Oral argument — Natl' Assoc. for Surface Finis v. EPA

K-12-1459 et al. National Association for Surface Finishing Petitioner v. Environmental Protection Agency et al. Mr. Salk for Petitioner AASF Ms. Chu for Petitioner's Clean Air Council et al. Mr. Weiner for State MEP Ms. Peoples for Respondent Good morning. I'm Jerry Salk for the National Association for Surface Finishing. And before I, I'd like to start with kind of an overarching big picture issue about this rulemaking that informs each of the legal issues that we've briefed. The 112-T6 issue, the 112-F2 issue, and the so-called non-PFAS issue. And the big picture is this, that in the 2010 proposal that EPA made on this proposal to revise air emission limits for chromium plating facilities, EPA estimated at that time that under the existing standards that were promulgated in 1995, 5,257 pounds of chromium were emitted nationwide. But EPA kept the record open because of some questions about the data that led to that estimate. In the final rule, having gathered additional data, revised its estimates, EPA estimated that the emission of chromium nationwide from the affected facilities under the existing standards was down to 956 pounds nationwide. It was not a change in performance. It was a revision in the estimate of how much chromium was being emitted. And that's what drove EPA to revise the limits downward. EPA realized at that time, based on the 956 estimate, and then a corresponding assessment of the risk that it was required to make under the benzene standard, so-called benzene test. EPA realized that many of these facilities were already performing at a level below the existing standard. And EPA decided to lower the standard on that basis and require all of the facilities. Are they not allowed to do that under D6 when there have been developments? When there's been progress? They are not allowed to do that on that basis. And I will turn to the D6 issue because there are no developments in the rulemaking record that EPA identified. But I wanted the court to understand because, as I said, this informs each of the issues, which I'll turn to now. But the estimate of actual emissions was greatly reduced so that EPA realized that many facilities were operating below the existing limit. They decided to lower it on that basis. They can't do that unless they – they can't revise the limit unless they satisfy D6 and satisfy F2. Now, your argument there is the absence of a development. That's exactly right. But the question is why isn't additional information about what is actually happening itself a development? The statute requires a development in practices, processes, or control technologies. That's explicit in the statute. And this court's precedent says that that's a core requirement of D6, that EPA identify a practice – a development in a practice, process, or control technology. Why isn't that a development in a practice? As a practical matter, the emitters are emitting a lot less. Yes. There's nothing in this record that shows an improvement in performance. The estimate was reduced. The performance of these facilities remained the same so far as this record shows. The same as in 95 because isn't the whole benchmark for this? They're looking in 2010 and in 2012 at 95 as the original back floor, right? They're looking at the performance of the facilities, yes, under the 95 standard. And you're saying there's no development in practice, process, or technology. EPA did not identify in the rule. Didn't they identify some add-on controls? The add-on controls that were identified were considered in the 95 MAC rulemaking. So they weren't new, is your point? Exactly. They were considered then. They were considered again in this rulemaking. Improved performance of existing technology? It can be, but there's no evidence in this rulemaking of improved performance. That's why I started with the big picture. What happens here is that the estimates of actual performance were greatly reduced. There is nothing in this record that shows a development. That's the core requirement, as this court has said. Just let me get that clear. So it's your position that the low level of emissions that they observed between 2010 and 2012 was actually the same as the level of emissions in 95. All along they were performing much, much better than anyone realized. It appears to be the result of the implementation of control technologies in response to the 95 MAC. There is certainly nothing in this record that EPA has pointed to that shows any improvement in performance, much less any development in control technology. As I said, EPA understands that that's a requirement. They said in the 2010 proposal that their technology review is focused on identifying and evaluating developments. And revealingly in their reply brief, excuse me, in the brief here, in EPA's brief here, they did purport to identify a development. They argued about improvements in mesh head performance. And as we responded in our reply brief, the record does not support that there is any improvement in the performance of mesh pad controls. And we cited some data on page 5 of our reply brief that showed that back in the 1990s, there were emissions levels at facilities controlled by mesh pads lower than what EPA cites in its brief as the emissions levels now under the so-called improved mesh pads. So your position is that if what industry is doing is implementing the control technologies required by the 1995 rule and therefore achieving reduced emissions, that that is not the kind of practice or process that would trigger further review? It's not a development. And there's nothing in this record. There's something in the EPA brief, an argument about a development, but we showed in our reply brief that the numbers don't show any improvement in performance of that. So under D6, they could not use the reduced actual emissions that they discovered because they revised their estimates. They couldn't use that under D6. That's not a finding of the development. There is no development. And now under F2, EPA, in order to enact these or adopt these lower emissions levels, was required to find that the lower levels were required to provide an ample margin of safety. In the 2010 proposal, this is a backstop to the MAP process. The MAPs are enacted in 95. F2 says as of eight years later, EPA has to look at this and see if a further reduction is needed for an ample margin of safety. In the 2010 proposal at JA26, EPA specifically found, quote, the current MAP standard provides an ample margin of safety to protect public health. Thus, we are proposing to re-adopt the existing MAP standard to satisfy 112F2 of the Clean Air Act. That's in the 2010 proposal, specific finding, the MAP provided an ample margin of safety. Subsequently, as I said at the outset, EPA reduced its estimate of actual emissions 80%. And in the final rule, it never found, and it couldn't find, on that record, couldn't find that the lower standard was required to provide an ample margin. It's obvious it provided an ample margin. I mean, the ample margin of safety test, or they would call it a test, is pretty loose. It hurts me to see why it shouldn't be the current facts as to costs and risks that are used to apply that test. And the current facts as of 2012 were understandings that a much lower level could be achieved moderately readily. Is that true? Well, it was being achieved, Your Honor, on the basis of the same MAP standards that were enacted in 1995 that led to the conclusion in 2010 that there was an ample margin of safety. And so, it seems obvious to me. Logic tells you that the lower standards provide an ample margin of safety. On the basis of the earlier data, yes. But is it an ample margin of safety on the basis of the current facts? Well, for much lower emissions, and correspondingly, EPA made findings and conclusions of much lower incidences of cancer, much lower risk levels. Mr. Stock, isn't part of what EPA is considering in doing an F2 analysis also cost? And so, it might say, well, given cost, you know, X is an ample margin of safety. And then if they think, well, the costs are less than we thought, and so let's capture more safety now with further consideration of low cost. It's really cheap to do a safer increment. That seems to be part of their position. Our position is that they didn't satisfy the legal requirement of showing that the new lower level was required to provide an ample margin of safety. But in addition, as you can see in addition... What if that weren't? What if we disagreed with you that they have to make a threshold necessity showing? Then let me turn to the next point, which is their analysis of residual risk was completely defective. Let me make sure I understand your point. Is your point that if there is an existing ample margin of safety, that there's no authority for the EPA to then lower standards to get a super ample margin of safety? That doesn't talk about a more ample margin of safety. At a minimum, they have to... So you're saying they're powerless to act if there's an existing ample margin of safety. They can't... As Judge Pillar was saying, what if they discover that it's a lot cheaper to do this than we ever thought before? So let's build even a greater margin of safety. That's not our record number one. I don't know that I'd say they're powerless. At a minimum, they would have to explain why the earlier finding and the existing limits do not provide an ample margin of safety before they have to pay a lower one. There is an existing ample margin of safety. But they've discovered that they can get even more bang for the buck. Are they without authority to do that? I'm not saying they're without authority. I'm saying they didn't do that here. I'm trying to understand the statute, and I want you to help me to do that. Would they have authority to do that? They have to show that the lower limits are required to provide an ample margin of safety, and they haven't done it here. If they show that the new standards are required to provide an ample margin of safety, then they can lower the limits. Just to be clear, I think what we're all trying to get at is why isn't that requirement evaluated in the light of all the data currently before the EPA? Yes, Your Honor, let me get to that, because we also argue that their assessment, their analysis, their factual analysis, if you will, technical analysis of residual risk is completely defective because my client submitted comments and actually tested the emission standards from 184 facilities. Submitted in our comments, EPA had projected estimates of emissions from all approximately 1,200 facilities that are affected across the country. We tested 184 – excuse me, EPA had projected from a data set of about 170 facilities, projected emissions estimates for all 1,200. We tested – we got actual data from 184 facilities and compared that to the output of EPA's model, the projections that they made. That data, which is laid out in our brief, our opening brief at 1415, I think it's mentioned in the reply brief, showed that of the 182 that we tested, EPA's projections, 62 facilities were closed. Sixty-nine or 38% no longer used chromium in processes subject to the rule. Forty-three provided emissions estimates that were lower than what EPA's model projected. And so over 90% of the projections, the output of their model that we tested with actual data showed that EPA's projections at those facilities, more than 90% of 184 were wrong. Counsel, nothing that I saw in the briefs addressed the question of the basic methodology employed by EPA in its projection. I mean, above all, I saw nothing to answer one way or the other whether it made a serious effort to compare apples with apples. In other words, look at 170 facilities and you get some numbers and then you project. It's only meaningful – I would think the projections would be only meaningful if you have a theory for extending the 170 – put aside the problems with the 170, the numbers for the 170, for projecting them to the remaining thousands and so on. And that isn't discussed by either side. I'm sorry, Your Honor, what isn't discussed? Well, you're taking a sample, right, and you're projecting it onto a large body of incidents. But what is it that indicates that the sample can be projected in a meaningful way? Suppose the – in other words, was EPA collecting data about emissions per unit of production? If it was unit per unit of production, how is unit of production defined? EPA's data, I believe it's discussed in our brief, was not emissions data. It was operational data from which they estimated emissions at the 170. But our point is that when we tested 184 of the outputs, more than 90% of those outputs were – of EPA's outputs from their model were overestimates of emissions. There wasn't a single example that we found that we tested where EPA's model had underestimated emissions. Over 90% of the outputs that we tested were – showed that EPA overestimated the emissions. And that – And did they not adjust for the overestimations that you identified? They adjusted those 173, yes. But that was, you know, 15% of the 1,200 that they were projecting. Did you propose an alternative methodology for – I mean, I think your point, as I take it, is when you show that the data that the agency relied on is this broadly and frequently flawed in these ways, that it's incumbent on the agency not just to say, okay, we'll throw out the ones that you have identified as flawed and then proceed as they were proceeding. It's your position that they have to say, wow, there seems to be – to the extent we're using this as a projection, that projection infects the whole data set. That is our position, Your Honor, and that was – I submit to you that this court said as much in the Appalachian Power case. And so my question – the follow-up question is, did you have an alternative methodology that you said, look, this is the way you should run these errors back through your entire data set, or this is – instead of using Monte Carlo or whatever it was they were using, you should use, you know, Monaco. It's not, Your Honor, but it's the EPA's obligation to, as the court said in Appalachian Power, address this, quote, stark disparity between its projections and real-world observations. If we disagreed with your challenge, your D6 challenge, and we thought that the D6 review was adequate, but we agreed with your F2 argument, it wouldn't make any difference for you. I mean, it's the same – the same rule came out of both reviews, right? And so even if we agreed with you that the analysis under F2 was flawed in significant ways, the rule still stands, right? You've got to win on both D6 and F2, right? We do, Your Honor. I'd like to reserve some time, but I want to say one thing about PFAS, if I may, just to be clear about this. In the final rule, EPA – we have no problem with non-PFAS, with the ban on PFAS fume suppressants. Non-PFAS is fine under the existing standards, but EPA projected in its risk analysis, its F2 risk analysis, that when these new lower limits are adopted, an additional 224 pounds of chromium, the emissions will be reduced further that much. And EPA said that that further reduction that would be achieved would be in 90% of the facilities that are affected. Ninety percent of the facilities that would have to lower emissions to meet the new standards would do so using fume suppressants. That's how EPA was going to meet the projected reduction in chromium emissions that it says this new rule will create. Ninety percent of the affected facilities – this is at JA485 – would have to use fume suppressants. And there is nothing in this record that shows that non-PFAS fume suppressants, as we've in our brief shown, nothing that shows that non-PFAS fume suppressants can accomplish those emission reductions. So the F2 analysis, in addition to the data problems I was discussing previously, it's also deficient because it relies – the reduction that this rule is projected to achieve in chromium emissions depends on the use of fume suppressants. In 90% of the cases, there's nothing that shows that non-PFAS fume suppressants can meet those reductions. Mr. Stuck, I thought there was information in the record on that and that that was sort of the premise of the surface tension analysis. For example, at JA442, there's a description of why the lower surface tension creates less mist because there's less of an acceleration when the gases, the bubbles, reach the top. And so if you have a lower surface tension, whether it's a foam or a low foam, it's going to just project less of that mist into the surrounding environment. And so I thought that there was some clear basis for that second part of their assumption, which is as long as you have a chemical that reduces surface tension, you're going to have a lot of evidence. Yes, Your Honor. There's some evidence that PFAS fume suppressants – in the record – that PFAS fume suppressants are correlated with reductions in emissions. There's nothing that shows that non-PFAS fume suppressants meet the same – surface tension limits. The only evidence is about PFAS fume suppressants. And the best – excuse me. The best – well, one more point if I may. The best evidence that EPA, again, I think I submit to you, recognizes that they needed to show that non-PFAS fume suppressants could meet the emission limits because at JA507, they put in this study which adjusted, so-called normalized, the German study, which they didn't like, and the Danish study, which they didn't like. But that's because of ventilation, different ventilation characteristics. But I thought that your challenge was that all you could show with non-PFAS chemicals was that they did similarly reduce surface tension, but that I thought you were saying that the piece that was missing was to show that a reduction in surface tension with a non-PFAS chemical was going to actually lead to reduced emissions. And I thought that was in the record in the various discussions of surface tension and why that's a missed or non-missed. We don't think it is for the reasons we've put in our briefs. Your argument is that the relationship between surface tension and emissions shown for PFAS may well not apply to non-PFAS. Is that correct? There's no evidence that it does. In fact, the only change here is not just PFAS to non-PFAS. Remember, there's a change in the surface tension limits. But yes, that's right. The only evidence in the record is a correlation. As far as PFAS is concerned, the relationship between surface tension and emissions, there certainly are data in the record. There's some evidence, but not about non-PFAS. Ninety percent of facilities are going to have to use to sustain this rule. Right. Excuse me, I'm very sorry. The red line only applies to you. It doesn't apply to us. I understand that. I understand that. I'd be delighted to come back. We're having an opportunity to make more arguments. In the 2004 rulemaking, there was evidence that EPA said it didn't matter what kind of fume suppression was being used to lower the surface tension. It just seems like they're building on a record where it's well established that whether it's PFAS or something else that lowers the surface tension, that's the critical factor. Do you have any evidence to the contrary? I guess that's one. Was there anything proper to EPA to say, hey, question your assumptions here the way you did with the other evidence we were just talking about a moment ago, something to raise a doubt for them about that function? Your Honor, let me tread lightly, but I will say this. We did a study and submitted a study that shows clearly that non-PFAS does not meet the surface tension limits, but it was submitted after the record closed and has been not considered by EPA. In fairness, in response to your question, that's the answer, but it's not enough. Go back to 2004. I mean, that's what the EPA decided in 2004, that it made no difference. What's wrong with that? Did you challenge that? Is there something about the 2004 conclusion that's faulty? Your Honor, frankly, I don't recall specifically the 2004 conclusion, but I don't believe it was a conclusion that – I don't think it could have been a conclusion – that non-PFAS fuels suppression. Well, it only dealt with PFAS. That's one thing. That's the only study that they had. Yes. Well, that's the point. It certainly couldn't have shown that the new, much lower limits could be met with non-PFAS or even PFAS. I think it's safe. We're done now, at least for now. Thank you. We'll give you some time. Good morning. Good morning. Your Honor, may it please the Court, the National Air Toxics Protection for millions of Americans is at a crossroads. Will it follow the course the Clean Air Act directs, assuring that revised standards meet the same tests, the same usual framework as all other promulgated standards, or will it be whatever EPAs prefer, however weak? In the end, the result will decide whether Americans in some communities can be left behind as technologies, practices, methods of controlling some of the most toxic chemicals advance. Their standards will become like an artifact in a museum, gathering dust like the first telephone, the first telegraph. It's not the way American business works to allow competitors to get so far ahead in anything. And for cancer prevention tools, air pollution controls, it's not the approach Congress chose. The plain text of the law answers this question. It says that 7412C223 is clear. It says that its test applies to emissions standards promulgated under this subsection. There is no textual basis to read that provision otherwise. Well, they would say that a revision is not a promulgation and that that's enough ambiguity to allow EPA to read that differently from how you propose it should be read. Your Honor, they have no textual basis for making that argument. This Court's precedent has interpreted promulgation broadly as an umbrella that covers final action released, including rules and other kinds of determinations. And we've cited that precedent. The Court has recognized that revised standards are promulgated. In this record, EPA stated that it promulgated standards under 7412C as well as 7412F. So their argument conflicts with the record. The law is clear. It says emissions standards promulgated under. And the only exception provided is a clear one in 7412B5 stated for, you know, allowing EPA in certain circumstances in lieu of Paragraph 2 to apply a different test. Under Chevron, you know, the clear intent of Congress expressed in the text is supposed to be the end of the matter. And in the end, EPA has no textual defense to respond. To the extent that their textual defense is that under D there is promulgation and there is revision, and we think that the reference to promulgation in the D2-D3 context is not intended to apply to revision under D6. And if they think that's an ambiguity, is there anything about the treatment of that in the rulemaking that you would think would be inadequate under Chevron Step 2, assuming we were to get there? I mean, obviously, we have also the precedents, the American Battery and NRDC precedents. But there were, you know, that's another major hurdle to your argument. I know you know that. But, you know, two points. The law is clear, even if there were ambiguity in D6 alone, which I believe is the only kind of textual argument EPA tries to make in the record, read as a whole in context. That law is clear. Even EPA itself in this record, as we've cited in this joint appendix, recognizes that this action as a result of a D6 review is a continuation of the technology-based 7412-D process. And so even EPA's own interpretation is unreasonable because it conflicts not only with the plain text, but how it has interpreted it. I thought you had also made an argument that they don't engage in any decision-making worthy of deference because they're simply relying on NRDC and, you know, doing kind of an out-of-the-box conclusory legal analysis. Is there anything else that they said in support of their reading? That's correct, Jen. They have no interpretation in the record of D2, D3. They tried to point to the Coke ovens rule. That rule previously also had no interpretation of D2, D3. And this court does not generally consider grounds for agency action that are not presented in the record. It has not engaged with the text. EPA's lawyers provide additional arguments, but even under Chevron Step 2, none of those can escape the plain text. You know, the D5 provides the only exception from D2, D3. And their lawyers try, I believe, to surpass the Chevron Step 2 test to create an artificial conflict with D6, but there is none. They cite a number of arguments that are not in the record and have no textual basis. Okay, now can you deal with the rather considerable hurdle you have in that we've already answered this question? Thank you, Judge Griffith. The bottom line is this question has not received a merit hearing on the textual issue. The text was never discussed in Battery Recyclers. That decision decided the question for the first time solely based on a prior decision. And just to step back... But you think we got it wrong. I mean, I'm confused. You've asked us to overrule it, right? So you must think that we've ruled contrary to your interest for what would be the basis for us overruling it. Your Honor, Battery Recyclers... We have law about precedent and how we're to approach precedent. I think so. Yes, Your Honor. Battery Recyclers decided the question without addressing the issue. That's not considered the basis for disregarding a prior precedent. Your Honor... That may seem illogical or wrong, but it's definitely our case law. Your Honor... You don't assess the reasoning of the prior precedent. Environmental petitioners respectfully request the court to exercise appropriate procedure to reverse the Battery Recyclers' mistaken finding that the previous case had already decided the question. We used various procedures, including an iron footnote, to reach that conclusion. But the bottom line is the issue is so important because it affects so many people living near toxic sources. Is that the bottom line? So that's your strongest argument for why we ought to use the iron footnote procedure because it has broad effect in the country. I don't think that under our case law that that's why we use the iron footnote procedure. Well, Your Honor, the court has reversed precedent when it is incorrect on an important issue for many different reasons. Absolutely, and this court could propose, suggest in bank action in full if that is needed, that the iron footnote process has been used to overrule prior precedent. And in this case, Battery Recyclers was the first case that had the legal question before it of what test applies to revised standards. So at step one, it reached an erroneous conclusion because the NRDC case did not even have the legal question before that court. It was reviewing only the decision in an unusual bit of rulemaking not to revise standards. And so NRDC held only that when EPA, when there are no emission standards promulgated, EPA need not follow the floor requirements. C6 is a filter in that way. All standards go through the review process and EPA must decide whether or not to revise them. I understand your position. I guess what we want to be really clear on is it's not every day that someone argues to a panel of the court effectively, you know, take this en banc at this panel stage. And so I heard you say, well, you could effectively try to take it en banc within the court by doing an iron footnote. My sense is, and I'm not experienced with this, but that this is not the kind of thing that one would do with an iron footnote. It's a much bigger issue. So really what you're saying here is I recognize there's precedent. I recognize you have to follow it. But, hey, we're going to try to take this en banc. Is there anything, I mean, there's not any way that you see that a panel of this court could actually, you know, come out differently from association of battery recyclers. I mean, it was brief, even though I understand your argument for distinguishing NRDC, but once you get to association of battery recyclers, even though, again, it seems to treat NRDC as precedent and you're disappointed in the level of analysis, we as a panel can't really do anything, can we? Well, Your Honor, there is some intervening law that is persuasive on a similar question, the white stallion case. And so the court could consider that to be persuasive enough that it warrants another look at battery recyclers. Of course, the iron footnote process is something the court has stated it uses it, has used in the past for a number of reasons, but it is sort of unusual procedure and it can be used for any reason the court deems appropriate. In this instance, the full en banc court would not need to apply the resources it takes to fully evaluate the merits. An iron footnote would only be required to recognize that battery recyclers mistakenly read the 2008 NRDC decision in that that decision decided a different question of law. And as Judge Henry Friendly put it in 1979, if I can quote him, a judge's power to bind is limited to the issue that is before him. If petitioners in the NRDC case had attempted to make the argument that that should be reversed en banc in that case, because if emission standards had been promulgated, then they wouldn't be able to ensure EPA applied the usual test, that would have been so clearly outside the scope of that case. And that is an illustration of why battery recyclers was wrong, because there were no standards promulgated there. The language in D2-D3 was simply never triggered. It was never brought into play. This court generally, you know, it is an extremely unusual circumstance to ask this court to consider overruling prior precedent. This is a circumstance, however, where there is no statutory interpretation on the books. We're not asking the court to change its mind on what D2-D3 means. Neither EPA in the record nor this court has presented an interpretation. It's not a circumstance where we were able to have a hearing and the court issued a decision that explained why we were wrong. You know, we truly believe the text is quite clear, but we don't even have a decision from the court saying, well, excuse me, there are, you know, these are some circumstances. There's no reasoning. Generally, sorry, disguises is important to protect a prior panel's reasoning when it has really, you know, addressed an issue. But the space of the opinion in battery recyclers and the NRDC case is clear. It never grappled with this question. Your time has expired. We'll give you back some time. Thank you. Morning. Morning, Your Honors. I'd like to thank the court for the opportunity to address the importance of this case to the emission control efforts of state and local regulators. Those efforts play a key role under the statutory scheme that Congress designed in 1990 in its efforts to enforce stringent national controls over air toxic emissions. At the heart of that statutory scheme that Congress designed are the minimum stringencies in D2 and D3 at issue in this case. In establishing those minimum stringencies, Congress commands EPA to look to the successful results of state, local, and industry experimentation and to, at a bare minimum, identify the top performing sources and make sure that other sources ultimately catch up. So in that way, the Act depends on state experimentation to drive more stringent national standards and further its overarching goal. The Act also promotes state experimentation through this process by – in two ways. First, it discourages leakage of emissions by ensuring that when a state promulgates a standard or otherwise attempts to control emissions in its state, those sources do not simply move to another state because there will ultimately be a national standard applicable to those other states. And the second way that it promotes experimentation is by – Your argument then is that the massive effect of more stringent regulations doesn't trap people to a state that imposes more stringent regulations. I'm sorry. Your argument seems to be that there's a competitive disadvantage to a state that imposes more stringent regulations than applicable elsewhere. But why wouldn't the health advantages of that standard attract people to that state? Well, I think that our position is that the state efforts to control emissions stand on their own – they're valuable for their own protections that they provide to state residents. Yeah, yeah. Why doesn't that protection make the state more attractive? Well, we think it does in many ways. Okay, the state has made that trade-off and should be – should attract migrants on that basis. Well, I think that we would agree with that, but it's at the same time the case that Congress encourages states to – any potential con to doing that would be lessened. So you're saying that it may attract others, but it won't attract business. And, in fact, it'll put any business that's using these chemicals at a competitive disadvantage vis-à-vis businesses in other states that don't have to meet such expensive or high emission standards. I wouldn't say that's true in every case. It may attract businesses who want to locate in a state where the air is going to be cleaner and the people are going to be more protected. But certainly in some cases, businesses are going to face a competitive disadvantage from more expensive control standards. What about the cost? Can you talk a little bit about the separate issue of the exclusion of the California data that – you know, not the en banc issue, but the other issue with the EPA's exclusion of the California data? They estimated that, you know, most of California's achievements were due to use of HEPA filters and that that was a technology that they just seemed to be vastly too expensive to impose nationwide. And I wonder what your response is on that. Well, we didn't join that. Okay. You don't have to. But, you know, our position is that those technologies are not so costly that businesses cannot adopt them. And, in fact, California's experience shows that businesses have achieved those reductions. And, therefore, Congress has to take them into account. I would like to briefly, as my time is running out, address a question that Judge Pillard asked earlier about the difference between revision and promulgation in 112B. That – we agree with environmental petitioners. That's a post hoc argument. But we'd also like to emphasize that if that were, in fact, the case, EPA would be able to use the revision process in a number of ways that it is – that it's prohibited from using the standard setting process in Section 112B. As we described in our brief, they could use it to backslide below the mapped floor. And although we didn't mention this in the brief, they could also use it to – well, it would not be subject to the savings clause in 112B7 that prohibits EPA from – or prohibits any standards from weakening standards set under other authorities. Like D6 itself, like D10, D7 only applies to standards that are promulgated under 112B, not standards that are, quote-unquote, revised under 112B. So we would respectfully submit there is absolutely no difference between the two. And we take the court and ask you to remand the issue for EPA to follow 112, D2, and D3. Thank you. Thank you very much. Good morning. Good morning, Your Honor. May it please the Court, my name is Timoni Peoples. I'm here today on behalf of the EPA respondents in this matter. And with me at counsel's table is Jan Tierney from EPA. EPA initially promulgated technology-based MAC standards for chromium electroplating sources in 1995. In 2012, it revised those standards pursuant to both 7412D6 and 7412F2. Now, despite petitioner's challenges, that rule is valid and it's a reasonable exercise of EPA's Clean Air Act authority. Now, unless the Court prefers otherwise, I will begin with environmental petitioner's statutory argument. As you just heard, environmental petitioners argue that EPA is required to re-perform the entire MAC floor analysis under D6 review or revision. This argument warrants a little consideration by the Court because it's foreclosed by either collateral estoppel or stare decisis. With respect to collateral estoppel, all the requirements are met here. In my mind, you don't have to address collateral estoppel. I just think since there's a new party in, it violates due process to apply collateral estoppel against them. So I just think that's a non-starter. But we do have the precedence and that's a concern. But what about on the merits? Where can you, given Chenery and given the requirement that your analysis has to be something that's based in the record, I'm just curious where there's an analysis. What I've seen in the record is pretty much sort of circular, referring to what the D.C. Circuit has done, rather than taking on the question under the statute, under Congress's text, what the relationship is between D2 and 3 and D6. Well, Your Honor, this, despite Environmental Petitioner's argument, this is not a post hoc rationalization. In response to comments, EPA very clearly stated its position. And in response to comments in Joint Appendix 517-519, it said that it disagreed that a D6 review is a new review under D2 and D3. It went on to say that nothing in D6 required EPA to recalculate the MAC floor, that it read D6 as imparting EPA with substantial latitude in weighing a variety of factors, and that its position had been upheld in NRDC and was not otherwise being reopened or reconsidered in the final rule. Now, EPA also cited to a 2005 rulemaking where it set forth its interpretation in greater detail. So the administrative record sufficiently discloses the grounds upon which the agency based its decision. And Petitioner's suggestion of a post hoc rationalization should be disregarded. Now, turning to the merits of – Can you list the sites for the – that's in the – in JA 517-519, the cross-reference back to 2005. Correct. Okay. Thanks. Now, concerning the merits of Petitioner's statutory argument, as we discussed in our brief, the statute does not unambiguously compel their reading, and nor does legislative history. Nothing in D6 requires EPA to apply the MAC criteria in a review or revision. And as the Court had noted, the Clean Air Act has distinguished between promulgation and revision. Now, sure enough, those are terms that are used in ordinary language. EPA does promulgate standards. But within the context of the Clean Air Act, several provisions that are cited in our brief show that Congress clearly made distinction when it's talking about promulgating standards and revising those standards. And also, the Court in NRDC versus EPA, a concurring opinion of Judge Wall, does also support this proposition that there is a difference between promulgating standards and revising standards. Now, you mentioned to people how much more onerous it is to do – to redo the entire MAC floor. Is it much more onerous than doing a revision, or is it just taking a different approach to the data in terms of the best-performing 12 percent? It seems like it might not, in fact, be much other than that standard being interjected into it. Are there other things that make it just as a kind of a practical matter, administratively, so much more complicated for the agency? Well, I do know that under D2 and D3, to actually calculate the MAC floor and then go on to determine if a beyond-the-floor standard is appropriate is a serious undertaking. It's very time-consuming and very costly and requires a vast amount of data. In what way is it different from the analysis under D6 and or F2? Well, under D6, EPA also considers that data, but it looks primarily at whether or not there have been developments in processes, control processes – practices, processes, and control technologies. And it also goes on to consider the cost and technological feasibility. So, again, it is a comprehensive analysis, but it's not as comprehensive or voluminous amount of documents considered as in a MAC floor analysis or a beyond-the-floor analysis. So we just feel that the given – In what way is it different? I mean, you said there's more to the MAC analysis. What is the more to the MAC analysis? Well, in the MAC floor analysis, EPA has to gather a variety of data and determine the best – I guess the 12 percent – what the 12 percent – best 12 percent of the sources are doing and then go on to determine whether or not other standards are cost-effective or achievable. I'm sorry, yeah, cost-effective or achievable under D2. And it's a serious undertaking. It's what led Congress to amend the standard back in 1990 because it took EPA so long to go through that process. What about the environmental petitioner's concern that your reading of the statute really does seem to allow for backsliding or really doesn't have, in fact, a substantive standard under D6 at all? Well, I believe that the backsliding argument was raised by Skidamichi. It was not raised by environmental petitioners, and so we don't consider that argument probably before the courts. But I will say with respect to – I'm sure it would be a concern of your agency, though. It would be a concern for the agency, but we don't believe that D7 requires an alternative interpretation. I'm sorry, what was the other point that you asked about that they raised? I was just wondering whether it's problematic from the perspective of the EPA that read as you read it, it doesn't really have a substantive standard that would have this ratcheting forward effect that one might think the Clean Air Act embraced. It just says, you know, check in, look how people are doing, and if you think it's necessary, revise. But there's not – I mean, the way that the parties pushing for the redoing of the MACFOR are characterizing it, at least. They're saying that, you know, once you drop out the MACFOR, there isn't an environmental standard. And thank you for repeating that. We disagree with that interpretation. In addition, this court in NRDC itself noted that the core requirement of D6 is a determination of whether there are significant developments in practices, processes, and control technologies, and certainly EPA's interpretation tracks that statutory language. It looks at the factors specifically enumerated in D6. But actually, I take it particularly from the footnote at page 48, if you're brief, that the development was a – that EPA found was a development in terms of new data about the situation. The development under D66. It wasn't a change in technology. It wasn't a change in illnesses resulting from chrome emissions. It was simply new data. Yes. About what the previous rule has produced. The new data, significantly new and improved data in 2012, showed that sources were achieving significantly lower emissions. And I just want to make a clarification. EPA used actual emission data for this analysis in order to show that there were developments, not emission estimates. It was actual data from sources. And EPA, once it analyzed that data, determined that there were significant improvements in the control efficiencies of these technologies. We used – Control efficiency or just the results? In other words, is there an analysis to show that the plants were employing different technology or advanced technology? Yes. Advanced beyond what was expected in terms of the earlier rules? Yes. What EPA determined was that there was an improvement in control technology performance. And how it determined that was that it – It's just the results. The results were better than EPA expected. Well, no, I wouldn't say that. But certainly that there were significant improvements, certainly over time, given since the time that the initial stands were first promulgated in 1995, manufacturers have improved their technologies. And so the technologies are able to achieve much lower emission reductions than before. I guess it seems like – I mean, it's a little ambiguous. You do say that the control efficiencies of composite mesh pads have greatly improved. And, you know, so it's an existing technology that's being used in a cleaner way or more efficiently, or there's some reference in the record to training. So I guess we're just trying to get a sense of the data. I thought the data that you used was more output and that there was some amount of inference and also some amount of projection across the broader range of plants. Is that right? There's inference about improvements given the lower emissions, and there's projection across a broader number based on the data set that you had. Correct, yes. The EPA used the estimated emissions to, in other words, to determine feasibility and economic burden for sources nationwide. Yes, sources nationwide. But it actually relied on actual emission data. Right. And that was partly because you didn't want to put the burden on industry or the states or anyone to collect a whole lot of data. You said, let's use what we have and what the states have and we'll try to work with that. That's correct. Can you respond? I'm still having a lot of trouble figuring out why the California data was not considered by EPA. It just seems like if you have a situation where you're looking to say, what's feasible that's really cleaning up the air, and California is a leader, I just don't really understand why you would say, well, let's look at what they're doing. And let's include that as a push factor for the rest of the country. Sure. To be clear, EPA did consider the California emission data. It was a process whereby EPA identified certain developments and then went about determining if those developments would warrant revision of the standard. So in considering the development in HEPA filters, EPA did consider the California standard where those filters are primarily used. But EPA determined that to adopt the California standard would essentially require all sources nationwide to adopt or use those HEPA filters, and that undertaking would be significant. It would impose a serious economic burden, especially to small businesses, and that it would not be necessary under D6 or required under F2. And so it went on to then consider other developments and determine whether or not those developments would warrant a revision. I mean, to put it simply, you know, what California is doing is just not economically feasible everywhere else. To be clear, I read the rulemaking as finding that it would have to change the implied cost-effectiveness level to embrace the California approach. Yes. Well, the cost of adopting the California standard was significantly more than the cost to adopt the standards that were the revised standards. And so when EPA looked at its analysis under D6 and F2, it just simply determined that that standard wasn't required under F2 to provide ample margin safety, or it wasn't necessary under D6. The cost-effectiveness standard, as I understand it, is not exactly a standard. It's a measurement, like a thermometer measures temperature, but it doesn't say how much is too low or too high. EPA just says that. That is correct. There's no rigid line. There's not even a formula. That's correct. A non-rigid formula. That's right, because each situation is different. But certainly here, the costs were great. I mean, with the filters, we're talking about costs between $27,000 and $43,000 per pound versus $5,000 to $14,000 per pound for the development that EPA used to base the revised standards on. So it's an enormous difference. But EPA did consider a variety of factors. It was not just cost. It was emission reduction, the number of plants that would be impacted. Again, the impact of small businesses. And under F2, certainly EPA considered technological feasibility, uncertainties, and the risk reduction that could be achieved. And so it determined that it was just not appropriate to adopt the California standard nationwide. Can I ask you what's your position on the question that Providing Judge Griffith asked about, do you think if we were to sustain your standard under D6 that we need to reach the F2 questions? Well, EPA's rulemaking was under both D6 and F2, and it's obligated under the statute to have a one-shot look at the residual risks under F2. So if the D6 standard was upheld, then it would probably be our position that, but not the F2 standard, it would probably be our position that EPA would have to, under the statute, still do that residual risk analysis. So if we just said we don't reach it, then you would take that as you haven't fulfilled your obligation yet to do the one-time risk review. I guess if you said if the court were to say that it did not reach the issue, then maybe we would interpret that to be that it still stands. I mean, unless it was specifically vacated, we might interpret that to mean that it's still valid under F2 and D6. Thank you. Can you address the question of the projections that EPA was making from its limited data set? As I understand it, it turns out that there are a lot of flaws in the limited data set. And in response to that, EPA corrects with respect to particular plants that NASF identifies as EPA having used erroneous data. And then what does it do? Then it completely reruns the Monte Carlo exercise, or what? Well, let me just get... I think it's now re-application, the verb re-applies, as it appears in the record, but rather externally. Sure. I'll try my best to explain it. So the Monte Carlo approach is a statistical tool. So it starts with a representative subset of facilities and data. And EPA does not believe that that subset here was limited. We believe that it was representative. But surely it was limited. It wasn't all the plants, right? It wasn't all the plants. It was short of all the plants, but the data was representative. And we feel that we had a really good number of data in order to run this. Okay. And then it turns out that a significant fraction of the plants covered have gone out of business and so forth. So then what exactly did EPA do? Well, for those plants that had actual data, that's put into the model. The model then will assign emission estimates for those plants where EPA actually didn't have actual data. So when industry came back with actual estimates or actual emissions for those plants that EPA did not have data for, EPA made that correction. For example, and I'm trying to be as clear as I can. So the plants that EPA thinks it knows, right, and then there's the plants that EPA knows it doesn't know and is making a projection about, right? Right. Okay. So as I understand, NASF brought in data about both sets. Is that correct? Mainly, industry brought in data mainly about the plants where EPA didn't have the actual data. And so where industry brought in that data, EPA made those adjustments and said, okay, we didn't have data. Before now, we have actual data. We're going to use that actual data. To the extent that the actual data are significantly different, very significantly different from EPA's projections, why doesn't that draw into question the whole exercise? Because it's a statistical tool, and it was never used to – Is it a good statistical tool? We believe that it is. We believe that it is. And I will point out also that industry does not – it says in its reply brief, it does not critique the Monte Carlo approach. Right. So if it does not critique the Monte Carlo approach, then it cannot critique the resulting projections that the Monte Carlo approach puts out. Well, it critiques the inputs, right? And where EPA received valid data from industry, it made those corrections. What industry's main gripe is, is that EPA did not adopt the broader trends of its data to its model. And EPA explained why it declined to do so. Some of industry's data were vague. It was unclear. It was unreliable. And EPA reasonably declined to make the changes. It instead opted to rely on the valid Monte Carlo statistical tool, which, again, industry has already conceded it does not critique. And so we feel that that choice is not arbitrary or propitious. It's very reasonable. And also in the record, we went on to explain that even if we accepted industry's broad generalizations about the data, the risk results would have still warranted a revision under F2. And, Ms. Peebleson, you had an argument, I think, about why even a closed plant or older data about a plant might still be relevant. Can you remind us what that was? That was, I believe, we took out closed, under F2 or D6, because there was an issue of closed plants under both. I assume you're talking about F2. Yeah. Okay. Under F2, if we had specific information that a plant had closed, we did remove that plant from our risk assessment model. But sometimes industry has submitted data saying that a plant no longer does hard chromium operations. Well, that doesn't mean that it doesn't do decorative or chromium anodizing. So, again, some of their data was very vague, and they wanted us to adopt broad generalizations about closed facilities. Well, that didn't make sense because there may have been new facilities that opened up nationwide, and we had to take those into account, and they offered no data on that. So, again, we stuck with reliable, valid data that was not ambiguous, and we made the changes, accepted much of industry's proposed changes to our model, but simply reasonably declined to adopt the broader trends. And we feel that that is appropriately reasonable for EPA to do. So, unless the court has any other... Yeah, I do have a question. What is the basis for inferring that achievement of the specified surface tension criteria will satisfy the emissions goal, which I take it is the goal of the EPA in all this? As far as I can make out, you don't have data on that. You have the European tests, but in order to make them relevant, you have to extrapolate from what appears to be, and EPA expressly concedes, as far as the German data is concerned, is a radically different ventilation rate. Well, you're talking about the Danish study, and that study we feel does support EPA's finding that non-PFAS fuel suppressors can meet the revised limits. Now, what EPA did was a normalization method, and it didn't... It should be... It's not complex. It's not novel. EPA does that quite often. It did that throughout this rulemaking, and basically all it had to do was you had emissions data that was in the Danish study that had to adjust the ventilation rate from the Danish study to ventilation rates that were more... What does this adjustment involve? Well... I didn't see to me that the EPA presented any significant dependence of this extrapolation. Well, if you're trying to compare the data in the Danish study and to determine whether or not non-PFAS fuel suppressors can achieve emissions in the United States, you had to adjust the ventilation rate to a ventilation rate that was more likely to be used by sources in the United States, and that's all EPA did, is just adjust that ventilation rate. And once it was adjusted, then you can compare the two sets of data. And EPA then determined that once you adjusted the data from the Danish study, sure enough, if those emissions didn't meet... I thought the result was when you factored in ventilation, that when you come to the United States, ventilation is not required here. Ventilation is not required. Well, let me back up a little bit. So when you're using a fume suppressant to meet the surface tension standard, then ventilation is not required. Ventilation is only required when you're trying to measure emissions from a tank. And so oftentimes facilities use a fume suppressant in combination with other controls to meet the emission limit, and so that's when the ventilation rate will come into account because you need ventilation to capture the emission. And so in the Danish study, there was ventilation, and therefore there was a ventilation rate. So EPA just had to adjust that ventilation rate from the Danish study to using a ventilation rate that was more likely to meet... Did EPA ever describe what the adjustment method is? I'm sorry? Did EPA ever describe what the adjustment method is? Yeah, it's described in the Joint Attendant at 508 through 510 and also at 633. And to what extent are you also relying on the Atatek study? Yes. The Atatek study also showed that non-PFOS fume suppressants performed just as well as PFOS fume suppressants, and I want to make a point... Is that on surface tension or emissions? Surface tension. I'm sorry. I'm circling back on the non-PFOS issue. It's not on ventilation. Well, the study did say that non-PFOS products were developed and are on the market that are just as effective in controlling emissions. That is in the Atatek study at 446 to 447. I'm sorry. I thought Atatek was about surface tension and not emissions. Right. That was my interjection. I'm sorry. I confused things because she was addressing Steve's question. I'm sorry, Judge Williams' question about ventilation and how the EPA adjusted the ventilation under the Danish study to the ventilation under the U.S. conditions, and I interjected and I think confused things by asking also about the surface tension question under Atatek. But the Atatek study does support both. It does also discuss surface tension and also discusses non-PFOS products that were after testing on the market and they're commercially available and they're just as effective as PFOS. At what? 446 to 447. No, no. Oh, I'm sorry. Effective at doing what? Oh, controlling emissions, controlling chromium emissions. And also in addition to the Danish study and the Atatek study. I rarely see that at 446. It's the top of the 446, the second paragraph, and it's the last two sentences. It says that both products can fully comply with PFOS legislation. Yes, but that sets up for firms an alternative benchmark in terms of surface tension. The claim by NASF here is that there's a gap between EPA's target in terms of emissions and its measurement device, its alternative measurement device, which is surface tension. I don't see anything addressing that. Well, I thought it was, and this would be helpful, Ms. Peoples, if you could address this. I thought it was the bottom paragraph on 442, which describes in terms that even a judge can understand the most common mis-suppressants are surfactant-based. They reduce the surface tension. This is a two-fold effect on the generation of mist, reduces the size of gas bubbles. The smaller bubbles travel more slowly. The lower surface tension reduces the energy with which the resulting droplets are ejected into the air. Together, both of these effects can reduce the emission of droplets and, therefore, mist generation. When a mist suppressant is properly applied, this has been found to be sufficient to reduce exposure to employees. So it's making the connection between surface tension and emission, and I thought that was an important piece of data for the more general point, at least from my understanding of the punitive gap in the data. That is correct, and that is what EPA did also rely on was the relationship between the surface tension and the emissions and used that as a basis for the data. Help me understand. I was under the impression that all of that analysis about surface tension was using PFAS, that we don't have any studies about the effect of non-PFAS on emissions. We have no studies about... We have studies about PFAS and surface tension and emissions. We have those, but we don't have it for non-PFAS and emissions. I may be wrong. I'm not making a point. I'm asking a question. Right. Well, so the relationship between the surface tension and the emissions, EPA describes, does not really depend on whether or not the fuel suppressant is PFAS or non-PFAS. I didn't ask the question. The question is how EPA knows that. Well, that's based on studies that are in the record. Now, that's the... What is equated to the effect of PFAS and non-PFAS? It's just the way that it simply works. That, again, if you have a lower surface tension, then the bubbles will become smaller. They'll rise to the surface more slower and burst with less energy, and they also create less emissions. You're saying surface tension is a proxy for emissions, right? If you can reduce surface tensions in all instances, you're going to lower emissions? Is that the point? I don't want to say all, but I'm definitely saying that, yes, if you have a reduced surface tension, then yes, the idea is that the emissions will be lower. Whether that's PFAS or non-PFAS. Correct. Correct. Is this study... I mean, I see on the top of 447 just the only text on that page. Comparison tests between current leading PFAS products have been made and the newly developed non-PFAS products performed just as well. I thought this study is referring to non-PFAS as well as PFAS. That is regarding the difference between... 447. It's not just controlling emissions, but certainly the surface tension standards and its relationship with the emissions are discussed in another case. I did want to also raise the point that the record does include information showing that facilities using non-PFAS fume suppressors can achieve the revised emission limits, and that is in the Danish study. We did the normalization approach. We adjusted it so that we can compare that data with the emissions that are likely to be in the United States. Also, the Minnesota facilities, we have data from Minnesota facilities where PFAS has been banned since 2008. They stated that they had the same or better control efficiency after making the switch. That was meeting the surface tension. No, that was overall. We asked them to let us know if there were any significant differences or changes after switching from PFAS to non-PFAS. The overall majority of those facilities said, we don't see any difference. In fact, some of them said we actually have better control efficiency with non-PFAS than with the non-PFAS. Let me ask you about JA447 that Judge Pillard directed your attention to. It says, comparison tests between current leading PFAS products have been made and the newly developed non-PFAS products performed just as well. What does that mean? Performed surface tension? Performed emissions? Both? What does it mean, performed just as well? I was under the impression that that was referring to surface tensions only. It's not referring to emissions? That's certainly possible. Yes, I'm sorry. That's probably true. It's referring to surface tensions. Performed just as well in that it was reducing surface tension. Yes. Again, there's no difference in the performance of PFAS and non-PFAS. That's the question, yes. Is it possible that you can have surface tension reduced using non-PFAS but it not affect the emissions significantly? Is that possible? Industry does not dispute that the non-PFAS can achieve the surface tension standards. They don't dispute that. It's the emissions, right? It's the emissions. The question is the extension. There is some data from that study we were talking about earlier, page 444, indicating at least in some cases it didn't, right? You said 444? Yes. What was cited in the industry's reply and what that discussion concerns is the problems that the testers encountered during the testing of those products. But when it goes on to conclude at 446, the top of that, after extensive testing of non-PFAS human-suppressing candidates, a family of products was finally developed, thoroughly field-tested, and brought to the market. So there were problems encountered during the testing, but after that, sufficient products were eventually brought to market. So is it... The conclusion is stated in terms of complying with PFAS legislation, which is not defined and may not relate in any way to emissions. But again, industry does not dispute that non-PFAS human-suppressants can achieve the surface tension standard. And certainly if the surface tension standard is achieved, then the source is in compliance. That's right. But the argument of NASF is that compliance standard is not achieving, may not achieve, EPA's stated goal here. It's not that they can't comply. It's their compliance doesn't achieve what EPA says it wants. Well, there's two standards. I understand that. So if you use... EPA is not interested in surface tension alone. It's interested in emissions. So the argument is that even though the firms can comply with this alternative standard, that doesn't mean that the standard is reasonable because, they say, may well not accomplish what EPA says it wants to accomplish. Well, we just don't... We just believe that that argument's not supported by the record, that the PFAS ban is supported. Even industry itself conceded that the PFAS ban was a reasonable requirement that's protective of the human health and the environment. Yeah, but the question is what the impact of imposing that ban is on emissions. But if a source uses fume suppressants to comply and it maintains a surface tension level, then it is in compliance with the rule. I understand that it's not compliant. Right. The question they're challenging the link between that compliance mode and the Stesha objective of reducing emissions. And I guess your answer is it's like if you put a pot of water on a hot plate over 200 degrees or on a flame over 200 degrees, it's physics, it's going to boil. And I guess our question is, is that true of surface tension, that if the surface tension is low, that there's going to be less emission for the reason that this narrative describes? Or is there some, you know, decoupling of that function? And if so, what is it? The record supports EPA's finding that the relationship between surface tension and emission is not a function of whether or not you use the PFAS, fume suppressant or non-PFAS. I thought you were looking to the data study. Because of the relationship, it doesn't depend on whether or not the fume suppressant is PFAS or non-PFAS as long as that surface tension... The data from which EPA drew that inference was PFAS data. Well, it was also in the 2004 rulemaking where EPA talked about variations and said that different parameters and said that none of those... But 2004, that was PFAS data as well. As far as I know, and correct me if I'm wrong, we don't have any study of showing relationship between non-PFAS and emissions. We don't have that. Am I right or wrong? Is Minnesota not evidence of that? Minnesota data includes non-PFAS fume suppressants. And also, as I said, the Danish study does also include non-PFAS fume suppressants. Does the Minnesota study go to data? Does those data go to emissions? I don't believe... I'm trying to recall, but I think that we just asked them to let us know, give us information on whether or not there were significant changes after switching from PFAS to non-PFAS. And they indicated that they had not encountered any changes. You asked a big question and they gave you a big answer. Well, if there had been a big difference, then we believe they would have let us know, but they didn't. And many of the facilities indicated that they had actually received better control efficiencies after making the switch. So that study, then you have the Danish study, the Agitech study, and you mentioned the German study. The German study results were less clear concerning environmental products. But with respect to PFAS, non-PFAS, it also supports EPA's finding that non-PFAS performs just as well as PFAS. The discrepancy in ventilation rates, even EPA acknowledged, were fairly wild, 200 to 1. Correct. Yes. Correct. Unless the Court has any other questions. Thank you. Thank you very much. Let's give Mr. Staff five minutes back, and we'll do the same for environmental petitioners. Wait until you get to the microphone. No, come on up. Don't start. You can use the microphone. Thank you. On the PFAS issue, I want to get to other things, but there is no data in the record establishing that the non-PFAS fume suppressants can meet the new emissions limits. It is a matter of inability to meet EPA's objectives. What about the Minnesota test? The Minnesota test was, I believe, surface tension. There's qualitative statements that are not data that may drift into something else. But it's also not just a matter of achieving objectives. It's a matter of the belief. Help me with the science here. Is the science here that if you do reduce surface tension, that you're always going to lead to reduced emissions? That's the concept. That's how fume suppressants work. But EPA needs data to support conclusions. And it's not just a matter of achieving their objectives, your honors. It's a matter of the validity of the residual risk study that supports the F2 conclusion. Because that study, the EPA's conclusion is that there are going to be 224 pounds of reduction in emissions. 90% of the facilities that are affected by this new rule, in other words, that are not already in compliance with the new limits, are going to use fume suppressants to achieve that. So 90% of the facilities that go to the 224 reduction are going to use fume suppressants. And there is no data in the records that shows that 90% can accomplish that. But is there a... It's a little difficult because the Danish and German data appeared after the rule might have been disclosed, right? Or at least after you... No, no, the... It appeared later in the day. Late, but the adjustment memo, the normalization memo was very late in the day. Okay. But does any of this cited data showing that different methods of changing surface tension are likely to have an effect on emissions? In other words, surface tension X arrived at by one method may produce one type of emissions called surface tension. Suppressant Y will have... Even getting the same surface tension... There are things in the record that show that there is not an equivalence between PFAS and non-PFAS, at least with respect to surface tension. And I'm afraid I can't do better than that. On emissions, I'm not sure about a method, whether there's anything in the record that shows affirmatively. But again, I would say that it's not our burden to show a discrepancy. There is... I believe there is data that shows that on surface tension, there's not an equivalence, but I think that's the best I can do for you. At least now. You made clear somewhere in the record that you were drawing a question. The equivalence in terms of moving from surface tension to emissions? Yes. That's in the record. I'd like to go back to something. Back to the opening point that I made. What's the best site for that? Where in the rulemaking you objected to this assumption? I'm confident it's cited in our briefs. I'm very sorry. I don't have it in my hand. But I would like to go to something right back to the outset, because I think it's very important and there may be some misunderstanding about it. When EPA revised its estimate of emissions from 5257 down to 956, it was on the basis of new data, but not new in the two-year period from 2010 to 2012. EPA completely changed its estimation methodology. It abandoned the methodology it used to get to 5256, which was calculations, and it did then in 2012 rely on an actual data set. But that data set extended back to the 1990s. The footnote that Judge Williams mentioned says that the emissions data used in its analysis overwhelmingly post-date the 95 emission standards. That's right, but some of it was back in the 1990s and reflected data about the implementation of the controls in response to the 95 map. I just want to be clear. This revision in estimates was not because they gathered new data in 2012. From 2012, that showed much lower levels. It was a whole new study and a whole new estimation method that shows, we believe, and I don't think I've heard any response to this, shows that those lower levels that EPA now estimates under its new estimation method based on emissions data from the 90s from the 2000s, that low emissions level has been in place so far as the record shows since the controls were implemented. But that turns on the legal question of how far EPA can go in defining a new development. There is no development on this record, Your Honor. That's an interpretive question. There's no improvement in performance. The only thing they cite in their brief is the mesh pad's improvement, and we've shown you the data that shows that back in the 1990s, the mesh pad performance was lower than what they say it is today. So, you know, we've got it in our briefs and it's an important issue, but either there's no development that has been identified either in performance or certainly in technology or processes, which is what is required. Okay. Do you have any further questions? No. Okay. Your time is up. Thank you. Thank you very much, Mr. Staff. Thank you, Your Honor. I'd like to offer five very brief responses and questions and one conclusion. First, we don't contend, as EPA argued, that D6 requires this every time EPA performs a review. It must follow the D2, D3 requirements because they say so when emission standards are promulgated. It kicks in once those standards are promulgated. EPA's reasoning is post hoc and the circular points, EPA points noted in the record, address only D6 and NRDC, which analyzes and addresses only D6. There is no statutory interpretation or analysis of D2, D3, which is the key language that is used to answer the question in this case. Even if it were included in the 2005 rule, which it's not, it can't sustain this one. On Judge Pillard's question, if it's more onerous and serious undertaking, every rulemaking is, EPA proposed two supplemental proposals here. The masked analysis in many ways is simpler and cleaner. It doesn't drag EPA into the quagmire of the cost issue, Judge Williams noted, but the difference is that EPA simply wants more discretion. It wants to take back the discretion Congress recognized it could not use under the toxics program because it had abused that and failed to regulate effectively. The overall questions regarding the removal of those emission data from California are so central because they highlight the problem when the statutory text is not applied. What happened in this case is that EPA then set standards based on what the worst performers achieved. They took out 15% of the industry, even more than D2, D3 refers to. They took emission data out that would have allowed them to perform the clear analysis this court and EPA are well familiar with. They used a test that has no test that they want to describe as such. In conclusion, we respectfully request that this court should not leave battery recyclers as the last word on important statutory text of the Clean Air Act governing the protections of over 187 toxic pollutants, over 150 source categories, thousands and thousands of sources that millions of Americans live near for the life of the act. To leave it as the last, should not leave it as the last word on that text that it radically changed without analyzing or addressing that text. An extremely unique, excuse me, a unique and unusual circumstance. If there are no further questions, thank you. Thank you very much. Thank you counsel. Case is submitted.