Oral argument — Airmotive Engineering Corp. v. FAA

Case No. 16-1356, Airmotive Engineering Corp. at Elle Petitioners v. Federal Aviation Administration. Ms. Ferguson for the Petitioners, Ms. Marin for the Respondent. Ms. Ferguson, good morning. Good morning. May it please the Court, my name is Laura Ferguson. I'm arguing on behalf of Airmotive Engineering Corporation and Engine Components International, whose part is subject to the Airworthiness Directive at issue. This Court should reverse and remand the Airworthiness Directive because the FAA's conclusion that Petitioners' engine cylinders present an unsafe condition is not supported by substantial evidence and is not the product of recent decision-making. The part at issue is a cylinder used on reciprocating engines which are installed in small airplanes. When this type of cylinder fails, it causes an approximately 20 percent loss of engine power to the affected engine. Most of the cylinders at issue in this case were installed on dual-engine airplanes where the loss of a single cylinder would have an even less significant impact. Under the FAA's own risk assessment policies for small airplanes, a partial loss of engine power is not an unsafe condition that would warrant an Airworthiness Directive. When the FAA issued this proposed rule making the Airworthiness Directive, it was met with overwhelmingly negative reaction. In the end, the FAA received over 650 comments, and by its own acknowledgment, 99 percent of them were opposed to the Airworthiness Directive. Well, that's understandable, isn't it? Well, you know, a number of the commenters were pilots. Pilots' associations. I do, I do, I do. It wasn't 100 percent right. And despite... So you said just then, as I thought you said, loss of power is not a basis for an Air Directive, period. The FAA, in its own risk assessment policies, does not treat it as a hazardous event. Under the small airplane risk assessment methodology, there's hazard rankings that go from 1 to 5, and it doesn't treat, and only those that are ranked 3 to 5... I'm with you, but what about the other evidence? What about the other? So what happened then is when there's this negative reaction, the FAA assembles this challenge team, and the challenge team does a risk assessment under this small airplane risk assessment, SARA system, that calculates risk based on number of fatalities. So this challenge team risk assessment, which is the only risk assessment we have in the administrative record, bases its conclusion that an Airworthiness Directive would be warranted solely, entirely, based on linking the cylinders, the ECI cylinders at issue to two fatal airplane accidents, a 2002 New Hampshire crash and a 2010 accident in the Bahamas. The problem here is that those accident reports were not included in the administrative record. They were not put on the record until three months after the final rule was issued, and one month after the petition was issued. So at that point, did you move for reconsideration? We did not move for reconsideration, no, Your Honor. There was references to the fatal crashes and the final rule, and if you actually look at the reports, the New Hampshire crash report does not assess causation. The Bahamas report does assess causation, but identifies multiple pilot errors that led to the accident. So what we have here is a condition that causes partial power loss to an engine if it occurs, and that, under the FAA's own policies, didn't warrant an Airworthiness Directive, so to create a basis for this comment. So fires, we can talk about fires, for example. The final rule says cylinder failures can cause fires. Isn't there evidence that this separation type of failure can cause fires? Well, there was only two examples they provided. How many do they have to provide? Well, of the two, in a 30-year period, neither of them involved the cylinders at issue, but one of them was apparently a 1987 accident where there was supposedly a fire, but we have no documentation of that in the record to know if a cylinder even caused that. It seemed for the moment that there were not established to be these cylinders. Nonetheless, there would be evidence that a failure by separation can cause fires. Certainly, which is why it's surprising that if the FAA was looking at 30 years and wasn't just focused on the ECI cylinders, they could only find two examples of any fires. In both cases, the planes landed safely, and we don't have documentation as to one of them to even know if the cylinder caused the fire. And I would also add that under the FAA's own risk assessment methodology, it's only an uncontained fire that is a sort of unsafe condition, a hazardous event. So fires can happen, but there's no sort of containment system. So even though they could only identify two instances in 30 years, the plane landed safely. And if it was fires they were worried about as the unsafe condition, then they had to assess the likelihood that those fires would be caused by a cylinder failure, and they didn't do that. So is your position that the agency could not issue its advisory despite the evidence it gathered because essentially it would have to be able to find a 100 percent risk before issuing the advisory, and a direct cause? What we are arguing is that the FAA has risk assessment methodologies for determining when airworthiness directives are warranted. The regulation says that an airworthiness directive should be issued when there's an unsafe condition, and so the FAA has developed policies and methodologies for deciding when particular hazards create a sufficiently unsafe condition to issue an airworthiness directive. And they didn't follow those here. And so what did they not do? What they did was wrongly attribute two fatal accidents to these cylinders, which then created a hazard. That's why I asked you, are you saying they could only issue the advisory if they could show a direct cause, as opposed to a precipitating event that leads to other things that create problems that lead to... You're speaking particularly about the two accidents. I'm speaking in response to your arguments. Well, I mean, certainly the fact that a part is installed on an airplane that has an accident doesn't mean that all the parts that are on the plane are suddenly associated with an unsafe condition. No question about it. That's not their agency's position either. Even if you were to assume that the two fatal accidents were caused by these cylinders, and as I say, those accident reports weren't part of the record, and the New Hampshire report doesn't assess causation, but even if you assumed that, then under the FAA's own methodology, they were to have assessed the likelihood that cylinder failures would cause fatalities. And they didn't do that. They assessed the likelihood that a cylinder would fail when the FAA's own methodology said that you're supposed to assess the likelihood of the hazard's effect or outcome. So they did their own risk assessment methodology wrong in terms of assessing likelihood. Well, that's because you're reading that as a direct cause requirement. And the agency isn't, right? Well, I mean, there has to be some way to test whether a part... Something goes wrong in the plane. Right. All right? The cylinder separates. It cracks. Then true pilots have been able to land safely. No one is injured. Right. But it could lead to other things. And who knows what the weather is, when the conditions occur, et cetera. And so this mosaic of evidence is enough to put out an advisory. Well, but... And furthermore, the industry itself makes some changes. And since 2009, there haven't been any more incidents. Suggesting, perhaps, that the advisory was warranted. No? Well, certainly the initial... The cylinders that are subject to the airworthiness directive were manufactured to the required specifications. That doesn't mean that improvements can't be made. Right? So, I mean, I drive a car with blind spot detectors. That doesn't mean that just because that safety technology, that improvement's available, that all the other cars that don't have that should be taken off the market. So I don't think it's reasonable to penalize them for it. Well, just because the transportation department puts out an advisory, saying you ought to be concerned about your airbags, and that adversely affects the airbag manufacturer. I mean, that's what I'm trying to get at. But there needs to be a conclusion that the part at issue, without regard to whether it could have been subsequently improved, or whether other people that made the part had a similar failure rate, whether the part at issue fails at a particular rate, and is associated with particular hazard outcomes that are severe enough, and are at risk of happening frequently enough, to warrant an airworthiness directive. So when the FAA attempts to support its rule by citing these potentially hazardous situations, they talk about fires, but they have no evidence of fires, except for these two. I thought you admitted that they had evidence of fires that may have been caused by the separation failure. I said except those two that I mentioned. And I ask you how many they have to have, and I don't think I ever got an answer to that. Well, it's not clear that the two were caused by the cylinder failures, at least as to the 1987 one. We have no idea. But if it's just one fire in 30 years caused by any cylinder failure ever, and that was a contained fire where the plane landed safely, then that's not a reason to impose an 88. You said they only found two, and I'm trying to figure out how many they had to find, and I don't think I've heard any answer to that question at this point. Well, it depends on whether you're going to use the small airplane risk assessment methodology. That would require associating them with fatalities. There were no fatalities associated with those two fire events. And if you use the Rule 8040.4a Safety Risk Management Policy, you use that matrix, which we included on our brief, where you calculate the likelihood of the fire occurring, and then there's a hazard rating based on whether it caused serious injury or fatalities. Here, they didn't cause that, and the probability would also be so low that it wouldn't put you in that red zone, which would support an airworthiness directive. So you would say they have to be in the red zone to have an airworthiness directive? Is that what you're saying? Certainly, yes, one that pulls products off the market. This airworthiness directive imposed an $88 million cost of compliance for cylinder failures, which have generally been considered to be very manageable. I think that's a lot more accurate than I was asking for. Would your answer yes to the question that I asked? Yes. Sorry, Your Honor. Now, as I read your conclusion, you're asking us to remand the matter to the FAA for a new risk assessment, one which follows the agency's own risk assessment methodology, and would be supported by substantial evidence. That's... We'd like to be able to put in our orders on remand exactly what we're telling an agency to do. Right. What is it that we would say they're supposed to do so far as complying with their own risk assessment methodologies that they didn't do in this case? Well, I don't think it's a reasonable inference from the facts to have attributed the fatalities to the ECI cylinders. So, if they have to take those off the table, they have to assess... Tell me how we would word the order that tells them how they're supposed to comply with their own methodology. That the agency should assess the risk without attributing these two fatal accidents to the cylinders, because there's not substantial evidence to support the finding that these cylinders cause fatalities. And... And I will say we take a sort of a model, the D and F Afonso Realty v. Garvey case, where the FAA had decided that a structure near an airfield created a hazard, and didn't use its own sort of hazard handbook, and followed that methodology. And this court said, we're not just required to take your say-so, FAA, that something creates a hazard. You need to follow your own methodology, you need to show your work. And remanded for them to do just that. So, in that case, we would have support to follow as a model. In addition, I just want to emphasize that the FAA, at the very least, needed to articulate a methodology, and show how the facts and the records supported this hazardous finding. And initially, the FAA, in the final rule, said we're following this Rule 8040 methodology that leads to the matrix, which we included in our brief. In the government's brief, they say, well, no, no, we're not really following that. We use this other methodology, which is this SARA methodology that the challenge team used, which turns on fatalities. So, whichever method it is, both hinged on attributing fatalities to the ECI cylinders. And also, the FAA failed to properly assess likelihood, because they looked at the likelihood of a cylinder failure rather than the likelihood of a hazardous event. So, let me ask you, just following up a little on judgment, Satel's question. I thought part of your argument here was that, even though this was a long-drawn-out consideration by the agency, and two supplemental notices were issued, various documents were made available. Yes. And these are my words, not your words. Some of the critical documents were not made available. The most critical. During the notice and comment period. You say the most critical. Right, the two accident reports. All right. And that's why I asked you, did you seek reconsideration? And you said no. Okay. So, apparently, I'll strike that. So then, I thought, in your reply brief, that you took, reply brief at 19, you pointed out that the government's statement in its brief at 35 was wrong. Where the government said, at 35, that the directive was consistent with the risk assessment using a risk matrix outlined in Order 840.4A. And the FAA brief says the FAA is not required to use that particular tool, and did not do so here. So, I thought your point was that, but in the notice of proposed rulemaking, the FAA said it was going to use the 840. Right. And how can it change now? So, those are sort of two procedural points I'm getting at. One is that data, record, reports, analysis, on which the FAA relied, was not made publicly available prior to the final rule. And then, second, its brief has acknowledged that it didn't follow the methodology it stated it was going to follow. Am I correct? That's right. So, the final rule was very clear on the point that it said, FAA Order 840.4A requires a risk assessment methodology as outlined in the order. We performed the risk assessment required by FAA Order 840.4A, and concluded that this AD was necessary. But, the problem was that the FAA never provided that risk assessment methodology. It sort of summarized it in very broad terms in the final rule, in the error-worthiness directive, but never provided it. It did cite to this challenge team assessment in the final rule as a confirmatory of its approach. But then, when the government filed its brief, it wants to distance itself entirely from this 8040.4A methodology, which is what the final rule itself said we're following. So, that is very much a problem. That's one of the examples of why we don't have reasoned decision-making here. To this day, standing here, I don't know what methodology the agency is saying supports its rule. Okay, but my follow-up to you, and then I'll stop, was in your own reply brief, you state, the final rule explicitly stated that its unsafe condition finding was based on Order 8040.4A's risk assessment methodology. Right. So, where's the conflict? I'm saying what's unreal. The government seems to be distancing itself from that. Well, but we have to look at what the agency said in the final rule, right? So, if the agency is relying on that Order 8040 methodology, then there should be a remand because, among other things, that methodology was never provided. What methodology? The risk assessment methodology done under that order, under the safety risk management policy that's reflected in that order, that assessment was never provided, was never shared, was never included in the record. Order 840.4A. Yes. That tells you what the methodology is, right? It tells you how you would do a risk assessment. Tells you what the methodology is. And the FAA, according to your own brief, says they applied that here. So, Judge Pantel was asking you, as I understood his question, what didn't they do here? So, they didn't provide the risk assessment, but based on their summary of the risk assessment that they did, they erred by ranking the cylinder failures as creating a hazardous condition because they linked it to fatalities, and they erred in applying that methodology by assessing likelihood based on likelihood of the cylinder failures rather than based on likelihood of the hazardous event, which is what their policy requires. And I see my time is nearly up. So, unless there's other questions. All right. Thank you. Ms. Myron? May it please the Court, Laura Myron for the FAA. I'd like to begin by noting that at the time this Errorworthiness Directive was issued, there was an overwhelming consensus that the separation of these cylinders posed an unsafe condition. There were reports, and this is in the record as well, reports from FAA safety inspectors, reports from the NTSB recommending corrective action in response to the separation of these cylinders. There were multiple service difficulty reports. Those are in the JA as well. There are a number of examples of them starting at 851 and more in the administrative docket as well. And there are mandatory service bulletins that are issued by the manufacturer itself detailing continuing problems caused by the separation of these cylinders. And we're not talking just about a small diminishment of power. We're talking about a violent occurrence where a piece of an engine comes apart in flight. It causes at minimum 20% power loss, but there are myriad other consequences that result, including loss of engine pressure, engine failure, damage to other parts of the engine, including the exhaust pipe. In some instances, fuel line separation, which can cause fuel to leak into the engine, which operates at temperatures above the auto-ignition point for engine fuel and can cause fire, as is documented in two instances referred to in the record. There are instances of metal contamination in the engine in increased vibration in airplane engines, which can cause, as is also noted, additional damage to other parts, including the propeller. The FAA has always considered this to be a very serious safety hazard. It has issued a number of airworthiness directives previously dealing with the failure of these, both ECI cylinders at issue here and also other kinds of cylinders on the market. To suggest at this point that there's not substantial evidence in the record to support the conclusion that this poses an unsafe condition is plainly incorrect. The agency is required to make two determinations for an airworthiness directive to be appropriate. The first is that the part at issue presents an unsafe condition. And the second is that that unsafe condition is likely to occur in other parts of the same type. And there's certainly substantial evidence in the record before this court and before the agency that would confirm both of those conclusions as appropriate in this instance. The agency did multiple risk analysis. It included in the administrative record sufficient evidence of reports from field inspectors, from pilots, from the company itself, detailing issues and problems that were continuing to occur with this particular part as a follow-on to its previous airworthiness directive that addressed this very same failure in parts of the same type number with a segment of the serial numbers at issue here. If I may just address briefly some of the procedural questions that this court had regarding the record here, I would encourage you to take a look both at the accident reports, which are in the administrative record at JA-52 and at JA-275. These are the crashes? Yes. I read both of those. I don't think that counsel misrepresented what they said. I didn't mean to imply that they misrepresented what was in them, just to suggest that they are evidence that supports the FAA's conclusion that this does create... Wait a minute. Madam, you either need to stop shaking your head. It's very distracting, or you need to leave the courtroom. Go ahead. I only meant to suggest that both of them contain evidence that would substantiate the FAA's conclusion. In both reports, witnesses account evidence that would contribute to a conclusion that the airplane had suffered a significant mechanical malfunction. In the New Hampshire report, it refers to the engine sputtering and backfiring. In the Bahamas report, it refers to white smoke trailing from behind the engine, and then goes on to conclude that... Excuse me, goes on to detail that... Lots of pilot error. Certainly, there's pilot error, but the fact that there was pilot error at a moment, as the report notes, of an unenviable emergency, does not undermine the agency's conclusion that the major mechanical failure was a precipitating event of that accident. And both of those were identified by number in the agency's rulemaking prior to the issuance of the final rule. They were publicly available at the time, and the comments that the agency received reflect that commenters had considered the accident and were suggesting to the agency that they understood what was going on at the time. So there's nothing to suggest there was procedural impropriety in the rulemaking at issue here. What about the other argument that's made about the risk assessment itself? Certainly. So I would also point the Court to Order 80404A, which is in the Joint Appendix at 570, and it details essentially a standardization and some guidance for risk assessment by the FAA across the various directors. Right. And what it says is risk assessment... I agree it tells you what to do. The question is, did you do it? Yes, Your Honor. And if you did it, did you make that available so that interested persons could comment on it? Yes, Your Honor. The risk assessment that is required is one that, as I said, takes account of severity and likelihood of outcomes. There are risk analysis in the Joint Appendix at 18, 19, and 352 that satisfy that standard. They were made available on the rulemaking docket in June of 2015, significantly prior to the final rule and with opportunity for additional comment. I take petitioners to be suggesting that what is required is use of the risk matrix that's in the appendix to 80404A, and that is not what the order requires. The order presents the risk matrix as a tool and says that it can be used, but other risk analyses may also be used. For example, an evaluation against the likelihood of a fatal outcome, which is the kind of risk analysis you see in the SARA risk analysis at 352 and in the challenge team's risk analysis at 18, 19. So just so I'm clear, that's what you meant on page 35 of your brief? Yes, Your Honor. When you say that, well, you know what you said, but it's this risk matrix that you're not required to use it and you did not do so here. Correct. But the agency did explain that the rulemaking, the determination in the rulemaking, the determination was consistent with the risk matrix, and that the risk analysis it had undertaken was consistent with 8040.4A more generally. But it also noted, as I might point you to at JA32, that the agency had also relied on the extensive evidence in the record from service difficulty reports, mandatory service bulletins, various recommendations from FAA safety inspectors and the NTSB that suggests that a significant problem was continuing to occur with these ECI cylinder assemblies and that some action was warranted. And if I might, to the point about how many accidents is necessary, no accidents are necessary in order to conclude that a part presenting a significant safety hazard merits an airworthiness directive. But in this instance, the agency relied on two fatal accidents, which the NTSB also noted resulted from the failure of these cylinders, and that is a significant amount and indicates a serious problem was ongoing with these cylinders. And if the Court has no further questions? I do have one question, and this is the certification process. Is that done at a general level? In other words, do you ever revoke the certification that a part has? So the certification process is at the outset. An airworthiness directive is the remedial measure when it becomes clear that there's an ongoing safety hazard with various parts. Does that automatically revoke the certification? No, Your Honor. But it is also the case that a part in use that is in conflict with an airworthiness directive is prohibited. So even though it doesn't revoke the airworthiness certificate, it is a mechanism that the agency uses to ensure various parts come off the market or various parts, which is one particular remedy that the agency might choose, or they're used in a certain manner. So an airworthiness directive might suggest various mitigating measures that pilots must use in order to continue using a part, various other compliance mechanisms of that sort. Okay. So is this true that this cylinder that's been certified can continue to be manufactured by other manufacturers, and it's only AEI or ECI and Airmotive that have manufactured it in such a way that you've issued this airworthiness directive? Yes. The airworthiness directive applies only to parts with this particular serial number and part number that are manufactured by ECI. It does not apply more broadly to comparable parts. Okay. And the more comparable parts might be made with a different process with respect to how the separation, the two parts of the cylinder are put together. Yes, Your Honor. And also the changes made by ECI in 2009, which resulted in significant reduction in the instance of failure, demonstrate that perhaps there was something in the manufacturing process or that was changed along the way that contributed to the failure, and so other parts, comparable parts can continue to be manufactured without violating this airworthiness directive. Okay. Any more questions? Any questions, Your Honor? Okay. Thank you. Does Ms. Ferguson have any time? Ms. Ferguson does not have any time. Why don't you take a minute if you want to. Thank you. The government refers to multiple risk assessments, and I just want to make clear that the two risk assessments they cite are that challenge team risk assessment, which is a few pages and depends entirely on these two fatal crashes. The other is a one-page worksheet, which is based, again, on the same two fatal crashes. That's at JA352. So there is no other risk assessment methodology in the record. There's no risk assessment methodology that's consistent with the Rule 8040 approach. In terms of whether the accident reports were publicly available, the Bahamas accident report was not publicly available until after the petition was filed. They provided the number, the incident number, but the report was not publicly available. And in terms of whether cylinder failures were inherently violent. What does that mean? If you had asked for it, you could not have received it? I think we would have had to have asked for it from the Bahamian government and see whether they would provide it. But the FAA cited it, right? But since these reports are the very thing that drove the two-risk analysis, it's just, I think, very irregular and not consistent with including on the record the basis for the decisions. You have not included those on the record. And I think it's because they're not supportive. In terms of whether cylinder failures are- Different from our cases where the agency is hiding the information. Certainly the report wasn't available. In other words, they cited the report. But the public couldn't really fully evaluate whether those reports supported the FAA's position. But there's no evidence that a request was made to the FAA and it refused to make the report available. No, I don't, Your Honor. But in any event, they don't support the ultimate conclusion reached. So, again, I would ask the Court to follow the approach in DNF-Afonso Realty and reverse the remand. Thank you. Thank you.