National Emission Standards for Hazardous Air Pollutants: Coal- and Oil-Fired Electric Utility Steam Generating Units: Final Repeal

The EPA proposed to repeal the requirement that sources must use PM CEMS exclusively for demonstrating

compliance with the fPM emission standard, as well as the adjusted QA criteria,\99\ and to return to the previous regulatory language from the 2012 MATS Rule and 2020 Final Rule that allowed owners and operators to demonstrate compliance using either quarterly stack testing, PM CPMS, or PM CEMS. The EPA solicited comment on the rationale that the higher costs for EGUs not currently utilizing PM CEMS, the availability of other air pollution control performance indicators that can inform operators of malfunctions, and the adequacy of current compliance options support repealing the requirement that all coal- and oil-fired EGUs must use PM CEMS (Question #3). The EPA also proposed and solicited comment on reinstating the low emitting EGU (LEE) program for fPM and non-Hg HAP metals, which reduces the required stack testing frequency for sources that demonstrated that their emissions are less than 50 percent of the corresponding emissions limit for 3 consecutive years (Question #4). Lastly, the EPA proposed retaining the updated fPM measurement requirements of allowing either an increased minimum volume per run or the collection of a minimum mass per run.\100\ The EPA solicited comment on these measurement requirements for fPM compliance demonstration, as the Agency believed retaining the additional option of sample mass would reduce measurement uncertainty and may reduce test durations and, therefore, reduce fPM testing costs (Question #5).

\99\ Emission standards are used to determine the acceptable tolerance interval when correlating PM CEMS. In the 2024 Final Rule, the EPA instructed the use of 0.015 lb/MMBtu, instead of the finalized more stringent limit of 0.010 lb/MMBtu, when developing PM CEMS correlations to ease difficulties correlating PM CEMS. See 89 FR 38535 (May 7, 2024).

\100\ For coal- and solid oil-fired EGUs, the 2024 Final Rule required a minimum catch for fPM of 6.0 mg or a minimum sample volume of 4 dry standard cubic meters (dscm) per run. Requirements for IGCCs included a minimum catch for fPM of 3 mg or a minimum sample volume of 2 dscm. There were no changes to the minimum catch and sample volume requirements for oil-fired EGUs.

Commenters provided both supportive and opposing arguments regarding the EPA's proposed compliance demonstration requirements for fPM. With several minor, technical exceptions, comments received on the proposed repeal of the PM CEMS compliance demonstration requirement for fPM did not result in a change to the position the Agency set out in the proposed rule. Therefore, the EPA is repealing the requirement to use PM CEMS for compliance demonstration with the fPM emission standard and restoring flexibility to owners and operators to choose between the fPM compliance demonstration options as proposed. The EPA is also reinstating the LEE program for fPM and non-Hg HAP metals as proposed. The EPA received comments that supported retaining the flexibility of a minimum volume per run or minimum mass per run sampling requirements but argued that the updated minimum volume per run sampling requirement would result in longer test runs and impose significant burdens on operators. Based on these comments, along with an additional review of the accuracy of PM CEMS at low levels, the EPA is finalizing as proposed but with minor technical revisions to the sampling requirements based on the purpose of the fPM testing. If PM CEMS are used for the compliance demonstration, owners and operators are required to follow the updated sampling requirements for minimum volume per run or minimum mass per run, as proposed in the 2025 Proposal. For all other compliance demonstration options, owners and operators are required to collect a lower minimum sample volume as originally required in the 2012 MATS Rule. 2. What is the rationale for the EPA's final approach and decisions on the compliance demonstration requirements for the filterable PM standard?

Upon further review, the EPA concludes that mandating the use of PM CEMS and removing previously available compliance alternatives was not “necessary” pursuant to CAA section 112(d)(6). As discussed in section III.A of this preamble, the EPA is finalizing the proposed repeal of the more stringent fPM emission standard from the 2024 Final Rule and returning to the fPM emission standard set in the 2012 MATS Rule and reaffirmed in the 2020 Final Rule. Therefore, the EPA's conclusion in the 2024 Final Rule that the costs for PM CEMS are commensurate with the costs for stack testing \101\ no longer applies, because longer duration runs that increase stack testing costs are no longer necessary to demonstrate compliance with a lower fPM. Many sources also use the stack testing compliance option to efficiently merge their PM and HCl testing programs into a single testing mobilization to test for those pollutants in conjunction, possibly yielding further cost savings. Further, the Agency finds that although the 2024 Final Rule invoked CAA section 114(a)(1)(C) as offering additional authority for the PM CEMS requirement,\102\ the provision is equally applicable to the alternative compliance demonstration options restored in this final rule.\103\

\101\ 89 FR 38535-37 (May 7, 2024).

\102\ Id. at 38535.

\103\ See CAA section 114(a)(1), (a)(1)(C)-(D), 42 U.S.C. 7414(a)(1), (a)(1)(C)-(D) (authorizing the EPA to require source owners or operators to “install, use and maintain such monitoring equipment, and use such audit procedures, or methods” and “sample such emissions” as required by the Administrator).

The 2024 Final Rule requirement to use PM CEMS to demonstrate compliance meant that up to two-thirds of EGU owners and operators would face higher compliance costs than when allowed to use quarterly stack testing or PM CPMS. As shown in more detail in the RIA for this final rule, the EPA estimates a cost savings of $2.9 million per year related to the repeal of the PM CEMS requirement, after the two-year exemption period (2030 to 2037); the estimated cost savings during the two-year exemption period (2028 to 2029) is $1.2 million per year. While the EPA concluded in the 2024 Final Rule that the use of PM CEMS would allow for more efficient pollutant abatement and increased transparency of EGU emissions,\104\ the Agency no longer believes that those advantages outweigh the increased cost of PM CEMS compared to the two other compliance options (i.e., PM CPMS and quarterly stack testing) that were determined to be appropriate for demonstrating compliance with the fPM emission standard in the 2012 MATS Rule.

\104\ 89 FR 38535 (May 7, 2024).

The EPA noted in the 2024 Final Rule that CEMS enable power plant operators to quickly identify and correct problems with air pollution control devices.\105\ However, there are other ways that owners and operators can become aware of air pollution control malfunctions without employing PM CEMS. For example, operators at EGUs with an ESP can instantaneously track and record opacity, secondary corona power, secondary voltage (i.e., the voltage across the electrodes), secondary current (i.e., the current to the electrodes), spark rate, and alarm and fault indicators to ensure proper functionality of the ESP in real time. For EGUs with technology such as FFs, bag leak detection systems (BLDS) and parameters like pressure differential (i.e., pressure drop), inlet temperature, temperature differential, exhaust gas flow rate, cleaning mechanism and cycle operation, and fan current and speed can serve as reliable indicators of process operations. These indicators also provide valuable data for analyzing trends and making informed

decisions about operational improvements and investments. As noted earlier in this preamble and in the 2024 Final Rule, a large majority of sources are already reporting measured compliance data showing fPM emissions that are below the previous fPM standard of 0.030 lb/MMBtu (via either PM CEMS or the stack testing compliance option), which further illustrates that the various options for demonstrating compliance with the fPM standards have been appropriate and effective.\106\

\105\ Id. at 38536.

\106\ Additionally, all fPM compliance data can be accessed by the public via the EPA's Web Factor Information Retrieval System (WebFIRE) at https://cfpub.epa.gov/webfire, which maintains the availability and transparency of fPM emissions.

Therefore, the EPA is repealing the requirement to use PM CEMS for demonstrating compliance with the fPM emission standard, as well as the adjusted QA criteria,\107\ and returning to the previous requirement that allowed owners and operators to demonstrate compliance using quarterly stack testing, PM CPMS, or PM CEMS. This provides greater flexibility to owners and operators and reduces the compliance burden, while still assuring compliance with the fPM emission standard.

\107\ New PM CEMS installations must follow Performance Specification 11 (PS-11), which requires the development of a site- specific correlation curve to relate PM CEMS readings to the PM reference method values. Emission standards are used to determine the acceptable tolerance interval when correlating PM CEMS. In the 2024 Final Rule, the EPA instructed the use of 0.015 lb/MMBtu, instead of the finalized more stringent limit of 0.010 lb/MMBtu, when developing PM CEMS correlations to ease difficulties correlating PM CEMS. See 89 FR 38535 (May 7, 2024).

The EPA is also reinstating the LEE program for fPM and non-Hg HAP metals, which reduces the stack testing frequency for sources that have demonstrated that their emissions are less than 50 percent of the applicable emission limit for 3 consecutive years. Sources that had previously demonstrated that they qualify for LEE status would not have to re-demonstrate that qualification. In the 2024 Final Rule, the EPA found that the optional LEE program was “superfluous” due to the PM CEMS requirement and the revised fPM emission standard.\108\ However, as the EPA is repealing those requirements, reinstating the LEE program for fPM and non-Hg HAP metals will further reduce the costs associated with stack testing for sources that opt in, while still assuring compliance with the emission standard.\109\ As mentioned earlier in this section, since LEE testing is only required once every 3 years once a source has attained LEE status, the ongoing LEE testing program is approximately 8 to 10 percent of the cost of a quarterly stack testing program.

\108\ 89 FR 38510 (May 7, 2024).

\109\ The LEE provisions are designed to ensure emissions are minimized. For example, EGUs equipped with a main stack and a bypass stack or bypass duct configuration that allows the effluent to bypass any pollutant control device are not allowed to pursue the LEE option under 40 CFR 63.10000(c). Furthermore, under 40 CFR 63.10000(c)(1)(i)(D), EGUs claiming LEE status may bypass a control device during emergency periods for no more than 2 percent of the EGU's annual operating hours.

Finally, the EPA is also updating the fPM measurement requirements that allow either an increased minimum volume per run or the collection of a minimum mass per run. As stated in section III.A of this preamble, a large majority of sources have reported measured compliance data showing fPM emissions below the 0.030 lb/MMBtu fPM standard. It is important that a sufficient quantity (i.e., mg of mass) of fPM be collected during these fPM test runs to allow for the accurate measurement of PM emissions, especially when the testing is being conducted to correlate or certify a PM CEMS. For these reasons, we have modified the fPM testing requirement to collect either a minimum catch of 6.0 mg or a minimum sample volume of 4 dscm per run if using a PM CEMS for compliance, whereas all other compliance demonstration options will be required to collect a minimum sample volume of 1 dscm per PM test run. 3. What key comments did the EPA receive on the compliance demonstration requirements for the filterable PM standard and what are our responses?

Comment: Some commenters agreed that there are other air pollution control indicators such as opacity, ESP power, and baghouse differential pressure that are used to provide timely information on potential equipment performance issues or malfunctions. Additionally, commenters said that the PM CEMS requirement falsely assumed the data would provide a better indicator of control equipment performance, which may not always be the case since PM CEMS measurements can be influenced by a variety of factors. Commenters stated that sources are required to comply with limits at all times, including between performance tests, and that under 40 CFR 63.10000(b), EGUs “must operate and maintain any affected source, including associated air pollution control equipment and monitoring equipment, in a manner consistent with safety and good air pollution control practices for minimizing emissions.”

Commenters also stated that owners and operators have multiple tools beyond CEMS to identify malfunctions in air pollution control equipment. These commenters agreed with the EPA's explanation in the proposed rule that facilities equipped with ESPs can monitor parameters such as opacity levels which can indicate higher than normal levels of particulates in the exhaust gas; secondary corona power, secondary voltage, and secondary current (indicating the collection of particulates on the plates and wires) to verify proper operation; power levels to the rappers and vibrators (used to clean the plates and wires); and the continued operation of the ash removal system to prevent system backup. For units with FFs, commenters stated that operators can rely on, as indicators of control performance, BLDS; pressure differential (indicates a bag leak or excessive buildup of the ash layer on the filters); temperature differential (for optimal bag/ filter conditions); exhaust gas flow rate to detect unfiltered gas escaping the system; power levels and operations of the bag vibrators or reverse-air systems to ensure proper bag cleaning activity; fan current which can indicate plugged bags; and opacity monitors. According to commenters, these monitoring practices, which are already in use across the industry, provide meaningful and timely insight into equipment condition and emissions performance without necessitating continuous emissions data. Additionally, other commenters stated that there are sufficient compliance indicators in place to ensure that PM (and HCl) emissions remain low between stack tests, such as operation of scrubber technology.

Alternatively, other commenters argued the EPA did not provide evidence that other parameters can be a substitute for complying with the fPM limit or be used in ways to quickly identify problems with pollution controls. Commenters also stated that the EPA did not demonstrate that these alternative parameters will provide the same or similar benefits as PM CEMS. Commenters disagreed that other performance indicators are as reliable as PM CEMS to identify malfunctions with the same sensitivity and that there is no requirement to continuously monitor and maintain a record of each of these parameters. Further, commenters argued that while monitoring of operational parameters of control technologies may

reveal anomalous conditions broadly, it does not quantify the mass or concentration of increased fPM emissions.

Response: The EPA agrees that there are reliable performance indicators that are continuously monitored and recorded, which are used to provide timely information on potential equipment performance and control device issues or malfunctions. As discussed in section III.B of this preamble, the EPA noted in the 2024 Final Rule that PM CEMS enable power plant operators to quickly identify and correct problems with air pollution control devices.\110\ However, there are other ways that owners and operators become aware of air pollution control malfunctions without employing PM CEMS. For example, for proper process operations purposes of the unit, operators of EGUs with an ESP can instantaneously track and record opacity, secondary corona power, secondary voltage (i.e., the voltage across the electrodes), secondary current (i.e., the current to the electrodes), spark rate, and alarm and fault indicators to ensure proper functionality of the ESP in real time. Similarly, for EGUs with technology such as FFs, BLDS and parameters like pressure differential (i.e., pressure drop), inlet temperature, temperature differential, exhaust gas flow rate, cleaning mechanism and cycle operation, and fan current and speed can serve as reliable indicators of process operations. These indicators, which are routinely monitored and recorded regardless of any regulatory requirement, also provide valuable data for analyzing trends and making informed decisions about operational improvements and investments. Moreover, these indicators help ensure that EGUs are operated and maintained, including associated air pollution control equipment and monitoring equipment, in a manner consistent with safety and good air pollution control practices for minimizing emissions. As noted earlier in this preamble and in the 2024 Final Rule, a large majority of sources are already reporting measured compliance data showing fPM emissions well below the previous fPM standard of 0.030 lb/MMBtu (either via PM CEMS or the stack testing compliance option), which further illustrates that the various options for demonstrating compliance with the fPM standards have been appropriate and effective.

\110\ 89 FR 38536 (May 7, 2024).

Comment: Commenters agreed with the EPA that the costs of installing PM CEMS are significant and the 2024 Final Rule failed to articulate why such costs were justified as compared to the other compliance methods permitted since 2012. Commenters argued that the EPA did not consider the fact that certain CEMS technologies, based on their designs and models, might not have replacement parts readily available from the original equipment manufacturers; therefore, when a critical component of the CEMS breaks or needs replacement, repairs could require costly expenditures or even a total replacement of the CEMS. Commenters also provided examples of how stack testing or CPMS costs could be much less than the cost of PM CEMS. For instance, commenters said that some sources conduct quarterly testing for HCl along with fPM to reduce stack testing costs further by merging the HCl testing and fPM testing into the same testing program mobilization, which would be noticeably cheaper than testing for HCl and fPM separately. Commenters also noted that testing costs for EGUs that previously qualified for PM LEE status were much less than the cost of PM CEMS, by some estimates as much as 20 to 30 times lower. Lastly, commenters explained that CPMS provide lower testing costs, simpler procedures for establishing operating limits, and fewer operational burdens compared to PM CEMS.

Other commenters disagreed that the higher costs of PM CEMS justify repealing the PM-CEMS-only compliance demonstration requirement from the 2024 Final Rule. Further, commenters said that the benefits of reduced pollution through the use of PM CEMS would exceed the minor incremental cost between PM CEMS and stack testing.

Response: As discussed in section III.A of this preamble, the EPA is finalizing the repeal of the more stringent fPM emission standard. The more stringent fPM standard required longer duration runs, which would increase the relative costs of stack testing. Therefore, the EPA's conclusion in the 2024 Final Rule that the costs for PM CEMS are commensurate with the costs for stack testing no longer applies because stack testing costs are not increasing due to longer duration runs.

However, the EPA disagrees that sources that elect to conduct quarterly testing for HCl tests combined with fPM would greatly reduce stack testing costs. On the contrary, adding HCl testing to an fPM testing program (or adding fPM testing to an HCl testing program) would result in an increased cost for testing, as opposed to lowering the cost. While some cost savings would be realized by efficiently “merging” a fPM and HCl testing programs into a single testing mobilization to test for those pollutants in conjunction, this approach would not result in a lower cost than what would be realized if testing for PM only.\111\ Nevertheless, the EPA understands that testing for fPM and HCl would not double the cost of an “fPM only” test program, since the fPM and HCl testing would occur concurrently without needing separate testing mobilizations, equipment setups, and equipment tear downs.

\111\ Commenters claim that quarterly stack testing for fPM costs $57,100/year, while merging the fPM testing with the HCl testing into the same testing mobilization would lower the testing costs to $37,500/year, yielding almost $20,000/year in cost savings. The EPA disagrees that costs for merging the stack testing for fPM and HCl would be lower than testing for fPM only.

The EPA agrees that testing costs for EGUs that previously qualified for PM LEE status will be much lower than the cost of PM CEMS, yielding further cost savings. Since LEE testing is only required once every 3 years once a source has attained LEE status, the EPA estimates that an ongoing LEE testing program is approximately 8 to 10 percent of the cost of a quarterly stack testing program.\112\

\112\ Per 40 CFR 63.10000(c)(1)(iii), an ongoing LEE PM testing program is required at least once every 36 months (or at 1/12th the frequency (i.e., 8.3 percent)) of a quarterly PM testing program to demonstrate continued LEE status.

Comment: Some commenters argued that the EPA failed to adequately explain its changed view of PM CEMS in terms of superior accuracy, transparency, and pollution reduction afforded by using PM CEMS, asserting that PM CEMS are already widely used throughout the industry and that the costs are small compared to a facility's overall operating expenses. These commenters stated that the EPA did not demonstrate that the other compliance demonstration options will provide the same or similar benefits as PM CEMS. Commenters asserted that PM CEMS have the benefits of greater reliability and accuracy and that continuous monitoring allows rapid detection of pollution problems so violations can be prevented or quickly fixed. Commenters also asserted that real- time information on pollution has more operational relevance for plant managers than do stack tests because real-time CEMS data allow managers to find the reasons for problems and provide learning that can have significant long-term benefits. Specifically, commenters stated that in the 2024 Final Rule, the EPA found that requiring the use of PM CEMS would provide 35,040 15-minute values for each EGU during a 1-year period, which is 243 times more information than is

provided by quarterly stack testing under the 2012 MATS Rule. Commenters also asserted that continuous monitoring can reveal wide variability that is obscured with stack tests that occur once a quarter, once a year, or even less frequently; and that continuous monitoring allows an automated response that can fix the problem before noncompliance occurs, or an alarm that notifies relevant personnel that a problem is occurring. Commenters also asserted that continuous monitoring is also a deterrent to negligence and fraud; when companies know that increases in pollution can be detected in real time, they are less likely to engage in risky or prohibited practices. Commenters also asserted that CEMS monitoring data allows the facility, neighbors, and regulatory entities to see if the facility is complying with the standard, as it does not rely on an outdated monitoring method and assumption that emissions will consistently remain at the level found during an unreliable stack test.

Further, commenters asserted that State regulatory programs rely on PM CEMS data to effectively limit emissions. For example, the Illinois Control Board recently relied on PM CEMS data in promulgating emissions limits for industrial facilities, including coal-fired power plants, during period of “start-up, malfunction, or breakdown.” Commenters said that this analysis would have been difficult to perform if only stack testing data were available, which would not comprehensively capture emissions levels during atypical startup, malfunction, or breakdown periods.

Response: The EPA recognizes that PM CEMS may have certain advantages over periodic stack testing in some situations. However, as discussed in section III.B of this preamble, pursuant to CAA section 112(d)(6), the EPA must consider cost in deciding whether it is “necessary” to revise the requirements and has broad discretion in selecting reasonable compliance demonstration methods under CAA section 112 and 114(a)(1). Stack testing under MATS has been shown to be less costly than operating and maintaining a PM CEMS on an annualized cost basis.\113\ Moreover, sources subject to both an applicable PM and HCl standard under MATS may use the stack testing compliance option to efficiently merge their PM and HCl testing programs into a single testing mobilization to test for those pollutants concurrently, possibly yielding further cost savings. The EPA also notes that stack testing for fPM under MATS continues to be required on a quarterly basis, which is more frequent than typical stack test programs which are required at either annual or multi-year frequencies.\114\ Finally, there is no indication in the record that EGUs are in noncompliance with the fPM standard. The record demonstrates consistent overperformance of the standard by a large percentage of the regulated community. In light of this, it is reasonable to continue to provide flexibility, and it renders the additional cost of mandating PM CEMS unnecessary.

\113\ Memo from Barrett Parker, EPA to Docket ID No: EPA-HQ-OAR- 2018-0794, “Revised Estimated Non-Beta Gauge PM CEMS and Filterable PM Testing Costs” (December 21, 2023).

\114\ For example, industrial boilers subject to 40 CFR part 63, subpart DDDDD, are required to perform fPM testing on either an annual or 3-year frequency, depending on the PM emission rate during testing. For stationary reciprocating internal combustion engines subject to 40 CFR part 63, subpart ZZZZ, testing for carbon monoxide (CO) is required every 8,760 hours or 3 years, whichever comes first. For stationary combustion turbines subject to 40 CFR part 63, subpart YYYY, testing for formaldehyde is required on an annual basis. However, testing for fPM is not required for stationary combustion turbines under 40 CFR part 63, subpart YYYY, or 40 CFR part 630, subpart KKKK, and is instead at the discretion of the delegated authority.

Comment: Regarding stack testing for fPM, commenters generally agreed with the EPA's proposal to retain the option of allowing either an increased minimum volume per run or the collection of a minimum mass per run, since that option would provide owners and operators greater flexibility that may reduce measurement uncertainty, lower test durations, and therefore lower the fPM stack testing costs. However, commenters requested that the proposed updated minimum volume per run of 4 dscm should instead return to the 1 dscm level of the 2012 MATS Rule. Commenters said that a 4 dscm minimum volume per run for all compliance demonstration options would require longer duration stack test runs of approximately 9 hours, posing larger costs, emissions, and operational difficulties for units.

Further, commenters argued that if the EPA finalized the repeal of the more stringent fPM limit and the PM-CEMS-only compliance demonstration requirement, the increase to 4 dscm minimum volume for each fPM stack testing run would no longer be prudent for all compliance demonstration purposes. While commenters stated that higher sample volumes could be useful to reduce measurement uncertainty for sources operating near emission limits or with low-level test measurements, individual owners and operators are best able to make such decisions based on the unit-specific compliance strategies. Commenters who disagreed with the EPA's proposal to retain these updated sampling requirements argued that more mass or volume does not create a more accurate correlation with PM CEMS, as the overall shape of the correlation curve is defined by mid- and high-level test conditions that have sufficient fPM mass due to artificially detuned conditions. A few commenters asserted that a minimum mass option is not feasible, as the amount of PM mass collected on the filter during stack testing is not known until after the run is completed and the filters are dried and weighed. While historical test results should enable sources to reasonably estimate the mass that will be collected during each fPM stack test run, commenters argued that sources should not be required to repeat test runs based solely on the collection of less mass than expected.

Other commenters argued that the updates to the sampling requirements in the 2024 Final Rule are necessary to ensure reliable test results, particularly at the low levels of fPM many EGUs are measuring. However, commenters expressed concern that the updates to the sampling requirements combined with the repeal of the more stringent fPM emission limit could result in emission spikes during stack testing. Commenters argued that the lower fPM emission standard combined with the sampling requirements in the 2024 Final Rule would have required plants to maintain the low emission rates they have been demonstrating and better protect surrounding communities.

Response: The EPA agrees that sampling requirements that allow either an increased minimum volume per run or the collection of a minimum mass per run would provide EGU owners and operators greater flexibility. The EPA also recognizes that a 4 dscm minimum volume per run would require longer duration stack test runs and that those longer test runs are not necessary for all fPM compliance demonstration options.

However, the EPA disagrees with the assertion that more mass or volume does not create a more accurate correlation with PM CEMS. While the mid- and high-level test conditions during PM CEMS correlations or verifications will most likely use the minimum catch value of 6.0 milligrams (mg) per run, it is still necessary to accurately measure values for the low-level test conditions. The EPA agrees with commenters that historical test data provides sources with a reasonable indicator and estimate of the potential mass that would be collected during each test run, therefore making it unnecessary to utilize the

minimum volume requirement, especially at the mid- and high-level test conditions.

After considering comments, and because the EPA is finalizing the repeal of the PM CEMS requirement and again allowing other fPM compliance demonstration options, the EPA is also finalizing minor technical revisions to the fPM sampling requirements. Specifically, the EPA is finalizing the requirement that allows EGU owners and operators using PM CEMS for compliance to collect either a minimum catch of 6.0 mg or a minimum sample volume of 4 dscm per run, in order to provide additional testing flexibility while also ensuring that a sufficient PM CEMS correlation sample is obtained. For EGU owners and operators using any other compliance demonstration option, the EPA is finalizing the revised minimum sample volume of 1 dscm per PM test run, which is considered sufficient for a representative gravimetric assessment of a source's PM compliance status.\115\

\115\ More specifically, the 4 dscm sample volume (or 6.0 mg sample catch) requirement would apply to any test associated with PM CEMS testing (e.g., Performance Specification 11 (PS-11) correlation curves, relative response audits (RRAs), and response correlation audits (RCAs)), whereas the 1 dscm sample volume requirement would apply to quarterly PM compliance testing. PM LEE testing programs would also be based upon the 1 dscm sample volume requirement, yet the required minimum sample volume for LEE testing will continue to be increased nominally by a factor of 2 (i.e., at least 2 dscm), per 40 CFR part 63, subpart UUUUU, table 2.

In the proposed rule, the EPA proposed to repeal the Hg standard for lignite-fired EGUs in the 2024 Final Rule and to return to the 4.0 lb/TBtu emission standard promulgated in the 2012 MATS Rule and retained in the 2020 Final Rule. The EPA solicited comment on whether there is sufficient data demonstrating that the standard can be met by lignite-fired EGUs with a range of boiler types and variable fuel composition (Question #6). The EPA also solicited comment on other achievable and cost-effective Hg standards for lignite-fired EGUs that are based on developments in practices, processes, and control technologies that the EPA should consider as an alternative to a standalone repeal of the 1.2 lb/TBtu standard (Question #7).

Upon further consideration and after reviewing comments received, the EPA is repealing the 1.2 lb/TBtu Hg limit for lignite-fired EGUs that was promulgated in the 2024 Final Rule and reverting to the 4.0 lb/TBtu Hg limit that was set in the 2012 MATS Rule and retained in the 2020 Final Rule. 2. What is the rationale for our final approach and decisions on the Hg emission standard for lignite-fired EGUs?

In the 2012 MATS Rule, the EPA promulgated a beyond-the-floor standard for Hg for the subcategory of existing coal-fired units designed for “low rank” virgin coal (i.e., lignite) based on the use of ACI for Hg control.\116\ The EPA established a Hg emission standard of 4.0 lb/TBtu for lignite-fired utility boilers and 1.2 lb/TBtu for utility boilers firing all other types of coal (e.g., anthracitic coal, bituminous coal, subbituminous coal, and coal refuse).

\116\ 77 FR 9304 (February 16, 2012).

The 2024 Final Rule lowered the Hg standard for lignite-fired EGUs from 4.0 lb/TBtu to 1.2 lb/TBtu based on the EPA's determination that commercially available control technologies and improved methods of operation would allow lignite-fired EGUs to meet a more stringent emission standard. The more stringent Hg emission standard brought the requirement for lignite-fired EGUs in line with the emission limitation requirements of EGUs firing all other types of coal. In the 2024 Final Rule, the EPA reviewed coal composition information and concluded that the Hg content, the halogen content, and the alkalinity were similar between various lignite and subbituminous coals. In 2021, EGUs firing subbituminous coal emitted Hg at an average annual rate of 0.6 lb Hg/ TBtu with measured values as low as 0.1 lb/TBtu, which the Agency found demonstrated that EGUs burning subbituminous coal have utilized control options to meet the 1.2 lb/TBtu emission standard despite the challenges presented by the low halogen content in the coal (which results in difficult-to-control elemental Hg vapor in the flue gas stream).\117\ The Agency asserted that its cost-effectiveness estimates for a model 800 MW lignite-fired EGU using a range of sorbent injection rates to meet the revised Hg emission standard were lower or consistent with cost-effectiveness values for Hg controls that the EPA has found to be acceptable in previous rulemakings.

\117\ 88 FR 24880 (April 23, 2023).

After reviewing the revised emission standard that was promulgated in the 2024 Final Rule, the EPA is repealing the revised Hg emission limit for lignite-fired EGUs because the revised standard was based on insufficient data to demonstrate that lignite units can meet the lower limit over the range of boiler types and variable compositions of fuels used at lignite-fired EGUs. Commenters provided both supportive and opposing arguments for issues regarding the Hg limit for lignite-fired EGUs. Comments received on the proposed repeal of the Hg limit for lignite-fired EGUs did not persuade the Agency to change its position from that set out in the proposed rule.

While the EPA found that all 22 existing lignite-fired EGUs at 12 facilities would need to control their Hg emissions to 95 percent or less to meet an emission standard of 1.2 lb/TBtu in the 2024 Final Rule,\118\ the Agency did not demonstrate that this high level of Hg removal is generally achievable for all lignite-fired units in the source category while taking into account the wide-ranging and highly variable Hg content of the various lignite fuels. In fact, Hg emission rates reported in the 2024 Final Rule from units at 11 of the 12 lignite facilities were well above the final 1.2 lb/TBtu emission standard.\119\ The EPA instead primarily relied on the emission reduction performance of only two units (at the Twin Oaks facility in Texas) that achieved the revised emission standard.\120\ Between August 1 and September 19, 2023, a series of Hg emissions performance tests were conducted on Twin Oaks units 1 and 2. The average Hg emissions rates for the 30-boiler operating day performance tests were 1.1 lb/ TBtu for unit 1 and 0.9 lb/TBtu for unit 2.\121\ Further, in performance testing for the previous year (2022), the average Hg emissions rates for the 30-boiler operating day performance test were 0.9 lb/TBtu for unit 1 and 0.6 lb/TBtu for unit 2. However, these tests were conducted over a limited operating period and are not sufficient to establish that meeting a 1.2 lb/TBtu standard continuously is possible for all lignite-fired EGUs.

\118\ Since the 2024 Final Rule, the Martin Lake and Limestone facilities have undergone permit changes that no longer allow the burning of lignite coal.

\119\ 89 FR 38546 (May 7, 2024).

\120\ Id. at 38539.

\121\ Id. at 38540.

Furthermore, the Twin Oaks facility, constructed in the early 1990s, is one of the newest lignite units and uses a circulating fluidized bed (CFB) combustor, which affects Hg emissions. Conventional boilers use coal that is pulverized to a very fine particle size (i.e., powdered) to maximize combustion efficiency and to minimize unburned carbon. In contrast, the design

of CFB combustors permits the burning of larger-sized coal particles. Fluidized bed units typically operate at lower temperatures compared to conventional boilers and have longer fuel residence times. As a result, CFB combustors typically have higher levels of unburned carbon present in the fly ash. The unburned carbon particles can behave much like injected activated carbon sorbent and, coupled with the lower operating temperature and longer residence time, can promote more efficient Hg removal as compared to that observed from units using non-CFB boilers with conventional pulverized coal combustors.

Other lignite-fired EGUs that utilize a CFB combustor also had generally lower Hg emission rates. For instance, the 2022 measured Hg rates reported in the 2024 Final Rule for the Red Hills facility in Mississippi, which also employs CFB combustors, was 1.7 lb/TBtu, compared to a range of 2.5 to 3.0 lb/TBtu for other lignite-fired EGUs in the southern U.S.\122\ Additionally, the lowest 2022 Hg emissions from lignite-fired facilities in North Dakota were found at Spiritwood Station, which also utilizes a CFB combustor. In revising the Hg emission standard for lignite-fired EGUs in the 2024 Final Rule, the EPA failed to evaluate the achievability of the revised Hg emission standard by affected sources that are not using the better performing CFB combustor technology.

\122\ Id. at 38548.

In addition, the EPA assumed in the 2024 Final Rule that the revised Hg standard of 1.2 lb/TBtu could be met by injecting better performing powdered sorbents using existing sorbent injection systems without the need for equipment modifications or additions. However, industry commenters noted that existing equipment at lignite-fired power plants may not be able to achieve the 1.2 lb/TBtu Hg limit and that demonstration testing would be required to determine a sorbent dosage rate, guaranteed injection rate, and the emissions rate that can be achieved when considering the Hg content variability of the lignite. Commenters stated that modifications to Hg control systems may be required to meet the 1.2 lb/TBtu emission limit. The EPA did not consider such cost in the final analysis for the 2024 Final Rule.

In addition, the Agency did not sufficiently investigate the complex composition of lignite coals in the 2024 Final Rule, including the variability of the Hg content in the inlet fuel source and the corresponding reductions needed to comply with the 1.2 lb/TBtu Hg emission standard. In the 2023 proposed rule, the EPA explained how the halogen content of coal influences the oxidation state of Hg in the flue gas stream, and thus the partitioning of Hg into elemental Hg vapor, oxidized Hg vapor, or particle-bound Hg, which impacts the Hg control approaches.\123\ Lignite and subbituminous coals both have a lower halogen content compared to bituminous coals, and the Hg in the flue gas from boilers firing those fuels tends to stay in the elemental vapor state, which is more challenging to control. The EPA noted that pre-halogenated (typically brominated) sorbents have been effectively utilized to control Hg emissions at power plants firing low-halogen content subbituminous coals. However, the EPA also noted that lignite coals tend to contain higher amounts of sulfur (more similar to some bituminous coals), which, under certain circumstances, can result in the production of sulfur trioxide (SO3) in the flue gas stream. SO3, in turn, is known to inhibit the effectiveness of some sorbents that are used for Hg control. The EPA acknowledged the challenges with higher sulfur content coals but noted that bituminous coal-fired power plants found ways to overcome those challenges-- sometimes by utilizing newly developed sulfur-tolerant sorbents. However, while the EPA acknowledged the respective challenges that the halogen and sulfur contents of coal can have on Hg control in the 2024 Final Rule, the EPA failed to address the impact of lower halogen content coupled with higher sulfur content on Hg control for lignite- fired power plants. Subbituminous coals tend to have low contents of both halogen and sulfur, while bituminous coals tend to contain higher levels of both halogen and sulfur. In comparison, lignites tend to have low halogen content (similar to subbituminous coals) and higher sulfur content (similar to bituminous coals).

\123\ 88 FR 24875 (April 24, 2023).

Commenters also provided data challenging the assumed inlet value of 25.0 lb/TBtu used in modeling in the 2024 Final Rule. For example, historical data indicate that lignite seams near the San Miguel plant in Texas result in coal feeds that have an average Hg inlet content of 34.0 lb/TBtu.\124\ As a result, San Miguel would need to achieve an average control rate of 96.3 percent to meet the standard in the 2024 Final Rule, compared to an 87.8 capture percentage for the 4.0 lb/TBtu emission limit .\125\ Additionally, monthly fluctuations in Hg content could require even higher control levels at least half the time. Ignoring monthly variability not only leads to an underestimation of costs associated with Hg removal but also overlooks control device modifications and enhancements required to achieve pollution control levels exceeding 90 percent.

\124\ Document ID No. EPA-HQ-OAR-2018-0794-5965.

\125\ Document ID No. EPA-HQ-OAR-2018-0794-5965.

It was not necessary to revise the Hg limit for lignite-fired EGUs in the 2024 Final Rule. This revised emission standard was based on insufficient data, and furthermore, the EPA did not demonstrate that the high level of Hg removal it required was generally achievable for all lignite-fired units in the source category while taking into account the wide-ranging and highly variable Hg content of the various lignite fuels. In addition, the Agency failed to evaluate its achievability by affected sources that are not using CFB combustor technology and assumed the revised standard could be met by injecting better performing sorbents without equipment modifications. The EPA also did not sufficiently investigate the complex composition of lignite coals, including the variability of the Hg content in the inlet fuel source, and ignored monthly variability, leading to an underestimation of costs. For these reasons, the EPA is finalizing the repeal of the Hg emission limit for lignite-fired EGUs that was promulgated in the 2024 Final Rule--1.2 lb/TBtu--and reverting to the Hg emission limit--4.0 lb/TBtu--that was promulgated in the 2012 MATS Rule. 3. What key comments did we receive on the Hg emission standard for lignite-fired EGUs and what are our responses?

Comment: Some commenters agreed with the EPA's proposed reconsideration of the Twin Oaks Hg data that the EPA relied upon in the 2024 Final Rule. First, commenters agreed the performance tests were conducted over a limited operating period and not sufficient to establish a more stringent Hg emission standard continuously for all units. Commenters argued that this facility was not representative of the national fleet because lignite seams burned at Twin Oaks differ from lignite at North Dakota lignite facilities and at other Texas lignite facilities. Second, commenters agreed that the CFB boiler design at the Twin Oaks and Red Hills facilities promote more efficient Hg removal compared to units using other types of boilers.

Other commenters disagreed with the EPA's proposed reconsideration of the Twin Oaks and Red Hills Hg data, arguing that the facilities' use of baghouses is an additional reason for highly effective Hg capture. These commenters asserted that while both facilities are equipped with ACI and would have very low SO3 present in the flue gas due to high free lime content, the presence of a baghouse makes the ACI, as well as any intrinsic capture of fly ash, much more effective. These commenters stated that if an ESP was installed instead of a baghouse, Hg capture would be more difficult. However, commenters asserted that every lignite unit is already configured in a manner for potentially higher Hg capture. These commenters stated that lignite facilities equipped with baghouses include Antelope Valley, Coyote, Spiritwood, Twin Oaks, Oak Grove, and Red Hills, and that some of those facilities also have an upstream dry scrubber (Antelope Valley, Coyote, Spiritwood) that helps make Hg capture with ACI more effective. The remainder of the lignite facilities (Coal Creek, Leland Olds, Milton R Young, Limestone, Martin Lake, and San Miguel) installed ESPs followed by a wet FGD, which enable additional Hg capture beyond what is achieved with ACI because wet FGD removes oxidized Hg very efficiently.

These commenters further asserted that higher Hg capture is possible for lignite plants with pulverized coal boilers (i.e., non-CFB units) and pointed to the Conemaugh power plant in Pennsylvania as an example. The Conemaugh facility is equipped with an ESP and wet FGD and burns bituminous coal. These commenters attempted to calculate the Hg content of coals burned at Conemaugh using EIA-923 data from 2016 to 2022, finding an average Hg content ranging from approximately 10 lb/ TBtu to over 50 lb/TBtu with a standard deviation ranging from near zero to almost 40 lb/TBtu.\126\ These commenters also asserted that most bituminous and subbituminous units receive coal from multiple mines and therefore face much greater variability in Hg content than lignite units, which are mine-mouth and only receive coal from one mine. Using the inferred Hg content information, commenters estimated that the Hg capture rate at the Conemaugh facility exceeded 95 percent every year, demonstrating that higher Hg capture at lignite units with lower Hg variability is possible using this configuration.

\126\ The survey Form EIA-923 collects detailed electric power data--monthly and annually--on electricity generation, fuel consumption, fossil fuel stocks, and receipts at the power plant and prime mover level. https://www.eia.gov/electricity/data/eia923/.

Response: The EPA agrees that baghouses are another technology for efficient Hg capture. However, the Agency did not demonstrate that higher Hg removal is achievable for all lignite-fired units once the wide-ranging and highly variable Hg content of the various lignite fuels is taken into account. As noted in the 2024 Final Rule, bituminous coals from Pennsylvania exhibit large Hg content variability similar to that of lignite, but bituminous coals also have higher natural chlorine content than lignite coal, which aids in the Hg removal efficiency.\127\ Therefore, the Agency does not agree with the commenters' example that a facility burning bituminous coal using a pulverized coal boiler demonstrates that a similar lignite-fired unit can achieve a similar level of Hg control.

\127\ 89 FR 38543 (May 7, 2024).

Comment: Commenters argued that the EPA failed to adequately justify returning to the 2012 MATS Rule Hg emission standard for lignite-fired EGUs in the proposed rule, as power plants that burn lignite coal represent a disproportionally large share of Hg emissions across all coal-fired EGUs. Specifically, commenters cited the 2023 proposal, in which the EPA provided that 16 of the top 20 Hg-emitting EGUs were lignite-fired and that lignite EGUs were responsible for about 30 percent of all Hg emitted from all coal-fired EGUs in 2021, while generating about 7 percent of total 2021 MW-hours.\128\ Commenters noted that the 2012 MATS Rule resulted in a 90 percent reduction of Hg from power plants, but few reductions came from plants that burn lignite coal. The commenters explained that lignite-fired EGUs are concentrated geographically in North Dakota and Texas, which increases the cumulative burden of such pollutants on surrounding and downwind vulnerable communities.

\128\ 88 FR 24876 (April 24, 2023).

Response: The EPA took all relevant comments and information, including information referenced above, into consideration when deciding whether to finalize the proposed repeal. Since the 2024 Final Rule, the Agency obtained updated information on the fleet of power plants burning lignite coal, finding that the Martin Lake and Limestone facilities (both in Texas) are no longer permitted to burn lignite and are now subject to a 1.2 lb/TBtu Hg emission standard. As mentioned earlier in this preamble, the 2020 Residual Risk Review found the residual risks due to emissions of air toxics to be acceptable from the coal- and oil-fired EGU source category and determined that the current NESHAP (as promulgated in the 2012 MATS Rule) provided an ample margin of safety to protect human health and prevent an adverse environmental effect. Risk from near-field deposition of Hg to subsistence fishers was evaluated, using a site-specific assessment of a lake near three lignite-fired facilities.\129\ The results suggest that methylmercury (MeHg) exposure to subsistence fishers from lignite-fired units alone is below the RfD for MeHg neurodevelopmental toxicity or IQ loss, with an estimated hazard quotient (HQ) of 0.06. In general, the EPA believes that exposures at or below the RfD are unlikely to be associated with appreciable risk of deleterious effects. The EPA reaffirmed its 2020 Residual Risk Review, which showed that emissions of HAP from coal- and oil-fired power plants have been reduced such that residual risk is at an acceptable level and provides an ample margin of safety, in the 2024 Final Rule.\130\

\129\ . Document ID No. EPA-HQ-OAR-2018-0794-0070

\130\ In the 2024 Final Rule, the EPA stated: “In the 2023 Proposal, the EPA determined not to reopen the 2020 Residual Risk Review, and accordingly did not propose any revisions to that review. As the EPA explained in the proposal, the EPA found in the 2020 RTR that risks from the Coal- and Oil-Fired EGU source category due to emissions of air toxics are acceptable and that the existing NESHAP provides an ample margin of safety to protect public health.” 89 FR 38518 (May 7, 2024).

Comment: Many commenters agreed that the EPA did not appropriately consider costs for lignite EGUs to meet a revised Hg emission standard in the 2024 Final Rule. Commenters stated the cost to comply with the revised Hg emission standard depends on the amount and type of sorbent required and ACI equipment additions or modifications, and since the amount of sorbent needed to achieve a more stringent standard is unclear and unit-specific, the sorbent cost cannot be reliably calculated. Commenters also disagreed with the EPA's claim in the 2024 Final Rule that SO3-tolerant sorbents could be used at lower feed rates to achieve greater Hg capture. Commenters stated that the EPA underestimated costs in the 2024 Final Rule by not considering the costs of modifications to the Hg control systems to meet a more stringent Hg emission standard, especially for units equipped with an ESP instead of a FF. Commenters stated that compliance costs cannot be accurately estimated because no lignite EGU has demonstrated that the revised Hg

emission standard can be met on a continuous basis. Commenters stated that these deficiencies in the Agency's cost analysis (e.g., failure to include annual capitalized costs for the Hg control system, updated sorbent costs, and costs based on theoretical sorbent usage) resulted, for example, in at least a $2.6 million underestimate for Milton R Young Station's Unit 2.

Other commenters claimed that the EPA provided no evidence in the proposed rule, and did not appear to conclude, that costs of possible modifications to Hg control systems render the Hg standard either not “achievable” or not cost-effective. These commenters stated that although the EPA asserted in the proposed rule that it did not previously consider the cost of possible modifications to control systems to meet the revised Hg standard, the Agency had in fact considered these costs in the 2024 Final Rule and found them to be reasonable.\131\ One of these commenters quoted a portion of the 2024 Final Rule in which the commenter believed that the EPA considered the costs of potential control system modifications and found that the need for “significant additional capital investment is unlikely.” \132\ Commenters asserted that the EPA stated in the 2024 Final Rule that the Agency expected sources to “be able to meet the revised emission standard using existing controls (e.g., using existing sorbent injection equipment)” and that “if site-specific conditions necessitate minor capital improvements to the ACI control technology, . . . any incremental capital cost would be small relative to ongoing sorbent costs accounted for in this analysis.” \133\

\131\ Document ID No. EPA-HQ-OAR-2018-0794-7609 (citing 89 FR 38508 (May 7, 2024).

\132\ Id. (citing 89 FR 38549 (May 7, 2024)).

\133\ Id.

Response: The EPA agrees with the commenters asserting that the Agency did not appropriately consider costs associated with Hg removal and overlooked costs for control device modification and enhancement required to achieve pollution control levels exceeding 90 percent within the 2024 Final Rule. In the 2024 Final Rule, the EPA noted that pre-halogenated (typically brominated) and sulfur-tolerant sorbents have effectively utilized to control Hg emissions at power plants firing low-halogen subbituminous coals. However, in the 2024 Final Rule, the EPA did not consider that the SO3-tolerant sorbents had not been extensively tested on lignite-fired EGUs and thus the feed rates and associated costs are uncertain and therefore disagrees with those commenters that asserted costs for these sorbents are reasonable. Since subbituminous coals tend to have low content of both halogen and sulfur and bituminous coals tend to contain higher levels of both halogen and sulfur, whereas lignites generally have the combined characteristics of low halogen content and higher sulfur content, the EPA disagrees with those commenters that asserted associated costs to achieve greater Hg capture in lignite-fired EGUs were properly considered in the 2024 Final Rule.

A. What did the EPA propose and solicit comment on regarding the relevance of residual risk to technology reviews under CAA section 112(d)(6)?

The EPA is finalizing the position that the Agency may consider the results of the one-time residual risk review requirement under CAA section 112(f)(2) in determining whether it is “necessary” to revise standards at the conclusion of subsequent CAA section 112(d)(6) technology reviews. Under CAA section 112(d)(6), the EPA is required “to review, and revise as necessary (taking into account developments in practices, processes, and control technologies), emission standards promulgated under this section no less often than every 8 years” (emphasis added). As noted in section II.A.1 of this preamble, the breadth of the term “necessary” in CAA section 112(d)(6) authorizes the EPA to consider the costs of revising standards in addition to the emissions-reduction potential of developments in practices, processes, and control technologies. Given the high costs and potential technical feasibility concerns with implementing the revised standards promulgated in the 2024 Final Rule, the EPA proposed to find that the 2024 revisions were not “necessary” under CAA section 112(d)(6) and solicited comment on whether and how the extent of further meaningful risk reduction opportunities should be considered in making that CAA section 112(d)(6) determination (Question #8). As explained in section II of this preamble, the EPA found in its 2020 Residual Risk Review that the residual risks due to HAP emissions from this source category are acceptable and determined that the current NESHAP (as promulgated in the 2012 MATS Rule) provided an ample margin of safety to protect public health and prevent an adverse environmental effect. The EPA reaffirmed the 2020 Final Rule, and did not reopen any of the underlying findings or conclusions of the one-time residual risk review requirement for MATS, in the 2024 Final Rule.\134\

\134\ See 89 FR 38518 (May 7, 2024). The results of the risk analysis indicated that both the actual and allowable inhalation cancer risk to the individual most exposed was well below 100-in-1 million, which is the EPA's presumptive limit of acceptability under the Benzene NESHAP. The results of the chronic inhalation cancer risk assessment based on actual emissions, as shown in Table 2 of this preamble, indicated that the estimated maximum individual lifetime cancer risk (cancer MIR) was 9-in-1 million, with nickel emissions from certain oil-fired EGUs as the major contributor to the risk. Approximately 193,000 people were estimated to have cancer risks at or above 1-in-1 million from HAP emitted from four facilities in this source category--all from oil-fired sources in Puerto Rico. However, the 2024 Final Rule only required controls for certain types of coal-fired EGUs and would not impact emissions from these oil-fired facilities.

As previewed in section III.A of this preamble, the EPA concludes that subsequent technology reviews under CAA section 112(d)(6) may consider the results of the one-time residual risk review requirement under CAA section 112(f)(2) in determining whether revisions are “necessary.” Specifically, the EPA will generally place greater weight on the cost of revising standards when the results of the one- time residual risk review requirement indicate that cancer risk from HAP emissions are less than the statute's aspirational goal of one in one million. Under those circumstances, revisions will generally be “necessary” only when costs are at the low end of the range of acceptability. This interpretation follows from the term “necessary” in CAA section 112(d)(6), which gives the Agency discretion to consider relevant factors and information, including information in the record for the NESHAP under review.\135\ It is also consistent with the risk thresholds that Congress wrote into the statute, including the requirement to promulgate additional standards when cancer risk exceeds the aspirational goal of one in one million and the

presumptive unacceptability threshold of 100 in one million expressly incorporated as part of the Benzene NESHAP approach.\136\

\135\ LEAN, 955 F.3d at 1097 (parenthetical in CAA section 112(d)(6) points to non-exhaustive list of considerations); see also Michigan, 576 U.S. at 752-53 (the term “appropriate and necessary” in CAA section 112(n)(1) directs agency to consider all relevant factors in exercising discretion); Nat'l Ass'n for Surface Finishing, 795 F.3d at 9-11 (discussing the EPA's discretion to consider relevant information when determining relevant information is “necessary”); Ass'n of Battery Recyclers, 716 F.3d at 673-74 (similar).

\136\ CAA section 112(f)(2)(B), 42 U.S.C. 7412(f)(2)(B).

Commenters provided supportive and opposing arguments as to whether a technology review conducted under CAA section 112(d)(6) should take into consideration whether any meaningful risk reduction would be obtained from further reducing HAP emissions under the technology review given the results of the residual risk review. The EPA has taken these comments into account in finalizing its position on the question and summarizes and responds to many of the most significant comments in the following section. Additional discussion, including further comment summaries and responses, are available in the Response to Comment document available in the docket for this rule.

C. What key comments did the EPA receive regarding the relevance of residual risk to technology reviews under CAA section 112(d)(6), and what are our responses?

Comment: Commenters agreed that the EPA should consider the potential for and materiality of risk reductions when conducting a CAA section 112(d)(6) technology review, citing multiple reasons. First, commenters argued that CAA section 112(d)(6) allows the EPA to revise standards only when “necessary,” which should be determined by assessing whether new regulations protect human health and the environment. Commenters stated that in Michigan, the Supreme Court explained that the EPA must consider costs when it evaluates benefits in deciding whether the 2012 MATS Rule was “appropriate and necessary,” and that case should similarly require the Agency to conduct a cost-benefit analysis to any new revisions under CAA section 112(d)(6). Commenters listed previous rules where the Agency used the lack of meaningful risk reduction as a factor in its cost analysis, such as the Industrial Process Cooling Towers NESHAP RTR,\137\ Petroleum Refineries NESHAP RTR,\138\ Halogenated Solvent Cleaning NESHAP reconsideration proposal,\139\ and Organic Hazards Air Pollutants from the Synthetic Organic Chemical Manufacturing Industry NESHAP RTR.\140\ Commenters also noted that in the RTR for the Coke Oven Batteries NESHAP, the EPA stated generally that findings concerning risk that the EPA makes in a section 112(f)(2) determination may be relevant in making any subsequent 112(d)(6) determinations for the related 112(d) standard.\141\

\137\ National Emission Standards for Hazardous Air Pollutants for Industrial Process Cooling Towers, 71 FR 17729, 17731-32 (April 7, 2006).

\138\ National Emission Standards for Hazardous Air Pollutants from Petroleum Refineries, 72 FR 50716, 50730 (September 4, 2007).

\139\ National Emission Standards for Hazardous Air Pollutants for Halogenated Solvent Cleaning, 73 FR 62384, 62404 (October 20, 2008).

\140\ National Emission Standards for Hazardous Air Pollutants for Organic Hazardous Air Pollutants from the Synthetic Organic Chemical Manufacturing Industry, 71 FR 76603, 76606 (December 21, 2006).

\141\ National Emission Standards for Coke Oven Batteries, 70 FR 19992, 20009 (April 15, 2005).

Second, commenters noted that the residual risk of HAP emissions from coal-fired EGUs are minimal under the MACT standards from the 2012 MATS Rule. Specifically, commenters noted that the EPA's 2020 Residual Risk Review found that the maximum lifetime cancer risk from coal-fired EGUs ranged from 0.002 to 0.344 in 1 million, which, commenters asserted, are orders of magnitude below what commenters further asserted is Congress's threshold for deregulating the source category. Commenters argued that further reducing cancer risks that are already less than one in one million yields negligible benefits, if any.

Other commenters disagreed that the EPA should consider residual risk during a CAA section 112(d)(6) technology review, arguing that the proposed rule misconstrues the technology review process under CAA section 112(d)(6) and fails to apply the best reading of the statute. These commenters stated that the EPA must review any new developments in control technologies and explain why those new developments make it necessary to adopt new standards. These commenters stated that the proposed rule is arbitrary and capricious because the EPA failed to consider any developments in practices, processes, and control technologies in finding that rescission is necessary. In the proposed rule, the Agency stated that “[g]iven the high costs and potential feasibility concerns with implementing the revised standards . . . the 2024 changes were not necessary under CAA section 112(d)(6).” \142\ However, these commenters argued that the EPA added no new information or analysis concerning developments in practices, processes, and control technologies--the rulemaking record is essentially identical to that underlying the 2024 Final Rule, as the EPA recognized in its 2025 RIA. Instead, these commenters argued, the EPA now determines that the 2024 changes were not necessary in a conclusory manner, only citing changes in the Administrator's policy preferences at the EPA as “developments” rather than citing any factual developments in practices, processes, or control technologies at emissions sources.

\142\ 90 FR 25544 (June 17, 2025).

Lastly, these commenters claimed that CAA section 112(d)(6) does not allow the EPA to withdraw revised standards based on its claim that otherwise achievable controls produce no “meaningful risk reduction.” These commenters also argued that the EPA provided no reasonable basis to conclude that the 2012 MATS Rule still provides the maximum degree of emission reduction achievable under CAA section 112(d)(2), instructing the EPA to “tak[e] into account developments in practices, processes, and control technologies.” These commenters asserted that the EPA may not decline to make otherwise “necessary” revisions based on its appraisal of risk reduction.

Response: The EPA agrees that the Agency has an independent statutory authority and obligation to conduct technology reviews every 8 years separate from the EPA's obligation to conduct a one-time residual risk review within 8 years of setting the MACT. The D.C. Circuit has recognized the CAA section 112(d)(6) technology review and CAA section 112(f)(2) residual risk review are “distinct, parallel analyses” that the EPA undertakes “[s]eparately.” \143\ It would be inconsistent with the text, structure, and legislative history of the CAA for the EPA to conclude that Congress intended the statute's technology-based approach to be sidelined after the Agency concludes the risk review, particularly because technology reviews, unlike the residual risk review, must be completed every 8 years on an ongoing basis. In the past, the EPA has occasionally determined that additional controls were warranted under technology reviews pursuant to CAA section 112(d)(6) although additional standards were not necessary to maintain an ample margin of safety under CAA

\143\ Nat'l Ass'n for Surface Finishing, 795 F.3d at 5.

section 112(f)(2).\144\ The EPA has also previously stated that it “disagree[s] with the view that a determination under CAA section 112(f) of an [ample margin of safety] and no adverse environmental effects alone will, in all cases, cause us to determine that a revision is not necessary under CAA section 112(d)(6).” \145\

\144\ See, e.g., National Emission Standards for Hazardous Air Pollutants: Refractory Products Manufacturing Residual Risk and Technology Review, 86 FR 66045 (November 19, 2021); National Emission Standards for Hazardous Air Pollutants: Site Remediation Residual Risk and Technology Review, 85 FR 41680 (July 10, 2020); National Emission Standards for Hazardous Air Pollutants: Organic Liquids Distribution (Non-Gasoline) Residual Risk and Technology Review, 85 FR 40740, 40745 (July 7, 2020); National Emission Standards for Hazardous Air Pollutants: Generic Maximum Achievable Control Technology Standards Residual Risk and Technology Review for Ethylene Production, 85 FR 40386, 40389 (July 6, 2020); National Emission Standards for Hazardous Air Pollutants for Chemical Recovery Combustion Sources at Kraft, Soda, Sulfite, and Stand-Alone Semichemical Pulp Mills, 82 FR 47328 (October 11, 2017); National Emission Standards for Hazardous Air Pollutants: Generic Maximum Achievable Control Technology Standards; and Manufacture of Amino/ Phenolic Resins, 79 FR 60898, 60901 (October 8, 2014).

\145\ National Emission Standards for Hazardous Air Pollutant Emissions: Group I Polymers and Resins; Marine Tank Vessel Loading Operations; Pharmaceuticals Production; and the Printing and Publishing Industry, 76 FR 22566, 22577 (April 21, 2011).

However, the EPA also agrees with commenters' assertions that costs should be considered in relation to potential benefits when evaluating whether revisions are “necessary” under CAA section 112(d)(6). That concept is inherent in the EPA's consideration of standard cost metrics, including cost effectiveness expressed as cost per ton of HAP emissions abated. HAP emissions figures are important because of the health and environmental impacts they represent, and it is reasonable to consider such impacts when determining whether to regulate. As noted earlier in this preamble, the EPA has long evaluated cost effectiveness in the context of particular HAP because different HAPs present different physical and risk characteristics. That concept is also inherent in the nature of cost consideration. As the Supreme Court explained in Michigan, “[c]onsideration of cost reflects the understanding that reasonable regulation ordinarily requires paying attention to the advantages and the disadvantages of agency decisions,” and it would be irrational “to impose billions of dollars in economic costs in return for a few dollars in health or environmental benefits.” \146\

\146\ 576 U.S. at 752-53.

As explained in section III.A of this preamble, there are circumstances in which the EPA may consider risk in a CAA section 112(d)(6) rulemaking as part of its determination of whether revisions to emission standards are “necessary.” The EPA concludes that in the present rulemaking, it may consider the small remaining risk from the non-HG HAP metals emitted from the source category--the cancer risks are less than one in one million for every coal-fired EGU in the source category--in assessing whether the costs of controls for those pollutants are too high, such that a revision to the standards based on those controls is not “necessary.” This conclusion is consistent with previous rules under CAA section 112(d)(6) in which the EPA has considered risk.

In 2020, the EPA finalized electronic data reporting requirements of MATS, including requiring data availability in Extensible Markup Language (XML) format and amending the reporting and recordkeeping requirements associated with performance stack tests, PM and HCl CEMS, and PM CPMS.\147\ As a result, sources are required to use the Emissions Collection and Monitoring Plan System (ECMPS) Client Tool to submit all required reports. The deadline to meet changes in electronic reporting was December 31, 2023, which has since been tentatively extended to the 1st quarter of 2026.

\147\ 85 FR 55744 (September 9, 2020).

As part of this rulemaking, the EPA is finalizing two minor technical, non-substantive clarifications to the relevant electronic data reporting requirements, such as (i) removing references to “ECMPS” and replacing with “ECMPS Reporting Tool” and (ii) revising the XML file format requirement to any file format specified by the Administrator. The main effect of these minor technical clarifications is that they make clear that the EPA will accept MATS reports in both XML and PDF (or other) formats, as opposed to only XML. These minor technical clarifications will better clarify and enable the reporting of electronic compliance data, in light of the fact that some reporting aspects are not supported by XML.

The EPA is making these non-substantive clarifications including under the “good cause” exception to notice-and-comment rulemaking incorporated by reference into the statute in CAA section 307. Under section 553(b)(B) of the Administrative Procedure Act, an agency may forego notice-and-comment rulemaking when it “for good cause find[s]” that providing notice and an opportunity for comment would be “impracticable, unnecessary, or contrary to the public interest.” Here, providing notice is “unnecessary” because of the minor, non- substantive nature of the technical clarifications.\148\ As these changes do not alter the substantive reporting requirements, there is good cause to make them without prior notice and comment. The Agency emphasizes that these non-substantive clarifications are separate and independent from any other change made in this final rule.

\148\ See Mack Trucks, Inc. v. EPA, 682 F.3d 87, 94 (D.C. Cir. 2012) (“This prong of the good cause inquiry is `confined to those situations in which the administrative rule is a routine determination, insignificant in nature and impact, and inconsequential to the industry and to the public.' ”) (quoting Util. Solid Waste Activities Grp. v. EPA, 236 F.3d 749, 755 (D.C. Cir. 2001)).

This action is a significant action under E.O. 12866 section 3(f)(1) that was submitted to the OMB for review. Any changes made in response to E.O. 12866 review have been documented in the docket. The EPA prepared an analysis of the potential costs and benefits associated with this action. This analysis, Regulatory Impact Analysis for the Final Repeal of Amendments to National Emission Standards for Hazardous Air Pollutants: Coal- and Oil-Fired Electric Utility Steam Generating Units, is available in the docket.\149\

\149\ Docket ID No. EPA-HQ-OAR-2018-0794.

We present the estimated present values (PV) and equivalent annualized values (EAV) of the estimated cost savings of repealing the 2024 Final Rule in 2024 dollars over the 2028 to 2037 period, discounted to 2025. In addition, the Agency presents the assessment for specific snapshot years, consistent with historic practice. These snapshot years are 2028, 2030, and 2035. The power industry's cost savings are represented in this analysis as the change in electric power generation costs due to the repeal of the 2024 Final Rule requirements. In

simple terms, these cost savings are an estimate of the decreased power industry expenditures resulting from the repeal of the 2024 Final Rule requirements.

Under this final repeal, the 2024 Final Rule would no longer reduce emissions of Hg and non-Hg HAP metals as projected in the 2024 MATS RTR RIA.\150\ The potential benefits from reductions of HAP were not able to be monetized in the 2024 MATS RTR RIA, nor were potential impacts from the 2024 Final Rule requirement to use PM CEMS for compliance demonstration. See section I.A for more details of the final repeal of requirements.

\150\ “Regulatory Impact Analysis for the Final National Emission Standards for Hazardous Air Pollutants: Coal- and Oil-Fired Electric Utility Steam Generating Units Review of the Residual Risk and Technology Review” (Ref. EPA-452/R-24-005). Document ID No. EPA-HQ-OAR-2018-0794-6966.

Table 4 presents the estimated cost savings of this final action in 2024 dollars discounted to 2025. This table presents the PV and EAV of these estimates discounted at 3 percent and 7 percent.

Table 4--Present Value and Equivalent Annualized Value of Compliance

Cost Savings Estimates of the Final Repeal From 2028-2037

[Millions of 2024$, Discounted to 2025]

3 Percent 7 Percent

discount discount

rate rate

Present Value................................... 670 490 Equivalent Annualized Value..................... 78 69

Note that, unlike the proposal, the compliance cost analysis for the final rule reflects the recent Presidential Proclamations that temporarily exempt certain stationary sources from compliance with the 2024 MATS RTR requirements. Additionally, this updated analysis incorporates recent updates to state and federal legislation affecting the power sector, including the One Big Beautiful Bill Act (OBBBA) of 2025, and also accounts for large projected increases in electricity demand stemming from data centers and Artificial Intelligence applications.

The EPA is obligated to present the agency's best scientific understanding and the implications of that science when developing policies and regulations. However, the EPA's analytical practices may not have presented the full range of uncertainties and associated confidence level regarding the potential benefit estimates from reduction in exposure from fine particulate matter (PM2.5) and ozone. In addition, the science regarding the exposure, health effects from exposure and valuation of reduction in health effect are evolving with better data and methods, especially at low concentrations of PM and ozone. The EPA's use of benefit per ton (BPT) monetized values introduces additional uncertainty. Although developed as a screening tool when full-form photochemical modeling was not feasible, the BPT approach reduces complex spatial and atmospheric relationships and may be more suited to model emissions that are geographically more uniform and species better mixing, thereby adding uncertainty associated with those estimates. Some of the sources of uncertainties include the set of assumptions used in projecting the health impact of reducing particulate matter. These projections are based on a series of models that take into account emissions changes, resulting distributions of changes in ambient air quality, the estimated reductions in health effects from changes in exposure, and the composition of the population that will benefit from the reduced exposure. Each component includes assumptions, each with varying degrees of uncertainty.

In addition, the EPA historically provided point estimates rather than just ranges of emission-related effects or only quantifying emissions when monetizing proved to be too uncertain. Therefore, to address these concerns, the EPA is refraining in providing primary estimates resulting from changes in PM2.5 and ozone exposure resulting from changes in NOX emissions but will continue to quantify the emissions until the Agency if confident enough in the modeling to robustly monetize those impacts.

A more robust description of the benefits and cost of this rulemaking is contained in the RIA, available in the docket, and is consistent with Executive Order 12866.

The information collection activities in this rule have been submitted for approval to the Office of Management and Budget (OMB) under the PRA. The Information Collection Request (ICR) document that the EPA prepared has been assigned EPA ICR number 2137.13. You can find a copy of the ICR in the docket for this rule, and it is briefly summarized here. The information collection requirements are not enforceable until OMB approves them. OMB has previously approved the information collection activities contained in the existing regulations and has assigned OMB control number 2060-0567. The information collection activities in this rule include performance testing, continuous emission monitoring, notifications and periodic reports, recording information, monitoring and the maintenance of records. The information generated by these activities will be used by the EPA to ensure that affected facilities comply with the emission limits and other requirements. Records and reports are necessary to enable delegated authorities to identify affected facilities that may not be in compliance with the requirements. Based on reported information, delegated authorities will decide which units and what records or processes should be inspected. The recordkeeping requirements require only the specific information needed to determine compliance. These recordkeeping and reporting requirements are specifically authorized by CAA section 114 (42 U.S.C. 7414). The burden and cost estimates below represent the total burden and cost for the information collection requirements of the NESHAP for coal- and oil-fired EGUs, not just the burden associated with the amendments in this final rule. The incremental ICR cost savings associated with the final repeal of these amendments is $29.8 million per year.

Respondents/affected entities: The respondents are owners or operators of coal- and oil-fired EGUs. The NAICS codes for the coal- and oil-fired EGU industry are 221112, 221122, and 921150.

Respondent's obligation to respond: Mandatory per 42 U.S.C. 7414 et seq.

Estimated number of respondents: 201 per year. Each facility is a respondent, and some facilities have multiple EGUs.

Frequency of response: The frequency of responses varies depending on the burden item. Responses include daily calibrations, monthly recordkeeping activities, semiannual compliance reports, and annual reports.

Total estimated burden: 191,000 hours (per year). Burden is defined at 5 CFR 1320.3(b).

Total estimated cost: $76,800,000 (per year); includes $24,800,000 annualized

labor costs, $4,500,000 annualized capital costs, and 47,500,000 annualized operation & maintenance costs.

An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB control number. The OMB control numbers for the EPA's regulations in 40 CFR are listed in 40 CFR part 9. When OMB approves this ICR, the Agency will announce that approval in the Federal Register and publish a technical amendment to 40 CFR part 9 to display the OMB control number for the approved information collection activities contained in this final rule.