Oral argument — Samuel Masias v. EPA

Case number 16-1314, et al. Samuel Masias, et al. Petitioners vs. Environmental Protection Agency, et al. Mr. Yutaili, for Petitioner Samuel Masias, et al. Ms. Perfetto, for Petitioner Sierra Club. Mr. Lane, for Petitioner Kansas City Board of Public Utilities. Ms. Berman, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Ms. Langworthy, for Respondents. Good morning. May it please the Court, Robert E. Kiley, on behalf of the petitioners, which I'll refer to as the Colorado Springs residents. For areas other than Colorado Springs, EPA applied a relative standard for what data was acceptable to make a determination of attainment or non-attainment. They used a best available standard or a good enough standard, basically. That's consistent with this Court's previous rulings. For example, in Round 1 of the sulfur dioxide designations in the Treasure State case, the Court held that the monitoring data was robust enough, with enough being the key term, to generate reliable determination. Similarly, in the Catawba case, the Court held that best available information was what was required. You say, in your opening brief, Colorado Springs residents' argument is not that EPA's decision that the airport meteorological data is not representative is per se arbitrary. Rather, their argument is that EPA used a different standard for judging representativeness of meteorological data for Colorado Springs versus the four areas which EPA designated non-attainment. But your argument here seems to be going back to the argument that you say is not the Colorado Springs residents' argument, namely a claim against the EPA's finding well-known of non-representativeness of the airport data. You've really, at least to me, confused the matter. I don't think so. I think I was saying exactly that. The issue is clearly whether the data is representative, but it's relative to how EPA treated the four other areas. You say that's the only issue you're raising, but then you encountered through the government brief the problem that you had not raised that before the EPA. I believe that issue was raised. We cited to Modeler Gray, we quoted his comments in which he said that it's basically industry practice to accept nearby modeling as... He did not offer comparisons with specific places, did he? He did not, and I don't think the requirement is that the commenters identify every fact. I think the requirement is that issues be raised. I think that it's sufficient. The issue is whether the modeling is representative. Then we framed the issue as whether it was consistent with those four specific examples. It's true that the modeler did not provide those four specific examples, but I don't think the court has ever held that failure to identify every fact in the comments that you're going to use in the brief is waiving the issue. Did those four specific examples exist, or did you know the EPA's position on those four other jurisdictions at the time comments were being made, or was that only revealed when the final EPA decision? I guess it'd be fair to say if one had read all hundreds of counties that issue, they would have known EPA's preliminary position, but they wouldn't have known EPA's final position. I did want to reserve two minutes. It's fine. Thank you. Thank you. May it please the court. Lisa Perfetto for Petitioner Sierra Club. This case is about simple math with significant public health implications. In promulgating the sulfur dioxide air standard, EPA projected that national implementation would save up to 6,900 lives annually and prevent 54,000 asthma attacks. There's a very real cost to delaying attainment with the air standard, and this case is the court's opportunity to right EPA's wrong by ordering a designation of non-attainment for Gallia County. I'd like to make three main points. First, a basic mathematical fix to Ohio's modeling demonstrates non-attainment using even the most conservative adjustment possible. Second, that fix is consistent with the form of the air standard and shows that it's not necessary to rerun the modeling in order to determine that the area is in non-attainment. Finally, the slate dip in plant emissions during 2015 does not save the area from non-attainment. Regarding the mathematical fix, EPA approved Ohio's modeling with one exception. Ohio made an unapproved and unwarranted 38% discount. I'm not sure if my colleagues would agree with me, but I'm ready to assume you're right on the first two points. But your writing off the 2013-15 data as something EPA should not consider puzzles me in view of the statutory mandate that EPA make its decision on the basis of the available data. Sure, and the key fact here is that EPA simply didn't have a 2013-2015 modeling set that it considered reliable. It had Sierra Club set, but per EPA, that was unreliable. And significantly, Ohio had said that given the timeline by which EPA had to promulgate the designations, it simply wasn't going to be possible to prepare a 2015 emissions set to match what it had for 2012-2014. And so EPA was obligated to make its determination using the available information by July of 2016. I'm sorry, you're saying that the 2013-15 data had not been fully assessed? Yes, so what EPA had, it did have a 2013-2015 set from Sierra Club, but it found that unreliable. And then it basically just had the flat annual emissions from the two plants in 2015. It didn't have what it considered a reliable 2013-2015 modeling set. I mean, certainly if that were available, it would make sense for EPA to look at that, but the fact here is that it wasn't. And so, you know, in the world of emissions, new information, new data becomes available on a yearly basis, but you can't use the potential that some new data will arise to stall non-attainment findings. In your attack on the EPA's reliance on new data, you don't seem to stress doubts about the quality of the new data, but about the sort of much broader claim that the process will never end and no decision will ever be reached if EPA is allowed to rely on new data in reaching simply the conclusion that a particular area is unclassifiable. Yeah, it's both. I mean, one, if we waited for new information, we'd be waiting every year for more information to get the latest thing, and that would basically delay a final determination indefinitely and prevent the intended public health benefits from coming to pass. But it's also true that EPA didn't have the reliable 2013 to 2015 data, and that's why they needed to use what they had, which was a set from Ohio 2012 to 2014. It doesn't depend on – I would have to accept the idea that if clearly unreliable current data comes in, the EPA should not rely on that, even in declining to decide. But if plausibly reliable data comes in pointing to a particular direction, why is not simply a pause, classifying it as unclassifiable, a reasonable response? Because the statute requires that EPA acts, and specifically – requires that it act on the basis of available data, and it specifically gives it the option of unclassifiable. Yes, and unclassifiable is acceptable if and only if no data exists. No, no data. It's supposed to reach whatever conclusion is most reasonable on the basis of available data. Yeah, and the issue here is that there's simply not reliable 2015 data. To dig in slightly deeper, the reason that EPA is preferring Ohio's modeling, at least in part, is that it's using these specialized emission sets in which it had worked with American Electric Power, the then owner, kind of reassess what the continuous monitors had shown and make corrections where there had been errors. And that's part of why EPA likes the Ohio set more than the Sierra Club set, and that data set does not exist. And so it's impossible, really, to tell without having – without Ohio doing that set, whether Ohio's 2013 to 2015 set would be lower than its 2012 to 2014 set. Regarding – EPA must have some settled definition of what accounts as available data. It has to be available, reliable data. I assume that's implicit in the phrase available data. Yes. And so do they have a rule out there about when – whether this 2013 to 2014 information met their standards for what should be relied on? If you had come to them with that data, would they have said, that's good, we'll use it? Or would it have failed their test for available, reliable data? Well, EPA has conceded that, given the timeline for these designations, that modeling 2012 to 2014 is consistent with its guidance. And so it's not requiring the 2013 to 2015 data. And so here – I mean, what they have is simply unreliable, and it can't be used to sidestep a non-attainment finding fully supported by the data it did have, which shows non-attainment in all circumstances. One of the questions I just had generally was that, in this case, the EPA is to – Gallia County? Am I saying that right? Gallia County, yes. Said, I don't like your data, I don't like your data, I don't like your data. I think there might be some different numbers out there coming, so we're not going to decide anything. And what I was curious about is I didn't see – it wasn't clear – what is – is EPA's role that passive in this process? I get that it takes information from the states because it has a cooperative role with the states. Does EPA have an independent duty if there's data there to process it itself? Or is this whole process more passive, where if nobody gives it the good information, it just can decide something's unclassifiable? Yeah, the statute is clear that it's EPA that has the obligation to make the area designations. And, in fact, it has to do so even if the states fail to suggest any designation. And here, what they've done is essentially to throw up their hands in the face of a fix that shows that the information before them was enough to show nonattainment. Certainly, in other instances, they've gone far beyond that. They've done their own modeling. They've revised and rerun state modeling. And, to be clear, we're not saying they needed to do that. In fact, we're arguing quite the opposite. That is, the information that they had, they could take a look at it and say, well, it's showing that the Ohio modeling is showing that it's at least 197, which is over the nonattainment value. It may be significantly more, but that doesn't matter because the designation question is a basic yes-no. If it's over the standard, it's nonattainment, and that's true whether it's one microgram over or 200. Well, EPA says it would be really, really, really, really hard to do that math fix that you propose. And that's simply not true. You might need to rerun the modeling if you wanted to know precisely by how much Ohio's modeling is showing nonattainment, but that's not what we need to do for the nonattainment designation. We just need to know whether, at minimum, it's showing nonattainment. And here what we do is we take the very lowest background that Ohio used and we factor out that 38% discount using that, and so that's yielding the minimum increase. It may well be more than that. In fact, it's exceedingly conservative to assume that that minimum background was happening at all hours in the three-year period, but we're showing that even if that were true and that raises all of the numbers averaged by 1.6, we're still in nonattainment. I see I'm over time. Thank you very much. I'm sorry. Your Honor, did you have a question? No. Thank you. Dennis Lane. Good morning. Dennis Lane for the Kansas City Board of Public Utilities, which we affectionately call BPU, so I may slip into that. We challenge the EPA's ruling that it could not determine that the max was met for Wyandotte County. Just to give you some background, when Kansas came in, they found that Wyandotte County was what's called unclassifiable slash attainment. And EPA said, well, that's great, but you didn't look at Missouri, and we want you to look at Missouri because BPU's plant, Nearman, might have some effect in Missouri, particularly in Jackson County, Missouri. So BPU, because there was a public comment period, remodeled and addressed EPA's questions, particularly they... Originally, they had just done Wyandotte County when they had looked at the receptors, but they expanded it into Missouri and up to the north. And they looked at Jackson County in particular, and what BPU found, really the Trinity report found, was that in 2016, because the Veolia plant... Before you get any further into the details, could you address EPA's argument that your client likes standing? Sure, Your Honor. Yeah. So, I think... And their argument is that there's no practical difference between unclassifiable attainment and unclassifiable. Right. There's no practical difference between the two of them, so your client is not harmed by this designation. Right, Your Honor. I just want to point out one thing preliminarily, and then I'll get to your question. If you look at JA-24 and JA-28, there's footnotes on both those pages. It's the same footnote. This is the EPA guidance to the regions, and what it says in the footnotes... I better read it because I don't want to be off on this. And I'm reading from footnote 7 on JA-24. While states have and may continue to submit designation recommendations identifying areas as attainment, the EPA expects to continue the traditional approach, where appropriate, of using the designation category unclassifiable slash attainment. So, to start off your question, Judge Tatel, EPA seems to think that attainment and unclassifiable slash attainment is the same thing. What EPA is saying, to answer your question, is that if you're either in unclassifiable or attainment, what you have to do is prevent... I know it's PSB. It's either prevent serious or significant deterioration. So, yes, you would have to do that in either situation so things don't go bad. Obviously, if you're in non-attainment, you're going to have to do more than that. But I don't see that as determinative of standing. If we're in the unclassified... So, if we're in unclassifiable slash attainment, that, based on the footnote, means that we're essentially in attainment. And since BPU just put in $240 million worth of controls on Newman, they can say to themselves, Hey, okay, things are good. We're in attainment. All we have to do is make sure we continue that. We don't... But it is the case that under either of these... in either of these categories, the EPA is free at any time, on its own motion, to re-examine classification. Absolutely. Absolutely, Your Honor. So, does that mean no one has standing? Because EPA can always redesignate? Most of it would be. At stake is whether the difference between these two categories can give anyone standing. I don't... Yes, the difference between these two categories is unclassifiable, so you don't know what you have to do. You don't know whether it's non-attainment or it's attainment. And I'm sure you know, Judge Williams, that utilities plan for the long haul. You know, the amounts of money are very high. They have to plan, and if they're not sure that this is attainment, then they have to worry about, okay, are we going to have to re-engineer Newman? Are we going to have to add new controls? I do understand the need for long-term planning and the problems that it poses, but unless there's a real-world difference in the susceptibility to change, it's still hard for me to see a basis for standing. I'm sorry. I think if you consider the footnote that I just read, and you say that unclassifiable slash attainment basically is EPA saying you've met the NACs, which is the same thing as attainment, that is a different resolution from unclassifiable, which says we can't... How is it practically different? If they just do unclassifiable, does that mean that they come back and check again the next year, that you have any extra reporting obligations, that there's extra paperwork you have to do if you're going to expand the plant, or for all regulatory purposes, are you treated the same as attainment? Which is it? You know, Judge Millett, I guess the question is on which side are you on. If you're on EPA's side, they're going to say, well, you don't have to do anything. I'm asking you from your side. I'm going to tell you, on our side, you're going to be worried. Am I going to have to do something? Yes, am I going to have to remodel? Do I have to file a new application to shift from unclassifiable to attainment? Might I have to re-engineer the plant or add new controls? Those are different. Either way, you have to make sure you're not having any substantial deterioration in air quality. So even if you're in long-term planning, you know you can't make the air worse? I'm sorry. Go ahead. Under any designation, you have to file a PSP. So I don't see that as, to me, that's your minimum level of, you're going to have to meet that no matter what. And you don't have any obligation to make things better when you're in the unclassified. Like you would if you were non-attainment. That's correct, but that's why I'm saying, you're correct. What I'm saying is, if you're unclassifiable, the injury is you've lost the benefit of attainment. What's that benefit of attainment? The benefit of attainment is that the way you've, because we've re-engineered Murren, the way we've re-engineered it and we've put on air controls, that's meeting the NACs. And so we don't have to plan to do anything else. With unclassifiable... But if you engineer now to meet the NACs... Pardon me? If you engineer now to meet the NACs, you're in perfectly fine shape regardless of what your status is, as long as you're engineering to meet the NACs. Won't you be okay? No, because unclassifiable creates uncertainty for you as to whether what you've done... BPU thinks it's met the NACs. My whole merits argument was that we showed that we met the NACs not only in Wyandotte County, which nobody cares, nobody disputes, but also in Jackson County, Missouri. But now we don't know whether we've met the NACs, so we have to prepare for that possibility that they would find it's non-attainment. Could you have that same problem if you were in attainment? Would you just weren't sure how the measurements coming in from Missouri or elsewhere were going to happen two years from now, or what the standards for NACs might be three years from now? Would you not have that same problem? That would be a different case, and we would be here probably arguing that the NACs shouldn't be lower than 75 ppb, which is at the present level. And if they found we were in non-attainment, then we would come back if we thought they were wrong. Here, if you look at JA-243, you'll see a table. We're not close. It isn't a question like the last one, are we one micron over? I mean... Counsel? Yes, I'm sorry. Was there any empirical data on the frequency of EPA moves to change, on the one hand, an unclassifiable designation, and on the other hand, an unclassifiable slash non-attainment? No, it's unclassifiable attainment. Attainment, I'm sorry, attainment. Those are the two issues, the two categories we're talking about. Are there any data on frequency of sending those? If there are, I don't know. I don't know, then. Okay, thank you. Thank you, Your Honor. We'll hear from, let's see, EPA, right? Yes. Okay. Good morning, Your Honors. Amanda Berman on behalf of Respondent EPA. With me at counsel's table is Andrea Criller from the EPA. Petitioners challenged three out of 61 area designations that EPA made in July of 2016 in regard to the primary national ambient air quality standard for sulfur dioxide. In each of these three instances, EPA concluded that the available data did not show at that particular time that petitioner's preferred designation was appropriate, and so as Clean Air Act Section 7407D instructs, it designated the area unclassifiable. I'd like to briefly highlight the key reasons that each of those three decisions was reasonable, starting with the Colorado Springs area. Petitioners rightly assert that EPA regularly makes its determination decisions based on the best available data, but for the data to be the best available, it has to be good enough. Here, EPA reasonably concluded, and its analysis on this point can be seen in JA page 518 in the response to comments document, that the Colorado Springs Airport meteorological data wasn't good enough. It wasn't representative enough of conditions at the Drake facility to form the basis for a nonattainment designation. I just want to ask you, and this may be an overarching question, the same one I asked for the Sierra Club, and that is, I see a lot of EPA going, no, we don't like that, no, we don't like that, but making no apparent efforts to take what's before it and actually proactively get the information. You had, I think, Drake's coal-powered plant, correct? Or had been? In regard to the Sierra Club issue? No, no, just, sorry. Is the Drake plant coal-burning? I believe so. Okay. All right. How come there's no measurements? You're getting reports, presumably you're getting emissions from them. Why do you just get to throw up your hands saying we don't like your meteorological data and we're not going to make any effort to see based on what Drake, you know, Drake must be filing reports. Monitor somewhere else there. You've got a coal plant. Colorado Springs is not that big an area. You've got the Rockies blocking the air in. It should have been possible to get this information and we know EPA had a really, really, really, really long time to get this data. So how come it's okay now just to say I don't like your data? So I think we have to back up a second judgment. So this set of designations were designations for an area where EPA didn't have monitoring data that showed it clear non-attainment for the first round of designations. These are designations where EPA knew it was going to have to rely on modeling. And to have modeling that's accurate, you have to have several key inputs. You've got to get a representative set of meteorological data in addition to the emissions data. Now EPA has a lot of its own emissions data in addition to what states submit. It has the CAMD, which is an air monitoring database that shows emissions from plants. But here in regard to this area, the Colorado Springs area, what it discussed at length in responding to comments was that it didn't have meteorological data that was good. There was the airport data, there was a highway set that nobody argues should have been used because we lacked the right inputs for that. And to have modeling that's accurate, we've got to have the right meteorological data. Because I don't understand how the federal government cannot have reliable meteorological data when there's not only an airport there, but you've got the Air Force Academy flying all over the place there. They've got to have actually really good meteorological data in Colorado Springs. The federal government has to. And so how can it just be that we don't like what you presented and so we're saying we can't have any? That's what I'm not understanding about. I can't answer why there's not a station at the Air Force Academy. Traditionally, EPA has looked at nearby airports where it does have access to the data and it looks at are those good enough. If it doesn't have that, then the state and EPA can work to get an on-site meteorological station as has now been done in regard to this area. The problem is that we didn't have enough data from that on-site meteorological station at the time of designation to use that on-site data. And when we compared that on-site data to what we had from the Colorado Springs Airport, we saw some key differences. And this is where that area really was different from the other four non-attainment areas that petitioners point to. As you can see in the diagram that we included on page 22 of our brief, it's larger on page 59 of the Joint Appendix, these are very different areas topologically, topographically, sorry, to the immediate west of the Drake Plant. We've got the Rocky Mountains. We've got Pikes Peak to the immediate left, a very steep change in elevation very close. The airport is in a flatter area where you've got a change in elevation that's further away to the north. And so what EPA saw when it compared the wind roads was it had very different winds at the two areas. And that's critical when you're trying to determine if meteorology is representative because you're trying to discern how the plume, the pollutant plume, is going to mix, how it's going to move, and where it's going to settle. And so the meteorology really matters in that way. Now, petitioners argue that EPA has failed to articulate an objective standard, but that's because this is necessarily a qualitative multi-factor analysis of the very type that this Court approved of in the Catawba County case. EPA balances factors including distance from the meteorological station to the airport, wind speed, and direction. And so as this Court said in the Mississippi Commission on Environmental Quality case, EPA... It's also understanding why EPA is part of this whole process that itself was engaged and didn't have some obligation to get its own data instead of, again, waiting for someone to give data and say, no, no good, and then starting the process of collecting data. Well, we understand your frustration there, Your Honor. And EPA noted that this was going to be an issue when it promulgated the 2010 NAAQS. What it's done to fix that issue is it then promulgated the 2015 data requirements rule, which under that rule, states are now going to have to submit monitoring or modeling for all areas, which the statute didn't require that, prior regs didn't require that. So EPA recognized there was a hole, and it is fixing it moving forward. But, you know, under the consent decree that resulted in making these designations at this time, EPA had to work with the information it had, and it didn't have a representative meteorological stat for the Colorado Springs airports. So you get the data there or in Wyandotte, Kansas, or somewhere. Does the EPA have a practice or rules governing how it responds to... Imagine hypothetically the new data showed a different status, either pushing it into nonattainment or pushing it into attainment. Does EPA have a tradition of reacting promptly to that? Or does... How long do you wait before you do something with this new information? So under Section 7407D... C-3, I believe, the administrator or the state can re-initiate the designations process at any time. So if EPA saw that there was a big discrepancy between a current designation and the data that's coming in, it could use that process to do that, or somebody could petition EPA to use that process to do that. I do want to briefly address the waiver issue that we believe is present in regard to this Colorado Springs designation. In their reply brief, petitioners cite to a reference and a comment that EPA sometimes uses airport data from airports that are even further away. We don't think that that was enough to raise the question of disparate treatment, that EPA's treated this area differently than other areas. EPA explained why it considered other sets of airport data representatives at the proposal stage in regard to those four designations that petitioners point to. We identify the relevant joint appendix sites in footnotes 22 and 23 of our brief. So those designations with the... with the representative airport data from those other airports were pointed to by EPA at the proposal stage. They could have raised this issue, then they didn't. Under this court's decision in National Association of Clean Air Agencies from 2007, an objection has to be prominent and clear enough to put the agency on notice. We don't think that was the case here. Moving on to the Gallia County area. This was a close call, and we think EPA reasonably concluded it needed better data to figure out which side of the NAAQS the area falls on. Now, petitioners argue that EPA shouldn't get deference here because this is basic math. Do you need better data, or do you need to just rerun the data? Well, Your Honor, our position on whether EPA should have rerun the data is that it's not that simple. I tried to... It might be simple, it might be complicated math, but you started with we need new data, and their position is no. You've got the data, you've got it for the years you identified that you want submissions based on, and maybe it's a complicated math program, but we've got great computers these days. Why can't you rerun it? Tell me exactly how many days and how much resources it would take. Just doing their most conservative modeling. You don't have to get precise. You just have to see if they cross the nonattainment line. Do you have any specifics on that? Well, here's the specifics I can give you. So for EPA to do that, it would have had to recalculate the hourly concentration at each of the 34,200 receptors in the model. That's the data points in the model. So that's close to a total of 300 million data points for each of the three years. Are you taking into account the argument of the petitioners that all they're asking is for an increase, which would be minimal, minimal increase of the minimal background concentration, taking it from 1 to 1.61? So what we think their math shows, what their math shows is that if EPA redid the modeling, and we don't think that's an easy fix, we're not even sure we could do it. It depends on which modeling files and what set of raw data we got with the modeling. And again, that was in April of 2016, round two of the modeling, after EPA had already asked for more data, three months before the designations. But taking just their math, the reason we don't think it's good enough to support a nonattainment designation is because at best all it shows is that you're likely to end up right over the next. So the math they do isn't a specific, it doesn't point to a specific design value. I'm sorry, what do you mean by that? What do you mean by it doesn't point, you said it shows they're going to likely be over the next, and this is estimation, this is standards, and so that's the test. We tried to explain this in footnote 35 of our brief. So because if you back out the background concentration, the incorrect one that the state used, if you take out that value, first you have to segregate out the background concentration, figure out what the right number is, and put that back in. And that, if you do that for each data point, for each of those 300 million data points I just referenced, that could easily change what was the fourth highest data point each year. So their math doesn't show what the resulting design value would be. It doesn't tell us what the new average over three years is. What it shows us is that it's likely to end up right over the next, and we agree with that proposition. But the critical thing here from EPA's perspective is that the data before the agency also included actual emissions from a more recent three-year period, from 2013 to 2015, that were significantly lower than those used in the state's modeling, the same modeling we're talking about here. But that's changing the rules of the game after the fact. You said give us this time period. They gave you that time period. Someone came forward with the data. You said you've got this discount in there. That's no good. But you've got the data. And you're saying, yeah, it would be hard. It would probably require a computer, maybe some Excel spreadsheets. But I still haven't heard that we couldn't do it, and that, quite frankly, with computers today, we couldn't do it within a few days or a week. We have the data to make the decision. We just don't want to do it. Your Honor, my understanding is that, as I stand here today, the agency's still not sure it could do it based on the raw data files that was provided with the model. When somebody gives EPA modeling, it doesn't just hand them a program that they can push the button and redo it. It gives them a set of inputs. Sometimes those inputs are done in a way that you can back out the background concentrations. Sometimes they're not. Could you tell Ohio to redo it, get rid of your discount, and redo it? Well, Your Honor, we had already done that once. This time, round 2, we're in April 2016. We're 3 months away from the final designations. I think it's reasonable for EPA to say at this point, okay, you know, the state again gave us flawed modeling. We can tell, and we agree with petitioners, that we can... Would it have taken more than 3 months for EPA to have done the math? Your Honor, what I'm saying is I'm not sure we even had the raw data that would allow us to do the math. Would it have taken Ohio more than 3 months to do this? I can't imagine. I don't know, Your Honor. I don't know whether it would have. But what we think it's reasonable for EPA to say... How long time did you... You rejected their first model, sent it back, and then they sent you this one. What was the time difference between those 2 models? The first model was from 2015. I don't remember exactly when in 2015. This was in April 19, 2016, the absolute last day on which we said that Ohio could submit another model. So we really didn't have time to go back for round 3. But in any event, what EPA said is, okay, we're looking at this modeling, and we can see you might end up right over the next, but at the same time, we do have the 2015 actual emissions. And we go through those in Joint Appendix page 608. And that is data that's also before the agency. EPA had that later 3-year set. Did that later 3-year set tell you they were at attainment, or did you just go, oh, that's more data that raises a question mark? What we said is it pushes in the opposite direction. So if we've got this one flaw in the state's modeling that suggests that you might end up pushed a little bit over the line towards not attainment, we've got another issue with the modeling in that we think they used overly high emissions that pushes you back in the other direction. We don't think it was overly high emissions. You think something has changed in the operation of the plant. You're not challenging the accuracy of the data for the years it captured, are you? I thought you thought there was a change in the nature of the plant's operation. No, here we just, we had a later set of emissions, and there's a table on Joint Appendix page 608 that shows you the change in emissions each year. And even for some of the earlier years, EPA was already in that table substituting out the emissions Ohio gave it for its own because it thought Ohio's were off. So again, EPA's looking at how do we reconcile all these moving parts in the modeling the state gave us? And we think that with different data suggesting different conclusions, it was reasonable for EPA to decide that the most appropriate designation at that time was unclassifiable. As this court said in the city of Waukesha case in 2003, the resolution of contradictory data lies well within EPA's expertise. I'd like to move on to Wyandotte, Kansas briefly. Can you answer my question? Can you remind me what that was? Are there any data on the frequency of EPA taking the initiative to change unclassifiable as opposed to changing unclassifiable slash attainment? We don't have any data on that, but we don't think those data would matter anyway because as a matter of law, it's a blank slate every time. An unclassifiable designation doesn't signify one way or the other anything for the next time around. And the same is true for an attainment designation. So if we have that data and it showed that statistically unclassifiable designations move more often to non-attainment than it would likely be because from our view, the data they're dealing with was already in the middle. It was already borderline, whereas the data supporting an attainment designation is likely to have already been well in the other direction. But in any event, as this Court explained in its memorandum opinion in the Catawba County case... Wait, wait, wait. You said if they're in attainment they're likely to be well in the other direction? Couldn't someone just be over the line on attainment? They could be, but I'm saying that if there were data that showed it statistically more often, unclassifiable areas are designated non-attainment next time around, we don't think that's because the unclassifiable designation itself has that meaning. That's simply because the data are already in the middle where that's less likely to be true in regard to an attainment area. But, you know, we think this is governed by the Court's memorandum... So they really are in a less stable position then? So, I mean, I think the whole point of unclassifiable is actually not that their data's in the middle, it's that we just don't have reliable data to make a decision. So here, you know, moving to the data that we do have, EPA has long been clear that if you're going to use an emissions limit instead of actual... So, unclassifiable, does that mean there's actually data in the middle pointing both directions, or does that mean we don't have sufficient data to make a determination at this point? It can mean either of those things. The statute defines it as a limit of available information. Sometimes that's because nobody submitted any information, sometimes that's because we've got submissions that have internal contradictions, or we've got multiple submissions that point different directions. So, here, you know, the limit they point to, the emissions limit that they wanted EPA to use was not even enforceable at the state level until December of 2016, six months after the in, again, sorry to overuse it, the Catawba County case, this time in its reported opinion, the court agreed that emissions reductions that had been committed to, but were not yet in effect, were too speculative to form the basis for an attainment designation. We think it should reach the same conclusion here. And I see that my time is well over. If the court has any other questions, I'd be happy to answer them. Thank you very much. Let's see. We will hear from the intervener. Mr. Langworthy. May it please the court, I am Lucinda Langworthy, and I represent a respondent intervener. I think you said... Sorry. You're right. The Utility and Regulatory Group. I call it UR, or the group for short. Yes. In your order on Friday, you asked that I be prepared to address the group's standings. Let me start with that. We believe that UR standing is evident from the record and the briefs in this case. This is a rule about... You asked if Drake was a coal-fired power plant. All the designations in this rule are of areas surrounding coal-fired power plants. That's how the areas were selected for designation. UR is an association of electric utility generating companies. Most of the UR members own coal-fired power plants. Do they own them in these regions that are at issue here? Pardon me? Have you put into the record that you have members in the regions that are at issue here? We did not put that in the record. Keep in mind, when we filed our motion to intervene, we didn't know where the areas were, but I can tell you that we do have members in the areas at issue here. Specifically... All three? We are not discussing the Wyandotte County. We do have members... The Ohio and Colorado ones. We do. I actually developed and I'm prepared to tell you about the members in Gallia. If you want me to identify the member in Colorado Springs, I'll have to go back and do that in writing. But in Gallia County, there are two large power plants. One is Gavin. One is Kyger Creek. We have... The Ohio Valley Electric Cooperative and American Electric Power have ownership interests in Kyger Creek. Kyger Creek is part of the modeling of Gallia County. And in fact, Sierra Club said in their brief that they were seeking a designation of Gallia County to bring about emission reductions from Kyger Creek. So I think it's quite clear that we have standing there. We felt it was evident from the record and it was granted intervention. We are the coal-fired power plant industry. Yes, we do have specific members. And again, if you want... It might be good if you just put it into the record. Do you want a submission after this? It makes sense for you to have standing established in the record. But quickly. Yeah, no, I can get that. Do I need to address If you want to dispute it. Rules tell you to show your standing up front. Well, it's also true first of all that we didn't know what power plants were going to be the target when we moved to intervene. And the rules do say, the court, the case law does say that if standing is evident, you don't have to submit more to the record. But I'm happy to provide more in the record. Let me very briefly, if I may, touch on two additional points and I'll the Gallia County right now. First of all, Sierra Club didn't argue that EPA should fix the Ohio modeling until after, six months after, in fact, the final rule was signed. They first raised that issue in a petition for reconsideration, which EPA has granted. And of course they can challenge the Why do you call what the EPA did granting that petition? That petition said you got the numbers right now, run them right now, and it will show non-attainment. And their response to the petition for reconsideration said, we're going to get some more data and do something later. In my mind, I cannot conceive of how that is granting a petition that says you have the numbers now, run it now. I guess my point is not whether that is granting the petition for reconsideration and Sierra Club could challenge that separately. My point is they didn't ask EPA during the comment period in the record to do the fix that they're now asking for. They've asked them after the record closed, after the rule was final. Yes. But the question is whether it's a functional the exhausting question is whether it's a functional denial of their petition. And what I'm trying to understand is how it is not a functional denial of their petition to say we're going to get more data. And if it were a functional denial of that petition, they could have challenged the denial of that petition. They didn't do that. That's not in this case. The other point I just wanted to make, you heard about the sensitivity of modeling. Ohio, when they submitted the modeling that Sierra Club would like to see fixed, they said this modeling cannot be used to demonstrate nonattainment because we know the model predicts too high when we can't correct for low wind conditions. So they said you can't just use this modeling for nonattainment. There's a problem. The modeling is sensitive to the inputs and EPA understands that and legitimately doesn't choose to rely on it. Thank you. So who had how much time left? Okay. You can each take one minute in the order you originally appeared, if you would like. So, Mr. Ukaley. Thank you. EPA does have authority to gather on data. For example, Clean Air Act Section 114 provides them with authority. Or most ambient monitoring stations also have meteorological stations. So they could have gathered it that way. But that's kind of a red herring because there was the on-site meteorological data and EPA actually claimed in their arguments that there were key differences, but they never explained what those key differences are and there actually aren't. What's in the record is modeling from Maureen Barrett and Dr. Gray that shows that they got the same results using the on-site meteorological data as well as the airport data. The violating receptor was the exact same one that's out of thousands of receptors. So that shows that the difference in meteorological data was not impacting this and that's because this is a short-term emission or ambient standard. So most of the meteorological data is not relevant. Thank you. Council for EPA says that we do not point to a specific value, but we do and that value is what they're saying is that the value of the non-attainment is at least 197. In its brief, EPA provides an example and a footnote about why the non-attainment value may well be more than 197, but they do not and cannot show that it would ever be less. More generally, the Clean Air Act scheme only works if EPA has some minimum obligation to make designations that make sense. Here, they're potentially opening up a huge loophole to states seeking to avoid non-attainment obligations. They're essentially saying in their argument that states, if they want to avoid non-attainment, do some modeling. If it's showing non-attainment, just cut it by some number, submit that information to EPA. EPA will throw up its hands and reward the states with an unclassifiable designation, which is treated precisely the same as attainment. Thank you. Thank you. And Mr. Lane. Judge Millett, I misspoke. In the answer to one of your questions, I said that there was a table showing what the new numbers were. That's actually at JA 393. I think I told you 243, so I just want to correct that. That goes to some of your questions, Judge Millett. This is the Trinity report, which BPU presented to EPA. EPA issued a public notice on March 1, and we filed this report on March 31. That was the comment date. This redid all the numbers based on what EPA told us were the problems that they found with our earlier analysis. We ran the numbers, as you said. I didn't do it. I don't know how they did it, but they have computers, and they figured it all out. If you look at this table, you will see that the highest value is 129 at the bottom of the table. That compares to 196, which, again, I don't know how, but that's the next level. It isn't... You asked a question. When you're unclassifiable, are you really close or not? I don't think 129 is really close to 196. Thank you. Thank you very much. The case is submitted. Stand, please.