Oral argument — Environ Comm FL Elec Power v. EPA

Case number 15-1239 et al. Environmental committee of the Florida Electric Power Coordination Group Inc, petitioner, versus Environmental Protection Agency, and Michael S. Regan. Good morning, Council. Mr. Ezra, please proceed when you're ready. Thank you, and may it please the Court. The state implementation plan provisions at issue in this case date back to the earliest days of the Clean Air Act. They have adequately served to protect our air for nearly 50 years. And yet, in 2015, EPA decided that the provisions were inconsistent with the Act and needed to be removed. In doing so, EPA got the standard for calling a state plan wrong, and in any event, and as EPA has recognized in three subsequent notice and comment rulemaking, it issued a call that is indefensible on the merits. Let me start with the standard for calling a state plan, because I think that issue cuts across every other issue in this case. And under the statute, Section 110K5, a state plan is callable when the administrator finds it is substantially inadequate to attain or maintain the relevant national ambient air quality standard to mitigate adequately the interstate pollutant transport or to otherwise comply with any requirements in this chapter. And the EPA called all of the state plans at issue here on the theory that a provision is substantially inadequate whenever it conflicts with what EPA decides is fundamental provisions of the Act. But that's wrong, and it's wrong in two ways. First, it is wrong because the SIP call standard requires some factual analysis of real-world effects to find substantial inadequacy. And second, EPA's own fundamentality standard just cannot be squared with the text of the Act. Let me start with our primary submission, which is that a SIP call requires factual analysis, and we derive that rule from text, context, and purpose. So let me just jump straight into the text. On our view, we think that three textual features demonstrate that we are looking to factual risk to the requirements of the Clean Air Act when issuing a SIP call. First, the text requires substantial inadequacy, and on our view, that demands facts to differentiate near inadequacy from substantial inadequacy. And I think textually that comes from the definition of the word substantial, which means considerable importance, value, degree, amount, or extent. And we need context from facts to make that call. Without facts, we don't know whether any given inadequacy that is some legal problem with the plan is really important or valuable or considerable in amount. And I think an example really helps bring this point home. EPA says here that continuous emission limit is a fundamental requirement of the Clean Air Act. Thus, according to the EPA, any break in continuity is a substantial inadequacy justifying a SIP call. But I don't think that any normal English speaker would say that a nanosecond break in an emission limit for one source out of potentially hundreds of sources in a state is a substantial issue, at least absent some factual evidence that the break in time matters. Is that true across the board? Because it may be true that in some situations the nature of the error calls out for an analysis of substantiality along the lines that you're suggesting, which is kind of an empirical assessment of how much does it matter on facts on the ground. But it doesn't seem to me to be immediately apparent that the phrase substantial inadequacy just by nature can't be used in context in which you don't need a factual on the ground assessment. There could be legal inadequacies that are material ones and therefore are substantial. So I guess I'd say a couple of things. The first is I'm not sure how we would know that the legal inadequacies you point to are material ones unless we look to some factual evidence. And so I do think that we ultimately end up looking for facts. And the analogy I would draw is that I think one of the things that the substantial inadequacy language is doing is something akin to harmless error review. So when we identify a legal error in a trial, we don't then just say the trial needs to happen again. We say, let's go look and see if it really made a lick of a difference in the real world. And I think that's what substantial inadequacy is asking, is did this make a lick of a difference? But the other thing is, if substantial inadequacy really was targeting just pure legal errors, I think we would see it come up a lot in the Clean Air Act. Because the Clean Air Act is not shy about imposing legal rules. And we went through and we looked at the entirety of the Clean Air Act. And the substantial inadequacy phrasing only comes up five other times. And they're all in reference to the SIPP call standard. And so I think if substantial inadequacy was really how this act meant to describe legal violations, you'd expect to see it a whole lot more. That's an interesting point about using the same phraseology cutting across the statute. But let me just consider this one in isolation for the moment. And suppose, for example, that our decision in NRDC, the one that comes up in the affirmative defense context, as everybody knows, suppose that that goes to the Supreme Court and the Supreme Court agrees and there's no footnote to, and there's, in fact, a determination that applies to state SIPPs, too. And I know that all those are caveats that you all don't want to accept, but just bear with me for purpose of the example, right? And then the Supreme Court in its decision says, and this is a significant problem, because there has to be authority on the part of courts to figure out what the nature of the violations and the nature of the sanctions are going to be. And an affirmative defense, as by hypothesis the D.C. Circuit said, is incompatible with that. At that point, is your position that that still wouldn't be the kind of substantial inadequacy that would allow the EPA to call the SIPPs that contain the affirmative defense that the Supreme Court by hypothesis would have said is incompatible with the EPA, I mean with the CAA? That's right. And let me give two reasons for that. So the first is I think even if the Supreme Court identifies there is a legal inadequacy in the state plan, that still doesn't tell us whether the legal inadequacy is substantial. And so it may be a very important legal inadequacy. I think that would make it much easier for EPA to make the factual findings that it needs to make to call the state plan. But it wouldn't establish that the inadequacy is substantial. The second point I would make is I don't think that should be particularly surprising, because if there's a pure legal error in the plan, the EPA has the ability to fix that. It just doesn't come from Section K-5. It comes from Section K-6. So Section K-6 says that if EPA made an error in approving the plan, and again, under Subsection K-3, EPA approves the plan after reviewing whether they meet all the requirements of the act. So if EPA made an error in approving the plan, it can identify that error and just remove it out of the plan. So EPA under K-6 can fix pure legal errors that were apparent at the time of plan submission. But here, EPA did not want to use its K-6 authority. That's JA-25. And I think having chosen not to use its K-6 authority, EPA should not be allowed to water down the standards in K-5 in order to reach the same end objective. And we've talked about... Okay, and let me just ask one last question along these lines, which is, even if it's true that for a legal error to be seen to be material enough, because I think you're speaking in terms of materiality at some level, that there would have to be some consideration of the extent to which that has a practical consequence. It's not immediately apparent that there has to be an analysis in a backward-looking one. It could just be a sort of forward-looking one that just says, look, this is a major problem under the CAA because these kinds of errors... Let's just use the affirmative defense example. One could just hypothesize that, well, these kinds of errors having an affirmative defense is going to have some serious practical consequences in all likelihood going forward. And it doesn't seem to me that you necessarily would need a detailed assessment. That could just be a blanket determination that this is just a significant thing because of the practicalities that may attend it. So I think I agree with everything that you said. I don't think it needs to be backward-looking. I think the assessment can be forward-looking. I think, you know, if EPA wants to say take affirmative defenses, we've run some models. We've looked at the facts on the ground a little bit. We concluded that these make it so that you can't actually in practice enforce the emission limits. I think, you know, we'd probably be back here debating whether that's arbitrary or capricious, but that would satisfy the need to find factual findings. EPA would be assessing whether there was a substantial inadequacy. I think ultimately, like, this is the easiest case because we don't have to debate what level of facts you need to show substantiality. And, you know, courts have gone in circles about what materiality means, and depending on the context, that can be a hard question. Here, EPA entirely disclaimed the need for facts. So I think in many ways this is the easy scenario. Now, I do want to hit the two other textual features that I think are important in the need for facts. And the second one is that the statute requires that the administrator make a finding before calling a state plan, and then Section 110.828 requires that finding to be on the basis of information available to the administrator. And on our view, those terms signal that we're looking for facts. I think that's true as a matter of plain language. We call them findings of fact and conclusions of law, not findings of law. I also think that's true as a matter of structure. So the call provision's demand for a finding stands in stark contradistinction with how the Act describes all of the other legal requirements in place in the plan submission process. It's quite telling. Throughout the plan submission process, states have to meet a number of legal requirements, and the Act always refers to the plan as needing to meet those requirements, not EPA needing to issue a finding that the requirements are met. That happens time and time again. If you look at Subsection 110k1a, EPA is told to promulgate minimum criteria that any plan must meet. In 1b, EPA reviews those criteria to determine if the plan met them. In 1c, EPA treats as unsubmitted any plan that doesn't meet the minimum criteria. In k3, EPA approves plans that meet all of the criteria. It's only when we get to Subsection k5 we have a finding based on substantial inadequacy. Our view makes that difference meaningful. EPA's does not. And the third textual feature is that if you look at the first sentence of Subsection k5 as a whole, I think the immediate context makes clear that we're looking for facts. So that sentence gives three grounds for calling the state plan substantial inadequacy to attain the NAAQ, substantial inadequacy to mitigate interstate pollutant transport, and substantial inadequacy to otherwise comply with the requirements of this chapter. I think anyone agrees that the first two noun phrases, the NAAQ and the interstate pollutant transport, require facts to make the substantial inadequacy finding. That's the EPA brief pages 115 to 116. But that suggests that the third noun phrase also requires facts. Because if it were otherwise, the same adjectival phrase, substantially inadequate, would be having different modifying meanings based on which noun in the same list it was modifying. So in your view, the only prospective modification that would be authorized is one based on modeling. That would be state specific? So when I say modeling, I just want to clear up, I'm not sure that I necessarily need a formal model. You know, I think EPA could, depending on sort of make predictive judgments based on facts on the ground. I think that would ultimately need to be state specific, but I don't think that EPA would need to point to any specific facts in any state. So that may be a little counterintuitive. Let me explain what I mean. I think if EPA says, look, we've identified positives in, let's say, 10 states, that for whatever reason we think are really problematic and we have facts on the ground that show that. I think EPA could extrapolate from that and say, well, these other five states, we don't necessarily have facts, but because of the presence of the factual analysis we've done for some states, we can conclude that these are going to be a substantial issue. So we conclude that those provisions are substantially inadequate. So I don't think EPA. It seems anomalous to think that if EPA in its experience has, you know, developed an understanding that there's a gap as a legal matter in a SIP, you know, whether that's something that has been factually demonstrated or not. If it's, you know, and I guess going back to the to the chief's question, if it's something that everybody now understands is a problem, it's it seems like it's not very consistent with otherwise comply with any requirement of this chapter. Just feels like a little, you know, redundant. Well, let's allow, you know, let's develop state specific models or let's allow the next to be violated so that we can make a demonstration that's factual. Seems like it's incumbent or arguably it's incumbent on the EPA to nip that in the bud before the emissions exceed the max. So there's a lot there. Let me try to unpack it. I hope I can get to everything. So on the first point, I don't think that we would hamstring EPA from identifying what I think is the core of your question, where EPA has identified a legal issue and has determined that that issue is making a difference on the ground from from remedying that issue in other states where EPA is not necessarily doesn't necessarily have the facts. Because I think although EPA ultimately needs to make a state specific finding, I'm not saying that it needs to base that finding on facts that are specifically from that state. Although I would point out sort of as an aside, the EPA can always get that data if it wants it under 110P. EPA can demand whatever data it wants from any state in considering whether to revise the SIP. Now, the second point is I think baked into the premise of your question was the idea that we've already concluded that the legal error issue is actually having some impact on the ground. And that's our whole point. EPA has to find some impact on the ground. And then the third thing I'd say, and this goes back to my answer to Chief Justice Navasan, if EPA really thinks that the legal error in the plan is just something that it can't live with, EPA has that authority. It just needs to proceed under K-6 because EPA would say, look, there's a legal error. That legal error was there at the time the plan was approved. We made an error approving the plan. And so we are using our K-6 authority to remedy that. EPA just doesn't want to use its K-6 authority here because what that would mean is it would remove the provisions that EPA identified as erroneous from the plan. And EPA thinks that's a bad policy outcome. EPA wants to go through the SIP approval process because it wants to leave the plan in place during the transition period. And our point is EPA shouldn't be allowed to avoid K-6 and in doing so, water down K-5. But that seems a little anomalous in view of cooperative federalism. What you're saying is the EPA itself should revise rather than, as under a SIP call, leave some discretion to the state to figure out how it wants to revise. I recognize that that may be a little bit anomalous, although K-6 is in there and we've got to live with it. And K-6 does say if there was a legal error in approving the plan, then EPA can fix it. But I don't think it's that anomalous because I think that K-5 and K-6 are really targeting different situations. I think they're really targeting the situation in K-5 where EPA has properly approved the plan. And for whatever reason, it just does not work in practice. And I think under that sort of framing of K-5, it makes sense that there would be a demand for facts because you have to demonstrate why it's not actually working in practice. And K-6 is meant to deal with the situation where EPA really made an error that would have been apparent from the outset. Now, I think on the whole, text, context, and purpose all make clear that the SIP call provisions demand facts. And if I could just briefly move on to discussing why, even if the SIP call standard was right, I don't think the SIP calls were justified here. And let me start with the automatic exemptions. So EPA says that automatic exemptions are— Before you jump there, let me ask one follow-up question on the line of argument you were making. You contrasted K-5 and K-6, but what if you contrast K-5 and—this is long—A2H2? And A2H2 talks about revisions on the basis of information. So that seems to imply a textual—sorry, that seems to imply a factual finding, whereas K-5 doesn't have that phrase on the basis of information. So what do we do with that variation? Sure. So I think that this court's already been down that road in Virginia, the EPA, that the two provisions work together and you read them in part or material. So I think K-5 says we need a finding. I think finding incorporates facts. And if there's any doubt about that, I think it's just supported by A2H. I think that the court's already gone down that road in Virginia. Why is it supported by A2H instead of contrasting with A2H? Because I think that the two provisions are really going to the same direction. So A2H talks about when EPA has the authority to demand a SIP call in a different provision of the statute. So I think—I'm not sure how you would draw the textual line between demanding a SIP call under A2H and demanding a SIP call under K-5. EPA certainly hasn't said that they're different sources of authority. And, you know, I do think that's why in Virginia the court said that you read the two provisions together. Again, well, if I do—let's say I agree with you that find, when it's used in a legal sense, a specifically legal sense, requires the kind of factual finding that you're talking about. But in a general sense, people say find all the time, and they would apply it to the kind of legal question. And they would apply it to any kind of error. I found an error. It wouldn't necessarily be a factual error. And if—you know, usually when we prize a legal term of art over a general definition, we need some reason to do that. What's the reason to do that here? So I think the reason is the immediate context, which is that when the Act is referring to criteria that a plan just might—must meet that is legal criteria, and I think, you know, baked into your question, this is find a legal error, especially with that framing of the word find. It talks about the criteria that what the plan must meet. It doesn't use the word find. It doesn't say EPA needs to find that the criteria are not met. And so I think on that framing, we would then say, okay, well, then find clearly is looking for something different. But I guess the bottom line is, if all I had is find, I probably wouldn't be leading the argument that EPA needed facts. I don't only have find. I've got substantial inadequacy. I've got the rest of K-5. I've got the contrast between the standards for going through this plan approval process where EPA gets to compare them and say, are all the legal requirements met versus the standard for calling a SIP? I've got the difference between K-5 and K-6. I've got the purpose of this Act, which is meant to be cooperative federalism. And we think that demanding facts means that EPA has to show that putting the states through the effort of going through a SIP call is going to make a lick of a real world difference. And so, you know, I think when I've got all that, I think find helps me, but I don't need to rest only on find. If there are no further questions, I'll reserve the balance. I'll make sure my colleagues don't have additional questions for you at this point. Okay. Thank you, Mr. Esrey. Mr. Frey, we'll hear from you now. May it please the court. I'm here on behalf of the numerous manufacturing, agricultural, and electric power interests that we've referred to as industry petitioners. I hope I can bring some real world perspective to this discussion, which has been pretty theoretical so far. For over 30 years, EPA, states, and the courts have all recognized two undisputed facts. That emission limitations based on what pollution control technology can achieve must recognize and accommodate the limits of that technology. And that it would be fundamentally unfair to penalize facilities for times when the control technology cannot meet the emission limitations based on normal operations. Even though the facility has taken all reasonable measures to install and operate and maintain that control technology. But then in 2015, in response to a rulemaking petition and a consent decree, EPA told over two-thirds of the states that their EPA approved SIPs were substantially inadequate. Because they contain provisions addressing those realities, addressing those undisputed realities. Now, there hadn't been any change in EPA's recognition of those principles. EPA just decided that most of the EPA approved SIPs, in the various measures that they chose in their discretion to accommodate the limits of control technology, have failed to comply with what EPA now describes as fundamental requirements that it's identified. Requirements that EPA says have been in the acts since at least 1977. Now, that conclusion that over two-thirds of the approved SIPs are inadequate would be remarkable in and of itself. But it's made more remarkable by the fact that EPA has disavowed any need to show that the identified provisions are in fact causing air quality problems. And by EPA's consistent disregard of relevant case law and of key words in the statute. Failing to take a look at the effects of the SIP as a whole is really a central problem here. And the courts have told EPA over and over again that it must do so. The agency instead insists that specific SIP provisions have to be written a certain way. While at the same time admitting that if you change the provision, it wouldn't prevent the emissions that are associated with SSM events from still occurring because they are, by definition, unavoidable with reasonable precautions. So EPA's preferred approach not only elevates form over function, but it really won't further the goals of the Clean Air Act. And in many instances, it likely will ultimately result in less effective measures to minimize air pollutants. Now, let me... Mr. Frey, it seems like the basic premise of your argument today is that nothing can be done to mitigate emissions during SSM. And I recognize that our other decisions, for example, Sierra Club and U.S. Sugar were in the context of Section 112. But the sort of anomalous character of SSM events is common across emissions of HAPS and emissions that are addressed by emissions of the six criteria pollutants. So in those cases, we could know there is a requirement that there not be exemptions or mere reliance on a general duty provision in order for the emissions limitations to be enforceable. So I think the question is not whether different approaches are needed for SSM events than for steady state operation, but more, is there something that is enforceable where the source is saying or the state is requiring do X, Y, and Z and X ante before and X are exceeded, those limitations can be enforced? Well, respectfully, Your Honor, I don't quite agree with your characterization of the court's opinions related to standards for hazardous air pollutants. I think what those precedents read carefully say is that EPA has to have a basis for the requirements that it imposes, not that there can't be alternative requirements to a particular numerical standard that applies during SSM events. But in this case, we're talking about state implementation plans, which, again, the precedent is clear that EPA, to disapprove a SIP, to make a SIP call, has to look at the SIP as a whole. It's not just disapproving a portion of the SIP, which it can do when it's reviewing a SIP in the first instance. The SIP call has to look at the SIP as a whole, and the provisions that EPA called, that EPA said rendered, no, EPA said those provisions were substantially inadequate. Actually, the language they used didn't say the SIP is substantially inadequate, but those provisions that EPA looked at in isolation, first of all, by their own terms, a lot of times, provided the additional protections that you described. For example, in order to claim an affirmative defense or in order to have the director determine that a violation didn't occur, facilities, under many of these provisions, have to show that they took all reasonable steps to minimize their emissions, that they minimized the duration, so they have to shut down if that would reduce the emissions. There are provisions in many of these SIPs that say you can't take advantage of this SSM provision if there was an ambient standard violation that occurred and that this event contributed to. Perhaps the best example, I think, of the problem with EPA's approach is, in some cases, when the state commented on the proposed rule and said, well, look, you're just looking at this SSM provision, but we have another provision in another part of the standard. The state can still require you to CC emissions despite this. We have all these other provisions and EPA said, well, yeah, but they're not in the same section. They're not in the same paragraph of your regulation. That's why I'm talking about elevating form over function. The totality of the SIPs, and EPA explained this again in the actions it took in 2020, but as we explained in our brief and in our comments to EPA originally, the totality of the SIP is what EPA needs to look at, and that's what it failed to do in this case. Let me just talk about what the impact is of these SSM provisions and how, in fact, it's not – well, EPA says the potential impact justifies these SIP calls, even though it says it doesn't really need to evaluate the impact. The interveners say, oh, you can have a lot of big problems if you have an SSM event, so that's why you can't have an SSM provision. In fact, what we have to recognize is that, again, by definition, these are events that are not avoidable with reasonable precautions, and they're events that are occurring at a particular time, as Mr. Esrae alluded to, a particular time for a particular source. Can I just interject just before we get too far off? You had made the point that you understand EPA to be saying that it wouldn't consider controls that weren't in the same part of the state statute. I wonder, just for my assistance in evaluating that, I don't think you're brief sight to a portion of the rule. Where would you look in the rule to confirm that that was part of EPA's ground for the SIP call, that something wasn't codified together in one place? Let's see. Well, for one example, Your Honor, we cited two comments of a number of states pointing out that they had other provisions in other parts of their state implementation plan, and EPA did not. You're thinking of, for example, the North Carolina comments? Right, exactly, and the discussion in North Carolina when EPA in 2020 withdrew that SIP call, which is the latest promulgated interpretation of the Clean Air Act that we have from EPA, actually. EPA recognized there that what the SIP call was doing was really quarreling with the way that the North Carolina SIP was organized, rather than with all the details of the plan as a whole. But if it would be helpful, we can provide additional citations to that problem. I know this is a – by virtue of the way EPA did this, and they said, oh, we've got 36 states that are inadequate, it's hard to find all the individual statements. So if that would be helpful, we could provide a letter to the court after this on that point. So let me go back to the practical impact. Let me first say that although the intervenors say the serious environmental impacts justify getting rid of these SIP provisions, there's not in the record information that ties the called SIP provisions to particular air quality problems. And in any event, EPA disavowed that as a reason for these SIP calls. And so under SEC v. Shannery and Michigan v. EPA, we can only look at what justification EPA gave for the SIP calls. But in any event – So I take its justification to be the legal argument that the emission limitations and their effectiveness and enforceability is central to compliance with 110, and that if you think of it as, you know, that you have to have a fence to keep the sheep in, and they'll escape if you have gaps in the fence. You can build a fence of different materials for different parts of the terrain, but you have to have the fence. And when there's a hole, we can't be confident that the sheep won't escape. And that's what we have here. So I guess given that what 110 is about is the control of these criteria pollutants in order to maintain an act, the notion that there has to be case-by-case demonstration would very much bog down the enforcement. And I'm just not sure what in the act makes clear that that's what Congress has in mind. Well, I think what makes clear in this context is we're talking about SIP revisions. And so EPA is bound by the limits in Section 110K5. And that does require a finding that the SIP as a whole is substantially inadequate. And as Mr. Esrae suggested, substantial has to mean something. If it just meant is, you know, doesn't follow some EPA interpretation of some provision of the act, then Section – the SIP revision, the SIP call provision could be – should have been written very differently. Right. So I understand the position based on the SIP call authority. But just for the purposes of separating the arguments, if we assume that under the third type of rationale for SIP call authority, which has to do with compliance with the Clean Air Act, assuming that that doesn't require a factual showing or factual modeling that a particular state has or is likely to exceed the NAAQS, but allows pointing to a defect that – and I thought Mr. Esrae had a really interesting and helpful analogy in referring to, you know, that isn't harmless error for purposes of the requirements here. Why wouldn't it be sort of the poster child of something that we couldn't treat as harmless error if there were a gap in the very limitations on which the state and the EPA are counting for confidence that emissions are being controlled? Why wouldn't that be something non-technical, trivial, you know, kind of nitpicking, but actually something rather central? Okay. I think there are two aspects to the answer to that. One aspect is grounded in the SIP provision, what states are required to do, what states have discretion to do in their SIP. And it's clear that the states get to decide the particular mechanisms used to meet the Clean Air Act requirements. And so if a state and its package of requirements says, you know, there will be a limitation on this source and there won't be a limitation on another source, and under these circumstances, there won't be a limitation for this source, that's all permissible under Section 110. And I don't take EPA to disagree with you as far as you've just gone. Okay. So to say that because the package of provisions, of requirements, of measures that the state has adopted is substantially inadequate because, not because it's not protecting air quality, but because it doesn't include a provision written the way EPA believes it should be. That's not what EPA thinks it's doing. EPA is saying, we have steady state, and then we have SSM. And back in the 70s and 80s, when it was drinking from a fire hose, it was focusing on steady state controls. Those are in place. And now it's realized over the experience of decades that these SSM events can, in some cases, be belching a lot of criteria pollutants. They really are holes in the fence. The sheep are running out. And just conceptually, that's obvious. We don't need to model or go state by state. EPA is bogged down and way behind in lots of its challenges. So at least this one seems like an easy call. Let's close these loopholes. And it isn't saying how you have to do it. It isn't saying, except that they have to be enforceable. It's just saying we can't have periods when there's a blank check for emissions. There's nothing other than, I'm a good citizen, I'm going to try hard, in the SIP that EPA, the state, or citizens could turn to to say, no, actually, here, you haven't met the criteria. First of all, I have to take issue with that blank check characterization that I did before, because the SIP as a whole has various provisions that are enforceable that will provide protection. But let me give a great example, I think, of why just this theoretical inadequacy isn't really a hole in the fence that's going to allow sheep out, okay? The affirmative defense provision. EPA acknowledges, has acknowledged numerous places, that even if you don't have an affirmative defense provision, you're still going to have these emissions. So how is that a hole in the fence when a state decides not to impose a civil penalty when that particular mission occurs? When, if you didn't have that provision, in the SIP revision, in the SIP call authority, EPA has to identify to the states what they have to do to remedy the inadequacy. But in this case, taking the affirmative defense provision away doesn't prevent the emission from occurring. All it means is that my clients will be subject to citizen suits for events that they could not reasonably have prevented. And EPA's solution to that is, well, you know, make those arguments, those equitable arguments to the court. But it doesn't prevent the emissions. That's not a hole in the fence. Well, I think the assumption and the assumption in NRDC in the Section 112 context is that there are things that, you know, that the general characterization that these are unavoidable is perhaps not appreciating that with penalties, with incentives, there are ways to better maintain. There are ways to, you know, manage the timing of maintenance. There are ways to use cleaner fuel. You know, there are a number of different ways. And I think that's what EPA is looking for in, you know, something to govern these SSM periods. So let me respond to that in two ways. One, I think it's important to understand that the practical effect of what all this does. So if you take away an SSM provision, but you don't take away the fact that it's unreasonable to set a limit that can't be met with the identified technology, what that ultimately results in, and the states made this point in many of their comments, is the limit that applies during normal operations will have to be kind of a common denominator. And that, in fact, has occurred. When EPA got rid of its SSM exemption provision for hazardous air pollutants, some industries then sued to challenge their existing emission standards because they were now no longer consistent with what's achievable with available technology. So that's a really critical, you know, reality check, is that if you don't have these provisions, you're going to have higher emission limits. The other thing is these provisions... And so there are going to be times when you can have exceedances. And in setting the ordinary steady state standard, it has to take account of those. And so the steady state standard will have to be lower during ordinary operation in order to accommodate for the fact that there are going to be uncontrolled periods with higher emissions. So, I mean, you point to that as another way that theoretically these issues could be dealt with. And I think, you know, EPA here is saying, in keeping with 110, we're going to let the states take the first crack. We don't really know how they prefer to do this, but that there are ways that are more concrete and more enforceable than what's in place. But I don't want to be beating a dead horse, and I do see we've used up your time. Could I just respond? Okay, thank you. So I don't think it results in lower limitations at all because, again, we're in the backdrop here where EPA hasn't asserted and can't show that the status quo, in fact, is causing ambient air quality standards. So there wouldn't be a factual basis for lowering limits in a SIP. And that's different from hazardous air pollutant standards where Congress specifically in 1990 said, we've got to be more prescriptive. We've got to say exactly what the levels will be for these hazardous air pollutant standards. But the other thing is that the – and EPA has recognized this. They made this point when they were approving of affirmative defenses – is sources, if they know that they can potentially avoid a penalty, will attempt to reduce the impact of their emissions. They have an incentive. And, in fact, they have to have a plan. So there's a requirement to have an SSM plan that says you're going to do these things if you have an event. That doesn't exist if you don't have these SSM provisions to begin with. If the call is going to lead to higher emission limits, then why is the industry against the call? Well, because we think the existing system is working and has worked appropriately and is the right way to do it. So we're not advocating a higher, at least common denominator standard. We are advocating and have supported a standard where in these director discretion provisions, for example, it's the expert agency that looks at all the facts, looks at what the facility did in an event, looks at their plans and their reports, and determines that, yeah, indeed, that was something that was unavoidable. We think that's a more practical and more effective way to address the problem. And it's worked. It's not a problem. We have a solution in search of a problem in this rulemaking on a grand scale. Let me make sure my colleagues don't have additional questions for you, Mr. Frey. Thank you. We'll hear from EPA counsel now, Ms. Buckley. Good morning, Your Honors. May it please the court. My name is Sarah Buckley, appearing on behalf of EPA. I plan to address why EPA correctly concluded that the automatic exemptions, director's discretion and overbought enforcement discretion provisions and the affirmative defenses at issue here are inconsistent with Fundamental Clean Air Act requirements. My colleague, David Kaplan, will then address why EPA appropriately exercised its SIP call authority to address those legal deficiencies. In 2008 and in 2014, this court interpreted the relevant statutory language here and determined that when EPA establishes an emission limitation, these exemptions and affirmative defenses are inconsistent with Clean Air Act requirements. EPA is now taking the reasonable position that the same conclusion holds when states establish emission limitations in their SIPs. I'd like to take this in three parts. First, to address EPA's determination that exemptions and SIPs render emission limitations unlawfully discontinuous and unenforceable and that, as in Sierra Club, the general duty provisions that petitioners have cited here do not remedy that discontinuity. Then, I'd like to address the conditional exemptions like director's discretion and overbought enforcement discretion provisions allow state directors to unilaterally immunize excess pollution from federal enforcement. And finally, that affirmative defenses, as this court concluded in NRDC, unlawfully impinge on the jurisdiction of federal courts to determine liability and assess appropriate relief, no matter whether it is EPA establishing an emission limitation or states. So first, on exemptions, I think the court is fairly familiar that these provisions essentially say that excess emissions during certain times are not violations of an otherwise applicable emission limitation. A good example comes from the West Virginia Code that states that the visible emission standards set forth in Section 3 shall apply at all times, except in periods of startup shutdown or malfunction. So, can I start you off on this part of the argument just by trying to understand the context in one particular respect, which is, suppose that I thought that nothing in the act actually requires a SIP to contain, requires every SIP to contain an emission limitation. Just assume that for purposes of the question for now. And why does it matter if a particular emission limitation is discontinuous? Yes, Your Honor. Here, the important point is that these states have selected emission limitations as a tool for regulating here. And all of the provisions that EPA identified in this SIP call are emission limitations. EPA is simply saying that when you have selected an emission limitation as your strategy to attain the next or to meet other requirements of the act, that those emission limitations must be consistent with Section 302K and must be continuous. But why? I guess I don't understand why. Because unless you can say that the emission limitation under the language of the statute is necessary or appropriate to meet the applicable requirements, then why does it matter that there's an emission limitation that's discontinuous? I mean, for example, suppose that the state just thought in the SIP that, yeah, we have this thing, and you could call it an emission limitation. You could call it emission reduction. You could call it anything. We don't care what the terminology is. We just think that it makes sense to have this, and we don't think it needs to be continuous. We think it makes sense to have this including periods of discontinuity as is specified in our SIP, and you approved the SIP. And that's the grounds on which you approved it. But now you're saying, well, we have this thing that you think we called an emission limitation, and therefore it has to meet some definition of what an emission limitation is. But there's nothing further that EPA did to say, and that emission limitation actually is necessary to meet the applicable requirements of this chapter. It just said you have a thing you call an emission limitation, and it's discontinuous. Ergo, there's a problem without tethering it to whether it's, in fact, necessary to meet the applicable requirements of this chapter. Well, Your Honor, in some programs under the Clean Air Act, it may be the case that an emission limitation as the control strategy is required, for instance, to meet substance use standards under Part D, prevention of significant deterioration, et cetera. Also, if we were to take such a- It could be, and I'm sorry to cut you off, and I want to hear what the also is, but just to know, just so you know what I'm thinking. Yeah, I think it could be that an emission limitation is necessary. It very well could be. And EPA definitely could explain why a particular emission limitation in a SIP is necessary. I guess my question is, without that subsequent explanation, why is it enough just to say that an emission limitation is discontinuous? Because it leaves at least me wondering, why then have we established that there's a problem here? Because we don't know that this particular emission limitation is, in fact, necessary to meet the applicable requirements of this chapter. I think there may be a legal response and a practical response there. And the legal response first is that if we were to take such a capacious view of the term emission limitation, then we would effectively read out of the statute any meaning given to the other terms, such as a control measure. If emission limitation and control measure are the same thing, why did Congress use separate terms? And why did Congress place such emphasis on what an emission limitation must do? It must limit the quantity, rate, or concentration of pollution at a source on a continuous basis. The practical response is that oftentimes in determining whether a state's plan does meet the NAAQS or meet other substantive requirements of the Clean Air Act, the state and EPA rely on modeling. And that assumes that an emission limitation applies on a continuous basis. And because some of these exemption provisions, particularly director's discretion, we're not able to predict how often these events might happen and might excuse the emissions. And we don't know how much of the emissions might be excused when the provision is unbounded, as many of these are. I'm sorry, I didn't mean to talk over you, Chief. You go ahead, because you're on a line of questioning. Go ahead and finish it. One more, and then I'm sorry to stay on this. Do you acknowledge that a SIP could contain something that I'll call an emission limitation that's not necessary to meet the applicable requirements of this chapter? I think that a SIP could contain an emission limitation that a state has selected as its control strategy to address NAAQS attainment. But every emission limitation contained in a state SIP is there to address some requirement of the Clean Air Act. Where an emission limitation is not a part of the SIP, EPA actually recognized in this action and did not issue a SIP call for, for instance, sources that were not included in a state SIP, which is the state's representation of how it's going to attain the NAAQS and meet other requirements of the Clean Air Act. That was like North Carolina or something? It was a case with that, right? It was a South Carolina agency. I mean, I guess my understanding is that the states don't include limitations for fun. They include them because they're part of what's necessary and appropriate to meet the requirements of the Clean Air Act. Yes, Your Honor, that's correct. And when a state has selected that strategy as a part of, you know, encompassed in its discretion of how to attain the NAAQS, then that strategy must comply with the requirements of the Clean Air Act. Now I'm not sure I follow, because your answer to Judge Pillard seemed different than your answer to the Chief. The Chief said, could there be a limitation that's not necessary to attaining a NAAQS? And I wasn't sure if you said yes or no. And then to Judge Pillard, who said, I think, could there be a limitation that's not necessary, that's just there for fun, you said no. So is the answer no? Right, Your Honor, I'm sorry. I believe the answer is no, because the state has selected an emission limitation to meet some requirement of the act. That might not necessarily be one of the the stringency requirements like BACT or RACT, but it nevertheless has selected that strategy to meet, for instance, NAAQS attainment. If it does not need that emission limitation for NAAQS attainment, it hasn't submitted it in its SIP, and EPA has no qualms with whether or not that emission limitation has an exemption or not. Can we back up for a minute, Ms. Buckley? How do you respond to industry's claim that EPA just didn't initiate these SIP calls as a result of any pressing concerns regarding the adequacy of the affected SIPs to achieve the NAAQS, and the state says the SIP calls do not purport to improve air quality. They say in the SIP call, EPA did not set out to address threats to air quality. So is that right? And if so, what is the EPA doing here? Well, the EPA first line here is ensuring that state SIPs comply with the requirements of the act. Those requirements are directed at ensuring that air quality standards are met, that the health-based standards of the NAAQS are protecting health and are not jeopardized by, as you said, Judge Pillard, thousands of tons of pollution belching out of facilities without an emission limitation and without enforcement recourse. You know, the state director's discretionary provision here. You disagree that EPA is not setting out to address threats to air quality. In your view, they are. I guess maybe this is just about that, the point that Mr. Esrae was making, where the focus is not on demonstrating shortfalls in air quality in particular regions or states. It's looking at the air quality architecture that the act requires and seeing that it is in good repair. The legal architecture, the SIPs. That's correct as far as it goes, Judge Pillard, in that the legal architecture of the SIPs is important to achieving the substantive goals of the act. Now, industry says all the states here submit demonstrations, they've attained the NAAQS, and something you said about the assumption in modeling that steady state is the level of emissions. Is there not, when states are reporting their emissions, if you're seeking information from the states, does that not reflect also the uncontrolled emissions during SSM events? Not necessarily, Your Honor. The NAAQS are calculated through a complex group of data input that I honestly could not explain here today. But there's a couple of important practical points about these events. One, that these exemption events or malfunctions or the exercise of discretion, those may, in many cases, obviate record-keeping and monitoring requirements. So, the effect of the exemption, the effect of the exercise of a director's discretion would be to eliminate those requirements for the event that's at issue. And then, second, is that the nature of SSM is it renders some of the monitoring technology that is in place at these plants. It makes it not work. So, we are not getting data from these periods that could be put into that complex equation that comes out with NAAQS attainment. And how do you respond to Mr. Frye's basic point that these events are, by definition, unpreventable and unpredictable, and that the states do have provisions that they're going to do their best, keep them short, use best practices? What more does EPA want? I think I have three responses to that, Judge Pillard. First is that I think Mr. Frye is referring only to malfunction events, because startup and shutdown, although they are not steady-state operations, they are normal modes of operation for which you could write an emission limitation. EPA acknowledged in its rulemaking here and its preamble that states don't need to use the same numerical standards and have to be a continuous standard that is the same across all modes of operation. And that does not need to be a numerical standard either to the extent that that is not attainable during periods of startup and shutdown. So, for the M part of SSM, if it is the case that the events are truly unpredictable, cannot be prevented through the exercise of good practices, then this court has told EPA and NRDC that that can be addressed through the exercise of judicial discretion in weighing equitable factors in determining what appropriate relief might be. In addition, the court in U.S. Sugar acknowledged that for those malfunction periods, it is a reasonable approach to say that that can be dealt with through exercise of enforcement discretion. So, those are two ways of addressing this problem of malfunction events. I think there was a third part. They would be subject to enforcement permissions that they could not reasonably avoid. If the three parties that Congress has authorized to bring enforcement actions under the Clean Air Act, the state or EPA or citizen groups were to decide to bring an enforcement action, then they would be subject to enforcement. And they might have to pay penalties. It's a necessary evil of what you're doing and internalize the cost. Right? Internalize the cost or, you know, use even better operational practices. Make sure that you're... Right. But if they have processes that are not available, then just... Right. I'm sorry. I think I missed. I was just saying that... What Judge Miller was asking you. I was just saying that, you know, the penalties are in part, I gather, to create incentives for compliance to the extent that compliance is available, but potentially also to force the sources to internalize the cost because, you know, these are pollutants that are causing, you know, particulate matter and lead. And, you know, they're harming human life. And presumably the level of production should reflect the cost that it entails. Exactly, Your Honor. And I think there's maybe one more point on this, which is that EPA did not shut the door on the idea that a state could articulate an alternate emission limitation for a malfunction event, you know, that is sufficiently predictable and how it might play out so that the state could write a standard. Can you tell us a little more about that? Because there seems to be a certain area, and the state's brief spends a lot of time on this. There's a certain amount of sort of argument about characterization where EPA is saying these are loopholes, these are blank text, these are gaps in the fence. And the states and interveners are saying there's all kinds of, you know, different looking fencing across these gaps. It's, you know, all kinds of showings are required, but practices, minimizing the time, you know. So how do we assess that? I guess, you know, one question is, what's your position? But also, if there's deference, who do we defer to, the states or EPA? Well, in interpreting the requirements of the Clean Air Act. How the SIPs interact, you know, what they actually entail. Right. Well, EPA is exercising its expertise as the regulatory entity to determine whether these SIP provisions do, in fact, create a continuous emission limitation. And I think it's important to note with these general duty provisions that they are very varied, and so the problems with them vary as well. But if we're talking about a generic nebulous general duty that is located elsewhere in the Act, it is something that applies generically. And thus, it applies without regard to what the underlying emission limitation that is subject to the exemption is. So it does not purport to be a part of that emission limitation. I think it's important to note, too, that a general duty can coexist with an emission limitation, as it did in the emission limitation at issue in Sierra Club. But what EPA is saying here, consistent with what the court said in Sierra Club, is that there must be an emission limitation at all times. And thus, if the state believes that the kind of principles articulated in the general duty serve to fill the gap of that emission limitation, then the state can make that clear. So long as two other factors are present. So first, that it has to be enforceable. So in terms of things like generally use good judgment, you know, that's articulating a standard of care and not something that the source actually has to do and that, you know, a regulatory party, an enforcing party could say the source did not do that thing during this time. Therefore, they have violated their emission limitation. And then the second part is that these gap fillers, these alternate emission limitations, have to comply with any substantive stringency standard that might apply for that source and that pollutant. And these general duty provisions that apply generically, the state has not articulated that as a part of its, for instance, racked or backed standard for that source and that pollutant. And so, for those reasons, EPA found that the general duty provisions for each of these SIP called provisions, the general duty provisions that might apply in these SIPs, do not form an alternate emission limitation. And are there other states that have submitted, that aren't petitioners here that have submitted, for example, interpretive letters explaining these kinds of issues or does it have to be in the SIP? EPA's position in the rulemaking is that interpretive letters are acceptable, but they have to exist at the time that the SIP is presented to EPA for approval so that EPA and the public know that that is, that is supposed to be the meaning of the terms at the time. It can't, after the fact, come into legitimize a provision that might not otherwise meet Clean Air Act requirements. I took one of your general duty provision arguments to be that sometimes there will be, there will be specific stringency requirements and the general duty provisions would be insufficient in those instances. Did EPA discuss which general stringency, sorry, did EPA discuss which stringency requirements each general duty provision failed? No, Your Honor, because that would really depend on the source category that the emission limitation operates on and the pollutant being controlled by that specific emission limitation. And beyond that, the particular circumstances of each state, which criteria pollutants it is an attainment for or non-attainment for, how severe its attainment or non-attainment is, and then, you know, factors relating to new source review and modification and constructing new sources. That's helpful. Let me ask one clarification question about your answer to Judge Pillard's last questions. So let's say there's a most generic of all the general duty provisions and a citizen suit says that a polluter has violated that generic general duty provision. Can the polluter say to the court, well, Ms. Buckley told us that that's not an enforceable general duty provision? Well, no, Your Honor. EPA's assessment here is that the gap filling alternate emission limitation has to be legally and practically enforceable. And so in assessing whether on its face these provisions could be practically enforceable, EPA is saying, we don't think this makes the cut. If citizen groups want to go and make a case that it does make the cut and that they can articulate a standard for why a source has violated that, you know, this doesn't shut the door on that. But if they might, it sounds like now you're saying they might be enforceable, and if they might be enforceable and might not be enforceable, then the call starts to look kind of speculative, at least as applied to these specific provisions. I think EPA addressed that it thinks it is unlikely that these general duty provisions would be enforceable or would meet substantive stringency standards in order to serve as the gap filling measure. And I think that that conclusion was reasonable. And in addition, there are some, you know, perhaps clearer cut and dried unenforceability problems with some of the general duty provisions at issue here. For instance, that provisions that provide that you cannot violate the NAAQS are contrary to the concept that citizens can't sue a source directly for a violation of the NAAQS, which makes sense because a source's ability to attain the NAAQS or the source's role in attainment of the NAAQS is defined only by the control measures that the state applies to that source. And so in the absence of... But can a citizen sue the state for its SIPs inadequacy? A citizen can sue the state for not meeting a commitment in the SIP that was necessary to achieve the NAAQS. So if we had the same, you know, lack of continuity arguments and enforcement discretion arguments and affirmative defense arguments as here, if you were to lose this case and citizens were to turn around and say, look, these holes in the fence, these SSM, failure to apply enforceable standards to SSM is rendering the SIP inadequate, unlawful, in violation of the act, how would that work? It seems like given that the SIP applies to the state as a whole, it would be exceedingly difficult to identify SSM events as the reason for exceedances. It seems like there's incredibly complicated causal problems there. Can you walk me through how one might do something more concrete to address this problem or a set of things that would be more concrete, whether on the part of EPA or on the part of citizens, if you were to lose this case? Oh, I thought I was following. Let me see if this answers your question, Your Honor. EPA discussed the three factors that it used to assess whether general duty provisions adequately fill the gap. It says that J.A. 194 to 95 and 261 to 62. And therein it stated that the gap filler must be a clearly stated component of the emission limitation. It must meet applicable levels of control standards and must be legally and practically enforceable. EPA then also set out seven factors that it prospectively suggested that states should consider when developing an alternate emission limitation to plug that gap. What I'm saying is if putting aside this suit, if in fact the NACs were exceeded in a relevant area of the state and citizens who live there wanted to sue and say, look, this is not working. And if they identified, let's say correctly by hypothesis, that SSM events were a big contributor, how could that be proved in a citizen suit, given that there are myriad sources with myriad different periods of functioning? I'm just trying to imagine as a plaintiff, how would one even begin to scale that hill? I agree, Your Honor, that it is a difficult to perhaps impossible problem. And I think somewhat to this court's cases and EPA's guidance about transport and how states can show that particular sources in other states contribute to violations of their NACs and the complexity there. And the reliance on a number of factors. But I think that the important point here is that these difficulties show that the general duty provisions don't do what the Clean Air Act says that they need to do to constitute emission limitations. They are not enforceable and they do not provide a continuous reduction, continuous limitation on the quantity, rate, or concentration of pollutants. I see that my time has expired. Can I just ask you one question? So on the continuity point, if we, by hypothesis, if we agree with you on the continuity point, then that's alone enough on automatic exemptions, because that's the whole rationale for the SIP calls on the automatic exemptions, right? And then on the director's discretion exemptions, that's one component? Yes, Your Honor, the lack of continuity is sufficient for you to find in our favor on the director's discretion provisions as well, although EPA has additional arguments. One, that there's a particular enforceability problem here where a state can unilaterally immunize excess emissions from EPA or citizen suit enforcement. That's contrary to the multifaceted enforcement scheme that Congress set out. And second, that these director's discretion provisions that are unbounded, that essentially say your emission limitation is X unless we say it's not X, are essentially a revision of the SIP that has gone through a public process that's been approved by EPA. Is that because you're assuming that the director's discretion removes the counting of those emissions from what's reported as emitted under the NAAQS? Or something broader? I think broader, Your Honor, although that is a problem. As I noted, the monitoring and recordkeeping requirements might be waived during that time. But also, something has gone through the public process set out in the Clean Air Act. And now... Although the public process includes the exemptions. That's what you're looking at. Right. Well, I guess the fact that EPA approved the thing does not mean the thing is necessarily consistent with Clean Air Act requirements. And that's what EPA has found here. Because these are so unbounded that EPA is not able to properly assess how frequently the provision would be invoked, how much pollution the provision could potentially excuse. How practically what the worst case scenario would be under the exercise of that discretion. But that is so substantial that it could operate as a revision of the SIP. I guess what I don't understand about that is it seems like the argument would be that the possibility of an exemption, as I understand EPA's logic, the possibility of the exemption, if it's a broad ranging exemption, could enable the state to act in contravention of what's in the SIP. And it's just, I don't conceptually understand that when the possibility of exemption is in the SIP. Right. Well, I think the SIP is based on many assumptions about the emissions that could result. And without bounds on the exercise of the discretion, there's a measurability problem to determine how much emissions could result. So I thought it was actually helpful. It sounded like the chief was starting to ask about the, so there's the continuity, there's the continuity argument. There's the explicated statutory factors to determine remedies, arguments. And then there's the statute setting out a process and procedure for amending a SIP. And the continuity argument applies to all of the called provisions. Is that right? Including the affirmative defenses? It applies to the affirmative defenses, you know, insofar as it addresses an argument that the affirmative defenses are, in fact, an element of the emission limitation. So that then would be, you know, within state discretion. But our response to that is, well, if that's the case, then that causes a continuity problem because, again, it immunizes excess emissions to which that defense applies from enforcement. Okay, but if it's not an element of the limit itself, then the enforcement, the affirmative defense is invalid. As a function of Section 7413's grant of jurisdiction to the courts to determine. Right. 7413E, is that right? 7413B is the grant of power, and then the factors are later in that section. Yes. Right. And that's the rationale of NRDC. So I'm just trying to correlate, like, if we were to agree with you on continuity, we also need to reach the 7413B argument. Would we also need to reach the process for amending SIP's argument? I don't think it's necessary to reach that argument. And if we disagree with that argument, how does that affect, does that curtail your ability to prevail? No, Your Honor, because sort of regardless of whether it's conceived of as a revision of the SIP or not, it is still creating a discontinuity. It is allowing excess emissions over the otherwise applicable emission limitation and is preventing the enforcement of that provision by EPA or citizen suit. So the discontinuity is evident in that break in enforceability, too. So I guess discontinuity, if we were to agree with you, is enough for the automatic exemption. There's more because it's the only thing, as I think it's essentially the only thing that underlies the calls based on automatic exemptions alone. For director's discretion, there's a few arguments that underlie the call, one of which is discontinuity, and it seems like the point you're making is that one's independently sufficient. And then for enforcement discretion, though, and I'm not sure how much that applies independently of director's discretion, maybe only with one state or maybe there's more than that, but I didn't understand discontinuity to be part of that. Right, Your Honor, I think the problem is better characterized as enforceability in that case. It's similar to the director's discretion provision in that it purports to say that the emissions are excused from EPA or citizen suit enforcement. Federal or citizen, right. So that's a separate argument from discontinuity. That's what sustains what one would call enforcement discretion independent of director's discretion. And then the last category is affirmative defense. On affirmative defense, you've got your kind of just NRDC argument. Exactly, Your Honor. If that's enough for affirmative defense, then that is enough to sustain the calls on affirmative defense. Yes, Your Honor. And on affirmative defenses, you know, the important point is that there's nothing in 110 or in 7413 that suggests that the fact that the state has created an affirmative defense should lead this court to treat it any differently than they treated the defense in NRDC. Yeah, I was just trying to understand the overall architecture and how much the discontinuity feature affects the various calls. And how, because at some point you got to add up to 30, how many ever calls it was? 35, 34? I don't remember. 36 states originally. There are locality ones as well. I think it might have come out to about 45 originally. The accounting of that is in the 28th day letter that we filed yesterday. Okay, I have a sort of an obscure fact. And then I have a question related to what our court's remedy might look like if we go down a certain path. And then I have a third question that's more theoretical. And so I'll start with the fact one. Do any called states have at least two continuous emission limitations? If you don't know, that's okay. I don't know specifically, but I would note that the SIP call was directed at specific emission limitations. So I guess I would have to assume that the state has more than the, you know, between one and I think no more than 10 provisions that were called for each state. If that makes sense. I think so. The remedy type question is, let's say that we agree with you about the affirmative defense related to monetary penalty provisions. We think that they're illegal. But let's say we think the rest are legal. What should we write? What should our order look like? I suppose that would be an order denying the petition with respect to the, I believe, 11 states that had unlawful affirmative defense provisions and then granting the petition with respect to the other SIP calls. I guess what I'm trying to get at here is that the remedy would be SIP call specific. And the accounting of the state by state accounting is available. I did write it down in the proposal at JA-33 and in the final at JA-241. In the numbers that you're using, sorry to interject, Judge Walker, in the numbers that you're using, do we award the remedy to states that never petitioned? So my notes tell me that there were only seven jurisdictions that challenged affirmative defenses. I think that's correct, Your Honor. I thought you said 11. Seven petitions, but yes, 11 states. Some of those states, okay, this might start to get complicated. Some of those states have, as we stated in our 28-day letter, now submitted SIP provisions that have been accepted by EPA through a final federal register notice. So that knocks out, I think, at least three on affirmative defenses specifically. But just generally, we would only be referring to remedy with respect to the states that petitioned? I might throw this question to Mr. Fry as far as what remedy industry petitioners have asked for and where they have shown that they have standing to ask that remedy. Well, he's probably going to want as big as possible, so why shouldn't he get that? Well, I think he shouldn't get that as a matter of the merit. Because as a matter of the architecture of the issues, you're saying that the fact that the states aren't in here doesn't mean that the SIP calls aren't being challenged by industry. I take that as an important point. Yes. So you had, I'm sorry, back to Judge Walker's questions, one on FAC, one on remedy, and? I had a theoretical one. I think you would say that at least some of the called SIPs are not unambiguously inadequate. You're using your Chevron step to general agency discretion to, you're exercising agency discretion. Am I right so far? I think you're right, Judge Walker, insofar as EPA's application of its interpretation, which it believes is the best interpretation of the Act, application to particular SIP calls is entitled to deference. My question is, how can something be substantially inadequate when its inadequacy is ambiguous? And I hate to demur on that, but I believe that my colleague, Mr. Kaplan, is prepared to answer that question. Can I ask you one final question along the lines of trying to figure out which arguments affect which SIPs and what we'd end up doing? Just take the converse and suppose, I'm just trying to isolate what affects what. Then what? Then that would certainly resolve in our favor the director's discretion and overbroad enforcement discretion provisions. But I believe that there's an enforceability aspect to the exemption issue as well, in that if there is no applicable emission limitation at that time, then excess emissions or a provision that operates to state that excess emissions are not violations of the application. If there is an applicable emission limitation, then that means that there's nothing in place at that time that is enforceable. And under Section 110A2A, it does state that there must be an enforceable emission limitation. But I guess I didn't see in the rule anywhere or in your brief, as far as I recall, that the automatic, you're going to automatic exemption there, that that calls were grounded in anything other than discontinuity. But maybe I might have just missed it. Off the top of my head, your honor, I can't point you to the pages of our brief, but I think enforceability and the fact that it must be an enforceable emission limitation is a part of the package of our argument. It's just necessarily there. Yes. Can I ask you, this is really going back to the continuity argument textually. Under 110A2A, it talks about enforceable limitations and other control measures, means or techniques. And then in the definition, it defines emission limitations and emission standards to mean requirements, including requirements relating to operation or maintenance that assure continuous reduction and any design, equipment, work practice or operational standard. And I'm just wondering about the relationship, as you understand it, between design, equipment, work practice or operational standards and the other control measures referred to in 110A2A. And I don't want to go on too long, but just to highlight a little bit my question. And so the question whether continuity is really always required of the types of measures that are in there to limit emissions seemed to me at first quite ambiguous because 110A2A refers to emissions limitations, which are defined to include the continuity requirement, but it also refers to other control measures, which aren't defined to include the continuity requirements. I thought, oh, well, the other control measures don't have to be continuous. But then when you look at how that term is used, other control measures, means or techniques, parentheses, incentives, fees, marketable permits, auction of emission rights. It actually sounds like the other control measures are in a way overlays or ways of meeting the limitations, whereas in the definition section, you have both requirements relating to operation or maintenance, you know, a filter, let's say, but you also have these more qualitative requirements like design, equipment, work practice or operational standards. So that makes me think that everything actually is an emissions limitation under 110A2A and that the other control measures are things that may be ways that a SIP provides to ease or provide flexibility on how to meet that. Like you can't have a trading system unless you know what it is you have to meet that you're trading over. I know that was long-winded, but I think it's really possibly at the heart of the challenge to the continuity requirement as it applies to 110 as distinct from 112. Yes, Your Honor, and EPA tried to grapple with this question at JA 178-79 and 184. I believe 184 describes a portion of the CFR that discusses types, a non-exclusive list of types of control measures, so giving more content to that term. But the other important point is that an emission limitation may encompass a work practice, you know, something that isn't a numerical standard, but it must limit the quantity rate or concentration of that pollution on a continuous basis. And you can, those things can coexist. As you said, the emission limitation doing some work in the SIP and a control measure doing other work. For instance, you know, a state can have some sort of transportation management plan that would be a control measure. That helps it attain the net. That's where I'm confused because you're referring to control measures as encompassing what in the 302k definition is actually encompassed within emissions limitations. Design, equipment, work practice, or operational standards. And it seems to me just textually that you have emissions limitations in 110a-2a that includes numerical operation or maintenance standards like scrubbers or filters, and an emissions limitation is also a qualitative, encompasses a qualitative work practice design, like, you know, a clean fuel requirement, or, you know, limiting duration of a shutdown requirement or what have you. And so under that reading, emission limitations encompassing both types or all types, the continuity requirement applies, whereas if you're reading control measures to be undefined and potentially including the kinds of measures one might use in an SSM, it seems like you've textually given away your continuity argument. And so as I was reading just the text of it, it seems like control measures here, control measures means or techniques of meeting the emissions limitations would be marketable permits, auctions. You know, it's a different, we're on a different plane when we're talking about these other control measures. Right, your honor, and I think the response to that is that if states were to bring a plan to EPA that contained just control measures and not emission limitations, and said that this meets the requirements of the Clean Air Act, then EPA could assess that plan. But here the states have picked emission limitation as one element of their plan, and an emission limitation has to be continued. An emission limitation includes a design, equipment, work practice, or operational standard. That operates to limit the quantity, rate, or concentration of an emission, you know, on a continuous basis at a source. In other words, EPA presumes, as I think 7410HUA presumes, that every control measure is not an emission limitation. That there's a distinction between emission limitations and control measures. Yes, your honor. And then the definition in 7602K is a definition of emission limitations, not of control measures. Even though, as Judge Filler points out, that definition includes the terms design, equipment, work practice, or operational standards, which kind of sounds in, it could be seen to sound in control measures. But the framework of these provisions presuppose that there's emission limitations on one hand, and control measures on the other hand. And you could have a state that submits a SIP that doesn't include an emission limitation at all. Yes, your honor, to the extent that that SIP would then meet requirements of the Clean Air Act, yes. Yeah, but at least conceptually that could be approved. EPA could approve that. And if that SIP were approved, then a continuity argument would kind of not make any sense. Because continuity only relates to emission limitations, but I'm just hypothesizing a SIP that doesn't have an emission limitation to begin with. That's right. So what is a control measure, other than the ones listed in the parenthetical, what is a control measure that would not be an emission limitation? I'm thinking like a use of clean fuel during, alternative clean fuel during startup or shutdown. Is that an emission limitation or is that a control measure? I think EPA would conceive of that as an emission limitation if it has made a work practice component of that emission limitation. Right, because the clean fuel would be less emitting. What about a duration requirement? If you're doing startup, you have to complete it within two hours. Yes, your honor, I believe that could also be a part of the package of work practices that... Emission limitation. Yes, your honor. Please go ahead. I jotted down a few instances that were discussed in the CFR site that again is available at JA 184 discussing. I think the term is control strategies there, but you might think of emissions charges or taxes as a way of incentivizing different emissions or different pollution practices. Changes in schedules or methods of operation or transportation control measures. What do you mean by transportation control measures? I believe there's a case that I reviewed. I want to say it was city of Arvin that was really directed at a transportation control measure. I suppose the state could make commitments to incentivize public transportation over use of individual vehicles. Okay, let me make sure my colleagues don't have additional questions for you. Ms. Buckley, thank you. Mr. Kaplan, we'll hear from you now. My co-counsel has explained why the four types of SSM provisions do not comply with the applicable requirements of the statute here at issue. I will discuss why EPA has the authority to issue the SIP call in these four instances. I think the first point I need to make is that EPA here is operating under Section 110 K-5, but that the word substantially inadequate needs to be understood in the context in which EPA used it in each of these in the SIP call action here. That is specifically with regard to the third prong. That third prong, your honor, is noncompliance or compliance with requirements of the act. It's in this particular context that you need to understand and construe substantial inadequacy. You need to go one step further. You also need to construe all the terms together in context with what are the underlying requirements that are issued. We've just had a lengthy discussion about what those elements are of continuity, scope of the court's jurisdiction to resolve violations. I don't repeat them, but those are the particular underlying requirements against which the SSM provisions are being judged. Mr. Cowan, by jurisdiction, we don't really mean like subject matter jurisdiction in the sense that would not be subject to Chevron deference. We mean like authority. I think that's right, your honor. It's hard. I look at the NRDC decision, and I think that the word jurisdiction could be used in different ways. There is an argument that 113G is a vest of courts with exclusive, I'll use the word jurisdiction, it could be authority, power to make those exclusive judgments and that they can't be intruded upon. The key point, though, is that when you look at 110K5, and you need to look at it in the context here. I want to speak specifically to Mr. Esrae's argument that fact-finding was required as an element of the substantial inadequacy. Again, the reason his arguments don't fly is that they look at the words find or substantial inadequacy out of context. You need to look at them in the context of what's done here, and that is looking at the third prong, not the first or second prong, the third prong, and then you need to look at the underlying statutory requirements. When you look at those underlying statutory requirements, fact-finding of the sort the state argues that EPA should have undertaken is certainly just not encompassed within those elements. There's no basis by which the fact-finding that state urges is an element of making the decision as to whether or not there is legal deficiency with regard to these types of legal requirements. Mr. Kaplan, Mr. Fry had basically accused EPA of refusing to look at SIPs as a whole in making that inadequacy determination, and I wonder if you have any more response to that than you've given your brief. Yes, Your Honor. First off, I understand what they mean by looking at the SIP as a whole. I've identified three possible things that could have been by that. The first is that they're saying that EPA is looking at the SSM provisions only in isolation and didn't look at other provisions that might work with them. And there's nothing to that argument because EPA went through notice and comment, and states and other parties had an opportunity to say, no, this works in concert with this. And so EPA did look at provisions as they apply in relation to the other provisions in the SIP. The other argument, as I take it, could be that they believe that all elements of the SIP need to be noncompliant before a SIP call can be issued. I think that's nonsensical. That's their interpretation of as a whole because SIP calls are issued all the time where there is noncompliance with particular provisions. And not every element of the SIP needs to be out of compliance with the particular provisions. And almost every SIP case involves particular provisions. And that's the magnesium case, any number of cases, the Michigan case in this court. The third argument, and maybe this is the one you're referring to, Your Honor, is I think their claim that if an area is attaining a max, then you can't issue a SIP call for these kinds of legal deficiencies. And that's clearly wrong as well because attainment of max, although important, is not the only requirement in the statute. There are numerous other statutory requirements that need to be met. But for purposes of 110A2A, aren't they all directed at attaining the max, not super attaining? And I thought that your argument was somewhat different, that it's extremely slow and onerous and not required that a violation of the max be shown. But that one of the problems with the exemptions and discretion and so on is that in part, as Ms. Buckley said, they affect the reporting. So they affect what even is considered to be the performance of the sources within the state and that the failure to comply with fundamental requirements is fairly assumed to threaten the max. That's not your position? No, that is part of our position, Your Honor. And I guess I think I'm understanding you to have asked a different question now. The issue of the difficulty to discerning what the actual impact is in terms of whether an exceedance resulted in air quality from a particular source and that relates to a particular SSM provision, that's real. That's discussed at the preamble. And that is real. And Ms. Buckley went into that. But I think that there are numerous independent requirements in the statute that apply and must be complied, continue to be complied with, regardless of whether an area does attain the max. There are increment requirements. There are visibility requirements. There are stringency requirements. There's any number, and the list is quite long. And so beyond that, once you attain, there's a whole slew of requirements you have to follow to maintain the max. So it's not just attaining and then you get carte blanche. Not at all. The tips also have to show maintaining the max. And there's a whole slew of requirements in the statute set out to ensure that that occurs. And the PFD, the new source review, there's a number of things that apply in those regards. And so you cannot immunize noncompliant provisions, provisions in the SIP, these SSM provisions, that don't comply with fundamental Clean Air Act requirements just because there's been attainment, because there's all these other requirements that I've mentioned. And so it's really a non sequitur whether or not there's been attainment or not. Another point, Your Honor, is that Mr. Esrae mentioned that it's a question of a harmless error, that there needs to be facts because otherwise we're dealing with issues, potentially dealing with circumstances of a harmless error. And that's not the case at all, Your Honor. Here, it's pretty clear that each of the requirements that we've been discussing of the requirements of the act are fundamental. Whether or not you have a continuous requirement, whether or not you are intruding on judicial authority, things of that sort. And this, the court's case law defines that. And so you can discern from the text of the statute, as well as from its architecture, that these are not trivial matters. What's an example of a statutory misinterpretation that would be trivial? An instance of noncompliance? Well, when you say misinterpretation... Well, I thought you were saying in your briefing now that EPA can make a call when there's a fundamental legal error. And I guess I'm asking for an example of a legal error that's not fundamental. Your Honor, I understand your question, and I'll get to that first. But to be clear here, there was no allegation that the types of legal requirements here aren't fundamental, that they're just trivial matters. But to your question, I think that's something that EPA would have to look at and evaluate in context of the SIP provision itself. And so there may be some aspects of the statute that are in the statute but wholly discretionary. Those might be candidates, but I would not want to presume to make that judgment for the first instance here. It would be for EPA to... I assume there would be a myriad, you know, filing, format, timing, signature. I don't know. I mean, there must be so many things that just wouldn't be about the guts of the statute. It doesn't seem like an unreasonable question. No, for sure. I don't think it's an unreasonable question. I just can't identify here any one particular statutory provision. But there could be... Let me rephrase it then. Instead of asking for a specific example, let's imagine that we write a sentence in this opinion that says, when it comes to the difference between a fundamental legal error and a non-fundamental legal error, the EPA has given us no limiting principle. I'm sure you don't want us to write that sentence. So why should we not write that sentence? So the limiting principle here, Your Honor, is that the requirement in the statute that's issued needs to be fundamental to the integrity of the structure and the SIP process. And here, each of those statutory requirements that issue continuity, not intruding on the exclusive domain of courts, ensuring that EPA's enforcement discretion and citizens' group enforcement discretion is not being precluded by discretionary actions of the state, those all go to the fundamental structure of the SIP. But there may well be elements that are not fundamental to that. Possibly, for example, I believe there's a provision where states can voluntarily bump up themselves on their own at their own discretion to do so. But it's a provision I believe in the statute. That may be a candidate for one in which it's left to the states. It's not necessary. It's not fundamental. And so it falls on the other side of the coin. If it's not necessary, because I think the question here, I think, is are there legal inadequacies that are not substantial inadequacies? I think that's ultimately, if I'm understanding correctly, where we're going with this. And the part of the statute that this interrelates with is the call on the basis that the administrator finds that the SIP is substantially inadequate to otherwise comply with any requirement of this chapter. And so if it's a provision that isn't necessary anyways, then does it jive with the language that says to otherwise comply with a requirement of this chapter? Because I think it seems to me what would be relevant is something that is a requirement of the chapter, but the violation of which still is not a substantial inadequacy. That's at least what I'm interested in, and I may not be capturing exactly what my colleagues are. I think that I understand your point, Your Honor, and so all I'll say is that here, if there is a violation of a fundamental requirement, the inadequacy is substantial. And I think that you can draw from the statute text itself and its architecture in this context. Which particular requirements would fall on the other side of the line would be something that EPA would have to evaluate in the context of the SIP and the SIP provision at issue and how it interacts with that requirement? I still haven't really heard you answering Judge Walker's question. What have you told us that prevents us from, that doesn't leave us saying when it comes to fundamental legal error versus non-fundamental, EPA has given us no limiting principle. I guess your answer is this is so clearly fundamental that we don't need to explore the outer bound. Well, I believe that's correct, Your Honor, because here there was no claim that the underlying requirements at issue here are not fundamental. Well, there is an argument that you've invented a standard that is limitless. There is that claim in the briefing. Well, I think that they certainly challenged the standard by saying it's not a permissible construction. And I think we've resolved that by pointing out that you need to look at each of the terms in the text in connection with the third prong and the underlying requirements at issue. Where I'm having trouble answering is what is the example of a requirement that would not be a fundamental legal requirement? And the reason I'm having difficulty is that that issue didn't come up in this lawsuit. Nobody posited such an example. EPA was not dealing with such requirements in this case. And so EPA did not explore that other side of the coin of its analysis. It was not presented in the context of the rulemaking by any party. And so that's why I cannot identify a particular clear example that falls on the other side of the line. And I think it's something that the agency has to look at in context with the SIPP provision that essentially or does not comply with that kind of requirement that Judge Walker is asking about. Can I ask you to address the reliance of the state petitioners on, at least that argument, put aside whether it was in the briefs, but that argument on K-6? Your Honor, we've got a situation here where the law has changed significantly, where legal interpretations have changed over time. And even if you were to assume that EPAs could use K-6 in all the different circumstances, in the circumstances we have here, you're assuming that that's true, that in no way would preclude EPA's authority under K-5 to issue a SIPP call to remedy these types of defects that are here where there are legal requirements. And that is the way the statutory structure, the federal-state cooperative structure of 110 is challenged. I think the argument being made is that if we read K-6 to cover this, then there's no reason to read K-5 to cover it. It's a tool for understanding the scope of K-5. You may have a response that K-5 also covers it anyway. But I guess my question is, do you have a response that says that K-6 doesn't cover this? Well, I don't have here a response here that it doesn't cover it. I don't have really an answer to that question. So I'm going to assume that it does cover it for purposes of this argument. Although looking at the text, it seems to talk about, as if we're talking about errors that may have been made at the time, as opposed to errors that arose. They became obvious afterwards when the law changed, interpretations changed. New decisions come down from this court. We've got the Sierra Club decision. We've got the NRDC decision. We don't change the law, Mr. Kaplan. No, I take your point. I'm not trying to make a semantic point. But if that's what you're saying depends on, then I'm not sure I'm going to be able to go along with what you're saying. No, I agree. I take your point. But the agency does have to respond and accommodate interpretations that this court provides in its decisions. What was the remedy under K6? Because I had focused on this provision before, but it speaks in terms of the administrator may, in the same manner as the approval, disapproval, or promulgation, revise such action as appropriate without requiring any further submission from the state. What does it mean to revise such action? It looks as though it's approval. It looks like it's exactly revising its approval. And the provisions are substantially different because of that point. Here, EPA is invoking the main apparatus within the context of how 110 operates. And it's invoking 110K5 because that comports with the federal state process that Congress enacted. And the difference in remedy is significant because it gives the states an opportunity. It gives the states an opportunity in the first instance to elect how to best respond and fix the provision. Now, I understand. States might say, no dice. We're not interested. We're not going to fix the provision. And the statute also has the remedy for that situation. In which case, if the state selects not to fix the provision, then the statute requires that EPA fix the provision itself using a FIP, the Federal Implementation Plan. And so that is the dichotomy and the difference between these provisions. And regardless of whether K6 could or couldn't have applied here, it in no way eliminates the different kind of remedy, the different process that EPA went through here. And can I ask one more question? And then I want to make sure my colleagues have a chance to finish up any additional questions they have for you. And if this is a question that should have gone to Ms. Buckley, we can exercise our prerogative to ask her on mute. But my question is this. On the discontinuity portion of EPA's argument under 7410A2A, the point I think at the end of the day is that discontinuity means that the emission limitation does not ultimately constitute an emission limitation because of the discontinuity. And what I'm asking is, at the end of 7410A2A, there's the language as maybe necessary or appropriate to meet the applicable requirements of this chapter. So presumably there's some requirement of this chapter that's not being met. And I'm wondering, what's the requirement of the chapter that's not being met? Is it the definition itself that's not being met? Or is there some other requirement of the chapter that's not being met? I think that is a question, actually, that's best left for Ms. Buckley. Okay, sure. Then let me make sure that... Can you answer the question that I asked Ms. Buckley and that she told me to ask you? I can repeat it. How can a provision of a SIP be substantially inadequate, not just inadequate, but substantially inadequate, when its inadequacy is ambiguous? Oh, yes. You're right. I understand that question. And let me respond. This is a circumstance that two courts have looked at, the U.S. Magnesium case in the Tenth Circuit and the Georgia Power case in the Eleventh Circuit. And both have opined that the way to remedy ambiguity, where ambiguity could lead to an inconsistent construction of the statute, is to issue a SIP call. So those courts have already said that a SIP call is appropriate in this instance. And the way this works is that where the ambiguity exists, the focus has to be on the text of the SIP provision. And the question is, is that SIP provision clear so that it can be applied by third parties? Can it be applied once we leave here, or once EPA approves a SIP, can it be applied by industry so it understands what its obligations are, by citizens so they can understand what industry needs to do, and can it be enforced? And can courts review that? And so the key is to have clear provisions that are not ambiguous where you can have an interpretation that's contrary to the statute. Because when you have an interpretation that's contrary to the statute, the SIP is not going to do its job. The SIP will not be able to do its job. And so the Georgia Power illustrates this situation where a district court judge looked at an SSM provision and thought that it did not preclude enforcement by a citizens group. The Court of Appeals reversed and said, it looks to us like this text does preclude enforcement, and so I'm going to have to preclude the lawsuit. But the court said, and the remedy that's appropriate in this circumstance, and I'll quote, if the EPA believes that its current interpretation of the Clean Air Act requires Georgia to modify its SSM rule, the EPA should require the state to revise its SIP. And that's at 443F3 at 1355. And the court cites the SIP call provision 110K5. And so that is the appropriate remedy in this situation where you've got the kinds of legal deficiencies that we're talking about. And in this case, it's a small matter for – if a state's interpretation is that – if a state's interpretation is that we never intended that offending provision, it's a small matter for a state to then submit a SIP revision to clarify that so that the words of the SIP themselves are clear. Okay, let me make sure my colleagues don't have additional questions for you, Mr. Kaplan. Not for me. Thank you. Then if – I beg my colleagues' indulgence, but if I could just ask Ms. Buckley the question I should have asked her before, and we can leave the timer at zero because we're in overtime with you. Thank you. But, Ms. Buckley, so did you understand the question that I asked about the final part of Senate Report 1082A as it relates to the discontinuity argument? What is the applicable requirement of the chapter that's not being met? That's – I'm sorry. That the applicable requirement under the chapter as referred to in Section 7602K, I think that there are two ways of giving that content. One is that, you know, emissions limitations may have to comply with substantive stringency standards. And, you know, note that emission limitation is a cross-cutting term, so that could apply in the 112 context to, you know, the MAC standard and the 110 context to substantive standards like we've discussed under PSD, et cetera. But that the continuity standard is also a requirement under the chapter. And the legislative history of this provision makes clear that Congress chose to put this continuity requirement here to ensure that it was applied across the board to Sections 110, 111, 112, and elsewhere. Yes, the question is, is it sort of circular or bootstrapping to, you know, if you have 112 and you have a requirement to come up with NESHAPs for hazardous air pollutants, and what is the non-302K provision other than in the subset of SIPP situations? Not where you have a RACT or a BACT or a prevention of significant deterioration, but just in the ordinary attaining jurisdiction where a SIPP has been submitted in an effort to create a mechanism for maintaining the NACs, what is the requirement outside of 302K that is referenced by that last clause? I think that it could be read as circular in that it is referencing back to an earlier part of the definition. But I believe the court in—this court in Sierra Club articulated that that continuity requirement had independent force outside of 112. The court's statement was that there must be, you know, a 112-compliant standard in place at all times. And the at all times comes from continuity, not from anywhere in 112. But the continuity is a tool to meet requirements of the chapter, and I guess I was understanding it maybe sort of less precisely than Chief Judge Srinivasan as the requirement to have an effective SIPP in place. But that's, I think, a little fuzzy, and I think he's asking you for something more, maybe more precise. Right. Well, the continuity requirement also speaks to the ability to maintain, attain the NACs, to enforce compliance with the NACs. And, again, the legislative history of this section illustrates Congress's concern with ensuring that a continuous limitation be in place in order to achieve those substantive goals. And those substantive goals are, outside of the stringency standards we've discussed, they are also about direct compliance with the NACs. But the thing is that I don't think EPA's rationale and the rule was that the failure with respect to discontinuity is a failure of meeting the NACs. It's that the failure with respect to discontinuity is the failure to meet the definition of an emission limitation. Yes, Your Honor, in that the requirement of continuity is a legal requirement inherent to that type of control as illustrated by the provision. Then it does seem circular to me in the way that – that's not to say that circularity is necessarily – I mean, usually we presuppose a circulator is a big problem. I'm not sure that it is, but I'm just trying to understand whether there is something that's non-circular about it because it sounds like the part of the chapter that's being violated is the definition itself. Right. Why isn't it enforceable emission limitation? I mean, I thought the whole case was really about enforceability and that you're not going to be able to keep the sheep in the pen if there are gaps. If the requirement is not the 302K requirement, but the 110A2A requirement of enforceable emissions limitations, that an emissions limitation is only enforceable if it is where a state and EPA have deemed some limitation scheme to be necessary. Obviously, it doesn't have to have one, but where it's deemed it to be necessary and submitted a SIP on that premise that it's the enforceability and then maybe all the corresponding enforcement provisions like the remedial and others that are – what's necessary. Right, and I think that's why I tried to say earlier that there is an enforceability aspect to EPA's argument – EPA's continuity argument here. EPA's argument that the automatic exemptions are inconsistent with the Clean Air Act. That's the operation of 302K and Section 110A2A read together. And when there is a discontinuity, you have lost the ability to enforce the emission limitation that would – that the state's plan says otherwise applies. Can I take a – Go ahead, Mr. General, please. Can I take a shot at a sheep metaphor? Imagine that fence is defined as a fence with no holes. And the state submits a plan to keep the sheep in. And the plan includes a fence with holes, and it also includes a sheepdog. And here we are 50 years later. There's no evidence the sheep's ever got out. And along comes the EPA, and they say, well, you said you had a fence, and your fence has holes, and that violates the definition of the fence, so we have a problem here. Why is that not this case? Well, Your Honor, I think that we have to look at the requirement as articulated about that fence and read the lack of holes as a requirement for having a fence. Or maybe the sheepdog is another means the fence doesn't really have a hole effectively. The sheepdog is the design, equipment, work practice, or operational standard. And if the sheepdog is not there continuously, then there's a hole. If the sheepdog is there continuously, then we're good. Right? So the fence is a quantity rate or emission or a filter. The fence is a filter, and the sheepdog is a clean fuel requirement. And if you have the filter all the time and the clean fuel requirement only half the time, then no good. I think that your skill in manipulating metaphors is much better than mine on the fly. Let me, let me just, I don't, I'm not sure I caught all the nuances of the metaphor, but on what... Sometimes talking about it directly rather than by analogy is more illuminating. Sorry, it's my fault. No, I think it's all helpful, just to me at least, to bring it back to EPA's actual rationale in this case. And because it seems to me the question is, the applicable requirements of this chapter, are they the requirement itself that there'd be that emission limitation requires continuity? Or is there something else like achievement of the next or some other standard? And I get the point if it's the latter, because if it's the latter, then that seems kind of naturally the way to think about this. But if it's the former, it raises at least the possibility of a circularity question. And I'm looking at J184 in the second column in the rule, and about halfway down the column, the EPA says, Thus, at a minimum, EPA interprets the phrase as may be necessarily appropriate, which is the relevant phrase in 7410, which I guess is 110A2A to use the language, the nomenclature that others have used, 110A2A. EPA interprets the phrase as may be necessarily appropriate to include what is necessarily appropriate to meet legal requirements of the CAA, including the requirement that emissions limitations must apply on a continuous basis. So, the relevant requirement, it looks like, is the definition itself, not some other requirement or standard like a max that's not being met. It ends up circling back to the definition. Yes, Your Honor, and I think that that is consistent with the court's interpretation, again, in Sierra Club, that gave that continuity requirement independent force. This is the difficulty that the under 112 EPA has to promulgate, whereas here we're asking, what is the content of what needs to be in there? I actually had underlined not the part that the Chief just read, but the ensuing part, which talks about basic legal requirements, e.g., the requirements of 110A2A through M, substantive requirements of RAC level controls, whether the schedules and timetables for compliance are legally and functionally enforceable. So, I thought it was sort of the requirement to have SIPs, to have SIPs that meet the various special requirements that may be applicable to sources and to non-attainment areas and the like. And so, that's what I was taking to be what feels non-circular, but I'm not sure that that's satisfying to my colleagues. Right, and I think, certainly, the continuity requirement is also in service of those other requirements, again, in service of enforceability and in service of substantive stringency standards or substantive aims of the Act. Right, I guess my question about that is, that could well be the case, and just reading the statute tells you that one can recite that, but then there's no effort made to substantiate that, other than just to say that that's generally true, which, just to look at the statute, would tell you that that's what Congress thought. It's in service of these broader objectives, but the agency didn't do anything to substantiate that. I think because the agency thought it didn't need to, because the requirement that was most clearly an issue was the perceived requirement to meet the definition of an emission limitation, which itself requires continuity. So, I take the point that this was an add-on, but it seems like an add-on that's just a statement without much to bolster. I guess not much to bolster in the sense that there was not specific fact-finding, again, but I think EPA's assessment and analysis of other requirements of the Act that, you know, I think EPA reasonably concluded would be jeopardized by discontinuity. I think that's where those come into EPA's rationale here, and just, once again, I think that's consistent with what the House report said about the purpose of this term. Okay. Okay, I think I'll, at least for my purposes, I'll leave it at that, and hopefully we won't call you up for a third time. That was my mistake. Unless my colleagues have additional questions for you at this time, we'll go to Intervenors Council. Thank you for a second time, Ms. Buckley. Ms. Issaad, we'll hear from you now. Yes, thank you. May it please the Court, Andrea Issaad for Environmental Intervenors. SSM events release massive amounts of dangerous pollution into communities, and they happen frequently. So, I'd like to talk briefly about those impacts and then address, make a few additional points about some questions we've been discussing, and then emphasize the plain language of the Act to conclude. One example in the record released 156,000 pounds of sulfur dioxide into a Louisiana minority community where many of the children are living below the poverty line. Sulfur dioxide reduces lung function and sends children and the elderly to the emergency room. As we discussed, these events are not even monitored or reported in most states, so most communities don't even know what they're breathing in. So, states cannot say that they're adequately protecting the air. They're only saying that because they're ignoring these emissions. I was curious about the reason for what you just said, and I don't doubt that it's true that there's a lack of monitoring, but I guess, why is there a lack of monitoring? That might be a better question for industry. I can tell you that EPA has recognized that, I mean, as far as NACS monitors, that the existing NACS monitor network is insufficient and that significant air quality problems can avoid detection. I can tell you that many facilities turn off their air monitors during these events. So, the monitoring responsibility is industry's? It's not EPA or state or citizens who want to file a citizen suit? Well, it's certainly not citizens, but I think both EPA and states have complementary monitoring responsibilities, but that an event can escape detection if the wind is blowing the wrong way. You can't possibly have a monitor in every location. We rely on all kinds of NACS cases on the monitoring system that's out there. If it doesn't show exceedances, we rely on that. If it shows exceedances, we rely on that. I've seen these models and reams of documents that are lying. Of course, it's not perfect, but that's not what this case is challenging. This case is challenging the treatment of SSM events, and I guess it does seem more relevant to me the kinds of things you're saying about whether they are considered to be in other forms of monitoring. Ambient air monitoring through stationary monitors would presumably pick them up, but then you'd have this causal question of where did this excess come from? But there are other monitoring requirements, and if you could be more concrete about how those are evaded, like individual sources have to report, and maybe they don't report this amount, and whether there's something about that in the record could be helpful. EPA does talk about this in the record. I know the legal memorandum at page 23. Sorry, I don't have the appendix slide off the top of my head, but they talk about this monitoring problem. I'm sorry, I lost the first part of your question. To the extent that there are indicia that these exceedances, as hazardous as you or as dangerous, as harmful to health as they may be, are not always picked up and responded to through factual ways, as Mr. Esrae was really urging they would be, should be, or most appropriately. So the SIP provisions don't require, or they either don't require the monitoring or reporting, or they explicitly allow sources to avoid that responsibility during these events. I know of two states that do track and monitor, Louisiana and Texas, and that's where most of our data comes from. Really, the point I'm trying to make here is that there are real-world impacts, or there's real-world consequences of these events. Even though EPA was not required to make any findings to that effect, because the SIP call provision does not require it, EPA did acknowledge these consequences, that they adversely affect human health. EPA recognized that NAAQS violations occur. It looked at one specific example on the record. And that these provisions interfere with enforcement, and there's community impacts. So that's really the point I'm trying to make, and that this rule is extremely important to fenceline communities. And one additional point, that fixing these provisions will help, because it will provide the needed incentive for sources to make the investment in proper design and maintenance of sources. Good actors do reduce the frequency of events and emissions, just like a newer and better maintained car will operate better. So I'd like to turn to the plain language of the statute and try to quickly make the point that this court should deny the petitions at Chevron Step 1, because the statutory language is clear. Section 110, under Section 110, SIPs must include enforceable emission limitations to meet all requirements of the Act. We've talked about those requirements of the Act and the minimum stringency requirements. And I want to address the control measure discussion briefly to make an additional point that we don't believe it's possible to meet the Act's requirements without emission limitations. Because we're talking about requirements like NAAQS that have a one hour, they're set over one hour or three hours with no averaging. So this notion that a state could use just control measures and not include emission limitations at all seems really unreasonable reading of the Act to rename an emission limitation into a control measure to subvert the continuity requirements. So this point has been made, but reading Section 110 and 302K together, the SSM exemptions and director discretion conflicts with the plain language requirement here. And the general duty provision cannot legitimize them because the general duty is unenforceable. It's unquantifiable. There's no objective metrics for the court to evaluate. Affirmative defense provisions conflict with the plain language of Section 304 and 113 as the court found in NRDC. So in closing, the court should reject the petition to weaken the important public health protections promised to communities in the Clean Air Act and rule it Chevron Step 1 because the Act is clear and it will provide more certainty to the parties and conserve judicial resources. Thank you. Thank you. Let me make sure my colleagues don't have questions for you. You timed it exactly. That was pretty. Thank you. Mr. Ezra will give you your rebuttal time will round it up to the 7 minutes you asked for. And at some point, if you could, it doesn't have to be the beginning, but at some point, could you just remind me whether K6 was referred to whether that argument was made in the briefing? And if not, why not? But anytime you want to. Sure. Thank you. And may it please the court. Let's keep talking about willful sheep. So let's assume that we've got a rule that you have a fence that has no holes in it. I think under the statute, the question you ask is the state has a hole in its fence. Does it matter? The hole may be so small that no sheep could ever escape out of it. And then we would submit that you would never find that hole is a substantial inadequacy. And I don't think that EPA has given you any explanation say of how you would find that any violation of legal requirements in the act. Which falls short of substantial inadequacy. And I think that that just means that the point that we made in our reply that they're reading the word substantial out of the act is correct. EPA does not give you any metric by which to measure whether these plans were substantially inadequate. And I think it's important. I heard Mr. Kaplan say that the reason he does not think that they don't need to do that is that all of these provisions of the Clean Air Act are designed to make sure that the states are doing its job. But the comments that the EPA got, I refer you to page 490 of the Joint Appendix, an example from Louisiana, said we are doing our job. These exemptions are not impairing air quality in our state whatsoever. And the EPA dismissed them because they said that does not matter to the fundamental requirements of the act. I think the point could be made in a couple of other ways. We talked a lot about modeling this morning. At JA-233, EPA got a comment that said, look, states are modeling and sources specifically are modeling FSM events. And the EPA said it doesn't matter one way or the other because we find that there's a fundamental violation of the act. And so I don't think the EPA is giving you any reason to believe this tiny hole in the fence that is identified is substantially inadequate. But then we get to the question, I think you hit the nail on the head, Chief Judge Mnugosan, do you need a fence at all? And the answer to that is no. There is no general requirement in Section 110 that states have emissions limits whatsoever. Section 110 AQA says that states need to have enforceable emission limits and other control measures, means, techniques, et cetera, as may be necessary or appropriate to meet the requirements of this chapter. Thus, if a state determines that emission limit is neither necessary nor appropriate, then it need not have an emission limit at all. And that makes sense structurally. I think Ms. Buckley said the reason she thinks that there needs to always be a continuously applicable emission limit is this court's decision in Sierra Club. But in Sierra Club, this court was interpreting Section 112D, and that actually proves our point. That provision shows that when the act wants to have a nondiscretionary emission limit, it knows how to say it. 112D says that the administrator shall promulgate regulations establishing emission standards, emission standards, the statutory synonym for emission limit, for each category of hazardous air pollutants. That's not exactly the EPA's position, right? Because the EPA with South Carolina said, look, if this category of sources is not controlled, fine. We're not going to look at that as inadequate. And there isn't any dispute here that all of the states that were called submitted SIPs, and none of them have said that what they've put in those were not put in there in order to meet their Clean Air Act obligation. I mean, I understand your argument as sort of a language argument that in the abstract, if you had a state that said, look, our strategy in the SIP is we're going to stand up a go green campaign. We think the consumers in our state put a lot of pressure on the sources. The sources are really tech forward. They really want to use the best technology with the consumer pressure and the knowledge of the sources. They're going to do better than anything we have come up. We're writing that in our SIP, and we're submitting it to EPA. And I take it that your position is that in that case, unless they showed exceedances as a factual matter or as a matter of modeling, they couldn't issue the SIP call. And EPA's view is, actually, we have to be able to look at that and say, legally, we don't think that meets the Clean Air Act. But that's not the case at all. So I think it is. And let me try and explain why. So I think, you know, it is true, all of these states submitted to the EPA as part of their SIP submissions a program that they thought would be necessary or appropriate to comply with the act. And we said, what we want to do in our program is we want to have an numerical limit that excludes startup shutdown and malfunction events. Now, EPA says, okay, that is not an emission limit. And I think our response is, well, that's fine, because we don't need to have an emission limit at all. And I just point out that I think Ms. Buckley said today that that would be fine as long as EPA found that that type of SIP provision was necessary and appropriate. And at JA 179, EPA says that it is, quote, approved SIP provisions that impose various forms of emission controls that are not, by definition, emission limits. And I think the latest and greatest interpretation from the EPA on this, which is in their withdrawal of the North Carolina SIP call, took the same exact position. So I think everyone is taking the position that states do not need to have emission limits if you think that the lack of continuity in the caused by the startup shutdown and malfunction event means that we don't have an emission limit. All that means is that we've used some other control device. And if the EPA wants to challenge that, Judge Gillard, I think that the answer is they can say, well, you have not demonstrated, either at the submission stage or as part of a call, that you're doing something sufficient to maintain or attain the NAC, the Interstate Pollutant Transport, for any other control device. But then I think we get back to Chief Judge Van De Valken's question, which is talking about JA 184. EPA would need facts to make that shown. Now, I do want to say, even if you think that you've got to have a fence that sort of at all times complies, the sheepdog hypothetical is actually quite good because we have one by nature of the general duty clauses. And Judge Gillard, you asked my co-counsel where in the record you can find EPA's objection to the general duty clauses. It's at JA 185. And what EPA says is that the problem with these clauses are that they are often located in different parts of the SIP and are often not cross-referenced or otherwise identified as part of the punitively continuously applicable emission limitations. EPA's problem is that they don't have a cross-reference. EPA doesn't point you to any reason why cross-reference is necessary in the law. And I point you to the decision in Texas EPA out of the Fifth Circuit that says EPA does not get to call a SIP based on problems with drafting styles. So I think that there's no reason why the state general duty clauses cannot suffice here. And again, the only other thing Ms. Buckley said this morning was that there may be stringency problems. But I think, again, EPA would have to point to facts to show the stringency problems. And also, that call would look substantially different than this one because it would say, state A, you have a non-attainment area where you lack stringency. We want you to rewrite the plan there. I see my time has elapsed. I want to make sure I get to the K-6 issue. And if the court will indulge me, I would like to say one thing about affirmative defense. I'm not sure that we've gotten there yet. So, you know, Chief Judge Ranhoff, on the K-6 issue, I don't think we briefed it, although I do think it's in the administrative record. It's at JA-25 where EPA discusses why it didn't issue it. And I guess what I'd say is, you know, we tried to directly brief in the limited pages we had why we thought K-5 didn't apply. I don't think we can waive statutory interpretation. I do think it's another reason why you would brief K-5. I wasn't thinking of it necessarily in terms of waiver, although I think about that. I was thinking more in terms of if it was the first response you had, and I was just wondering why it wasn't featured if the negative fragment of K-6 is that K-5 should be construed to exempt these kinds of legal things that, under your theory of it, could be accommodated under K-6. Sure. I mean, I guess I don't think that it's our sort of affirmative best argument. I take the point that both Mr. Kaplan and I think you made, that you could read K-5 and K-6 as being overlapping. I think there's reason not to do that, given, you know, substantial and binding and all the reasons we've discussed this morning. I think that the real question, and I think what I was trying to respond to as well, it seems a little odd that EPA has no ability to control sort of obvious legal errors. I forget if I got that question from you or from Judge Ballard. I think it's a good question, and I think our answer to that is K-6. Can I ask, I'm going to follow up, and I know you have one on affirmative defenses, and we'll let you make that point, but I want to follow up on the discussion about emissions limitations and their necessity, or lack thereof, and how that might implicate this case. So, suppose you have a SIP that only consists of emissions limitations, and the whole rationale for why the SIP should be approved is, look, we've got these great emissions limitations, and you ought to approve it because these are the gold standard. And then, in that hypothetical, could EPA call the SIP based on a recognition after approval, post-approval, that there's a discontinuity problem? I don't think so, and the reason I don't think so is that even if EPA identified that as being a problem, that wouldn't actually show any legal error in the SIP because the SIP would still have all of the control mechanisms it has. I think whether or not they're labeled an emission limitation or not, EPA could still judge whether they're sufficient to attain or maintain the max or any other ambient air quality standards when EPA is assessing the SIP in the first instance under K-3. And so, I think the way I would conceptualize that problem is, if EPA finds everyone thought this was an emission limitation, we've now discovered that it's not, I think the question they should ask is, well, does that make the SIP insufficient? Do we think we need an emission limitation to be necessary or appropriate to meet the other requirements of the Act? In other words, your point is, if the statute doesn't require the thing that the EPA defines as an emission limitation in order to gain SIP approval, then the fact that the thing that undergirds the SIP doesn't constitute an emission limitation doesn't, in and of itself, provide grounds for calling the SIP. It could be an ingredient of an argument that the SIP should be called because of some consequence that comes about because of that. But would it be the failure to meet the definition of an emission limitation in and of itself that would be grounds for calling the SIP, even in the case in which the entire SIP is composed of things that the state itself is contemplating constitute emission limitations? I think that's exactly right. I don't think we would say that the state sort of mislabeling the content really impacts the evaluation or the legal sufficiency of the SIP whatsoever. I mean, if that's the case, then it's really just a matter of saying what we thought, you know, in the 1970s was an emission limit is really a control device. I'm not sure why that would be a basis to reject the SIP, much less one to call a SIP as being substantially inadequate. Okay, and I think you wanted to make one point about affirmative defenses, and we went over time with EPA counsel, so we'll let you make that point. Sure, and much appreciated. I know we've done quite a lot. So I think, you know, if you look at NRDC, and I think that the arguments we've gotten today recognize that EPA's sole basis for the call here is the decision in NRDC. There are two relevant paragraphs, and Chief Judge Schoonemason, you know this much better than I would, but they're in part four of the opinion. The first one deals with the court power arguments. But NRDC wasn't only about the court's power, it was also about the EPA. So the very next paragraph after that deals with EPA's argument that it thought that it can impose affirmative defenses under Section 301. The court rejects that argument because EPA's general rulemaking authority was not specific enough to underlay affirmative defenses. I think we have a specific rulemaking authority. I just want you very quickly to two sources. One is Section 110A2C that allows us to create a program to provide for the enforcement of the measures outlined in 110A2A. I think program to provide for enforcement is quite a broad term. It would include things like allowing only injunctive relief of non-interior damages. And then the second one is 110A2A that we talked about a lot this morning that allows states to design enforceable emissions limits as may be necessary or appropriate to meet the requirements of the Act. We think inherent in that delegation is the ability of the states to define how those emission limits will be enforced in each circumstance. That delegation seems peculiar to me because parallel to the state's authority to come up with the package of regulations that will suffice at the federal level, the EPA has general regulatory authority. The notion that EPA doesn't have analogous authority, I'm not entirely digesting that point. EPA does FIPS. EPA regulates in all kinds of ways under the Clean Air Act. Right. So I guess what I would say is in NRDC, the only authority EPA pointed to, and I think it's because it wasn't a FIPS case, it's a 112 case, was its general regulatory authority. And what the court said is in light of the enforcement provisions, we don't think that that sort of very generic ability is enough. But I think our point is we have much more specific ability. And I do think that the structure of NRDC makes clear that there's no sort of impingement on court power to have an entity that has the ability to create affirmative defenses, to exercise that power to create affirmative defenses. And I'll just say one last thing. I'd encourage you to read EPA's reasoning in its last noticing comment rulemaking on this issue, which was in the withdrawal of the call in Texas, where EPA essentially adopts our view of this case. And, you know, to the extent you think that there's really any Chevron issues going on in this case, I think those issues would favor our interpretation, because in the Texas call, in the North Carolina withdrawal, and the Iowa withdrawal, which is the last noticing comment rulemaking that EPA has issued, EPA adopted all of our interpretations, especially on that. Thank you. Let me make sure my colleagues don't have any further questions for you, Mr. Esrae. Thank you. Thank you, counsel. Thank you to all counsel. We'll take this case under submission.