) in determining that such co-firing could not be the BSER for medium-term coal-fired steam generating units.

\254\ For details on the analysis, see memorandum entitled Trends Relating to Fossil Fuel-fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

i. 2024 CPS Requirements Reduce the Availability of Natural Gas for Other Purposes

The EPA now finds that the potentially large demand for natural gas associated with 40 percent co-firing in coal-fired steam boilers is unreasonable and could produce a significant adverse consequence related to energy requirements. As explained in the June 2025 NPRM, the EPA believes that coal-fired steam generating unit capacity and generation will now continue to comprise a substantial portion of the

nation's electricity supply.\255\ A number of coal-fired steam generating units are delaying or canceling scheduled retirements in light of increasing electricity demand, the changes in tax incentives for various types of electricity-generating resources in the OBBBA, and Administration actions to support the continued operation of coal-fired capacity.256 257 Furthermore, the EPA's modeling projections show a substantial capacity of coal-fired steam generating units operating past 2032.\258\ The EPA's updated modeling estimates that 100 GW coal capacity will be operational in the 2035 model run year,\259\ whereas the EPA projected 52 GW of coal capacity would exist by 2035 in the 2024 CPS analysis.\260\ While it is the case that not all 100 GW would necessarily be subject to the 40 percent co-firing based standard, it is very likely that the amount of capacity that would be so subject is significantly higher than the EPA projected in the 2024 CPS. Because much more coal capacity is anticipated to remain operational than was previously projected in the 2024 CPS modeling (almost twice as much in 2035), it is reasonable to expect that more units could be subject to a standard of performance based on 40 percent co-firing with natural gas. Thus, the total volume of natural gas that sources would need to implement co-firing could be both substantial and greater than previously believed.

\255\ 90 FR 25752, 25772, 25774 (June 17, 2025).

\256\ D. Proctor, U.S. Coal Plants Get Reprieve as Market and Policies Change, Power (February 6, 2025). Available at: https://www.powermag.com/u-s-coal-plants-get-reprieve-as-market-and-policies-change/.

\257\ See, e.g., U.S. Department of Energy, 2026 DOE 202(c) Orders. Available at: https://www.energy.gov/ceser/2026-doe-202c-orders.

\258\ Under the CPS, coal-fired steam generating units operating past 2032 and choosing to permanently cease operation before January 1, 2039, would have had a standard of performance based on 40 percent natural gas co-firing.

\259\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

\260\ U.S. EPA. RIA for 2024 CPS. Document ID No. EPA-HQ-OAR- 2023-0072-8913. Table 3-14.

More specifically, using the latest available data from the EIA, the U.S. electric sector comprised approximately 174 GW of coal-fired EGUs in 2024, which collectively consumed approximately 6.98 quadrillion Btus of energy. According to operation planning data reported to the EIA, 146 GW of this 174 GW coal capacity in the U.S. electric sector are expected to remain in service through 2032.\261\ If these 146 GW of units maintained their 2024 utilization levels and co- fired 40 percent natural gas, they would draw more than 2.3 quadrillion Btus of natural gas. For context, the entire U.S. electric sector consumed approximately 13.9 quadrillion Btus of natural gas in 2024 (i.e., implementing the co-firing standard would require an increase of 17 percent of 2024's total gas consumption in the U.S. electric sector).

\261\ The survey Form EIA-860 collects generator-level specific information about existing and planned generators and associated environmental equipment at electric power plants, including scheduled retirements. U.S. EIA. Form 860 data. Available at: https://www.eia.gov/electricity/data/eia860/.

As noted above, there are many critical demands for natural gas other than for electricity generation, including for industrial uses, residential use, and as a transportation fuel. According to the EIA, total demand for all uses of natural gas was approximately 33 quadrillion Btus in 2024. Furthermore, the EIA projects that the demand for natural gas, driven by domestic consumption and liquefied natural gas exports, will grow both in the near term \262\ as well as in the long term.\263\ EIA forecasts record high industrial and power sector natural gas consumption by 2027 264 265 and projects a 30 percent increase in LNG exports by 2027 as five new LNG export projects begin operation.\266\ This increasing demand stresses supply, resulting in an increase of projected costs of natural gas. Henry Hub gas prices in 2024 dollar-years are projected to rise to $4.12/MMBtu by 2030 and $5.26/MMBtu by 2035.\267\ In most model run years, EPA's analysis indicates that the cost of natural gas would be higher with the 2024 CPS in place as compared to the scenario absent the requirements of 2024 CPS.\268\ Also, updated EPA projections show substantial increases in natural gas combustion turbine generation when compared to prior EPA projections conducted for the 2024 CPS.\269\ Using a large volume of natural gas in coal-fired steam generating units when there are already increasing demands on the natural gas supply (from, among other things, higher domestic demand, including due to increased combustion turbine buildout, and greater liquefied natural gas exports) further exacerbates the potential for adverse energy impacts, as evidenced by the natural gas price increases associated with the 2024 CPS requirements. Therefore, the EPA finds that diverting the volume of natural gas necessary to support 40 percent natural gas co-firing from other uses is unreasonable because it could result in a significant adverse impact on the energy system, such that 40 percent natural gas co-firing is not the BSER.

\262\ U.S. Energy Information Administration. EIA expects higher wholesale U.S. natural gas prices as demand increases. Available at: https://www.eia.gov/todayinenergy/detail.php?id=64344.

\263\ U.S. Energy Information Administration, Annual Energy Outlook 2025. Available at: https://www.eia.gov/outlooks/aeo/data/browser/#/?id=13-AEO2025&cases=ref2025&sourcekey=0.

\264\ U.S. Energy Information Administration. U.S. industrial natural gas consumption expected to hit records in 2026 and 2027. Available at: https://www.eia.gov/todayinenergy/detail.php?id=67686.

\265\ U.S. Energy Information Administration. Natural gas for power generation flat this summer, record high expected in 2027. Available at: https://www.eia.gov/todayinenergy/detail.php?id=67725.

\266\ U.S. Energy Information Administration. U.S. natural gas exports to grow nearly 30% by 2027 as LNG facilities ramp up. Available at: https://www.eia.gov/todayinenergy/detail.php?id=67484.

\267\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

\268\ Id.

\269\ Id.

Comments: Some commenters stated that using large quantities of natural gas to fuel steam generating units is extraordinarily wasteful at a societal level. A commenter also argued that the adverse impacts on the energy system are especially unreasonable when compared to the relatively small CO2 reductions--16 percent--that are available from 40 percent natural gas co-firing.

Other commenters expressed concern that the EPA's objections to the 40-percent co-firing standard based on gas supply were unsupported. Commenters asserted that the EPA lacked data in the June 2025 NPRM to sustain the argument that there is insufficient natural gas supply to support the 40 percent co-firing standard.

EPA Response: The EPA agrees with commenters who stated that the impacts of 40 percent natural gas co-firing on the energy system are unreasonable, especially given the relatively small amount of CO2 reductions available. The Agency also notes that the Agency has broad discretion in weighing the statutory considerations under CAA section 111(a)(1) to determine the BSER.\270\ The rationale included in the June 2025 NPRM and further corroborated by the arguments and data included in this final rule justifies the EPA's finding that 40 percent natural gas co-firing would have adverse consequences for the energy system: the co-firing standard could strain the supply of natural gas and reduce its availability for other purposes. Although several commenters cited projected natural gas use and supply figures from the 2024 CPS to support

arguments that ample gas would be available for co-firing, such comments do not consider the changed circumstances of the energy sector. The EPA cited these changed circumstances in the June 2025 NPRM and, in this final rule, is adding further information that has become available since the June 2025 NPRM's publication in Spring 2025. As discussed in this preamble, the EPA's updated analysis shows that the total volume of natural gas that affected sources would require to implement the 2024 CPS's co-firing standard is 17 percent of total power sector gas consumption in 2024. This increase is over 20 percent higher than the percent increase estimated under the 2024 analysis using the same approach (i.e., assuming all units active in 2032 co- fire 40 percent natural gas and maintain 2024 utilization levels). This substantial increase in natural gas demand would place upward pressure on natural gas prices.

\270\ See Lignite Energy Council, 198 F.3d at 933 (“Because section 111 does not set forth the weight that should be assigned to each of these factors, we have granted the agency a great degree of discretion in balancing them.”).

Moreover, the increase in natural gas demand from co-firing units would now be occurring in the context of significantly higher natural gas prices. The EPA's updated analysis projects that natural gas prices will increase significantly more than the Agency predicted in its the 2024 analysis of the final 2024 CPS (50 percent higher in 2035 than what was projected under the 2024 analysis), meaning that the impact of the 2024 CPS on the price of natural gas would be greater. The updated analysis projects that, in a scenario with the 2024 CPS remaining in place, the Henry Hub price of gas would increase nine percent in 2035,\271\ while the 2024 analysis projected that the Henry Hub price of gas would have increased only three percent as a result of the 2024 CPS in the same year.\272\ In sum, the anticipated constraints on natural gas availability combined with substantial increases in natural gas demand mean that diverting a large volume of natural gas to 40 percent natural gas co-firing in coal-fired steam generating units would further exacerbate the already-strained natural gas market.

\271\ See memorandum entitled Trends Relating to Fossil Fuel- fired Electric Generating Units in the docket for this rulemaking (Docket ID No. EPA-HQ-OAR-2025-0124).

\272\ Docket ID No. EPA-HQ-OAR-2023-0072, Table 3-12 in the 2024 RIA.

ii. Inefficiency of 40 Percent Natural Gas Co-Firing

The EPA is further concluding that the energy requirements associated with 40 percent natural gas co-firing in a steam generating EGU are unreasonable because such co-firing would be an inefficient use of the comparatively constrained (relative to previous assumptions) availability of natural gas, particularly compared to use in a combustion turbine. This is a relevant consideration because the two types of units provide the same product--electricity--and because they are being covered by a single regulatory regime.\273\ The EPA therefore believes that it should attempt to optimize the use of natural gas amongst the affected sources. Applying 40 percent natural gas co-firing would result in a decrease in boiler efficiency by approximately two percent (to a total boiler efficiency of less than 40 percent) due to the higher hydrogen content of natural gas relative to coal. In the 2024 CPS, the EPA argued that this decline in efficiency could be partially offset by decreases in auxiliary power demand related to coal handling and emissions controls but acknowledged that uncertainty remained about whether this offset would be true in all circumstances.\274\ In the EPA's unit-level cost analysis for the 2024 CPS, the Agency assumed a two percent decrease in boiler efficiency would result in one percent overall heat rate penalty for the unit.\275\ Thus, for a theoretical coal-fired steam generating unit with a heat rate of 10,000 British thermal units per kilowatt-hour (Btu/kWh), the EPA estimates that the heat rate increases when co- firing to 10,100 Btu/kWh.

\273\ See 40 CFR part 60, subparts TTTT and TTTTa.

\274\ 89 FR 39798, 39895 (May 9, 2024).

\275\ See spreadsheet entitled Unit-Level Cost and Reduction Estimates for Natural Gas Co-firing Final Rule attached to Document ID No. EPA-HQ-OAR-2023-0072-9095.

Comparatively, the use of large amounts of natural gas for combustion in combined cycle EGUs is more efficient. New natural gas- fired combined cycle EGUs generally have an operating efficiency of greater than 50 percent. If the natural gas that would otherwise be used for co-firing in a steam generating EGU were instead used in a combined cycle unit with a heat rate of 6,700 Btu/kWh, the combined effective heat rate of the coal-fired EGU and gas-fired combined cycle unit would be approximately 8,400 Btu/kWh, which is substantially less than the coal-fired steam generating unit with 40 percent natural gas co-firing. In addition, effective combined heat rates would be lower if the natural gas were used in a simple cycle combustion turbine EGU.\276\ As described in section IV.A.2.b.i of this preamble, the availability of natural gas is projected to be relatively constrained due to a combination of increases in domestic demand and increases in liquefied natural gas exports. Given these circumstances, the EPA believes it is reasonable to consider the relative efficiency of natural gas use amongst regulated sources when determining BSER. When looking at the affected source category as a whole, the EPA finds that impacts on the energy system of 40 percent natural gas co-firing are unreasonable because it is significantly more efficient to use natural gas to generate electricity in a combustion turbine, rather than co- firing natural gas in a steam generating boiler.

\276\ For a representative combustion turbine with a heat rate of 8,700 Btu/kWh, the combined effective heat rate would be approximately 9,500 Btu/kWh.

Comments: Some commenters supported the EPA's proposed determination that the Agency can or must consider the relative efficiency of co-firing natural gas in a coal-fired steam generating unit versus in a combustion turbine. One commenter stated that co- firing natural gas in a boiler is less efficient than burning natural gas in either a combined cycle system or a simple cycle turbine. Another commenter noted that although choosing to co-fire may be reasonable for individual units due to source-specific circumstances, co-firing is not a reasonable practice on a fleetwide basis.

Other commenters refuted the basis of comparison between heat rates of natural gas co-firing units and NGCC due to the technological differences between an NGCC and a steam unit. Some commenters questioned the relevance of this comparison and asserted that the EPA should consider whether natural gas co-firing reduces emissions from the affected units, not whether a different type of unit can use natural gas more efficiently.

EPA Response: The EPA believes the relative efficiency of steam generating boilers and combustion turbines is relevant because the energy requirements factor of CAA section 111(a)(1) requires the Agency to consider adverse impacts on the energy system.\277\ The interplay between different potential uses of a limited fuel to generate electricity is pertinent to this inquiry. Moreover, the EPA believes this is a relevant consideration because coal-fired steam generating units and NGCC units were addressed as part of a common regulatory framework under the 2024 CPS.\278\ Thus, the Agency's

regulation should acknowledge the competing uses of natural gas by the affected sources. Directing natural gas to be used in a less efficient manner, i.e., in a steam generating boiler, is not a reasonable use of this resource. This is especially the case when the availability of natural gas is anticipated to be relatively constrained due to multiple competing uses, both within and beyond electricity generation. Furthermore, the EPA has considered the decreased efficiency of a coal unit co-firing 40 percent natural gas relative to the CO2 reductions available from such co-firing and has determined that the adverse impacts are not reasonable given reductions in emission rate (e.g., lb CO2/MWh-gross) of only 16 percent.

\277\ See preamble sections II.C.3 and IV.A.2.b for discussion of the EPA's interpretation and application of CAA section 111(a)(1)'s “energy requirements” factor.

\278\ See 40 CFR part 60, subparts TTTT and TTTTa.