# Oral argument — Keystone-Conemaugh Projects LLC v. US Environmental Protection Agency > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/argument%3A88812 ## Record - **Collection:** Oral argument - **Document type:** Oral argument - **Published:** November 6, 2023 ## Text Our next case is Keystone v. EPA 22-3026. Mr. Schaffer. Good morning, sir. Good morning, Your Honor. May it please the Court, my name is Joseph Schaffer and I represent Petitioner Keystone-Conemaugh Projects LLC. Jesse Walker of Pennsylvania DEP has four minutes of the remaining petitioner's time. May I reserve two minutes for rebuttal? Yes. This case is about limits. In one respect, it's about limits on emissions of nitrogen oxides, or NOx, that EPA set in a federal implementation plan for four Pennsylvania electric generating stations. But in another, more fundamental respect, it's about the limits of EPA's authority under the Clean Air Act. We are asking this Court to vacate the federal plan at issue here because EPA exceeded those limits. How did they exceed these limits? And what are the limits as you understand them? Well, the limits start first with Section 110c1 of the Clean Air Act, because Section 110c1 establishes three ways, and only three ways, in which EPA can promulgate a federal plan. So this is a procedural challenge at the moment you're talking about? This is not a procedural challenge, Your Honor. This is a statutory challenge. My point is, let's separate out what you think was wrong with the procedure EPA followed from what you think is wrong with the substantive limits that it adopted. I will agree in that respect. If the Court refers to it as a procedural challenge, yes. This is not addressing the substance of the limits that EPA adopted. It's addressing the power of the EPA in the first instance to publish or issue the federal plan with those limits. And here, there are only three ways in which EPA has the authority under the Clean Air Act to publish or issue a federal plan. And the only one that EPA invoked here was its authority to issue a federal plan after disapproving a state plan. So it seems like the heart of your complaint is that the presumptive overall cap was knocked out and then EPA charged straight in and adopted a series of individual site-specific ones. So why was that jumping the gun? Why was that procedurally improper? Well, to do that, for EPA to adopt any type of limits, it had to properly invoke its authority under Section 110c1. So it's not a matter of EPA not liking the presumptive approach and deciding to adopt a different approach. It's that EPA had to do one of three things, either find that the state had failed to submit a plan, that the plan was administratively incomplete, or that the plan was disapproved. And here, EPA invoked its authority under Section 110c1b. Right. They disapproved the state implementation submission in whole or in part. And the issue here is that EPA disapproved the wrong plan. There were only two plans here, state plans, that EPA conceivably could have disapproved in order to give itself the authority to promulgate a federal plan. The one that EPA selected was a state plan the PADEP had submitted in 2016. And this Court will recall that that was the subject of litigation, a petition for review, that was filed in this Court and that was decided in the 2020 Sierra Club decision. And the consequence of that 2020 Sierra Club decision was vacater. And vacater wipes the slate clean. Consider one of the definitions that Black's Law gives for vacater is annulment. And so consider it this way. If you annul a marriage, you don't put the proposal back on the table for one of the parties to reject. You wipe the slate clean. But what EPA did here after Sierra Club is it treated the vacater as if that 2016 state plan were still pending in front of it to act upon. And it wasn't. The slate was wiped clean. And it was incumbent upon PADEP in the first instance, under Section 110c1, to take the first step. And EPA didn't wait. What it did is it reached out to that 2016 state plan that was a legal nullity after this Court's Sierra Club decision, and it proposed to disapprove that. The whole plan was a nullity? Only parts of it? Only parts of it were a nullity. That's true. But those were the parts that EPA proposed to disapprove in order to give itself the authority. We're dealing with a two-year clock here. Maybe we set the clock too short or something. But they're racing against the clock. Well, EPA was under a deadline from this Court. But the existence of a deadline does not excuse EPA from complying with the plain language of the statute. And the statute here requires EPA to or limits EPA's authority, rather, to issue a federal plan only if it takes one of those three preparatory actions. Why can't it treat our vacatur of the part as the equivalent of disapproving the state implementation plan submission in part? Well, first, EPA didn't do that. If it had done that, then it would have had no need to take the action to propose and then finalize disapproval, and an appellate court can't rely on an argument for rulemaking that EPA did not raise below. But in the second instance, the plain language of Section 110C limits the preconditions to action taken by EPA, not by this Court. And so the action would need to have been taken by EPA, and it wasn't here. The only state plan that EPA could have disapproved here to give it the authority to issue the federal plan was the 2022 revision that PADEP had submitted. And that state plan has still not been acted upon. EPA has never disapproved it. And for EPA to have promulgated the federal plan here, it would have needed to do that. Your Honors, I'd like to turn to the second issue that I want to raise with the Court today. And it assumes that EPA had the authority to issue the federal plan. Now, it didn't. But even if it did, EPA imposed NOx limits here based on projections of selective catalytic reduction or SCR usage that are contradicted by EPA's own fact findings. And this is a substance issue, Your Honors. So let me start with an analogy. If you know that your car returns better fuel economy during highway driving than city driving, you wouldn't estimate your city fuel economy by using a data set weighted towards highway driving. The return of result is completely divorced from reality. In the context of these stations, base load operations are much like highway driving. They have high heat inputs and steady state operations. Cycling operations, by contrast, are more like city driving, lower heat inputs and more frequent start-ups and shut-downs. Everyone here acknowledges that these stations currently operate as cycling units. And they have for about the last five years. Everyone also acknowledges that cycling units will, at least occasionally, perhaps even often, need to operate below the minimum heat inputs for their SCR systems to function. As a result, you would have expected EPA in this federal plan to develop NOx limits by estimating how frequently cycling units, not base load units, but cycling units, can operate SCR equipment. But that's not what EPA did. What EPA did is take a 10-year data set from 2011 to 2021 that included multiple years of base load operations. And this is most apparent, Your Honors, in pages 511 to 515 of the appendix, where EPA's technical source document, technical guidance document, shows the graphs with the hours that these stations spend above SCR thresholds and below SCR thresholds for the past several years. That 10-year threshold included multiple years of base load operations, approximately five for each of these stations. So it was already weighted towards base load operations. And what EPA then did is select the third highest SCR usage rate from that data set, or for Kahnemuth, the second highest rate. And given that EPA had already staffed the data set towards base load operations, selecting the third highest frequency of operations gave EPA an estimate of SCR usage that was biased towards base load operations. It's not representative of cycling operations at all. And so EPA staffed the data set. It cherry-picked the variables. It committed the same errors that this Court found required vacator in its Sierra Club decision in the same results required here. Your Honors, I have a few minutes left, and so I just will briefly... I have a question. Yes, Your Honor. About Montour. Montour seemed to suggest that the EPA's decisions, which you believe warrant vacator and remand, are in fact defects that courts typically find curable. Why isn't that applied to the other three? Yes, Your Honor. And that goes to remedy, of course. So Montour relies on the allied signal factors for that, Your Honor. And they say that the allied signal factors look first to whether a defect could be cured on remand and also to the disruption from vacator. And here the deficiencies were not curable on remand for two reasons. First, EPA acted contrary to law when it issued the federal plan here because it didn't have the authority. That's an incurable defect. It cannot be addressed by additional fact-finding or support in the record on remand. But there's also a fundamental error here with how EPA approached the calculation of these limits for these three other stations. It affects the frequency, which I talked about. It also affects the rate for reasons explained in our briefs. And, again, that's not something that can be cured by the addition of additional explanation into the record, which is something that's typically the case when remand is ordered without vacator. What is it you're asking us to do? We're asking you to vacate the federal plan here and to remand to EPA for either the approval of a compliant state SIP provision, and there is one pending in front of EPA that it can and is required to act upon under the Clean Air Act. That's what we're asking this court to do. If there are no further questions, I am at time. Thank you. Thank you very much. Mr. Walker. Good morning. May it please the Court. My name is Jesse Walker, and I represent the Intervenor Commonwealth of Pennsylvania Department of Environmental Protection. I have two points I'd like to make today for the Court. Those points are EPA's FIP has imposed RAC standards Pennsylvania cannot explain and has circumvented Pennsylvania by ignoring its SIP submissions. The first point regarding EPA's FIP issuance is that Pennsylvania is forced to issue permits incorporating the FIP as RAC that lacks the legal and technical basis to defend them because EPA is imposing a novel strategy that departs from its longstanding RAC guidance. This is problematic because Pennsylvania has to submit the permits as SIP provisions and is stuck holding the EGUs to a standard that is not RAC. It's based on EPA's interstate transport methodology that has never been used before for Clean Air Act RAC purposes and departs from the agency's RAC standard. Pennsylvania must do so or be subject to Clean Air Act sanctions. EPA's contention on page 26 of its brief that it did not violate the Clean Air Act by usurping Pennsylvania's emission control choices because Pennsylvania still has to submit a SIP is hollow and self-serving. This is because Pennsylvania must adopt the FIP control limits in its subsequent RAC submittal. This problem is compounded because EPA erred in issuing a FIP that disregarded the technological limitations of the pollution control technology despite EPA's own admission of those limitations and changed current and future EGU operations, joint appendix 154 and page 44 of their brief. Did not perform an evaluation of existing source equipment degradation at the three EGUs, which is inherent in the case-by-case analysis and instead substituted with a third best ozone approach and omitted, and this is important, omitted consideration and analysis of EGU operating data outside the ozone season for 2011 to 2021 reflecting those technological limitations, which results in improperly skewing the data towards lower emission limits on a year-round basis. That's at Pennsylvania's brief, pages 27, 28, joint appendix 691. The second point regarding SIP inaction is that EPA should have acted on Pennsylvania's SIP submittals instead of circumventing Pennsylvania. The facts here are distinguishable in this case because Clean Air Act section 110C1B requires a disapproval and both Sierra Club and EPA's after-the-fact August 2022 disapproval only involved Pennsylvania's presumptive RAC regulation. Pennsylvania submitted case-by-case RAC SIPs pursuit to a different lawful EPA SIP-approved case-by-case regulation, which this court acknowledged in Sierra Club, but EPA never acted on those submittals. EPA issued the after-the-fact August disapproval of only the presumptive RAC regulation and then its FIP, which it claimed was case-by-case RAC using a methodology inconsistent with its longstanding case-by-case approach. EPA's brief on page 27 notes that Pennsylvania SIPs remain pending before the agency, and if EPA approves them, they would replace the FIP. But there has been no action by EPA to do so before the imminent deadline under the Clean Air Act just weeks away. There was action the first time around, the first time around. So I find it interesting that you and your friend on your side are focusing all on activity in the last year, but why isn't it enough that, you know, we had a SIP and we had FIP before the 2020 decision in Sierra Club in this case? Sure. Thank you, Your Honor. Yes, so it's Pennsylvania's position that the scope under 110C1B, the disapproval, was to the presumptive RAC regulatory provision, and you were correct earlier when you said only parts of that. That was the provision that was struck down. And then Sierra Club also acknowledged the case-by-case approval, so Pennsylvania moved to issue a SIP pursuant to its lawful SIP-approved regulation at which EPA never disapproved any case-by-case. So it's Pennsylvania's position that, like Oklahoma and Arizona, which EPA relies on in its case, that EPA should have acted either before or concurrently, consistent with its own agency practice, as those cases referenced, to take action on Pennsylvania's SIP. Let me ask. This is something that's not really part of the merits here, but what happened? Wasn't it Pennsylvania working with the EPA on their SIP, and then all of a sudden communications broke down? Thank you, Your Honor. Yes, so what had happened was after the Sierra Club decision, Pennsylvania, as this Court would charge Pennsylvania with submitting a revised compliant SIP or EPA to issue a FIP, that must be in accord with the agency's RAC standard. And so what happened, Pennsylvania proceeded consistent with 12999 to have the operators submit a SIP revision to Pennsylvania. Pennsylvania found upon reviewing the operators' SIP revisions that those were inadequate, and then commenced working to develop a SIP revision, and cooperatively worked with EPA for a number of, period, over 20 meetings, over 1,000 man hours. And quite frankly, the timing issue had come up where EPA, they say in their brief that they were essentially developing the FIP with a radically different strategy at the same time they were working with Pennsylvania on its tiered emission limit approach, which is an alternative emission limit approach different than what was rejected in Sierra Club by this Court. But so basically, they had, and I believe it was late February, informed Pennsylvania that they were going to issue the FIP, and we submitted our SIPs in time, but they just sat on them and did not act on them. That's what happened. Was that because the time, the original opinion was too short? I think that we could have used just a little bit more time, yes. Yes, Your Honor. These facts all demonstrate that EPA should have acted on Pennsylvania's case-by-case submittals either before or concurrent with the FIP, and as I mentioned, EPA relies on multiple cases in their briefs, Oklahoma, Arizona, showing that it's common agency practice for it to do so. In closing, Pennsylvania had its emission control choices supplanted by EPA's SIP inaction, and it faces incorporating FIP limits that it can neither explain nor defend as racked. You're going to hear from EPA's counsel that they had to do something to satisfy this Court's order in Sierra Club by August 27th, despite submitting an extension request with this Court for itself. But the FIP does not satisfy that order because it is not racked. For the reasons explained today in those in Pennsylvania's briefing, intervener Pennsylvania DEP requests that this Court vacate the FIP. Thank you. Thank you. Mr. Adkins? Yes. Good morning, and may it please the Court. My name is Brandon Adkins of the United States Department of Justice. I represent the respondent, United States Environmental Protection Agency. I'd like to make three points for your honors. First, EPA had statutory authority to promulgate a federal plan upon disapproving portions of the 2016 state plan. EPA's action was also reasonable, and it was consistent with this Court's remand instructions in the Sierra Club case. Second, the federal plan reasonably addresses the problem of the power plants at issue in this case, hovering at temperatures below which they can operate their SCR controls, and thereby creating excess NOx emissions. And third, the federal plan reasonably reflects EPA's careful consideration of decades of operating data for the particular sources at issue here today, showing that those sources are capable and have achieved the federal plan limits, or the limits that are set in the federal plan. We ask that this Court deny the petition. All right. So focus on your first point. Your friends on the other side hang their hats on. There was no action on the most recent plan. Talk about the partial disapproval of the 2016 plan and why that suffices under 110C1. Sure. Under 110C, we believe the text of the statute is unambiguous. Upon disapproving a state plan, the agency is obligated to promulgate a federal plan at any point within two years, unless two things happen. The state corrects the deficiency and EPA approves that corrected state plan. It's true that the state submitted a corrected state plan after EPA proposed a federal plan and before the agency finalized that federal plan. The agency has not yet taken action on that plan. It will in due course. But just because the state submitted the state plan doesn't extinguish or limit EPA's authority. Okay. So I think I'm with you on that. But I think the more troubling thing here is this is a scheme of cooperative federalism. All the cases say that it's supposed to work this way, hand in glove, give the state the first shot. You disapprove this presumptive overall cap, but then charge straight to these individual site-specific limits without going through giving the state a shot at what to do with these particular facilities. So why is it that we shouldn't say that the changes you make after the partial disapproval are limited to the scope of the things that you partially disapproved? Because your changes are broader than what you disapproved. Sure. I'll address this question. I don't think the changes were broader than what we disapproved, what the agency disapproved. But this is an argument that the state raised for the first time in its reply brief, frankly, where the state conceded two things. They conceded first that the federal plan set source-specific emissions limits. That was an argument that we heard from the petitioners in prior briefing. Oh, this wasn't a source-specific limit. So that's one thing that they conceded. Second, they conceded that EPA has some authority under the Clean Air Act when disapproving a state plan to promulgate the federal plan. Their issue is that because the agency promulgated source-specific limits, they exceeded that authority because the 2016 plan set categorical limits. It was a presumptive approach. There's no support in the statute to support Pennsylvania's argument. There's nothing in any of the case law to support Pennsylvania's argument. The source-specific limits that are set in the federal plan are for the same sources and under the same RAC standards of the disapproved portions of the state plan. EPA also acted reasonably here. In month 21 of the 24-month revamp period, EPA proposed the federal plan. On that day, the state still had not submitted a revised state plan. It's true that the agency advised the state in February 2022, with only six months to go before the end of the remand period, that it would then focus its resources on pursuing the federal plan and finalizing a federal plan. But I think this is an instance where the agency was actually operating quite reasonably and waited quite a bit of time, worked with the state up until that point, before pursuing the federal plan. Just from a boots-on-the-ground perspective, so you have this pending SIP now, right? Correct. What happens to that? So the SIP's pending. The agency will take action on the SIP. The deadline to do so, the SIP came in in two waves. The first deadline is towards the end of this month. The second deadline is towards the beginning of December this year. Is it possible the EPA is going to implement the SIP? Approve the SIP? That is one possibility. The agency has not yet proposed how it will act on the SIP. So it could approve, it could disapprove. It could approve in part or disapprove in part. But there is no question that this Court's decision on the federal plan would set the bar for what is wrapped for these sources. With respect to my second point, the federal plan reasonably addresses the issue with these sources hovering at temperatures below which they can operate their SCR controls. There's no question here that the power plants are capable of very low NOx emissions when they're using the controls that are already installed. What EPA and the state observed in the operating data for these sources is that sometimes they use their controls and sometimes they don't. And they don't or they observed instances where they didn't at the same temperature thresholds where they could have been operating those controls and thereby causing excess NOx emissions. EPA's weighted rate approach attempts to encourage the operators to use the controls that are already installed while accounting for the cycling that was observed in the more recent operating data. Because the federal plan does not set a threshold or different tiered rates for whether the units are operating at high demand or low demand, it closes the gaping loophole that this Court identified in the Sierra Club opinion. With respect to my third point, the federal plan is based on EPA's very thoughtful consideration of decades of operating data for each of these sources. The petitioner, by contrast, has not pointed to one ounce of data to contradict those findings. EPA reasonably looked at ozone season data to set a rate that applies year-round. The RAC standards must apply year-round. EPA used ozone season data because it reflects a time when the sources, it shows what the sources are capable of achieving when they're using these NOx controls. Similar sources in neighboring states have achieved similar NOx reductions year-round, and EPA looked at non-ozone season data, too, and confirmed that the sources can use their SCR controls outside of the ozone season and achieve similarly low rates that are set in the federal plan. For all those reasons, as well as those stated in our brief, we ask that the Court deny the petition. I'm happy to answer any other questions. Are you still working with the state on their SIP? My understanding is that the EPA has not been working with the state with the state plan, but it is pending before the agency, so there's really, it's in the agency's court, so to speak, to take action on what's been submitted now. You say you're not working with them, so it's under your consideration is what it is. That's correct. Okay. Thank you. Thank you very much. Mr. McFedrin. McFedrin. Rhymes with excedrin. Good morning. May it please the Court. Charlie McFedrin for Sierra Club. EPA's rule offers significant emission benefits, 6,000 tons per year of NOx reductions. That's in Table 3 of the final rule. Sierra Club supports EPA's action and urges the Court to deny the petitions for review. 17 Pennsylvania counties in which 8 million people live fail to meet the ozone health standard. We have an urgent need to protect the public health by taking action to reduce pollution from our biggest polluters, including the plants at issue here. The Clean Air Act gives EPA broad authority to issue SIPs, to issue FIPs, excuse me, and to deal with SIPs, sorry. State plan, regarding state plans, EPA can approve state plans, can deny state plans, can revise state plans, or can issue its own plans, as we heard here about a federal plan, regarding RACT, R-A-C-T. The Act requires EPA to issue a federal plan within two years after disapproving a state plan. Counsel for EPA took the Court right to the language of 110C, 7410C of the Clean Air Act. The Administrator shall promulgate a FIP at any time within two years after the Administrator disapproves a SIP, in whole or in part, unless the State corrects the deficiency and the Administrator approves the plan or plan revision before the Administrator promulgates such a FIP. That is the exact timeline that has occurred here. EPA has acted to propose a FIP before there was even a State submission in May and June of 2022. EPA finalized its disapproval in August and then issued a FIP. EPA has followed the letter and the spirit of 110C. And I have to say, EPA supervision is required in this statute. We talk about how the States are important, how the States have the lead on these issues. Yes, but EPA management is critical. And that's why the statute reads the way it does. That's why 110C, and I'm also talking about 110K, provide these various tools for EPA to supervise State action. And EPA action here has been essential. Since the remand of the prior rule in 2020, we've gotten more data on emissions levels, NOx levels from these plants. EPA action here will decisively reduce pollution in a way that will help people in those 17 counties I mentioned, the 8 million people. EPA's plan is based on a comprehensive review of data, the details of which are spelled out in the record. It takes the state-of-the-art pollution controls that these sources already have and requires them to be operated more frequently. It provides flexibility by averaging over 30 days, averaging emissions to even out emission spikes. And it provides for a daily emission limit that's based on maximum levels of operation. And in its final rule, EPA provided even more flexibility, providing a plant-wide average based on an average of weighted rates at the various units of the plant. State-of-the-art controls, 6,000 tons of reduction, flexibility. The court should deny the petition for review. Thank you. Thank you. Thank you. Judge Restrepo, you asked what happened here. And Judge Bevis, you asked a question about cooperative federalism. And that's really what's at the heart of this case. My friend on the other side told this court, as of the date of disapproval, the state had not submitted its SIP, the 2022 SIP provision. Well, the reason for that is because EPA told PADEP not to do so. That's at page 699 of the record. It's in PADEP's comments. And it's unrebutted by any other evidence. Now, the question here is, why would EPA tell the state, don't submit a SIP? The state has a statutory obligation to submit a SIP, even if there's a FIP that's in place. And, in fact, the state faces significant sanctions, the loss of significant funds, if it does not have an approved SIP in place. So why? Why would EPA tell the state not to do that? And the answer, I think, is in a comment that my friend made during oral argument. That is, the FIP sets the floor for whatever rack these sources must meet. That's the issue. The reason why EPA told PADEP not to submit a SIP is it wants to set the floor. It wants to control the manner of emissions and the way that PADEP regulates these sources. And it cannot do that under the Clean Air Act. It's prohibited. It's inherent in the cooperative federalism structure of this statute that the state has the first opportunity to determine the manner of emissions in its jurisdiction. As Bethlehem Steele Rigorsich said, Seventh Circuit case from the 1980s, the state proposes, EPA disposes. What EPA is doing here is taking away the ability of the state to propose. It is federalizing the Clean Air Act, and that, Your Honors, is why vacator is essential here. Because if this court does not vacate, it will give EPA a blueprint in which it can federalize the Clean Air Act and take away from the states the congressionally delegated power to determine in the first instance how it should control emissions in its jurisdiction. Nothing further, Your Honors. Are there any questions? No. Thank you. Thank you. Counsel, we thank you all for your arguments and your briefs. It's been really helpful, and we will take this case under advisement. We're going to order that the transcript be prepared, and counsel should speak to the clerk with respect to the cost and the preparation. And lastly, split the cost of the transcript. Thank you, Your Honor. --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/argument%3A88812. Public record. Not legal advice.