Oral argument — Troy Chemical Corporation v. EPA

Case number 14-1290, Troy Chemical Corporation Petitioner v. Environmental Protection Agency. Mr. Goldberg for the petitioner, Ms. Chen for the respondent. Good morning, counsel. We'll hear from Mr. Goldberg on behalf of Troy Chemical Petitioner. Good morning, your honor. Thank you very much. It's great to be able to see everyone. This is an appeal from the inclusion on the national priorities list of the Pearsons Creek site in Newark, New Jersey. EPA listed that site based on two factors. First, an asserted threat to 0.15 miles of wetland frontage along Pearsons Creek. And second, an asserted threat to the human food chain based upon fishing observed at a pier in Brooklyn, 13 miles away from the presumed point of release across Newark Bay and New York Harbor. It's important to recognize that EPA doesn't dispute. For this site to qualify to be listed, both factors must be scored in the way that EPA scored them. If either of these two factors fails scrutiny, the listing must be vacated. On the rulemaking record, neither factor justifies the score that EPA gave it under the hazard ranking system. And that's what I will explain this morning. Turning first to the wetlands issue, the principal issue before the court is a straightforward question of administrative procedure. Did EPA provide adequate information at the proposal stage for Troy and others to meaningfully comment? Suppose we agree with you that the disclosure was inadequate. Don't you still have to show prejudice, which is to say some possibility that greater disclosure might have impacted the bottom line? Yes, Your Honor, and I have no doubt that we can make that showing. What would you have done differently? You knew exactly where the wetland was. The only thing you didn't know was the precise location of the soil bores and you didn't have the pictures. So let's assume you had both those things and we're still in the rulemaking and you're formulating your comments. What would you do differently to make your case that this is not a 0.1 mile wetland? First, Your Honor, we would have looked at the way in which EPA viewed the New Jersey determination that there is no wetland along the frontage of Pearsons Creek. That is both mischaracterized and inadequately explained in the documents that EPA presented. Second, with the additional information provided in the 90 pages of support that EPA provided with the final rule but not at the proposal, we would have been able to, first of all, verify whether each of the 18 flag locations and each of the soil borings, in fact, one, had wetlands vegetation and two, that the frontage is continuous. It's really important to recognize that in this situation. Yes. Well, you have, I mean, you now have the photos. Let's just talk about those. You've seen them. Is there anything in the photos that would strengthen your case that this is not a contiguous wetland over that tenth of a mile? No, I don't believe there's anything in the photos but, Your Honor, you know, because this record has been closed, we haven't gone to consult our what's in those photos. And most importantly, the EPA needs to demonstrate that there's at least 0.1 mile of wetlands frontage. There's plenty of evidence in this updated record indicating that the creek is filled with debris, the flow is impeded, there's been industrialization and disturbed soils. We didn't have any of that back in the time of the proposal and, in fact, we might, given this record, have commissioned our own wetlands experts to go verify EPA's finding. That we could do based on the supplemental record. We couldn't do that based on the fuzzy pink line on the map that EPA provided at the time of the proposal. Yes, correct. Wouldn't you have had more incentive to seek to do your own study and to seek to gain access to the property at issue previously on the record that you had, which is less than on a record that has more? You know, Your Honor, I think that in this setting what EPA is trying to do is reverse the burdens in the rulemaking. It's EPA's burden to move forward. We, in fact, went to EPA and said, hey, is there more information? And we were told, no, there isn't. And so we had to assume that the complete record that EPA relied upon was what was in the rulemaking. When you say that they said, no, there isn't, where is that in the record that they, that they have that representation? On page 77 of the joint appendix, footnote three to our it's also mentioned in our briefs, Your Honor. You know, it's EPA's burden to move forward and to provide adequate evidence to support its proposal. And what we said in our comments was there's not enough information to delineate the wetlands provided. Why doesn't the guidebook put aside the procedural error just in terms of whether their determination is supported by the record? Why isn't the guidebook sufficient support? They have nine soil borers, they examine all of them, they're experts on the scene, they explain what they see, and they show three positives which cover the most of the segments at issue, and they explain that the negatives are up the slope from the wetland, which seems to suggest they're coming back negative not because the vegetation is changing every hundredth of a mile, but because those borers are too far up the hill. I mean, wouldn't we say that's good enough if we were just reviewing the adequacy of the bottom line judgment? Under the circumstances, Your Honor, I don't believe that is adequate because what we're looking at is a wetland that barely exceeds the point one minimum distance. Having three borings, EPA itself makes clear it's not adequate. They rely on 18 flag locations, but the flag locations the record at the final rule makes clear only looked at vegetation, not the three characteristics that are required in New Jersey, hydrology, geology, or soils, and vegetation. And so it requires the 18 flags as well as the borings, and we have no idea where the flags are, or if the vegetation at those flags, or the soils, or the hydrology meet the requirements in the guidance. And so... Is there any reason to think that vegetation would shift? We're talking about a hundred, two hundred feet on the margin here. I'm not a scientist, but just intuitively you would think that you're not likely to see huge variances, you know, one football field down the river, the creek. With deference, Your Honor, I think that's a question that our experts should have an opportunity to respond to. They should have an opportunity to look, especially in what EPA recognizes, and Troy recognizes, is an urbanized area where there's a lot of disturbance along the creek, where there's debris, and there's all kinds of industrial activity, and a finding by the state that there is no wetland. But the finding has expired, right, by its own terms? Well, the finding expired, right, the finding was, would have allowed construction without violating the Clean Water Act, in other words, without being considered to be filling wetlands, up to 10 months before EPA's field work. I mean, in theory, it's expired, as EPA points out in this brief, but that's not, you know, if Judge Katsas' question, how quickly will things change? Over 10 months, the idea that, well, yesterday, 10 months ago, there wasn't a wetland, and today there is a jurisdictional wetland, to me, is a very difficult leap for- But how can you rest an argument on a determination that has no legal validity at the time of the comment, at the time of EPA's posting? I'm not, we're not suggesting that it's about the legal validity, Judge Wilkins, we are suggesting that the observations by New Jersey, the recognized expert on wetlands in that area should be addressed by EPA. Were EPA to, you know, have said, well, we've looked at this, and, you know, that New Jersey has missed something, maybe things have changed over time, but what the record says is, ah, we don't have to worry about that, because it says there's an open channel. I mean, it seems to me that if New Jersey had issued a letter that said that this is a wetland, and EPA relied on it, you would be here arguing that it has no legal relevance at all, because it had expired. And I would think that that would be a pretty good argument. So, I'm just, I'm just trying to get to the nub of the issue here, which is that, which is that I don't see, it would be reasonable for an agency to ignore that letter under the circumstances, wouldn't it? Well, again, Your Honor, the NPL listing process, as this court has recognized, involves the application of the hazard ranking system to particular facts. The state of New Jersey made a finding about facts. That finding, you know, may have expired by the time of EPA's to consider the factual predicate that formed the basis for the New Jersey determination. Mr. Goldberg, I'm not clear about what prevented you or your experts from taking soil samples, or core samples, or examining the vegetation, or noticing that there was a great deal of debris that had accumulated over the years in the creek. What prevented you from, your experts, from making that case? So, we would have had to obtain access, but that may have been doable. I can't say that we were absolutely 100% prevented from... Is the creek a navigable waterway? Given the Clean Water Act definition, it probably is as a practical matter. Who owns the creek bed? I don't know. I mean, below Troy, you know, when it crossed Troy's property and it was still a creek, there was an easement in favor of the city of Newark. The reason I'm asking that is because, if it is a navigable waterway, and I assume it's a tidal creek, the state and private property owners could not prevent your experts from traveling down that creek in a boat and getting out into the mud and taking soil samples. Well, you know, your honor, the record is pretty clear that the only way to get access is through the property south of Troy, through 429 Delancey. That's why EPA's contractors had to return to the site with EPA representatives to use EPA's authority to get access. I'm not saying it's impossible as to why we didn't. Again, this is a matter of the rulemaking process and EPA cutting square corners with its rulemaking. It was EPA's burden in the first instance to do more than paint the line on a map and say, now you go prove we're wrong, fellas. And, you know, that's the scenario that I think, Judge Randolph, you're spinning out here, that we may, you know, once EPA says this might be a wetland, then it becomes the burden of the responding parties to prove that it's not. Again, this is a notice and comment rulemaking. EPA needed to provide adequate information in its proposal to support the position that it wanted to take and to allow interested parties to comment. And that's our position. And so we didn't seek access at the time, although now I can assure you that if this is, if the court vacates the decision and EPA decides it wants to continue to list the site and it decides it wants to continue to list the site based upon wetland, we'll have people in waders or a boat going to verify whatever the current state of that wetland or the creek banks are and to determine whether or not it is in fact a wetland. Given your time is about to expire, do you want to say anything briefly about your other issues? Yes, your honor. The other issue raises a separate administrative law question, but one that's no less serious. And that is where there's record evidence, contrary to the assumptions in the hazard ranking system, EPA is obligated to consider that evidence. In this case, Troy presented evidence in the record, including all kinds of technical documentation indicating that it is implausible for contamination from Pearsons Creek to be a factor at a fishery 13 miles away on the American Veterans Memorial Pier in Brooklyn. EPA never responded. Mr. Goldberg, may I interrupt you there? In your reply brief, you seemed to indicate that you were challenging the validity of that regulation. Correct? The regulation, the 15 mile limit. We see there being a hierarchy of ways the court could decide this case. One is to say that the EPA's interpretation of its regulation is contrary to the purpose of the HRS. It's contrary to the preamble of the HRS. And because this court in multiple decisions, including several cited in our briefs, have said that assumptions must yield where there is record evidence to the contrary. In the event that the court finds that EPA is correct, as it argues on page 25 of its brief, that this is the only permissible interpretation, that the HRS needs to be applied literally and wrotely in every situation, then as applied, we believe this factor of the HRS is contrary to the underlying statute and therefore, as applied in this case, is invalid. Okay. So EPA says you're too late. The regulation was promulgated in 1990, I think, and you had 90 days. Did Troy own this property at the time this regulation was promulgated? Yes, your honor. The regulation was promulgated in the mid 80s and Troy came into existence and owned its property in 1980. So what's your argument regarding why you didn't have to comply with the 90 day review period? Fair question, your honor. I think there's this situation fits squarely within the line of cases by this court, recognizing that later challenges can be recognized. Most recently, the court recognized it in the NPL context in the is not something that could have been contemplated at the time. And so we feel that there's ample basis for the court to look at the validity of this factor of the HRS as applied in this case under the genuine parts and other precedent for allowing later challenges. All right. Judge Katsas, Judge Randolph, do you have any further questions? No. All right. Thank you. We'll give you some time on reply. Mr. Goldberg will hear from Ms. Chen. Thanks. Good morning, and may it please the court. Sue Chen for the United States. There's no dispute that there's at least 7,300 pounds of mercury at the Pearsons Creek site. And in scoring the site, EPA disclosed detailed documentation supporting the wetland score, and it also accounted for potential mercury migration in threatened fishing grounds. First, the wetland. And I'll talk first about the disclosure aspect and then EPA support for the wetland delineation. At proposal, EPA put Troy on notice of the wetland that's the basis for the wetland score. It provided sufficient factual detail for informed commenting by disclosing the wetland's location and length. It also disclosed the logbook and data forms, which provide detailed documentation of the area's hydrology and vegetation and soil conditions. There's no requirement to also disclose that documentation in photographic form. Why not give them the location of the soil bores? Because those GPS coordinates add no meaningful information that could affect the score. What matters... Why not? This case is on the razor's edge. We're talking about this designation rises or falls on 300 feet of difference. There's a difference between 0.1 miles and 0.15 miles. So let me... Presumably you think there can be variance within a range of a couple of hundred feet, because I assume you didn't put in nine soil bores for nothing, and six of them are coming back negative. Two points on that. The first is the margin of error is actually not as small as you think, because you need only a 0.1 mile stretch in the zone of actual contamination. We have 0.15. So I'm sorry? The difference is 0.05. Right, but on the map, that line can be up to a third shorter, and it will not make any difference to the score. The second point is what matters for the wetland score is that 0.15 mile stretch in the contaminated zone, and that information was disclosed at proposal. There was a map at JA 36, which is reproduced in our brief. It shows you the wetland in pink and the contamination in yellow. There's a logbook and the data forms, and the logbook at JA 493, along with the data forms, tell you that there are three locations. There's three soil sample locations on... And six or seven that are negative. Right. Wouldn't it, in that context, where you're talking about what seems to be a pretty small amount of land, as far as geology or hydrology goes, and different results as you move down that segment, most of which are negative, it would seem like the difference between a few hundred feet here or there could be critical. So, I think there's a misunderstanding about how the soil boring locations are used. So, let me skip ahead and talk about the basis for EPA's wetland delineation, but I do want to return to the GPS coordinates. So, EPA's basis for concluding that the wetland exists and, you know, its continuity is a holistic is that the wetland scientist is on the scene. He sees something that looks kind of like a wetland, and he's going to take a closer look. He's going to look at the hydrology. He's going to look at the vegetation, and he's going to take a soil sample to check out the soil conditions. And, you know, let's say he decides, based on his professional judgment, that that location is a wetland. The next thing he does is he's going to look around and see how long does this 30 feet look like the location that he just said is a wetland, then he's probably not going to do anything. But, if at 35 feet, things start to change and look different, he's going to take a closer look, and that might involve taking another soil sample. So, the fact that... That fact cuts against you because that expert making all of those judgments that you samples. He is taking different samples to see where the wetland ends. He needs to take samples, both in the area that he thinks is a wetland and areas that he thinks he's not. That's not because the main goal of the wetland delineation is to figure out where the wetland begins and ends. He can't take soil samples from just what areas he thinks are wetlands because that doesn't tell him about where the wetland ends. So, I mean, Troy is talking about how these soil sample locations are not next to each other. Well, they're not supposed to be. This is... That's not how the wetland delineation works. No, I understand that, but it seems like his expert judgment is he of hundreds of feet. As you say, it might have been his expert judgment that he looks at one point in a tenth of a mile segment and he knows that vegetation doesn't change within 0.05 of a mile, so he says, good enough, I'm done. But he didn't do that. He kept testing different bores and the results came back mixed. Six of them were negative. Right, and those areas are not part of the wetland, but in the area that he saw were wetland, he put down wetland flags and, I mean, that's the area that he's decided are wetland. The soil borings that showed up where he concluded is not a wetland are not part of the pink line that you on the map. Seems like it's a pretty marginal determination where the detail might have mattered. And he took account of the details. The whole point of this exercise is to figure out... Yeah, but your friends on the other side couldn't... didn't appreciate the full significance of... didn't have all the facts that bore on, you know, what do you make out of the fact that six out of the ten or nine soil bores are coming back negative? And I understand your theory. Your theory is, well, those spores are too far up the hill from the stream, but maybe not. Maybe they're coming back negative because the wetland areas, instead of being 0.1, are 0.01. But there's... Well, you look at the data forms that tell you the wetland conditions in those areas. There's also the logbook at JA 492 that summarizes the wetland scientist observations about that wetland along the creek. And you see the pink line that is drawn up there based on the GPS coordinates that EPA collected at the scene. So that's how you know that EPA determined this is a 0.15 mile stretch. What in the original disclosure described how the EPA determined that the 1.15 mile area was actually contiguous? So if you look at the map at two places. If you look at the map at JA 36, the key tells you that the wetland is shown in pink, and the line that represents the wetland is a continuous pink line. So that tells people that EPA thinks that there is a continuous stretch of wetland. Then you look at JA 492, which is in a logbook that talks about a small fringe wetland at the base of a steep slope along the east edge of Pearson's Creek. And then it describes where that is on the 429 Delancey property. The other thing I'll note is that, again, this is a wetland delineation. My question to you was what in that original disclosure tells why and how EPA determined that it was continuous? It's based on the data forms for the soil sample locations, and also the logbook which tells you about that wetland along the east edge of Pearson's Creek. That's the documentation. And again, the context is important because if EPA had seen discontinuities, it would have noted that. It noted the places when it took soil sample locations that are not wetlands. But going back to the GPS, I just wanted to make clear that you know from the logbook that there are three soil sample locations that are on the wetland. They're SB 1, 4, and 7. So by that fact, you know that these three locations are on the pink line that represents the wetland in the map. And so all that the GPS coordinates tell you is their exact placement on the pink line, which doesn't help you with disproving the wetland, and it does not help you with the GPS coordinates. It likewise shows the six other coordinates seem to be right at the stream as well. They're very close, and I do want to take this opportunity to respond to Choice's argument about SB 8, which they say lacks wetland characteristics, but is shown as being on the wetland in this map at JA 379 that you just pointed out, Judge Katsas. And setting aside the fact that Choice waived this argument by not including it in the opening brief, SB 8 is actually a good illustration of another reason why GPS coordinates are not helpful here. If you look at the data form for SB 8, which is at JA 508 to 509, it tells you that this location is quote five feet from creek on steep slope. So what happened here is that the horizontal distance between SB 8 and the wetland was pretty close, and that's why on the map it's showing up as being on the wetland. But in reality, SB 8 is actually sitting above the wetland on that steep slope. The problem of course is that GPS coordinates don't tell you elevation, just longitude and latitude, and so when it's trying to represent this three-dimensional world on a two-dimensional plane, sometimes it's here, it can be misleading. But we do have that three... Yeah, I understand that, and it is a plausible reading of the guidebook that the results that are coming back negative are coming back negative because the soil bore is elevated from the stream. But that doesn't seem self-evident, and you were the target of this designation. I would think you would surely be interested in exploring whether those results are coming back negative because the bore is higher than the stream bed, or whether they're coming back negative because the vegetation is just changing. Well actually, you do know that information because there are data forms for every single soil-boring location that describes where that location is. It also documents in a lot of detail the hydrology and vegetation and soil conditions, and you know, so that gives you a lot of information about why that area might not be a wetland. Turning now to the food chain thread, unless there are any more questions on the wetland. Okay, EPA accounted for potential mercury migration over this 15-mile target distance limit in two ways. First of all, in the response and comments, EPA explained that during major storms, Pearsons Creek flows very fast and the mercury-contaminated sediments are not contained, and so there is the potential that the creek is washing out those sediments into Pearsons Creek and the rest of the estuary, which in the middle of a major storm is not going to be sitting still either. Troy's calculation at JA-96 ignores movement in a channel. They're assuming that so EPA has shown potential mercury migration in this limit. The other way that EPA has shown that is that its regulations on scoring the food chain threat reflect the agency's judgment that mercury release can potentially migrate 15 miles downstream and threaten the fishery. Back when EPA promulgated these regulations, it had to decide where to draw the line and to make that decision, one of the factors it considered was how far a contaminant can travel downstream before being diluted to a point where it's no longer important in the risk analysis. Does it matter what the contaminant is? It does, and so elsewhere in the score, there is this thing called the persistence factor for each substance that accounts for how likely it'll degrade in that among the most persistent factors that EPA encounters at these sites. So there's not going to be a lot of degradation over the 15 miles. And let me just respond to Troy's point about challenging... Excuse me, this is the regulation 4.1.3, is that right? So there are two regulations. The actual food chain score is 4.1.3.3.1. Do you realize that in your addendum with that part 500 regulation, the only thing you left out is 4.1.3? I am pretty sure we included the exact regulation. We might not have included the 4.1.3, which is the more general regulation, but we'd be happy to file that as a supplement. I can get it. It's annoying. I had to get it off of CFR. I am very sorry, Judge Randolph. But Troy is saying that somehow it is challenging the regulations, but they have not petitioned for review of the regulations. And in their opening brief, they certify that the ruling under review is the listing decision. They didn't mention the regulations. So you can't even consider those regulations at this point because they're not for you. And I just want to make some general observations about the hazard ranking regulations, which Troy says they're very formulaic. They're very mechanical. And yes, those regulations can be formulaic, but that's a virtue and not a bug because the statute at 42 USC 9605-88A tells EPA to add sites to the national priorities list based on relative risk. So EPA has to consider a site's risks relative to other sites. I mean, that's how you know what your priorities are. And that's why it's important to use the same formula across different sites. If you fiddle around with the formula, if you say that, well, this site we're going to consider fisheries only two miles downstream, but at that site, it's going to be 15, then you really compromise EPA's ability to do the relative risk assessment that the statute requires. I'm happy to answer any more questions. All right. Judge Randolph, Judge Katsas, anything else for respondent? No. All right. Thank you. Thank you. Mr. Goldberg, we'll give you two minutes for rebuttal. Thank you very much, Judge Wilkins. Just a couple of critical points that I think would be worth the court keeping in mind. First of all, with respect to the wetlands delineation, it's important to recognize that the scientists who went to visit the wetlands in no place either assert that there's 0.15 miles of wetland or that those wetlands are continuous. Neither the handwritten notes that are the field log nor the 90 page document, which incorporates those notes along with other documentation that EPA provided at the final actually have the scientist making that finding. EPA makes that anchors without providing any indication of why that pink line is where it is. The 0.15 miles shows up in point at page 58 of the joint appendix, which is separate from the initial field notes. This idea that, oh, there's 0.15, it's this pink line. Well, wait a minute. I think Troy is entitled to ask and this court is in a position to compel EPA to make sure that it provides enough information to know where that line is anchored and why it is where it is. On the food chain point, I just point out that the reflexive application of the HRS is the same for the discharge from Troy as it would be for a mercury fever thermometer breaking and being observed flowing into the creek. EPA certainly isn't arguing that it, because of the HRS, has to automatically give that thermometer break a score of 20 on the HRS, but that's the result of its reflexive application of the hazard ranking system. And that's why it needs to be able to consider additional record evidence as Troy has submitted in this case. All right. Thank you. We will take this matter under advisement. We'll call the next case, please.