Oral argument — State of New York v. EPA

19-1231, State of New York et al. Petitioners v. Environmental Protection Agency, and Andrew Wheeler, in his official capacity as Administrator of the U.S. Environmental Protection Agency. Mr. Wu for the Petitioners, Mr. Berman for the Interveners, Ms. Spence for the Respondents. Mr. Wu, whenever you're ready, please proceed. Thank you, Your Honors. May it please the Court, Stephen Wu for the Petitioners. I'm reserving two minutes for rebuttal. I'll be dividing my time with my friend Mr. Berman for the Interveners, who will be addressing the specific question of EPA's denial as it relates to electric generating units subject to the now-invalidated cross-state update. I want to spend my time on three reasons that EPA's denial should be vacated and remanded. First, EPA improperly relied on the purported adequacy of other cross-state air pollution rules that this Court has now found to be inadequate. Second, EPA improperly faulted the Petitioners here for failing to meet what it deemed to be their initial burden under Section 126. And finally, EPA disregarded relevant evidence about ongoing air quality problems in the denial. To begin with the first defect in the denial, EPA expressly based its denial here on the assumption that the cross-state update rule and the closeout rule had fully resolved the problem of interstate transport of ozone precursors from these upwind sources. And this is true both at the Step 1 analysis and at the Step 3 analysis. At the Step 1 analysis, the denial expressly incorporates the analysis from the update about what it calls the 2023 Future Analytic Year. And it bases its finding that New York and other Petitioners would not have downwind air quality problems on an assessment of whether in the year 2023 there would be issues. But in Wisconsin, this Court squarely held that 2023 or similar years was the wrong year to look at because the downwind areas are required to comply by 2021 or earlier times. And the failure by EPA to adopt that statutory deadline in its analysis here is a reason to overturn their finding that there would be no downwind air quality problems as of the correct future year. Now, EPA has claimed in a footnote to this denial, which came out after Wisconsin was finding a problem with downwind air quality. But that is incorrect. I mean, what they refer to in the footnote, again, is the concept of the Future Analytic Year. There is no analysis here of the correct Future Analytic Year, which is 2021. Instead, all of the analysis centers around 2023. And that choice of an incorrect year, that is two years past when the downwind states have to come into attainment, is a legal error and the basis for reversal. At the Step 3 analysis, EPA also incorporated and relied upon both the update and the closeout here. And I think this is clearest to see at 56089 of the denial, which is at JA51. In the litany of reasons that EPA gave at the Step 3 analysis, one of the ones they relied upon was to say that the state's peer failed to meet their burden to get relief under 126B, in part because they did not establish that there were necessary additional source-specific unit-level emissions on top of what was established in the update. In other words, the baseline that EPA chose to determine whether the petitioners had provided sufficient evidence to show relief under 126B was the baseline established in the update and later in the closeout as well. But again, this part of that... Sure. It's the paragraph that is in the middle column and then the right-hand column that begins while EPA acknowledged the CAFRA update. Got it. And it's the end of the paragraph on the right column where it says the petition has failed to demonstrate it's necessary to implement these additional source-specific and unit-level emissions. And so... Mr. Wu, this is Judge Griffith. May I ask you a question to help me better understand how 126B works? The language of 126B speaks of a source or a group of stationary sources. Now, that doesn't sound like 350 sources from nine states to me. Well, the language allows sort of a range of options here. And prior... What's the group? The group here is 350 from nine states? You think the language supports that sort of... Yes, it would. I mean, the group here are... What does group mean? Is it anyone you list? Well, here a group of sources identifies upwind sources that the petition identifies as contributing to downwind non-attainment or interference with the 2008 ozone acts, the 2015 ozone acts. And so it is targeted in the sense that it is talking about a specific form of interstate transport and identifying the sources that contribute to our downwind non-attainment problems. Now, this is not... And you're reading... It would be interesting in your reading of how 126B is supposed to be a good neighbor provision. In your view, is there any meaningful difference between 126B and the good neighbor provision, other than who initiates the process? That in 126B, a state can... A governmental entity can start it, it can initiate it, and then in the good neighbor provision, it's EPA. But other than that, is there any meaningful difference to you in your reading between 126B and the good neighbor provision? Yes. And the differences were recognized by this court in the Appalachian Power cases. 126B, in contrast to 110, allows for a different schedule for compliance and a more expedited schedule for compliance. It also empowers EPA to provide for source-specific and short-term limitations that don't have to be the equivalent of a complete regional... How do you respond to the EPA's argument that that shorter timeframe actually cuts against you, that that shorter timeframe suggests that what your suit is really about is just trying to get them to do what you want them to do under the good neighbor provision, and that 126B is intended for something far more targeted? And it gets to the burden of proof at some point, because then the argument, as I understand the saying, is then the burden is on you to come up with showing the problems, and that the shorter timeframe supports that reading. How do you respond to that? Well, there's a couple of responses. I mean, one is that this case in particular is an inact one for EPA to be complaining about the timeframe here. The information that EPA should have collected already as part of its obligations under the good neighbor provision. And what was clear from the legislative history of 126B was that Congress intended it to be an independent mechanism for the states to seek relief that is parallel to the good neighbor provision, but that is meant to compel the agency to act when it has failed to do so, which is precisely this context. So, you know, whatever might be true about other instances, this is a context where the same information EPA claims is missing is what it is supposed to collect under the good neighbor process, and what this court has now recognized it should have collected in order to ensure that downwind states could meet their own obligations by the relevant statutory deadlines. In addition, there's a mischaracterization of what this petition is seeking. I mean, EPA has characterized this petition as seeking the equivalent of what is going on in the prior New York case. But that is incorrect. What we are seeking here is not a global transport rulemaking, and therefore we would be subject to the standards applicable to that. What we're seeking is to curve emissions from upwind sources that our analysis has demonstrated, you know, result in nonattainment or interference with our maintenance of the 2008 and 2015 standards. And it is perfectly compatible with that argument for EPA to provide, as we argue both in the petition and in our briefs, short-term and source-specific controls, even if it is not controls necessarily that extend across all 350 sources. And I guess this is the last thing that I would say about the 60-day deadline, which is if EPA conflates here the time that it would take to impose controls beside the actual level of controls across all of these sources with what we are asking for, what the statute envisions within the 60 days, which is just a finding of significant contribution, and therefore sort of placing these sources in the regulatory bucket, allowing EPA or requiring EPA to consider what emissions controls should apply here. For that sort of discrete analysis, the 60-day period should be enough time for EPA to determine in this context that these sources are significantly contributing to our downwind attainment problems, and therefore should be subject to more controls. But doesn't the 60-day limit suggest that they are right on the burden of proof here, that if someone wants to take advantage of this more expedited process, then the burden is on you to show them where the problem is and whether there is a cost-effective solution to it. That seems to me a pretty powerful reading of the time limits. Well, we don't dispute that the states have some initial burden in bringing a 126B petition, but the problem with this denial is in how EPA defines what that burden is. Now, there's many things that EPA says in this denial, but one way of understanding what they think is missing and what the states have to establish is we have to come in with something that EPA has never done, which is a complete regional transport rulemaking that uses the analysis across many sources and many states. But their response to that is that's because you've identified 350 sources. You've asked for too much. Stick to a group, right? Right. But, I mean, number one, we stick to a group in the sense that we are talking just about interstate ozone transport here, and we are trying to identify all the sources that contribute to our problem, but I think more fundamentally, you know, EPA's complaint about the number of sources here is in a way a red herring because one of their grounds for finding this petition inadequate is that we don't identify more sources. We don't analyze more sources. I mean, one of the things they say at 56090 of the denial is that we were required to conduct an analysis that compares controls for these sources to, and I quote, other unnamed sources in order to do what they think is necessary under step three. In other words, that we were supposed to look at more than just these 250 sources and explain why the controls we are proposing are cost effective compared to that. And I think that line in particular highlights the degree to which what EPA has defined as the relevant burden here is not what this petition establishes, but instead the states having to come in with a ready-made solution for a regional transport rule. And EPA is claiming the entitlement to deny a petition on the basis of our failure to do, again, what EPA's responsibility really should be here. And one reason that that's an improper burden to impose upon the states is because of what this court recognized in Appalachian Power, which is that 126B is meant to be independent of 110, that what EPA does when it does these regional cross-state transport rulemaking is sort of one thing. But what the state court entitled to ask for under 126B is a separate remedy that adopts when I was coming out. This is Jeff Shreve. This is Jeff Shreve. Can I just ask one question about the overall architecture of your argument? So we're talking now about the relationship between 110 and 126 and the burden that's imposed upon EPA. The first set of arguments that you started out with, just on following the way they interrelate, the first set of your arguments that you started out with has nothing to do with this, as I understand it, because your point is even if you're wrong on all this, then there's still something wrong with the order because it inappropriately relies on pre-Wisconsin conceptions of the rule? Yes, that is correct. That is correct. The way that we understand them relating is that the defect you just identified is sort of a complete reason to vacate and remand the rulemaking. And what I'm now engaging in is the question of if you ignore all of that, why it's still the case that EPA's finding of a burden on the state is invalid for other reasons, even taken on its own terms. I didn't mean to sidetrack you, but you can finish that response and then we'll see if my colleagues have any further questions before we move to intervenors. Sure. And I'll just very briefly say, I mean, the reason that we are focused on the improper burden defined by EPA is that we are concerned that EPA will just rely upon this same invalid reason on remand if it is able to fix the defects in its reliance on the pre-Wisconsin, pre-New York precedents. And so we do think it is important in this context to clarify that EPA cannot require the state to establish a complete regional transport rulemaking in order to obtain any relief under 126B. I'll reserve the remainder of my time for a moment. Wait, I'm sorry. I apologize. I had some questions too, if that's okay. It's Judge Millett. So you agree that the states have some initial burden and that the line here is that they just went too far. And I'm hoping you can help me pin that down. Starting from the bottom right corner on JA-50 and going through there and through 52, there's a list of things that EPA says should have happened. I'm trying to figure out which ones you think is part of your burden, where this line is between your burden and their burden. And so the first thing they have, little ruminant one, is verifying that the name sources whose emissions are those from the most recent emissions inventory continue to emit nitrous oxide at the same rate or continue to operate. Do you agree that's part of your burden? I apologize. Let me try to flip to that. Sorry. It's JA-50, the very bottom right corner, 50-51. There's four things, four ruminants there. Okay. So you agree that's your burden, that first one. That's pretty big. That's just how to come up with the sources continue to emit at the rate or continue to operate. Yes. I mean, yes. And we've identified those sources in the... We identified the emission rates of those sources in our petition. Could you verify that they continue to operate? Because the Kentucky MFS brief says some of that, at least theirs, don't. Right. I mean, I think part of the answer to that is that there was a vast gap in time between when this petition was filed and when EPA actually acted on it. So the time it was filed, was that all accurate and up-to-date? Yes, I believe that is correct. And so the false test for not confirming it, either that or the fact, I think is really on EPA. I just want to make sure, though, that at the time you filed, that that was all up-to-date. Yes. Okay. And you agree that's part of your burden? Yes, that's right. Okay. And then Roman 2... I'm sorry, just because I'm really trying to understand how much you're supposed to do on your end and how much they're supposed to do on their end because everyone's sort of just complaining about the other but not helping me figure out where the fine is. What about ruminant 2? Well, I do think the rest of them are more appropriately placed on EPA here. With one caveat, which is we do in the petition provide a method of determining cost-effective controls, which is sort of what ruminant 2 is talking about here. We do not go source... We agree. We do not go source by source and say, you know, we know for this plant what machines they have installed and what controls they have already done. And that is in part because of the difficulty of states obtaining information for upwind sources that are located outside of our jurisdictions and where EPA has the ability to obtain that information. But if I could also answer your question in a slightly different way, one of the frustrations with this denial is that it actually does not identify the standard that it thinks states have to establish. What it does in the list that you identified here and it does in the next page and yet another list of information that they claim is missing, it's just repeatedly sort of talk about information that EPA wishes it could have had to conduct what it deems to be a regional transport rulemaking. And I do think our global answer to a lot of these is it is unreasonable to impose that burden on the states or unreasonable to deny the petition on that ground. One, because much of this information, including the Romanesque you're talking about, are not easily available to the states. And then two, because for much of this information, it goes to a standard the states shouldn't have to meet, which is something like a regional transport rule. And then three, I should add, is much of the information they claim is missing is also information that is relevant at the 126C stage when you are deciding which specific controls to impose upon these sources. You said not easily. Go ahead. I'm sorry. I finished your sentence. I'm sorry. Go ahead. Sorry. No, I really didn't mean to interrupt your thought. I was going to say that you said not easily. Some of this information is not easily obtainable or is it not obtainable at all? Well, much of this is not obtainable at all. And in particular, the source-specific information where we're able to know what exactly they're it's impossible for the states to obtain. And EPA's sort of identification of some public sources is not the equivalent of the type of analysis that they think is necessary under their improper standards. Part of your burden is to show that there are upwind sources that are contributing significantly to your, in this case, non-attainment. And contribute significantly, obviously, when the EPA is coming up with rules has the meaning it's been given by EPA and upheld at EME Homer. But what I'm trying to understand from you then, how much are you supposed to show to establish not just that there's these sources in these other states that are emitting pollutants, but that they are contributing significantly to our non-attainment? How do you show that? How much do you think you can show? What do you think the law allows EPA to require you to show in that regard? Right. Well, I don't know that we have a position on the minimum level of information that's required. Although, what I can say is we think the evidence provided in this petition is enough. And the reason that it's enough is for the initial stages of this process, we sort of used the same method that EPA has used in its prior cross-state analysis to identify sort of, first, the states that are significantly contributing to downwind attainment problems. And then, two, to identify the sources that are the major contributors to that problem. And our petition conducts an analysis, which is similar to sort of the air and wind analysis from prior rules, that sort of traces the problem to individual named sources. And then finally, our petition gives a basis for believing that these sources have available cost-effective controls that should make them part of the regulatory bucket of sources for which EPA should determine emission controls under 126C. And the basis for the petition showing on that front is that New York requires its own sources under a parallel analysis to adopt controls to reduce ozone emissions in a cost-effective way. And we have shown that sources upwind from us emit at greater rates than New York sources are required to do and could substantially curb their emissions if they were to either adopt control technologies at the cost level that we have here or to reduce their emissions to that level. Now, is that the same as, you know, a plan of all 350 states to figure out what their emissions controls are? I mean, the answer is no. I mean, but that is in part because 126C envisions that EPA will actually do some analysis and work in response to this petition. I mean, the statute is not sort of a remedy that the states can have directly against sources. It is a request for EPA to exercise its regulatory authority over sources under its direct regulatory ambit, and we think what that entails is that EPA, number one, has some duty, not a complete duty, but some duty to conduct a further analysis, and it's particularly appropriate for it to do so using information that is uniquely within its control. And to put some of that, to shift some of that burden back onto the states is an unreasonable ground for denying a petition that otherwise provides substantial information about the reasons that we have downwind nonattainment problems. You had a response to a question, I think, from Judge Griffiths about, you know, the short time limit in the 126B context. He said, look, here EPA either has or should already have this information because it's been for quite some long period of time now engaged in trying to make a transport rule that would govern this big neighbor problem. But what if a petition of this magnitude came in early on, before EPA would have been in the position of collecting this type of information? So you wouldn't have that argument, which is a case-specific argument, I think. How then would EPA within its, let's even assume, eight-month time frame, would it really be able to do everything? If it didn't already have it in hand, would it be able to do all the analysis that a petition of this magnitude requires and the hearings and everything in eight months? It may not in a different case. And so it might be that there's either a pollutant-specific or sort of case-specific reason for denial in that circumstance. But I do think it is important that this is a situation where at the time of the denial, EPA was already years late in collecting this information and coming up with a regional transport rule. And I think that's true for two reasons. One... No, I'm just trying to understand where... I'm sorry. So I get the arguments about EPA's delay on this particular area, this particular... where you all want a rule and the rule... the partial rule was brought down and all that, and it's just been going on forever and ever. But we're trying to understand what the statute means and how it assigns burdens. And so is it your position that if EPA gets to the end of... just as a matter of law, in any case, if EPA gets to the end of that eight-month... you know, the 60 days and the six-month extension, what we'll call an eight-month period, they get there and they still just can't get the information they need? They haven't been able to get it all. They just haven't had enough time. But I hear you say that they can then deny it because they didn't have enough time to act, even if the petition met whatever burden of proof they required. They couldn't act on it. That's a sufficient ground for denial of petition under 12060? Well, I don't think we would say that. And all I was answering was that... I thought you just said that when I asked you if a petition of this magnitude had come and then the point when you couldn't say that EPA should already have all this information at hand. Well, I guess I would answer... Go ahead. Oh, I guess I would answer this way, which is that, again, in a different context and a different pollutant, you know, how EPA defines the standard in that case may depend upon sort of what the delta is between what is presented in the petition and what EPA has to do. And what I'm not certain about in answering you right now is whether there is sort of a global way of defining that standard. I'm not talking about different pollutants. I'm talking about this exact, you know, ozone pollution that just, you know, if your petition had come in, you know, 2010 or 2009 with this exact same content. So I'm not talking about different chemicals. And they said, thanks for calling us to our attention. Big problems. We're wrestling with them. There's no way eight months is enough time. Eight years hasn't been enough time. Yeah, well, so let me answer this in two ways. I think the first answer is we would not take as given that EPA is right in saying that it's not enough time. And I think part of the problem here is that, as this court recognized in Wisconsin, you know, administrative infeasibility is not really an excuse for not complying with various statutory deadlines. And so if EPA were to come in and say, you know, in the ozone context, we need much more time to do this, I think a perfectly valid response is, you don't have that time. And more importantly, we can't wait for that time. I mean, these are not – this is a delay that results in increasing harms for the downwind areas. But the second thing I would say – Oh, go ahead. Go ahead. Well, I apologize. But the second thing I would say is that this is not an all or nothing proposition. A 126B petition identifies sources and seeks to curb emissions from those sources. EPA sort of improperly thinks here that it either does something like a global solution or does absolutely nothing. And that's not what 126B envisioned. There's no reason that EPA could not adopt sort of more limited solutions here. For instance, we propose short-term daily emission limits on certain identified sources as opposed to a global solution. And there is not an explanation for why, and in your hypothetical, may not be an explanation for why more targeted relief could not be available even within the 60 days or eight-month period that would meaningfully address the problems that the downwind areas are facing. And that kind of what I'm calling a partial solution fits within the purpose of 126B. And the reason that statute is phrased as one that can be about even just a single source or a group of sources, and the reason that there are these short-term deadlines is because it recognizes and Congress recognizes that downwind areas have urgent needs to control pollution. 126B gives them a mechanism to obtain that even if it is not a complete solution and perhaps would be appropriate to have a partial solution, especially in areas where EPA has great difficulty coming up with a comprehensive solution to a larger problem. Okay. Thank you very much. Thank you. I'll reserve my time for rebuttal. Thank you, Mr. Liu. Mr. Berman? Thank you, Your Honor. May it please the Court, Joshua Berman with the Sierra Club, on behalf of Petitioner Interveners. Before I jump into the issue of the specific sources subject to the cross-state rule, I just wanted to pick up on one thing on the line of questioning from Judge Millett. I think if you look at the Joint Appendix pages 50 to 51, the list of information that EPA identifies, that list is conjoined by an or, not in hand. And I think that sort of highlights the arbitrariness of EPA's denial here. EPA fails to identify what sufficient information would have been from New York. And I think that, again, the Court's job here is not necessarily to identify, here is exactly how the burden should be allocated. The burden allocation issue is certainly a thorny one. I think the question here is whether EPA's denial was arbitrary and capricious. And because EPA is short of, you know, EPA in places seems to suggest that New York needed to submit an entire transport rule and identify not just the cost-effective emission reductions available from the sources it identified, but identify how those compare in cost efficacy and control to sources not identified in the petition. So that kind of gets past this question of the 350 sources. But EPA's problem here is that EPA fails, short of that, to identify what an actual reasonable burden that a petitioning jurisdiction could in theory meet that would satisfy the agency. The agency says different things on different pages. The list that Judge Millett identified is conjoined by an or. It's not entirely clear even what that list is supposed to mean. And the fact that the agency says different things in different places, I think, highlights the arbitrariness of EPA's denial. I'm happy to pick up on that, but I was going to turn to the EPA's response regarding certain electric generating units subject to the cross-state update. Before you turn to that, would you be able to help me understand? I don't know how your view is the same or different than Mr. Wu's, but what in your opinion is the initial burden that a 126B petitioner needs to meet in showing in particular that not just identifying that there are sources upstate that are spewing out a lot of pollution and the jet stream is bringing it over to us, but that those sources, those specific identified sources, are contributing significantly. What do you think that burden should be? Sure. And I guess, again, I would flag that I think that goes beyond what needs to be. I know the judge is interested and I will answer your question, but I think that goes beyond what the court needs to determine here, again, because it is clearly EPA's burden here to at least identify what the threshold showing needs to be, and EPA hasn't done that. EPA has said various things in various places that are contradictory, and so that to me makes the whole denial, the whole basis of the denial arbitrary. Yeah, but you agree that there is some burden on the 126B petitioner, right? No doubt, and I think what is quite clear is that EPA hasn't. What is that? I don't want to say, I don't know exactly what the line is, but I think what New York has done is clearly sufficient because New York has shown that if you use EPA's sophisticated source apportionment modeling tools and the sources that New York has identified, their collective contribution to ozone levels in New York, in the non-attaining areas of New York, exceeds the 1% threshold that EPA, in all of its prior transport rules, has said is the threshold for a significant contribution. So New York has identified sources that collectively impact its ozone levels by more than one part per billion, more than the significant threshold EPA has relied on in all its transport rules. So I think that is a sufficient showing. The question is whether the source or group of stationary sources contributes significantly to non-attainment or interferes with maintenance. What New York has done here, I think, is clearly sufficient. Again, though, I don't know that we need to go as far as to say this is the exact burden, because I think EPA truly has the burden to identify what the threshold showing needs to be, and EPA has not done that. Okay, thank you. Okay, sure. I'll turn to those specific sources. So New York's petition identified a number of sources that are failing to fully implement control strategies that EPA has found to be highly cost-effective. This includes electric generating units that have installed catalytic controls but are failing to optimize them, and those that lack basic combustion controls. EPA, nevertheless, denied New York's petition for these units because it claimed that these control strategies were already accounted for in developing the update rule. But EPA's reasoning is flawed. First, the update rule has been found deficient, and a deficient cap-and-trade scheme cannot preclude further relief from underperforming units. This court recently invalidated the update rule precisely because it failed to require upwind states to eliminate their significant contributions by the relevant attainment date, or indeed by any date. And the update rule does not even purport to address the more protective 2015 ozone standard, which will require additional emission reductions. Consequently, the fact that units subject to the update rule continue to operate without employing control strategies that EPA has deemed highly cost-effective means that further relief is available from these units, including through Section 126. Moreover, meeting the update rule caps does not imply that any particular control strategies have been fully implemented, and this is an important point. The rule does not cover all sources of electric generation. It only covers coal units. Emissions from coal units can decline simply because the units are running less, a phenomenon that's discussed by industry interveners in their briefing here, and cleaner generation sources are running more, not because the units have actually implemented new or improved controls. The record shows that many units continue to fail to fully implement basic control strategies. There's no barrier to EPA granting the expectation as to these units and requiring them to actually achieve the highly cost-effective emission reductions that EPA previously identified. How many of the units at issue, the 350, are electrical generating units that fall within the argument you're making to us right now? So EPA, there are approximately 130. If you look at a number of places in the brief where EPA says that it lacks sufficient information for the 220 non-electric generating units. So we're talking about 130 units that are subject to the rule, and then there's a subset of those that have installed but are failing to fully implement controls that EPA has previously identified as highly cost-effective, and those are identified in a couple different places, Joint Appendix page 490, Joint Appendix page 543 in particular, also Joint Appendix 491. So we're talking about several dozen units that have installed high-end controls but are failing to optimize them, or units that have persisted in not installing basic combustion controls, two strategies that EPA has previously found to be highly cost-effective. How do you know that they have them and aren't using them or haven't installed them? It sounds like are you able to get that sort of source-specific information in other states that New York says it can't get? So electric generating units are distinct from some of the other source categories that are identified in your expectation in that electric generating units are subject to very detailed monitoring and reporting requirements, and all of those data are then published on EPA's Clean Air Markets database. So in New York's petition, it used then-current data through the date of its petition to show exactly what was happening during the ozone season at all of the electric generating units for the prior three ozone seasons. That was then supplemented and made more current. Again, we discussed there was a long delay before EPA actually responded to the petition. That was then supplemented in additional, more current information using the 2018 ozone season was added in New York Department of Environmental Conservation's detailed comments, and that's found in Joint Appendix 489 to 491. And then additionally in the record, there's additional information about what was happening during the 2018 ozone season in Joint Appendix page 543. Thank you. Thank you, Council. Thank you. Ms. Spence, we'll hear from you. Good morning, Your Honors. Good morning. May it please the Court, my name is Tamara Spence, and I'm here on behalf of EPA. I'll go into the issues concerning the burden in Step 3 in detail, but before I do, I want to make two points up front. First, petitioners make this extraordinary concession. They made it in their reply brief, and they made it again here this morning, that what they're seeking to do is to force EPA to fulfill its duty in the federal good neighbor plan context. This case isn't about EPA's duty to issue a federal good neighbor plan. This case is about EPA's review of a state petition under 7426. Petitioners had their day in court about the federal good neighbor plan for the 2008 standard. It's on remand, and EPA's working on it, and it's not before the Court today. Second, many of petitioners' arguments this morning don't really go to the ultimate basis for denial. The basis for denial is because the state didn't meet its burden under Step 3. We heard from New York Council this morning that the petition was actually based on the preexisting good neighbor plan, and it's not. I can point you to specific statements in the record where EPA says expressly, repeatedly, that the reason it denied the petition is because New York failed to meet its burden under Step 3. Can I ask a question about the failure to meet the burden, which is it seems to me that, at least in some situations, it's going to be very difficult for a petitioning state to have the information that you appear to require for them to get home on a claim under 126B. Can you just address that reality that a petitioning state often will lack a means to obtain the information? Well, they have means to obtain a lot of the information. They could have done a lot more than they did. So, what we point to in the brief, and respondent interveners point to some of this information, too, a lot of the building blocks for a Step 3 analysis are publicly available. What they would need to do with that information is to do the analysis, put it together, and come up with, you know, some kind of basis to make a cost-effectiveness decision, and that's usually in the form of, you know, a numerator and a denominator, you know, cost of controls as compared to tons of NOx reduced and the downwind improvement that you would get from that. For the EGU sources, the vast majority of that information is out there. What a state would need to do is the analysis. For the non-EGU sources, there is information out there. It's a little bit less complete than for the EGU sources, but there were things available that they could have done. We know that New York is able to do air quality modeling. They do it for their own SIP, and they, in fact, provided modeling. They just didn't provide the modeling that EPA would have needed. So what would happen in this situation? I'm not necessarily trying to say that either side is necessarily right or wrong. I'm just trying to figure out practically what happens. So if a petition state submits a petition like this and then says, we've obtained as much information as we can get our hands on, it's true that for our inquiry to be completely rounded out, we'd need the following additional information, but we just don't have the means of obtaining that. You do. Our petition rests on the idea that once you look at that, we think you'll find that there's a problem vis-à-vis cross-state pollution. It's possible. What EPA has said is that they're looking for a colorable cost-effectiveness analysis. If a state were to come forward and say, here's the information I have. We think this supports the conception of cost-effectiveness. Yes, EPA would look at that. One of the things that we heard from petitioner interveners is that, well, EPA hasn't told us how we have to do that. Well, EPA has given them options and also said, you know, we would consider any alternative analysis. Here, petitioners haven't done that. They haven't provided a cost-effectiveness analysis at all. How can you say that with respect to the electric generating units? I mean, it sounds like some of your analysis, they've given a lot of concrete information, and then EPA just disagrees, you know, like the RACT, R-A-C-T, using that as a standard. But disagreeing doesn't mean they haven't met their burden of coming forward with just the sort of information you want. So why isn't what they came forward with on the electrical generating units at least sufficient? Just to meet their burden. Not to necessarily persuade you or prevail for you, but just to meet their burden. What more do they need to do on electrical generating units? We heard this morning, petitioner interveners talked about 130 electric generating units. This actually breaks down into two categories. So the 350 sources total, you've got three buckets. The non-EGUs, the EGUs that were fully addressed in the update, there's about 83 of those. Those are the sources with catalytic controls. And then the 47 that don't have those installed. So the arguments that petitioners are making about, well, you've already made a cost-effectiveness determination in the update. So in the update rule, EPA did make a cost-effectiveness determination about those 83 power plants with catalytic controls. But the update also has a remedy for that. What petitioners provided here was just to say, okay, well, it looks like there are certain units that aren't running those controls. And that was true in the record. There were about seven sources in 2018 that were not running their controls. And we know that because of the rate that they were admitting at. If it's operating above 0.20 pounds per million BTUs, we know that they weren't operating their controls. EPA's response to that is, well, that doesn't tell me that those violations aren't being remedied. They are being remedied because we've remedied it with a cap-and-trade program. So you haven't provided anything. Isn't this precisely what the point of the 126B petition is? I mean, you complain about how sweeping it is. But now we're talking about a portion of this. Let's pretend all the petition did was say, as to those 47 sources that are not using the controls that EPA found to be cost-effective in the update rule, they're not using them. Maybe they're doing cap-and-trade, but you know what? It's not working. But it's exactly the targeted, not rulemaking. They're not looking for a general transport rule. What they're saying is the system that's out there is not working because they are contributing significantly to our non-attainment. That sounds like precisely what 126B is for. And is it an answer to 126B that we're letting them? It's okay that they contribute significantly to your non-attainment as long as they're engaged in cap-and-trade? I'm not sure how the two intersect. I just didn't see that explanation in your decision. But if that's your explanation. Is that your explanation? So the 47 that do not have selective catalytics installed, those sources, I think, have to be entirely separate from the ones that EPA deemed to violate the good neighbor provision because of the ability to run SDR. What EPA has said as to those, and you can see this, there's a discussion of it at JA-54, where EPA says, well, the cap-and-trade program gets the total down. So that was the remedy for those violations. They were deemed to violate. We implemented a remedy. And this doesn't mean the remedy is not working. The remedy is working. I'm trying to ask, I mean, it may work for sort of your transport rule, but I thought their argument was it's not working for New York. They're pushing us up just to credit for these purposes their assertions. They're pushing us over non-attainment here. Whatever they're doing up in their state, they're killing us down here. So here's what we don't know. We don't know that just because there's a fleet-wide average that can be achieved, we don't know that these specific sources can actually achieve the rate that New York is proposing. So that's part of their burden failing here, is they didn't show plant by plant by plant what would happen if you impose those controls one at a time? That's their burden? What EPA has said their burden is, is to say for the sources that are subject to the petition, what controls can they implement? What's the cost of those controls? What's the emission reduction potential from those controls? And then what's going to be the downwind air quality improvement from that strategy? Or an alternative analysis to show that. But the state hasn't done. Okay, go ahead. Please finish. What did the state not do here? What the state didn't do was support a cost-effectiveness finding. So even for the specific. They said they used your cost-effectiveness finding that's already in the rule, in the update rule. They used your cost-effectiveness finding. EPA's cost-effectiveness finding was to a fleet-wide average. It was not as specific sources. Ms. Spence, let me, maybe you answered this and I didn't catch it. But Judge Millett asked you, are they supposed to do that on a site-by-site basis? And I didn't hear your answer to that. The answer is yes. Yes, okay. Thank you. Yes. And the information to make that kind of reticulated analysis, you think that's all in the arsenal of the petitioning state? For the EGU, yes. For the EGU, yes. A lot of those building blocks are publicly available. If New York could do the analysis for that, they certainly are able to do the modeling to show what the downland impact of the reductions they're asking for would be. I noticed there was a comment that Mr. Berman made about, you know, well, some of the core plants are just operating less. I'm not sure I see how that would be a problem, because if the source is operating less, you know, that seems like it might be harder to show that it would be cost-effective for that source to, you know, to implement new controls. That's just a side note there. I'm sorry, I interrupted your question. Not at all. Can you just help me understand what the burden is that's articulated in this 50-52? Because you begin with a sentence, the EPA begins with a sentence that says, the petition could have included one or more of the following potential analyses. And then there's four Romanets. And they said we did one. And it just says one or more. They said we did Romanet one. Do you disagree that they did Romanet one? And how are they supposed to knit? Let's just get that question. Sorry. Did they do Romanet one? I think that's questionable, Your Honor. They certainly didn't say. Is there a finding that they didn't do one of the four? I'm not sure whether EPA thinks they need to do them all or not. You begin with one or more. They tell me they did number one. You're not sure that they didn't do number one. So what's the problem? If you can't show that they didn't do number one, then they've met your test. I don't read this, Your Honor, as laying out a specific test. I read this as sort of laying out the categories of things EPA is looking for. Wait a minute. What does that mean? I don't understand what that means. It's not a test. It's just the categories of the information they want. That sounds like a test or a list of requirements. And the commission says, here's the problem with the petition. Petitions should do one or more, and they did one. Let's give them that they did one. They might say they did more, but let's say they did one. And now you're going to say, well, that's not really the test. Then what is the test? What is the test? How are they supposed to know what exactly they're supposed to do next time? Where do I look in this Step 3 analysis for precisely what it is that they need to come back to you with? So item number one is not talking about cost. If you go up that same paragraph. I'm not asking. Okay, go ahead. Okay. So you go up the same paragraph, and it says, New York has not sufficiently developed or evaluated the cost and air quality factors that the EPA has generally relied on in Step 3, has not described or conducted any sort of multi-factor analysis to determine whether cost-effective controls are available at the named sources, and has not provided any alternative analysis that would support a conclusion at Step 3 that the named sources will significantly contribute. That's what EPA is talking about. EPA is looking for cost-effectiveness. So item one, well, you have to get the baseline in order to know what reductions are achievable, but the next part is what reductions are available from these sources and what do they cost. If you look at the good neighbor provision, it says it talks about amounts, amounts that significantly contribute. There's no way under EPA's test that relies on cost-effectiveness to say what's significant if you don't know what can be cost-effectively achieved. So there has to be some consideration of cost. I guess I'm even more confused than before. So what is the point of the sentences with the four Romanets? Do one or more of these. What is that supposed to mean in the list of things at the beginning of the paragraph? So that's referring to the factors that EPA was looking for in order to identify significant. So it doesn't mean they have to do all of them, but that one, if that's true for EGUs, and I'm not certain that it is, but I can presume that it is. Let's assume they did emissions inventory. If you want to talk about your CGs, you can do that. Okay. Sure. But what that doesn't do is talk about cost. I thought you just said that's what the point of this list was. And you all said one or more. I'm not making that up. That's how you're supposed to identify significant. That's what you say right there. It seems to me like both from the ruling itself or the order itself denying the petition and then the arguments and stuff, they're all over the place on what it is they want. And you say, well, we've listed some factors. You can do these or you can do something else. But then, you know, we need to have this information. And when they say we can't get some of this information, they go, well, well, then maybe you don't have to do everything. But over on 52, you say, you know, I'm sorry, 5051, you say, look, you've got to do source comparisons, which isn't one or more. You've got to do those. And then over on 52, you say you've got to do source comparisons to figure out significant contribution. You just have to do that. And then you say on 52, source comparison necessarily involves. So can you just answer me this question? Is it the position of the EPA in this order that the state must do comparisons to other unnamed sources? And when they do that comparison, they necessarily have to do the things listed on page 52 in the left-hand column, this source comparison necessarily involves, and then it identifies a number of things. Am I right in reading the order that way? That is absolutely not EPA's position. It is not EPA's position that a petitioning state would have to do a comparative analysis. What I think EPA means by this list, when they say or, is the last item on the list is basically would be a cost-effectiveness analysis that's comparing relative costs to other potential sources. So that would be one way. And Long Island Springs also has the relative to other sources requirement. So they don't have to do either. So your position is just to be clear that in their petition to show a significant contribution, they do not have to compare to unnamed sources. Is that how you read this? Yes, they do not have to compare. And item number three is not comparative, to be clear. It is describing downwind air quality impacts of controlling the sources in the way proposed. Relative to other sources. What does relative to other sources mean? I'm sorry, I don't mean to take up too much time here, but I'm really just confused as to what they're supposed to show. So what does relative to other sources mean there in Roman Springs? I think all that means is if you do a model that says, well, the only thing that I've changed is what these sources are doing, what's my downwind impact? I think that's what that means. You think that's what it means, but you're not certain? That's how I read it. The downwind impact is part of the cost-effectiveness. How can they show cost-effectiveness? How would you envision someone showing significant contribution or cost-effective? This was just significance in the rule. You're saying this is not a test for cost-effectiveness? How are they going to show cost-effectiveness and significance both without doing a comparison to other sources? I know you say you don't have to do it, so how could they do it? I would be speculating because petitioners haven't presented a cost-effectiveness proposal here. What EPA said is that if there was a – They say they have because we've relied on your own judgment. So it's not that they haven't. You just say what they did isn't good enough, which EPA often does and is often entitled to do. But when you say something's not good enough, you're supposed to say what you needed to do. And so if you don't think they had to do these things that are listed here on 50, 51, and 52, then how are they supposed to know what they need to do other than – and they say we've done – you say they're done cost-effectiveness. When you say they're wrong, isn't EPA supposed to say with some specificity or some level of guidance exactly what they need to do to make it right? I think they should know. What EPA has to do is consider a proposal in front of it. So the proposal here, they did not make a cost-effectiveness proposal for non-EGU sources. Okay, but what about for EGU sources? Did they do one for EGU sources? For the EGU sources, for the 47 that do not have SCR, there is no cost-effectiveness proposal. We don't know if longer-term strategies would be cost-effective there on this record. And then you have the 83 sources. What do you mean exactly by cost-effective? That's done without comparison to other sources. So then what does it mean? You've already – you said you've made this sort of global conclusion about what installing, say, catalytic converters can do. How would it not be cost-effective at any given – at a specific plant has excessive emissions to put a catalytic converter on? Is the point that it won't reduce at some point? The EGU sources, there are two buckets. Bucket one, there's 83 sources that have catalytic controls already installed. I'm talking about the 47 that don't. Yeah, I'm talking about the 47 that don't have what the EPA has already said is a cost-effective solution to contributions. Now I understand your question. Yeah, it's based on the wrong premise. So EPA has not made any finding that installing catalytic controls is cost-effective. The update did not consider the installation of catalytic controls. What the update applied to was sources that already had catalytic controls, and EPA found that it was cost-effective to operate them as a fleet. EPA did not analyze the update. Do you distinguish between installing and turning them on? Yes, yes. Installing things doesn't do anything. You've got to turn them on. So I don't think anyone says installing is going to stop pollution. You've got to turn it on. I'm very – sorry, I'm not – you're more participatory than me, but I don't think for a minute that New York would be so silly as to say, we just want you to install them but not turn them on. So they're talking about the exact same cost-effectiveness, and that is turning them on. Except that there's Category 1, sources that have them and haven't turned them on. Category 2, sources that don't even have them. EPA did not analyze the possibility of installing and then turning on catalytic controls. So now I'm really missing something. Why on earth would EPA care whether people install things they aren't using? I'm not sure that's the right consideration. What EPA looked at in the update was do sources that already have catalytic controls, is it cost-effective for them to operate them? Operate. Yes. Okay. Yes, and that's what EPA found in the update. EPA said as a fleet, that's cost-effective. Operate is just that once you operate something, it creates cost. I take it that's where. Yes. In fact, I believe that was $1,400 per ton of NOx reduced. That's where the cost-effectiveness finding in the update was. There is no cost-effectiveness finding in the update for installing new controls for sources that don't already have them. But for the 83 that already have them in and aren't turning them on, what about those? What more are they supposed to show us for those 83? So for that 83, when we're breaking this down, what EPA said was we've already implemented a remedy for that. That's the cap-and-trade program. All right. Yes. So there were violations and they were remedied. And the remedy is working, and you haven't, just because a small handful of them were not turning them on, that did not in and of itself establish a new violation of the good neighbor provision because that's how cap-and-trade works. It sets a cap for the fleet. And that's being met. In fact, the states that are subject to the update budget are below the budget. That just sounds to me like a marriage ruling on a petition, not a burden problem as to those. They said, look, we're just saying the good neighbor, sorry, 126B doesn't apply if the sources you identified are engaged in cap-and-trade. Is that right? Is that what you're saying? Or what EPA said? The way EPA said it was I'm not saying you could never have a new good neighbor violation for sources that are already covered. What we're saying is you haven't given us anything new to make an additional finding. That's how EPA said it. So they sort of couched it as a burden for those sources, but I think you could look at it either way. And the new thing isn't that, hey, you're continuing to cause nonattainment? That's not a new thing? It's not new as to these specific sources. So what we don't have is the cause. We don't have the connection between, okay, these specific small handful of sources, if they do this on a source-specific basis, what results does that get you downwind? Let's say, for example, as Mr. Berman suggested, that certain coal plants are just running less. Let's say that some of them are retired because since this denial came out, several of these sources have retired. That is still a way to keep these sources under the cap. I'm sorry. I think I've taken up way too much of your time. No, that's okay. Let me just make sure my colleagues don't have any additional questions before we give you a chance to conclude your segment. Jordan, please, yeah, can you please wrap up and we'll give petitioners a rebuttal time. Yes, Your Honor. I would like to wrap up with one final point on the burden. I'm sorry. I don't mean to cut you off if you're not done with one of your responses, so please just finish your response and wrap up. I'm not trying to arbitrarily cut you off. No, I think I've sufficiently answered the questions about the 83 sources. I just want to wrap up with a response to a question that I believe Judge Millett asked of petitioners, which was about if 60 days is not enough in order to make the finding that's requested, or if 60 days plus the optional six-month extension, if that's not enough to make the finding requested. If EPA does not have a basis to support a finding of a good neighbor violation, then under 126, EPA has to deny it. There is a separate process for EPA to affirmatively assess all the upwind states that are potentially contributing to downwind nonattainment, and that's the FIT process. That's happening separately. It's on remand, and EPA is working on that. If there are no further questions, we would ask that the Court deny. I apologize. Can I just add one more? Didn't the update recognize that EGUs may need to actually do more reductions to comply with the good neighbor provision so that what the update was requiring might not be enough? If it did, did that not apply to the cap-and-trade ones as well as others not involved in cap-and-trade? This goes back to the three buckets. There are the non-EGU sources. EPA did in the update say those sources might need to do something. There are the 47 EGUs that have not installed catalytic controls or non-catalytic controls. It's possible that those sources will need to do something else. When the update said that I'm only talking about EGUs, may need to make more reductions to comply with the good neighbor provision, it was specific only as to those who have not yet installed catalytic converters, or was it that specific, or was it talking about the EGUs? There's the words that EPA used, and then there's what it meant. The words EPA used was reductions available after 2017. What EPA was talking about was things like installation of controls because that takes longer. The sources that are operating SCR, that's pretty much all the source can do. They were able to do it before 2017, and EPA said that as a fleet, they can do it before 2017, and that's what was already covered. So there is a category of EGUs that are potentially subject to more controls on the remand. And that would be the 47? Yes. And the 47, okay. All right, and that's part of what New York is targeting here, right? The 47? Yes. Okay, thank you. And the denial is that EPA doesn't have the basis to make the findings here. Got it. Okay. Thank you. Okay, so if there are no further questions, we would ask the court to deny the petition for review. Thank you, Ms. Spence. Thank you. Mr. Liu, we'll give you your rebuttal time. Thank you, Your Honors. I just have a couple of points to make. One is on EPA's argument that what is missing and the reason the states didn't meet their burden here was some failure to establish cost effectiveness. And there is a lot that is baked into that term, and it has become less clear, I think, over the course of this argument. I mean, EPA has said at this argument that a comparison is not required under its analysis of what the burden is. But the denial says otherwise. And I think the leftmost column on JA52, which begins with EPA disagrees with the commenter's assertion, goes into that in great detail. It says, for example, apportioning responsibility for emissions reductions across many sources in many states is part of the four-state framework. And then later in that paragraph says that the problem with the petition, with our petition, is that the absence of that comparative information means EPA cannot determine whether the sources we have identified have cost-effective emissions compared to other unnamed sources. So I think this highlights the difficulty that we have had in pinpointing what it is that is missing, assertively missing from our petition. But the concrete omission that EPA appears to have identified as a denial, but may not be insisting on anymore, is the absence of analysis across the entire region for unnamed sources. We have argued that that's not required by 126B because 126B does not require regional transport rulemaking. If EPA is now changing its mind about whether that is a reason to deny the petition, that itself is a basis for a remand here. Separately under cost-effectiveness, I think EPA has also acknowledged here that at least for non-EGU sources, there is not readily available information, maybe not even for EPA, on what is, in order to determine cost-effectiveness as EPA has understood it. And again, this highlights the unreasonableness of EPA relying upon this lack of information to deny relief to the petitioning state. If the information is not available even for them, to say that the state had to come with this information from out-of-state sources is an arbitrary and capricious basis. And then my very final point is on the EGU sources that were the subject of some discussion. Again, what I think the argument here has highlighted is the degree to which for the EGU sources, EPA continues to rely on the adequacy of the update for, as counsel said, providing a solution for the cross-state air pollution problem. But this court in Wisconsin rejected the update as being adequate there. It did not say that it was adequate or inadequate only as to some sources rather than others. What it said was the analysis EPA used to determine adequacy was improper. For EPA to come in and say that it was proper and was the basis for denying the petition sort of confirms the degree to which their reliance on that now-invalidated rule makes their denial here also unreasonable. Thank you. What is your answer to their argument that what was missing for the electric generating units was source-specific modeling? Well, I think our answer to that is that the update itself, up until the point it actually imposed controls, did the analysis that is needed for whatever burden is necessary here. For the EGU, to be clear, EPA identified in the update those sources as significantly contributing under the four-step cross-state framework. Later, it made a distinction between sources with FDRs and those without. But there is a significant contribution finding here. For EPA to now say that the states had to do something more to establish that finding is just incompatible with what the update itself found. And it is, as I believe you articulated earlier, that might be a merit-based ruling later to say that you need to do more at the control stage, for example. But it is not a failure for the states with that to satisfy their burden at the initial stage in order for EPA to conduct a further analysis. Okay, if my colleagues... I'm going to make sure my colleagues don't have any further questions. If not, we'll thank you for your argument. This morning, I'll counsel. We'll take the case under submission.