Oral argument — U.S. Department of Labor v. Tampa Electric Company

Final case of the day, U.S. Department of Labor v. Tampa Electric Company, 21-11681. Now the courtroom is getting really lonely. All right, Mr. Gilliland is here for the petitioner, Mr. Saffer for the respondent. Mr. Gilliland, you're going to start with 12 minutes. Yes, Your Honor. Very well. Good morning, Your Honors, and may it please the Court. The Commission majority in this case committed legal error when it decided that the cited provision of the HASWOPER standard did not apply to the ammonia release in this case. In doing so, it substituted its judgment for what were sufficient measures for the actual requirements of the standard and failed to defer to the Secretary's reasonable interpretation of the standard. Can I ask you a quick question before we get going? And again, this is remedial. The standard, this arbitrary, capricious, and abusive discretion, or otherwise not in accordance with the law standard, what does that last bit mean, otherwise not in accordance with the law? Everything else, arbitrary, capricious, abusive discretion, sounds really deferential. And then the last bit, otherwise not in accordance with the law, sounds sort of de novo-ish. Well, Your Honor, in this case, that portion, in our opinion, would be one of the reasons to reverse in that the Commission's failure to defer was not in accordance with Martin v. Osherick and Kaiser v. Wilkie. Okay, fair enough. You can proceed. This was a point of curiosity. The Commission's determination that the response was not an emergency response was an error for three reasons. First, the response came from outside the immediate release area, contrary to what the Commission found. The ammonia release was uncontrolled at the point that it was releasing from the sump freely into the air. And third, the release of an unknown quantity of ammonia from an unknown location where TICO at the time knew that they already needed to call 911. Can I ask you a question? So as I read the reg, the first sentence states a rule, and the next two sentences state exceptions to that rule. Is that fair? Is that a fair way of… Correct. So the first sentence in the rule, it says an emergency response… is a response effort by employees from outside the immediate release area, as you say, or other designated responders to an occurrence which results or is likely to result in an uncontrolled release of a hazardous substance, right? Correct. In the first exception, the subject of control is the substance. But in the statement of the rule, the subject of control is the release. It says in order for this to be an emergency response at all, within the meaning of the rule, it has to have—the occurrence has to have resulted in or be likely to result in an uncontrolled release of a hazardous substance. So it's not enough just for you to say, well, once the ammonia is in the air, the substance is not controlled. The question under the rule is whether the release was controlled. Why wasn't this release controlled? Are you understanding the distinction that I'm drawing? I do understand the distinction. I don't believe that that specific distinction was OSHA's intent in the standard. Well, that's okay. I'm just reading the language of the standard. It says that the release has to have been uncontrolled to fit under the rule. Before we even get to talking about exceptions, in order to fit underneath the rule, the release has to have been uncontrolled. And so I need to know why this release was uncontrolled. Given the fact that you've got this sump, you know, sort of first they filter it through the sump, and only if the sump gets oversaturated then does the valve open and the ammonia dissipate or sort of evanesce into the air. Why isn't that controlled? I do think arguably the release was controlled at the point that it went into the sump. The question is, at some point it became the release was not controlled when it released it out into the air. So what does it mean? So this is the discussion I want to have. What does it mean then for a release to be controlled? That you've got to have someone kind of like hawking it all the time and like fine-tuning the mechanism to make sure they're sort of actively steering it at the time? Or I think Tampa Electric's position is this was a controlled release because the release, you know, occurred according to plan, basically. Yeah, sure. What does it take for a release to be uncontrolled, in your view? Well, I would say that there are a number of ways it could be controlled in the sense that the guidance talks about, you know, there are times where a liquid is released but is immediately absorbed, right? Or if there's sort of local exhaust ventilation, the release would then immediately go into the ventilation, right? And that would be controlled. But so in your view, any release of a gaseous substance into the air is uncontrolled? Any release of a gaseous substance into the air is uncontrolled. That doesn't necessarily mean that the standard applies because of the third exception, right? Right, but within the meaning of the rule itself, any release of a gaseous substance into the air is uncontrolled. No matter, even if you've got like some guy at a computer who sees it going through the sump and then this other thing, and then he says, oh, it's about to hit the valve, and I'm going to kind of like play with the trigger to let like a little bit out, then a little bit out, then a little bit out, then a little bit out, that's uncontrolled. I just want to make sure that I understand. That's an uncontrolled release. I guess what you're saying is because the substance itself is uncontrolled, the release is uncontrolled. Well, that would mean any release is uncontrolled, wouldn't it? Yeah. Yes. Yes. Position. Yes, any release. Any release at all. Of a gas that flows through the air. In this circumstance, any release immediately is uncontrolled. It would be uncontrolled. It would be controlled at the point that they shut off that valve. But again, that doesn't necessarily mean that the standard applies. Well, the release is still uncontrolled. You shut off the valve and you stop further uncontrolled. You stop the release, yes. Well, what's out there is already uncontrolled. Correct. Okay. The point I think you're driving at is why include the word uncontrolled? If any release is uncontrolled, I say uncontrolled. Your Honor, any release of a gas freely into the air is uncontrolled. There are, as we said in our— This is an uncontrolled release of a hazardous substance. If every release of a hazardous substance is uncontrolled, I say uncontrolled. Your Honor, it's not our position that every release is uncontrolled. And as we said in our reply brief, like the example I just gave, a liquid that would be released and be immediately absorbed would be controlled. So even if all gases that were released freely into the air were uncontrolled, that wouldn't make the term uncontrolled superfluous in the standard. I understand. I understand what you're saying. Your Honor, I think that— Is there any circumstance under which you would control a release? I mean, that means you generate the release and you're controlling it while it's ongoing, I suppose, and then stop it. If I understand your question, Your Honor, a release that diverted— Well, I'm reading on to Judge Card's question. A release that diverted a— If you could have a controlled release and you could have an uncontrolled release, basically, what would a controlled release look like? I think a controlled release that diverted, for example, that hazardous liquid— Well, the ammonia gets out. How could you have a controlled release of ammonia? Down there where the sump is, for example. Your position, I think, as you understood earlier, is that there is no such thing as a controlled release of gas. Yes, Your Honor. In which event, you don't even need the term uncontrolled in the rule, in the standard. At least with respect to gas. Just any release. Well, Your Honor, I think as Judge Newsom has pointed out, the standard doesn't just apply to gas. It applies to hazardous substances in many forms. So that is the argument that we made in the reply brief, that even if there is no such thing as a controlled release of gas, there are—there may be controlled releases of other substances. Can I ask you one other question about the language of the reg? This one now is about the first exception. Yes. Okay? The one that reads, responses to incidental releases of hazardous substances. Where the substance can be absorbed, neutralized, or otherwise controlled at the time of release. So that doesn't—it seems like the argument that you've been making with respect to the first exception, in part, is that the substance here, the ammonia—let's assume we've gotten past the rule that this was an uncontrolled release. Now we're wondering whether the substance can be absorbed, neutralized, or otherwise controlled. And you say ammonia once released into the atmosphere is not—the ammonia was not controlled, right? Correct. What about the fact that the language says whether the substance can be absorbed, neutralized, or otherwise controlled? Like what about this fogger apparatus that at least was theoretically possible to kind of neutralize this ammonia once released? It might not have worked. It didn't happen, but it can be neutralized by that. I understand your question. I think the way OSHA intended can to be used in that sentence is the response that was whether the employees in that situation, in the immediate release area, can contain it specifically to that release. It just seems—I don't know—it seems odd. It seems like if that's what they meant, they would have said responses to incidental releases of hazardous substances where the substance was absorbed, neutralized, or otherwise controlled. Well, I guess there has to be a decision. A decision is made to clean it up or to have an emergency response. But also, I would also point out that if the court agrees with our position that the rovers did not come from within the immediate release area, or if the court determines that they weren't in the immediate release area at the time of the release, then that second exception doesn't apply regardless. I would just get to the point of the Commission's determination that there was no potential hazard. The Commission relied on the various mechanisms that were available at the SCID and on the employees' training, the use of the windsock, the availability of the fogging system, the emergency respirators. But first, none of those measures, except arguably the windsock, protects employees before they are exposed to the hazard, which is the intent of the standard. And more importantly, the basis of the Commission's argument is that TICO used alternative protective measures to what the standard requires. But neither the Commission nor TICO are entitled to decide that these other measures are sufficient and forego the use of SCBAs, which is what the standard requires. Can I ask you one more question? I keep saying one more question, but one more question. In the first exception, employees in the immediate release area or by maintenance personnel, why weren't the rovers maintenance personnel? Even if they did come from outside the immediate release area, why don't they qualify as maintenance personnel? There may be a perfectly good answer to this. I'm just not quite sure what it is. I believe that TICO hasn't argued that they are maintenance personnel, that they were designated as the responders. I mean, it may be that because TICO hasn't made that argument, we'd have like a Chenery issue or something if we sort of were interested in that. But I was just curious if the department had a position on whether rovers qualify as maintenance personnel. It seems like they would. The department's position would be that those were the designated responders and they were, even if they were maintenance personnel, they were responding in their capacity as emergency responders when they were dispatched from the control room. Okay. I see I'm over my time. Yep, very good. You'll have the full three on rebuttal. Thank you. Mr. Sapper. Thank you, Your Honor. May it please the Court, Arthur Sapper for Tampa Electric. Thank you. The Commission found that this was a controlled release within the plain meaning of the first sentence of the definition. And that finding is not only supported by substantial evidence, it's supported by all the evidence. So can I ask you a question just to kind of like play devil's advocate on the question that I was asking your adversary? So your position is that this was a controlled release because, and tell me if I'm putting words in your mouth, because it occurred according to plan, right? You had like designed the system to work this way. Ammonia goes into sump. When sump gets saturated, it releases gradually, right? I would say according to design and intent, Your Honor. Or to be more precise, design intent. Okay. So I guess what if your design was if the ammonia pressure gets too high, the pipe will explode. But it will explode sort of like this way instead of the whole way. No, Your Honor. It doesn't explode at all. No, no, no. I wasn't suggesting that that's how you designed it. Would that be a controlled release? If a different company was in here and said, look, the release of the ammonia occurred according to plan. Because we designed these pipes so that they would have like weak seams. So that if there was, you know, an over pressurization of ammonia, it would like sort of explode in one vector out of the seam. But it wouldn't like explode the whole plant or something. Does that make sense? Is that a controlled release? Yes. But I would also say that that does not mean that the employer has no duty. As we say on page, I believe it's page 34 of our brief. There are OSHA standards that deal with uncontrolled releases such as the hypothetical that Your Honor just described. But I thought you said that would be a controlled release because it occurred according to design and intent. I'm sorry. I'm sorry. It is a controlled release. But if you had but I'm sorry, there are OSHA standards dealing with controlled release. Got it. OK. Got it. I'm sorry. I misspoke. Yeah. So it's not like a free pass just because it's a controlled release. There's 1910-1000. There's 1910-134 on respiratory protection. There's 1910.1200 on training employees on chemical hazards. There is no free pass here. The only question really in this case is whether or not the rovers should have been equipped with spacesuits and SCBAs before they responded to something that they might and correctly did conclude did not require an emergency response. That's the only issue in this case. So what do you think about if we get if we get through sentence one and now we're talking about sentence two? I mean, do you really think the rovers responded from within the immediate response area? Yes, because OSHA has told us that it's the immediate response area in the ORC letters. Is that the one where it says that the immediate response area can be the entire plant? It's the one that said that the immediate release area can be the employees assigned work area. Can be, though. Can be, right? I don't remember the exact language. I mean, it essentially says if you assign an employee to a certain area and the release occurs there and he responds to that, it's a response within the immediate release area. And that interpretation letter was not just an interpretation letter. It was an interpretation letter that was vetted by the, quote, drafters of the standard. But you don't you don't deny, right, that as a factual matter, these rovers came from a quarter mile away. I don't deny that. And that they drove and it took them 10 to 15 minutes to get there. I doubt that it was really 10 to 15 minutes. All these times are just estimates. But they came from a quarter. Your argument, counsel, is if you had a plant that was, let's say, let's say a mile across because of the various outlying areas they had. And all you had to do not to worry about this is assign the rovers to the entire area. That means a mile away would be in the immediate vicinity. Yes, Your Honor. I would also say that that's not a. How about five miles? Yes. That's not. Plants have this building for safety reasons that far away. Five miles is in the immediate vicinity so long as you make it clear that the rovers are required to check on that area. I would say, Your Honor, that if that's in accordance with the ORC letters, if that's the employees assigned working. The ORC letters say may or can. They don't say all or less be. I think we're just asking you to exercise some common sense. I mean, of course, a rover might have response duties over a can, have response duties over an entire plant. But like the Hyundai plant that's outside Montgomery, Alabama, is like a city. I mean, it's ginormous. And the idea that the rover who exists on this side of the plant, that the immediate release area is on the other side of the plant, does begin to strain credulity. And again, Your Honor, it's not a free pass. The employer still has responsibilities under other. Well, the employer would probably have to have rovers stationed closer by. And because of the time limit, they don't have time to put the gear on. I mean, it's obvious if they're a mile or two away or the next block or some such thing. They're not in the vicinity for purposes of safety, it seems. Your Honor, an employer who assigned employees to come from several miles away would be probably in violation of a lot of ocean standards. Well, a common sense violation. It could be a general duty clause violation. But it's not an emergency response. Well, by the time they got five miles to the seat of the release, there's ammonia all over the place. They're all probably crawling around with the gear on, I suppose. I suppose, Your Honor, but again... That's not the case here. Again, Your Honor, it's not a free pass. The employer would probably be in violation of other requirements. The only question is, do you automatically mount... As a matter of curiosity, when you design the plan down there that have these rovers and whatnot, everybody knows where they are with respect to various potential disasters, I'll put it that way. Does OSHA pass muster on those things? Pass muster? Well, for example, OSHA realizes if they're this distance or some other distance, they've got to have on safety gear. I'm not sure I understand. The question is, does OSHA bless your plan to have rovers here, there? No. Do they know about the plan? There's no screening of the employer's plan in advance. We find out only when OSHA decides to second-guess us. So the employer reads the standards and does its own thing, is that it? We read the standards. We read the compliance directives. We read the interpretation letters, particularly those drafted by the drafter of the standards. Well, if the Secretary is correct in this case, in the next release of ammonia, you're going to have to have rovers closer on if the obligation to wear all the gear is eliminated. I'm not sure I understand. Follow me? Well, let's suppose they should have stopped and taken 20 minutes or whatever it is to put on the safety gear in this situation. Yes. Okay. So the next episode, I suppose Tampa Electric is going to have these rovers closer by these places where you can have a release in order to avoid the time that it takes to put the gear on, for example. If the citation is affirmed, that's perhaps one thing we could do. I don't want to say no. I do want to go back to the control release issue for a moment. Can I ask you one question about that, and then I want to hear what you have to say? So I think I may know the answer that you're going to give. So am I right to take away from a comparison of the first sentence, the rule, and the second sentence, the first exception, that the first sentence makes the subject of control the release, and the second sentence makes the subject of control the substance, and that we should take meaning from those sort of disparate treatments? Exactly. And that in this case, this is a rule case, and so we ought to be concerned about whether the release itself was controlled without respect to whether the substance, once it hits the air, was controlled. Exactly, Your Honor. I was going to say as part of my prepared argument that the Secretary has here the wrong noun. The Secretary is saying that the ammonia was uncontrolled once it got out into the air. It has nothing to do with the first sentence of the definition. It asks if the release is controlled, and of course it was. And all the evidence, not just substantial evidence, but all the evidence in this case, is that that's exactly what happened here. And the Commission also, by the way, observed that Osh's view, which confuses the nouns, would make nonsense of the first sentence of the definition. Now, there are uncontrolled releases, and the preamble is chock full of them. It's talking about tank cars and railroad cars where they, for example, derail, and so their tanks are punctured. The puncturing of the tanks causes an uncontrolled release, and then you get emergency responders. That's what the standard was intended to talk about. And the idea is that the tank puncture on the railway car crash is just accidental. Who knew this was coming? It's uncontrolled. It's not designed. But my view of control is the employer handling these gases has a plan of controlling them. Yes. Of handling them in all circumstances. Yes, of course. And so it's when the unintended circumstance arises, like my colleague says, the railroad car explosion or something. Exactly. That's what the HASWAPER standard was intended to do. Because the employer is not anticipating that in its design. Exactly. The HASWAPER standard was never intended to tell employers how to run a chemical plant. But it was, I guess your point is, it was designed to incentivize them to plan, you know, sort of strategically. For the unexpected, Your Honor. Strategically. Exactly. Got it. Okay. And this focus on the wrong noun. If you take a look at OSHA's reply brief at pages 14 through 16, it affects every single argument there. It afflicts every single argument there. This focus on the wrong noun. Also, I'd like to address the term immediate release area. Also, in the first sentence of the definition. Now, OSHA wants that issue to be decided by formula. Immediate release area. Respondents, quote, close enough to notice the release occurring. Words that are not in the standard, by the way. And it's built solely on a definition in an abridged dictionary. Well, there are unabridged dictionaries, Your Honor, and we consulted one. And lo and behold, they have a broader definition, giving you the plain meaning of the term. And, by the way, that definition matches exactly what the drafters of the standard put into the ORC letters. It could be in the vicinity. And that's what happened here. I might also point out on another point. Our rovers did exactly what OSHA told them to do. OSHA told us in the preamble, this is a performance standard. It gives you, quote, flexibility, end quote. In the compliance directive, it said we're not going to prescribe a formula. You have to decide on a case-by-case basis whether something is an emergency response. You have to consider certain factors. For example, the estimated exposure or degree of danger. Well, they estimated. The size of the spill, they knew that from the fact that there was only one alarm, alarming. The training or experience of the rovers. The personal protective equipment at hand. Well, we had escape pack respirators at hand at the skid. They took into account pre-established standard operating procedures. I'm quoting from OSHA's own compliance directive. Take all these factors into account. They even used the word factors. They even, in the compliance directive, reiterated that this is a performance standard that grants you flexibility. So the commission... Of course, Your Honor. I understand your position. It may be subparative. If you've designed a system where hazardous substance does exactly what you plan to do, and the system is designed so that someone will be able to handle it for five minutes, and then after that, everybody's on their own, it gets spewed into the atmosphere and run. Now, you're saying that that's not an emergency response from five minutes and one second on. It was the other. Oh, there was some. I'm sorry. And it's controlled for five minutes. But immediately thereafter, vast and hazardous amounts of ammonia spew into the atmosphere. That's just the way the plant thought they would design it because they never thought it would go that far. That's not an emergency response. I would say this was not an emergency response, Your Honor, because the system behaved exactly as designed. It was controlling the ammonia. Not after five minutes. Oh, no, Your Honor. It was controlling the ammonia at every point in time. Now, after the ammonia is in the air. You're fighting my hypothetical. Please don't do this. The system was designed so that there would be no control of the ammonia after five minutes. They controlled it for five minutes. But it was according to design, which, as I understand your argument, is all they have to do. Just control it the way it's designed to control it for however long it controls it. And it's not an emergency response. So after five minutes, or if it was designed for two minutes, there's a vast cloud of potentially lethal ammonia released. That's not an emergency response. I believe Your Honor is asking me a hypothetical. Am I correct? Yes, that's why I asked you not to fight the hypothetical. I see. I apologize, Your Honor. I did not understand that was a hypothetical. If the system is designed to control the ammonia for five minutes, and then it essentially gives up, and it spews the ammonia all over the place, then after the five minutes in that hypothetical, I suppose it would be an uncontrolled release. All right. So why is that? I'm sorry, but why? So I thought your position earlier was that so long as the plant, like, runs according to plan, it's controlled. But in Judge Karn's hypothetical, the plant does run according to plan. It's just that it's a bad plan. It's no longer controlling the ammonia? No, no, no. See, now you're using the wrong noun. The noun is release. The noun is release. The question is whether it's a controlled release under the rule. So it doesn't matter, your position is, whether the ammonia is controlled. It matters whether the release is controlled. And in Judge Karn's hypothetical, the release is controlled for a period of time, and then all hell breaks loose. But that's how the plant was designed to work. And so I thought your position was, right, wrong, or indifferent, that's a controlled release. If the valve stopped working after five minutes, if the computer turns off after five minutes, if there's nothing that's operating on the ammonia, if the sump pump, I don't know, collapses after five minutes, I suppose in that hypothetical, you wouldn't have anything operating on the ammonia anymore. And in that strange hypothetical, you might not have a controlled release at that time. Prior to that time, I suppose you would. What about if the whole thing about controlled release boils down to whether the plant functioned according to plan? What if there really was no meaningful plan? Like what if here, Tampa Electric had no sump? It's just like when the ammonia gets, I don't know, too highly pressurized or something, we just blow it out into the atmosphere, and that's our plan. I mean, on the strongest version of your argument, that's still a controlled release because that's how Tampa Electric set it up. On a weaker version of your argument, it might be like, no, no, no. The plan has to be like a meaningful plan to guide and direct the release of this stuff. To be as candid as I'm required to be, Your Honor, it would still be a controlled release, according to the nomenclature of American industry and the safety profession. If you look at the literature that uses the term controlled release, in that hypothetical, even if you had one pressure relief valve that all of a sudden opened up to the atmosphere, it would still be a controlled release. Well, the end result is, in that hypothetical, the tort law would take over, and Tampa Electric would be sued by all sorts of people. Again, it's no free pass, Your Honor. For failing to anticipate and design a system. It's no free pass, Your Honor. And I might point out that you don't design pressure relief valves to just spew out in the atmosphere. You put them up high. Right? There's a chimney requirement, in effect, to put them up high. Why? 1910-1000, which was not cited here, which is the non-free pass OSHA reg. It's not just a free pass. The only question here is whether or not you ring the alarm and you get everybody into spacesuits and SCBAs. That's the only issue. It is not whether or not you protect your employees in the plant. You do. We were just cited here under the wrong standard. And OSHA is trying to shoehorn this case into a standard that was never designed to regulate how you run a chemical plant. Okay. I think we've carried you about six minutes over, and I think we've gotten the relevant pieces of your argument. So unless there are other questions, we'll hear from your adversary on rebuttal. Well, thank you, Your Honor. Very well. Thank you so much. All right, Mr. Gilliland, three minutes. Thank you, Your Honor. On the question of whether other standards would apply if this were a controlled release, HAZWOPER has a very specific purpose, and that purpose is, as here, to protect responders who are coming from outside of the area who may not have full information on what the release is. And so the standard is designed to address that risk, and those other standards don't focus on that risk. Also, OSHA didn't intend – the purpose of the provision suggests that OSHA did not mean the reading of controlled, that Your Honor proffered, that it's controlled because they intended it to go through the sump and out the pipe, because the purpose, again, is to protect emergency responders. In that scenario, it's not controlled in any sense that's relevant to the people who would be coming to address the risk. The tough thing about that to me is that the second sentence is very – the first exception, the second sentence is very clear in making the subject of control the substance. And the first sentence, the rule, just doesn't do it that way. It does it different. It decides to make the subject of the control the release. So it's like the drafter knew how to make the subject of control the substance. They did it in the second sentence. They just didn't do it that way in the first sentence for whatever reason. So I just don't know that we can, like, ignore the language of the first sentence that makes the subject of control the release. So it's not enough to say, well, once ammonia is released, it's uncontrolled. I think, I mean, they might say true enough, but they would just say that's just not what the regulation requires. To control the substance, they said the regulation requires us to control the release. OSHA's position would be that if the release is not aimed at controlling the substance and it's just directing it out into the atmosphere, that that is not a controlled release or a control of the substance. Okay. I also want to point out that the Commission was very inconsistent on this point because, in addition to saying that it was controlled by the SCID's various mechanisms, and it's not clear whether they're talking about the sump or the fogging system, and it also, multiple times in the decision, the Commission said the rovers were, quote, able to quickly control and contain the release after ascending the SCID. But the whole point of the standard is to protect those workers before they expose themselves to that hazard. To the point, my colleague on the other side was saying that the employers consult the compliance directive in determining these factors. I would just point out that the compliance directive is very clear that if there is limited information on, to characterize the site, or if the incident commander can't characterize the site, that it is, explicitly is an emergency response and SCBAs must be warned. So we would ask the Court to, we believe the Court can decide the remaining issues in the case based on the facts of the record, as we explained in our opening brief. So we would ask the Court to remand to the Commission with an order to affirm the citation and assess the proposed penalty. Okay, very well. Thank you both very much. That case is submitted and will be in recess until tomorrow morning at 9 a.m. All rise. Hang on. Got it? There I am. There you go. I just don't want to roll too far. Recording stopped.