# Oral argument — Hearth, Patio & Barbecue Assoc v. EPA > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/argument%3A74299 ## Record - **Collection:** Oral argument - **Document type:** Oral argument - **Published:** January 28, 2021 ## Text Case No. 15-1056, Hearth, Patio & Barbecue Association Petitioner v. Environmental Protection Agency. Mr. Chung for the petitioner, Ms. Batt for the respondent. Mr. Chung, good morning. Good morning. Thank you, Judge Anderson, and may it please the Court. For nearly 30 years, EPA's air emission standards for residential wood heaters accounted for the emission test results will vary, sometimes significantly, when a single heater is tested repeatedly in one lab and when that heater is tested in multiple labs. EPA included several substantive protections in the original 1988 standards to avoid putting the burden of this variability entirely on manufacturers, which were almost all small businesses. In 2015, EPA arbitrarily removed those protections and in doing so violated basic principles of administrative law. Because the 2015 audit provisions are arbitrary and capricious, this Court should set them aside. I want to begin with the threshold question of ripeness. The issues in this case ripen for review upon the rules promulgation. This Court has repeatedly recognized that arbitrary and capricious challenges present purely legal questions. This involves the purely legal questions of whether EPA adequately explained its reversal of long-standing regulatory protections and whether EPA offered a reasoned basis for the 2015 audit provisions. These are fit for review now. There are two kind of interlocking issues that I see. One is ripeness and the fact that even though these standards have been promulgated for over five years, we have no indication that they have ever been implemented in the unlawful way that you allege. Secondly, we require as part of our rules, the federal rules of appellate procedure and our circuit rules, that standing be demonstrated in the opening brief and in the docketing statement. In your brief, your client, the association, represents members who are affected by the rule, but we have no listing of who those members are, no affidavit or declaration. And so even if we believe that the rule is problematic, we have kind of like no indication of who's injured and no indication of that an injury is going to happen. And the EPA says, well, there's kind of like this review or hearing procedure that might doesn't all of that auger towards dismissing this petition. So Judge Wilkins, regarding the first point in terms of whether EPA has implemented the audit provision in the last five years, well, in responding and attempting to respond to Hearth Patio and Barbecue Association's concerns, EPA's response was effectively, we don't plan to audit during the first five years, so there's no need to worry about that. But if that were good enough, then EPA would always be able to dodge review in these sorts of cases, because remember the Clean Air Act requires that challenges to nationally applicable rules, such as this one, have to be brought within 60 days. EPA could just sit on the rule temporarily and refuse to implement it, and it could always evade review. To your second point, and the question about the injury, well, the injury is self-evident from the record, Your Honor. The petitioner's members are the directly regulated entities, and this court has always- The members are, I mean, are they not retailers? Are they actually manufacturers who are subject to these, or are they retailers of Hearth Patio and Barbecue materials? So the petitioner's members include both manufacturers and retailers. This case is being brought on behalf of the manufacturers- I think the question here that Judge Wilkins is asking is, how do we know that? Is there anything in the record, any evidentiary material to meet your threshold burden to establish standing? So the materials in the record come from the comments that are in the joint appendix. On behalf of the manufacturers, the Hearth Patio Barbecue Association commented on the impropriety of the audit test provisions and how they failed to account for variability, and that's causing us injury now because our members have to spend tens, if not hundreds of thousands of dollars to develop products that not only meet- Before you go on to describe the injury, just, I mean, we're really at a very threshold question, or Judge Wilkins raised a threshold question about where in the record are the manufacturer members identified and their role in making the kinds of room heaters and central heaters that are subject to this rule? So in joint appendix 227 through joint appendix 233, we talk about the concerns that our manufacturers have with audit testing and- Right, but we understand the nature of the concerns. The question is the identity of your members. No members are named there. Am I right? Well, if the question is whether we specifically call out companies by name, then- Or have identified we are, you know, our members include people who manufacture what? Right. So in our opening brief, in the standing of statement, or sorry, the statement of standing, I believe it includes, and I'm, bear with me for a second. I believe it indicates that our members are the manufacturers who are directly impacted by both the standards and the audit testing provisions. Bear with me for a second. So that's on page 21. Our members include the companies that manufacture residential wood heating appliances that are regulated by this rule. And as I was initially indicating, the manufacturers have to spend money not to just meet the standards once, but they have to try to get far enough below the limit that they have some chance of surviving an audit. And on that point, it's important to emphasize that you can't really divorce the substantive standards themselves from the audit provisions for standing purposes. Our members have to design in light of both of those provisions now, and to be sure the provisions like audit testing requirements are substantive requirements. In cases like Appalachian Power, 208 F3rd at 1027, you know, this court has recognized that things like test methods and the frequency of testing for compliance are substantive requirements and that they impose obligations and duties directly on regulated entities. They can also affect the stringency of the limit. Yeah, that's an interesting question and something I was wondering about in reading the briefing. I think you just said you can't really divorce the substantive requirements from the audit, substantive standard from the audit requirement, but you're not challenging the substantive standard. So if I may put a finer point on that, you can't really divorce the emission limits, the particulate matter limits from the audit testing requirements. Both are substantive in this court's view, according to prior Clean Air Act precedents. And so yes, to your question, Judge Pillard, we are not challenging whether the particulate matter limits themselves are achievable on a one-off basis, but what we are challenging is the ability to audit and require manufacturers to meet those time and time again without any adjustment for variability during the audit testing. That's what we're challenging here. So regarding the merits, the challenged audit provisions are arbitrary and capricious for two reasons. First, EPA failed to explain why it was changing its long-standing positions on audit testing and variability, and it cannot now rely on belated justifications. Secondly, the justifications EPA now gives fall short because EPA did not actually build margins of variability into the standard, but even if it had tried, the claimed margins lack record support. So the Supreme Court recently reinforced in Regents that it is a foundational principle of administrative law that judicial review of agency action is limited to the grounds the agency invoked when it took the action. When EPA finalized the 2015 audit provisions, it never acknowledged, much less justified, why it was removing specific protections from prior rules that accounted for variability. These failures are evident in EPA's brief. In trying to show it had good reasons for changing the audit rules on pages 34 and 35, there isn't a single citation from EPA's rule, the preamble, or the response to comments. The reason is EPA did not explain in 2015 why it was no longer prohibiting audits on an interlaboratory basis, nor did it justify deleting the regulatory text that provided for adjustments to the emission limit during audits. Well, it does refer to, it says a reason that the different laboratories might produce different results is about financial incentives. So one way of understanding what EPA did here was it assumed there was variability, but when it circled back and was actually confronted with the 2010 Kirkheap-Ferguson material that Hearth, Petio, and Barbecue Association says is the substantiation of its concern, when the EPA looked at that study, it said, oh, actually, we don't think there is as much, we don't think this does substantiate the point. So variability was overblown, and another reason for variability was this conflict of interest. So we're going to put in place conflict of interest protection, and we're going to go that just as we've tested to set the initial emissions limit, we're going to assume that testing works for auditing. And that's, I mean, all of that is in the record and is an explanation for the new rule going forward. So to the first question about the conflict of interest, the discussion on its face is speculative. Essentially, EPA is trying to come up with Again, remember, the labs that conduct this testing have to pass EPA proficiency requirements. And so the explanation doesn't really make sense. If EPA is calling into question the proficiency of its labs that are accredited to conduct the testing, there are ways to address that. But speculating as to whether there are financial incentives to cook the books just doesn't, simply doesn't hold up. But more importantly, Judge Pillard, you indicated that EPA is essentially acknowledging or assuming that there is variability, and that's a really key point here. Because the absolute most that the record studies show is that EPA is disputing the claims about the magnitude of variability. But that's very different from saying that there is no variability. Now, remember, they finalized an audit provision that effectively says EPA is not going to consider variability at all during audit testing. If all of the record evidence uniformly confirms that variability exists, it is not reasonable to finalize an audit provision that ignores it. And the practical effect of this is obvious. The way the rules are set up, any exceedance of the emission limit results in revocation. Even if, for example, a wood stove is tested and it comes back during an audit at 2.1 grams per hour, that's above the 2.0 gram per hour stove, and that manufacturer loses its certification. How can that be a reasonable result when all of the record evidence shows that there is variability and EPA is not disputing that? All it's saying is maybe it's not as great as the Kirkheath-Ferguson study says it is. But again, that's very different from saying there's no variability. In addition... Can I ask a question though, if you gave the example, if a test comes back at 2.1, isn't there a process for the manufacturer to challenge that and it's not like an automatic decertification? Great question, Judge Wilkins. Here's the recourse that a manufacturer would have. It would have the ability to test two additional units, and if the average of the three tests comes in below 2.0, then that would be considered a pass. Or it can test four additional units and hope that all four tests come below 2.0. But again, plain language of the reg says that if there's an exceedance of the emission limit, EPA will start the revocation proceedings. So EPA's belated justification that the audit provisions, the fact that they no longer account for variability is no big deal because you'll have the ability to appeal. It's questionable how EPA could indicate that it could comply with its regulations if it somehow excused variability during the course of an individual audit. The exceedance means revocation. So it belies EPA's defense that maybe somewhere down the road it will take additional evidence of variability into account. That flies in the face of the regulation, which is why their ripeness defense doesn't really hold up. Can you at some point before this argument concludes identify someplace in the record before us where HBPA identifies its members or identifies a manufacturer who is a member of the association? Yes, I will do that during Ms. Poth's presentation. And Mr. Chung, there's no record of variability with respect to audits or testing of central heaters in this record. Yes, that is correct. And that was my follow-up point to the fact that EPA is bickering about the magnitude of variability. You're correct, Judge Pillard. All of what we've been talking about and all of the evidence in the record goes to Woodstove. So for two of the three appliance categories out here, what we have is a complete failure to analyze the issue altogether. There's no data on variability for furnace testing or hydronic heater testing, which is fatal to the claim that it built in Virginia. But doesn't that potentially cut the other way in the sense that there is no data that variability is a problem when testing the central heaters, which are more sophisticated equipment, their oxygen intake is more regulated, the fuel that's used to burn in them tends to be more standardized because it's not your little, you know, common room wood stove. It's a furnace. Respectfully, Your Honor, I think it actually cuts the opposite direction. So two things jump out. The first thing to keep in mind is that there are multiple test methods in play here for each category of appliance. There's both a particular matter measurement method and an operation and fueling protocol method. So there's variability in the particular matter measurement method, which is true across all three appliance categories. But with respect to the appliance fueling and protocol methods, I think it's important to distinguish between the engineering that goes into the firebox and the complexity of the operation and fueling protocol. So the firebox for the central heaters is bigger, which, if anything, introduces additional variability. So two things. One, the operation and fueling protocols are more complicated on the central heater side. On top of that, the more wood you load in, the more variability you could introduce, because how the fuel load settles as it burns, how it collapses, all of those are potential sources of variability. So the bigger the firebox, common sense suggests... There's potential, but we don't have any evidence that you've put in and comments. There's no evidence in the record on that whatsoever. So the loan comment, and this comes out in our reply brief, the loan data point on variability indicated that variability for one hydronic heater was large, was very large. And that site is in our reply brief. That is the only thing that is actually a data point showing that there's what happens when you test a hydronic heater more than once. But keep in mind, Your Honor, EPA wasn't regulating on a blank slate in 2015. It had been regulating wood stoves for almost 30 years. There was no reason for it to believe that when you test a central heater, there wasn't going to be variability or that it would magically disappear. It's incumbent on EPA to prove what has changed between 1988 and 2015, when for almost 30 years, there was a longstanding recognition that variability is a problem. And therefore, it built in an outright prohibition on doing interlab audits, for example. So it's your position that the variability of the audit testing for central heaters really follows from your position on the variability of the audit testing on the room heaters. That's in part true, but also take it from EPA itself. Look at the test methods in the rule, Joint Appendix 63 and 75, Section 15 of the two hydronic heater methods on its face say that there's too much variability to be able to determine the repeatability or the reproducibility of this method. That statement in the test method itself is fatal to EPA's claim that it built in margins of variability. Which page are you reading from? Joint Appendix 63 and 75. This is in the rule itself. Section 15 is entitled Precision and Bias. And the statement is identical in both test methods. It's two different test methods for hydronic heaters, because there are multiple types of hydronic heaters. But the sentence is one and the same. It's not possible to specify the precision of the procedure in this method, because the appliance operating and fueling protocols and the appliances themselves produce variable amounts of emissions and cannot be used to determine reproducibility or repeatability of this measurement method. This is not talking about particulate matter. Correct. This is the operating and fueling protocol test method. The particulate matter measurement method is 5G or 5H, which is a different set of problems. You said that there's variability in the testing method, but isn't one of the changes also that precisely in order to have more consistency? No, I would dispute that, Judge Pillard. EPA gave options for using different test methods for two reasons. One is within each appliance category, there are different subcategories of appliances. So in subpart AAA, for example, you have a test method for the stick-burning wood heaters, and then you have a separate test method for the pellet-burning wood heaters. For hydronic heaters, you have a default test method for hydronic heaters. Then you have a different test method for ones that have partial thermal storage, a third test method for ones that have full thermal storage, and yet a fourth test method for cordwood. Back up. My question was much simpler, which is that for the same heater, it was my understanding, and maybe this is a better question for EPA, but it was my understanding that under the 1988 rule that the laboratory could choose from among a range of different methods for the same device and testing devices. We no longer do that, right? So you're referring to the different options available for the particulate matter measurement method, which is 5G. Because you've said, just as a matter of advocacy, I'm finding this a little bit challenging. You've said they've done nothing to acknowledge variability, and I mean, it just seems like that's an overstatement. They have acknowledged it. They've improved cordwood testing. They've identified conflict of interest and addressed it. They've addressed the Kirkheap-Ferguson study in its substance. In the audit provision, they've acknowledged variability and allowed rebuttal of the evidence on which they've relied, and they've chosen more, and they changed the substantive standard from the proposed rule to the final rule to create a buffer to require less exacting performance on the part of the stove. So to say they've done nothing, I'm just having trouble understanding your response to what they did do. So the rules no longer protect manufacturers against the risk of variability, and that's because they now allow for audit testing at any lab of EPA's choosing, and there's no longer provisions that provide for adjustments to the emissions limit to account for inter-lab variability. EPA scrubbed those requirements out, which is to say that the audit testing provisions effectively assume that variability is zero. What about the building in of buffer at the front end in the substantive standard? So that is the margin defense that EPA is now raising. I think the margin defense suffers from numerous flaws. First of all, it is post hoc. Certainly with respect to furnaces, there's nothing in the record that resembles a statement that they built in a cushion into the standard. The brief candidate knowledge is that in 2015, EPA didn't even have data points showing that the number was achievable. If there's no data indicating that the limit itself is achievable, how is it possible that EPA could have built a margin or a cushion into the standards? The second point is the margins are set so low that there's very little room here for error here. How is a manufacturer supposed to build in a margin to an ever-shrinking number? You can't really slice the bologna any thinner. Thirdly, when EPA claims that it builds in margins, it refers to the same preamble statements over and over, which are at joint appendix 15 and 16. Those statements are vague at best. They cannot fairly be stated or construed as stating that EPA built specific margins into the emission limits because EPA never actually made findings about variability and did not analyze the expected amount of variability. All those statements say if we assume that commenters are correct that variability is a certain amount, that's not going to be a problem for the cleanest appliances. Paraphrasing what commenters claim is very different from making a finding on variability. Again, we have to take a step back and look at this against the regulatory backdrop of nearly 30 years. EPA knew variability was such a problem in 1988 that it dealt with it in multiple ways. It, first of all, differentiated between intra-lab variability and inter-lab variability. For intra-lab, it very explicitly said we are finding and building in a 1.0 gram per hour margin. Manufacturers, you are on notice. You better take that into account in designing your products. Don't design something that hits 7.5. You better come in under because you might not be able to survive an audit. On inter-lab variability, EPA committed to collecting data, doing a statistically sound analysis, putting those findings out for notice and comment, which it never did over the years. But because it never did that, there was an outright prohibition on doing any inter-lab audit testing for 27 years. That all gets swept under the rug effectively now because, again, there wasn't even a recognition in any of the joint appendix sites that EPA was changing the audit rules in a way that would lift the ability to do inter-lab audit testing, much less an explanation as to how it analyzed inter-lab variability. You have to view that against the lens of, you know, there was this long-standing commitment that it was supposed to study and do a fulsome analysis, and it never did that. EPA never fulfilled its promise to meaningfully analyze this issue in 1988. Fast forward to 2015, they throw a bunch of data points out there, but they never really explain how they got to the claimed margins that they're relying on. They never explain what particular data points they're emphasizing over others. A lot of this comes out in EPA's brief, but again, our view is that that's post hoc and cannot be considered under cases like Regence and Chenery. But, you know, again, the claimed margins, those statements on their face don't even make sense. If anything, EPA is implicitly conceding that there's no record support for the claimed margins because all they're doing is accepting at face value commenters claim that variability is, quote, no better than one gram per hour. By approaching the issue in that way, EPA is all but admitting it doesn't really know what the true amount is, or much worse, that variability can be significantly higher than one gram per hour. If it's higher than one gram per hour, that means that the cleanest stoves at the time, which EPA was holding out to be the best of the best, i.e. the one gram per hour stoves, would still fail an audit. So the margin defense makes absolutely no sense. And then the final point, I'm way over my time, but the final point is, again, for central heaters, remember, there's no data whatsoever for EPA to base a variability finding on. So how can they possibly have built a margin into the standard when there's no data indicating achievability for the furnace standards, and there's nothing, no data points on which to build a margin for the hydronic heaters? And on top of that, the test methods say, we can't really measure the variability here. Okay. You haven't objected to the achievability of the standards. So again, this is where, you know, achievability of a standard once is very different from achievability of the standard during subsequent audits repeatedly. Is it? Yes. Your Honor, in cases like Appalachian Power, again, you know, the frequency at which with you measure compliance changes the stringency of the standard because meeting a number one time is much easier than meeting that number repeatedly, especially when you know that there is inherent variability baked into this, right? EPA is never saying that a 1.0 gram per hour stove is going to hit 1.0 every single time, whether it's tested in one lab or whether it's tested in multiple labs, quite the opposite. All of EPA's data points and record sites all acknowledge that there's variability. It's just trying to dispute the magnitude of the variability. That's very different from saying there is no variability at all. All right. All right. Are there any more questions? We'll give you some time in response, Mr. Chung. Thank you, Judge Henderson. Yeah, if there aren't, we'll hear from the EPA, Ms. Bott. Thank you, Your Honors. Good morning and may it please the Court, Simi Bott for EPA. Your Honors, during rulemaking, HPBA advocated for a buffer to be incorporated into the standards because even the best performing heaters cannot attain exactly the same test results every time. EPA listed and set standards not at the lowest levels achieved, but at levels about twice as high. Those same margins of variability are incorporated into the audit test because it is exactly the same test as what manufacturers must pass originally to certify that their devices comply with the standards. What's more, during the audit process, manufacturers can rebut a failed audit test with any relevant evidence, including on variability. HPBA's challenge to the audit provisions should be rejected for three reasons. One, the standards themselves accommodate variability. Two, EPA's technical conclusions on variability are reasonable. And three, EPA will consider evidence on variability should manufacturers present it in an audit hearing. When you say EPA's technical conclusions on variability are reasonable, what are you referring to? I'm sorry. What are you referring to? What is EPA's technical conclusion and where is it documented? Yes, Your Honor. EPA's technical conclusion is that the margins of variability built into the standards are sufficient. They account for repeatability, reproducibility, reliability, precision. I think that you can find exactly those words on pages JA-287 and JA-297, which is in the response to comments. And the current position is that the margins are adequate? Yes, that is EPA's position. And as I said, those conclusions are based on EPA's evaluation of all of the technical data submitted, not just HPBA's study, not just Kirkheide-Ferguson, but also the analyses that were submitted by another manufacturer, Woodstock Soapstone, as well as a study that EPA sponsored at the Brookhaven National Lab. Those are specific pieces of technical evidence that EPA weighed. For example, in Brookhaven, EPA was looking at variability within one lab and found that variability was much less than HPBA estimates and also less than is incorporated into the standards. And then on interlaboratory variability, the Woodstock Soapstone study shows that variability is, again, much less than HPBA argues and also less than is incorporated into the standards. So for those reasons, EPA concluded that the margins and the standards are sufficient. How can you say it's less than what's incorporated in the standard when the variability was over 1.0 grams? Well, your honor, the variability was not over 1.0 grams in the Woodstock Soapstone study. Those heaters regularly came in under two grams per hour. And some of them, yes, came in around one gram per hour, but slightly over that. Maybe I'm thinking of the Puget Sound Clean Air Agency that the EPA relied on. It had imprecision rates higher than 1.0, right? Yes, your honor. Puget Sound didn't conduct its own study, but it analyzed HPBA's data. So yes, Puget Sound did conclude looking at HPBA's data that variability could be 1.0, could be higher than 1.0. And that's because, as Puget Sound identified, there are several flaws in HPBA's study, and HPBA's data is not applicable to the new heaters that are governed by these new standards. Go ahead. I'm sorry. I didn't mean to interrupt you. I think as Puget Sound explained, HPBA's data is based on these old heaters, which were much higher grams per hour particulate matter emissions. So their variability as expressed in grams per hour is also higher. When you're looking at these new heaters with lower emissions, their variability in terms of grams per hour is also lower. That's why I think the percentage variability is more informative. And when you're looking at that, that percentage variability that is adequately accounted for by these standards, and it takes into account that the grams per hour variability are just going to be less for these new heaters because their emissions are less than grams per hour as well. What about, oh, I'm sorry. Go ahead, Judge Pillard. I was just going to follow up on the Woodstock-Sopson study talks about minimal interlaboratory variability, but none of the studies, either on Hearth Patio and Barbecue Association's part nor on the EPA's part, address the central heaters. And I guess it's really a burden question. What's your position on that? Because they say, aha, you have nothing. They want numbers. They want tests. They want you to run something that shows here are the laboratories that test these boilers, and here's how they come back and audit, and here's the accuracy numbers. And we don't have anything like that. And I guess the question is, why isn't that fatal to this rule? Your Honor, yes, it is a burden question. It is petitioner's burden to show that this would be a concern. But more than that, what we have in the record is a statement by petitioners that they expected variability for hydronic heaters, which are a type of central heater, to be similar as precision for wood stoves, those room heaters. So not only is there just a lack of data presented by petitioners on this issue, they, in fact, agree that variability would be similar. That seems like a little bit of a stretch. I mean, that's the PowerPoint you're talking about? Yes. And, you know, they're talking about it in the context of, and I don't know if the record reflects where that was presented, but they're saying, you know, we think there's a variability problem, and we don't have reason to think that there isn't a variability problem with the central heaters as well, which is quite a bit different from saying, you know, we think that the precise problem or lack thereof that we think is manifest in the room heaters is, to the exact same extent, manifest in the central boilers. Well, Your Honor, the exact words that they used is precision is likely to be similar to the precision for the wood stove test methods. So I agree that the context is precision is a concern for both of these different types of devices, the room heaters and the central heaters, but they did say precision is likely to be similar. And as I said, there is really no data showing that precision is so much more of a concern for central heaters, nor does that really make sense. I mean, these are the same category of sources in the end. These are wood heaters, they burn wood, the sampling method, the particulate matter sampling method is the same. So it is really petitioner's burden to show here in this context that there would be some sort of greater concern for the central heaters than there would be for the room heaters. I think Your Honor, Judge Pillard had pointed this out before, but EPA didn't just account for variability when setting the rates of the standards, it also accounted for variability by improving the test methods within the standards. And yes, I am referring to the test methods for sampling particulate matter emissions. As petitioner has acknowledged here, there is one test method here, it is by petitioner's own calculations 10 times more precise than one of the other methods that was available under the that EPA has decreased the likelihood of variability in the different labs. I think petitioner has said an argument today that it is just speculation, that conflict of interest is a problem, but if you look at the Woodstock Soapstone study, in that study you will see an analysis that it is so unlikely under the old rules where there was no conflict of interest that the original certifying lab was actually using the same test method as the other labs that didn't have financial interest in the outcome of the certification test. It was less than a 1% probability that these tests are being conducted in the same way, which of course would yield greater variability. Can you just list for us the ways that you believe that the record reflects that the EPA responded to this issue? You mentioned conflict of interest, you mentioned narrowed specification of which test methods could be used, you mentioned the higher initial emission limit, you mentioned the ability in an audit hearing to rebut, including with evidence of variability. Are there other things you think that we should be considering? There's something I had about the improved cordwood test, about responding to Kirkheap Ferguson, but I don't know if you see those as separate. Your Honor, I think the four items that you mentioned are the critical factors, and I'd just like to elaborate a little on that last point, the audit hearing. I think Petitioner has neglected to explain the full context of the audit hearing. It's not just that manufacturers can test two or four devices and those devices would have to meet certain levels and then they would end up passing. That's not the case. The audit hearing provision allows a presentation of any relevant evidence. Manufacturers can present test results from their research and development phase, other test results they have. The manufacturers are a third-party certification process. They're supposed to look at their information every year, so there's a lot of data that they should have at their disposal that they can marshal in this audit hearing. EPA has provided options, the two and four devices that Petitioners have mentioned, which will automatically rebut a failed audit test, but if they have evidence on variability, they can present really any evidence on variability that's relevant to their device at the audit hearing should they eventually fail an audit test, but I think as Judge Wilkins pointed out earlier, there's really no evidence that these audit provisions have ever been a problem, not in the previous rule, not for this rule, and if there is a concrete problem with variability and how it's affecting the audit provisions, that can be taken into account at later time in a challenge to EPA's particular decision in an audit, but at this time, it's clear from the facts in the record that these audit provisions are reasonable. Your Honors, I am happy to detail the flaws in HPBA's study, but I don't know if that would helpful to the court. Let me ask, Judge Pillard, do you have any more questions? Did EPA rely, I mean, I don't recall seeing this in the brief, but one of the things that's somewhat striking in the regulatory material is the fact that Washington State and Oregon and Europe and Canada have standards, I think in, I'm not wrong, in every case, as are more demanding than the EPA standard, and is that support that you're relying on here? If not, why not? Your Honor, EPA did look to other standards that were in place when deciding what to set the standards at, but I think for the particular question that petitioners are raising here, whether variability is incorporated, that's hard to tell from just those numbers of the other standards because you don't know how many test runs were required, for example, if they were an average or, you know, what their allowance and their auditing provisions, if they had any more, so we're not specifically relying on that for this argument. And let me ask you, why is an audit, a performance audit, needed at all? I mean, I think this is in Mr. Chung's briefing, you know, what they would prefer is if you just looked at the product and made sure the product matched a prototype. Why isn't that enough if the product, I mean, in a way it's implicit in the initial one representative device from a model line is enough for certification? And why isn't that enough over time to look at one representative device three years later and make sure that it's materially the same device? Your Honor, of course, EPA is also looking at the devices themselves and making sure that they don't change over time, but there were other commenters and particularly the states and local governments that are represented by the group of amici here that were concerned about just the certification test process, that it's not being done by EPA, it's being done by the manufacturers, and they had concerns about, you know, whether there was opportunity for bias there despite the conflict of interest provision and whether there was need for more oversight. And EPA's had these audit provisions on the books since the very first standards for these wood heaters, so EPA acknowledged the concerns about the certification process and kept the audit provisions on the books. All right, Judge Wilkins? No, I don't have any other questions. All right, then we'll hear from Mr. Chung for two minutes. Mr. Wilkins, and for all of us, the information he asked for. Thank you, Judge Henderson. Regarding Judge Wilkins' question, so one of the main appendix sites to look to is the PowerPoint presentation that is at issue, which Judge Pillard rightly indicated should be viewed in context. That is a presentation that HPBA gave to, I believe it was EPA back in 2010, I believe. Don't quote me on the date there because unfortunately the deck is undated, but that indicates that, you know, the HPBA is, the member companies are the manufacturers, and they're concerned about not just achievability of the standard, but also the lack of precision in the methods. Give me a JA site. 202, I would say. Does it identify any specific member company? It does not call out companies by name because in context at the time, HPBA's concern was the industry as a whole was staring down the barrel of this variability problem. It does not claim them even as HPBA members. I'd have to go back and check the surrounding slide deck to see if there are indications of, you know, whether there's the express terms, you know, these are members of HPBA. Now that is in the HPBA comment letter, but I think unfortunately the excerpts from the comments are not in the HPBA. And so what you see from a very lengthy set of comments are the provisions specific to audit testing where, you know, HPBA had already laid out in advance, like, here's what the association is, here are what our members do. They make all three types of appliances. Could you pull the entire comment letter? And we can, I guess we could get it, but if you could pull the entire comment letter and send us the full thing, that would be helpful. Yes, I'd be happy to do that. So very briefly in response to Ms. Bott's point, I think, again, it is important to place this case in context and look at the regulatory history. For decades, EPA's regulations reflected the recognition that variability is a problem, and that is why there were limits on the ability to audit. Those limits are no longer there, and there wasn't an comments or the preamble as to why it was appropriate to lift those limitations. EPA, after the fact, is pointing to different studies in the record that provide that there is a dispute about the magnitude of variability, but there's never any data point that indicates that a test is going to come back identical every single time, and that's because that data simply does not exist. I think, you know, it is incumbent on EPA against the backdrop of a longstanding recognition that this is a problem to explain to the court, how did it arrive at the claimed margins that it now says it built in? Where is the explanation in the record as to how they got to 1.0 gram per hour on the wood stove side, for example, or where they arrived at the random percentage on the central heater sides? Where is the explanation as to why that is a reasonable number, especially in light of their longstanding commitment to conduct a statistically sound analysis of variability? It seems like a thin leg to stand on to point to a letter from a manufacturer that came in after the close of the common period, which is the Woodstock soapstone, by the way. It seems like a thin leg to stand on, given a long-standing promise to do a meaningful analysis of this issue, to say, oh, well, there's a manufacturer over here who claims that his data points are clustered around the same point, but how is that reasonable to rely on that manufacturer without an explanation as to why those data points are reliable, why that should be something that they can extrapolate across to the entire industry? So, Mr. Chang, the focus of your objection is the shift from same laboratory to any laboratory. That's what you're objecting to. That is a major part of the objection, but it is also the fact that in the prior standards, there was a comprehensive approach to variability, and that was building a specific margin into the standard themselves, plus guard rails on the ability to audit. There are no longer being no interlab audits. No interlab, but here they've built in a margin. You disagree about the extent of it, and then the second thing would be the interlab, that guardrail has been removed. That's the key, no? Correct, but I would disagree. You're right, but I would disagree. We don't agree that they built a margin in. We don't think they did that at all. The statements in the preamble don't bear that out, and again, how do you build a margin in when you have no data on the central heater side? So, part of our point is the standards are set so low that if you accept our argument that they didn't actually build a margin in, it's arbitrary because all of the record evidence confirms that variability is real. So, how can you finalize an audit provision that doesn't provide for two final points on Ms. Bott's assertion about EPA being able to take any evidence whatsoever into account at a later date? Again, how does that square with the text of the rule that says that any exceedance results in revocation? There's no indication in the rule that EPA could excuse an exceedance on the ground of variability, so would they even be complying with their own regulation if in a future audit they accepted evidence from a manufacturer explaining why their stove came back at, say, 2.2 instead of 2.0? Could they really excuse that by pointing to variability? That doesn't seem fair. The final point I will make is, and this goes to Judge Pillard's point about what a reasonable compliance program might look like. Coincidentally, in EPA's response to comments, they indicate that maybe a reasonable audit program and one that would be rarely invoked is one that would result in audits in the event of either a suspicion that the manufacturer is no longer making clones of the certified model or a suspicion that somehow a lab might have fraudulently doctored the books during testing. Well, that's at JA-289. Coincidentally, that's also the sort of program that HPBA asked for, the sort of audit provision that HPBA believed would reasonable. But that is very different from an audit provision that provides for no limitations on inter-lab audits, provides for no adjustments to the emission limit to account for variability. One that is based solely on the suspicion of fraudulent testing or a manufacturer no longer making clones of the original certified model is a very different looking audit provision than what we have here. All right. If there are no more questions, then your case is submitted. Thank you. --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/argument%3A74299. Public record. Not legal advice.