McLaughlin v. Union Oil Co.

POSNER, Circuit Judge. This case arises out of the explosion of a pressure vessel at Union Oil Company’s Chicago refinery in 1984 that killed seventeen workers. OSHA conducted an investigation, and, on January 11, 1985, cited Union Oil for three “willful” violations of the Occupational Safety and Health Act of 1970, 29 U.S.C. §§ 651 et seq., and one “serious” one. On the basis of these citations, OSHA assessed penalties against the company totaling $31,000, the maximum — 1041 — possible given the violations. See 29 U.S.C. §§ 666(a), (b). Union Oil contested the assessment and the matter was referred to the Occupational Safety and Health Review Commission, which adjudicates disputes concerning the application of the Occupational Safety and Health Act. After a hearing which lasted twenty-seven days the administrative law judge upheld most of OSHA’s citations but reduced parts of two of them (they were composite citations) from willful (maximum penalty $10,000) to serious (maximum $1,000). The resulting award of penalties totaled $22,000, which was roughly two-thirds of OSHA’s assessment. Both Union Oil and OSHA asked the Commission to review the administrative law judge’s decision. The Occupational Safety and Health Act provides that the decision (“report”) of the administrative law judge “shall become the final order of the Commission within thirty days after such report ... unless within such period any Commission member has directed that such report shall be reviewed by the Commission.” 29 U.S.C. § 661(j). The Commission is supposed to have three members but it had only one at the time, Chairman Buckley, and he declined to direct review. “[U]nder normal circumstances I would direct review____ Resolution by the Commission of the 15 alleged procedural errors and 28 alleged substantive errors would be most helpful to a Circuit Court of Appeals [sic] to which an appeal may be direct-ed____ But these are not normal circumstances for the Commission: We have lacked a full Commission since September of 1986, over 17 months; we have lacked a quorum to take official action since April 27, 1987, over ten months____ I am no longer optimistic that the Commission vacancies will be filled in the foreseeable future.” This statement was issued on the thirtieth day after the administrative law judge had submitted his decision to the Commission. Treating that decision as final by virtue of Chairman Buckley’s refusal to direct review by the Commission, OSHA petitioned this court to review the portions of the decision with which OSHA disagrees. Union Oil has cross-petitioned, and we begin with its arguments. The Commission had no quorum when it was asked to review the administrative law judge’s decision, and Union Oil argues that the absence of a quorum prevented the decision from becoming final and hence appealable. See 29 U.S.C. §§ 659(c), 660(a). The usual rule on quorums, see, e.g., FTC v. Flotill Products, Inc., 389 U.S. 179, 183, 88 S.Ct. 401, 404, 19 L.Ed.2d 398 (1967); Tagatz v. Marquette University, 861 F.2d 1040, 1042 n. * (7th Cir.1988); 28 U.S.C. § 46(d), made expressly applicable to the Occupational Safety and Health Review Commission by 29 U.S.C. § 661(f), is that a bare majority of the authorized membership is necessary for a quorum. Since Union Oil, although partially successful before the administrative law judge, was a bigger loser than OSHA and is therefore the principal petitioner in this court (though technically the cross-petitioner because OSHA filed its petition first), we are presented with the paradox of an appellant who is challenging the appeal court’s jurisdiction. The paradox is only apparent. If the administrative law judge’s decision is not final, then Union Oil is not required to comply with it, either by paying the penalties or, more important, by altering its safety practices to comply with the administrative law judge’s interpretation of the Occupational Safety and Health Act and of the regulations under it. An employer who fails to abate a violation after the Commission’s order upholding OSHA’s citation becomes final opens himself to an additional, and much more severe, penalty — $1,000 a day until the violation is abated. See 29 U.S.C. § 666(d). As is typical in labor disputes (the genre to which this case belongs), the employer, Union Oil, is on the side of delay. The general reason employers resort so frequently to delaying tactics is to make the laws for the protection — actual or apparent — of workers appear impotent. The particular reasons for delay in this case are not only a desire to stave off having to change safety practices but also the fact that Union Oil is embroiled in litigation in an Illinois state court-thirty separate law — 1042 — suits, we were told at argument — with manufacturers and other suppliers whom the injured workers and the survivors of the killed workers sued. The manufacturers and suppliers are seeking contribution from Union Oil toward the damages they have paid or may have to pay to the worker plaintiffs, and Union Oil is concerned about the possible impact on that litigation of a final determination in this one that it willfully violated the Occupational Safely and Health Act. Union Oil’s jurisdictional argument does not persuade us. Cf. Pennsylvania Steel Foundry & Machine Co. v. Secretary of Labor, 831 F.2d 1211, 1213-14 (3d Cir.1987); Marshall v. Sun Petroleum Products Co., 622 F.2d 1176, 1179-80 (3d Cir.1980). The statutory condition for the finality of an administrative law judge’s decision is merely that no member of the Commission direct, within thirty days, that the decision be reviewed — a condition satisfied here. There is nothing in the statute concerning the reasons no member might direct review in a case. One reason might be, as here, the absence of a quorum; even more dramatic would be a case where, because the Commission had no members, no member could direct review. It might be a scandal for the President or the Senate to prevent an important federal commission from operating by failing to fill vacancies in it (we are not told who was responsible for these particular vacancies), but it would not follow that decisions by the commission’s staff were unappealable. Congress wanted persons aggrieved by orders of administrative law judges of the Occupational Safety and Health Review Commission to be able to obtain judicial review even if the Commission declined — for whatever reason —to review the orders. (The worse the reason, one might suppose, the stronger the argument for judicial review of the staff decision.) Such declinations are, after all, common, even though in this case the reason was unusual. Both in Social Security disability cases and in unfair labor practice cases, appellate review within the agency (by the Appeals Council of the Social Security Administration, and by the members of the National Labor Relations Board, respectively) usually is perfunctory, so that in most such cases judicial review of the agency’s order is in effect judicial review of the administrative law judge’s order. See, e.g., NLRB v. Certified Grocers of Illinois, Inc., 806 F.2d 744 (7th Cir.1986). And, unlike those cases, persons aggrieved by orders issued by OSHRC’s administrative law judges lack even a formal entitlement to judicial review by the Commissioners; such review is strictly discretionary. Of course there is a difference between a case where no Commissioner thinks the administrative law judge’s order worthy of review and a case that apparently meets the Commission’s standards for review but the absence of a quorum persuades the remaining member that it would be futile to direct review by the Commission. And ours is the second type of case. Union Oil complains that it has been cheated of a level of review to which the statute entitles it, albeit a level of merely discretionary review — but probably discretion would have been exercised in favor of review had there been a quorum. In the unlikely event that this deprivation denied Union Oil due process of law (or some other procedural right) — despite the principle that there is no constitutional right to an appeal, even in a criminal case, see, e.g., Jones v. Barnes, 463 U.S. 745, 751, 103 S.Ct. 3308, 3312, 77 L.Ed.2d 987 (1983); Ortwein v. Schwab, 410 U.S. 656, 93 S.Ct. 1172, 35 L.Ed.2d 572 (1973) (per curiam); United States v. Puzzanghera, 820 F.2d 25, 26 (1st Cir.1987), and despite Union Oil’s failure even to argue that it has a constitutional right to anything more than a purely discretionary review of the administrative law judge’s decision — the company could, but it does not, cite the violation of its rights as a ground for our setting aside the administrative law judge’s order, which became the Commission’s order by operation of the statute. The denial of procedural rights could be a reason the order might be invalid, but it does not deprive the order of its finality and hence appealability. — 1043 — Shortly before argument we were told that a second member of the Commission had been appointed, so that the Commission now had a quorum at long last; that Union Oil had asked the Commission to reconsider Chairman Buckley’s refusal to review the administrative law judge’s order; and that the Commission had ruled (whether rightly or wrongly, compare United States v. Fornea Road Boring Co., 565 F.2d 1314 (5th Cir.1978) (per curiam), with Marshall v. Monroe & Sons, Inc., 615 F.2d 1156 (6th Cir.1980), and J.I. Hass Co. v. OSHRC, 648 F.2d 190, 193-94 (3d Cir.1981)) that it could not act on this request because jurisdiction over the case was now in this court, see 29 U.S.C. § 660(a). Union Oil asks us to remand the case to the Commission so that it can consider whether it wants to hear the appeal from the administrative law judge after all. We decline. The Commission’s quorum is a precarious one. The new member, Linda Arey, had only been given a recess appointment, and after the election the President-elect’s press secretary had announced that all recess appointees would be asked to resign on January 20. (The parties haven’t bothered to inform us whether Arey submitted her resignation and, if so, whether it was accepted and, if so, effective when.) And while two is a quorum, the Commission cannot act effectively when it has only two members unless both agree. Moreover, the Commission has a tremendous backlog because of its vacancies; we are told it may take two years for the Commission to become caught up. In these circumstances (which include Union Oil’s failure to keep us up to date on the quorum situation), arid considering the public safety implications of this case, we decline to consign it to á possible limbo by remanding it to this fragile agency. So much for procedure; let us turn to the merits. The case has a huge record full of technical mumbo-jumbo, but the essential facts are simple and the issues reasonably straightforward. The pressure vessel that exploded — an “amine absorber” used in petroleum refining — was a steel cylinder 53 feet tall and 8.5 feet in diameter. An amine absorber removes hydrogen sulfide, from a stream of liquified petroleum gas consisting of propane and butane. The process is carried out at high temperature and under great pressure. The vessel was designed to withstand pressure of up to 230 pounds per square inch and temperature of up to 140 degrees Fahrenheit, although the pressure and temperature in the vessel when in operation were lower— in the neighborhood of 200 pounds and 100 degrees. Because of the danger of serious fires at petroleum refineries, Union Oil maintained its own firefighting force. This force had either two or three components (Union Oil says two, OSHA three). One was a regular daytime crew made up of volunteers from among Union Oil’s employees who worked the day shift, which ends at 4:30. This crew was well trained, and well equipped with “firefighting personal protective equipment” (i.e., fireproof clothing). This crew was not involved in the disaster and OSHA has no complaints about its training or equipment. Next was the “shift fire crew,” a smaller, less well trained and less well equipped firefighting force that came on duty at 4:30 p.m. with the night shift. In addition to these two fire crews, operating employees of the company were instructed to remain at their posts in the event of fire and take such feasible steps as might be necessary to contain an emergency until the relevant firefighting crew (depending on the time of day) arrived. At 5:30 in the afternoon of July 23, 1984, employees of Union Oil noticed a vapor cloud fifteen to twenty feet long hovering next to the amine absorber about twelve feet above the ground. The cloud had originated from a crack a few inches long in the side of the vessel eleven feet above the ground. The shift fire crew was summoned; in the meantime the operating employees in the vicinity of the vessel tried unsuccessfully to open valves that would have relieved the pressure within the vessel and to redirect the flow of liquified petroleum gas so that it would bypass the vessel. The shift fire crew arrived, wear — 1044 — ing normal work clothes. The vessel’s operator told his supervisor that it would be necessary to shut it down. The supervisor agreed, and had taken a step toward the control center, when the vessel ruptured, emitting liquids and vapor which instantly exploded. The explosion hurled the top of the vessel more than a half mile, and, together with the ensuing fires, killed ten members of the shift fire crew, five operating employees, and two security guards, while injuring fourteen other employees. Investigation of the accident revealed that the crack that had resulted in the rupture of the vessel was due to “hydrogen stress corrosion cracking” in a microscopic area of the interior wall adjacent to a weld made in 1974. Chemical reactions within a pressure vessel can prevent hydrogen atoms from passing through the wall of the vessel as free atoms. Instead they clump into hydrogen molecules, causing the wall to harden. This hardening can disrupt the structural integrity of the wall and result in microscopic cracks that can expand to the point where, as in this case, the vessel ruptures. The area adjacent to a weld is particularly susceptible to hardening. The first issue on the merits is whether, as OSHA charged and the administrative law judge found, Union Oil’s failure to discover hydrogen stress corrosion cracking in the amine absorber before the disaster occurred violated the general duty of an employer under the Occupational Safety and Health Act to “furnish to each of his employees employment and a place of employment which are free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees.” 29 U.S.C. § 654(a)(1). There is no doubt that the hydrogen stress corrosion cracking in the vessel that exploded was a hazard, and a lethal one. Nor is there any doubt that the hazard posed by hydrogen stress corrosion cracking in pressure vessels used by the petroleum refining industry was well recognized long before the accident. And measures involving magnetic particle testing and angled ultrasound testing were available and used to find cracks caused by such corrosion while the cracks were still too small to be seen with the naked eye. Union Oil conducted visual inspections of the interior walls of its amine absorbers regularly, but when it had last inspected the one that exploded the crack that resulted in the rupture was not noticed and may still have been in its microscopic stage. Union Oil argues that until the accident hydrogen stress corrosion cracking had never been discovered in the heat-affected zone of (that is, the area immediately adjacent to) a weld in an amine absorber, a type of vessel in which pressure and temperature are lower than in vessels in which such cracking had been observed in heat-affected zones. Union Oil is correct that the fact a hazard is recognized in some circumstances doesn’t make it a recognized hazard in all others. But we cannot sav that the administrative law judge committed a clear error in finding that hydrogen stress corrosion cracking was a recognized hazard with respect to the heat-affected zones of amine absorbers. There was substantial evidence that such cracking was a pervasive hazard in pressure vessels and that vessels operated at the pressures and temperatures customary for an amine absorber were not thought to be immune from the hazard, either generally or in the heat-affected zones around welds. On the contrary, although no amine absorber had ever exploded as a result of hydrogen stress corrosion cracking, cracks due to this condition had been noted in the walls of amine absorbers in the years preceding the explosion. The vessel that exploded, moreover, had a long history of corrosion of various sorts, and there was expert testimony, which the administrative law judge was entitled to credit, that a prudent inspector would have searched the entire vessel with tools more powerful than unaided human vision for microscopic cracking. Had Union Oil conducted such a search it would have found the crack that caused the accident and would have repaired it. And it would not have made sense for Union Oil to inspect every part of the vessel except the heat-affected zone. The fact that no amine absorber had ever exploded was, we repeat, not decisive. Accidents are, fortunately, — 1045 — low-probability events, and it is desirable to prevent hazards before they produce disaster. Recognition of a hazard should not wait upon the occurrence of a fatal accident. See Brennan v. Smoke-Craft, Inc., 530 F.2d 843, 845 (9th Cir.1976); Simplex Time Recorder Co. v. Secretary of Labor, 766 F.2d 575, 588 (D.C.Cir.1985) The hazard must not only be “recognized”; it must be “likely” to cause death or serious bodily harm. These criteria, on which see generally Kelly Springfield Tire Co. v. Donovan, 729 F.2d 317 (5th Cir.1984), are not so distinct as may seem. If the risk to life or limb posed by a hazard is slight, the industry is unlikely to recognize it as a hazard — the hazard might be so slight that it had never caused an accident and was unlikely ever to do so. The converse is less clear. Even if a hazard is great, industry may not at first recognize it because it may be the result of new circumstances. Eventually there will be accidents, and then the hazard will be recognized, but not until the first accident occurs and its cause is pinned down. There is nothing new about pressure vessels, however, so really what Union Oil is arguing is that the explosion was a freak accident— the hazard that caused it was slight and therefore unrecognized. This raised a factual issue that the administrative law judge resolved defensibly. He found that the hazard of hydrogen stress corrosion cracking in an amine absorber had been widely recognized even though no such cracking had ever before resulted in a rupture and explosion, and that it would have been unreasonable for the owner of such a vessel to suppose that the heat-affected zone around a weld was immune to the peril and therefore to forgo methods of inspection that would discover microscopic cracking in that zone. Moreover, although pressure vessels, including amine absorbers, were not new, the particular problem of hydrogen stress corrosion cracking appears to have been relatively new in amine absorbers; cracks had first been noticed in amine absorbers only a few years before the explosion. The hazard was recent enough that it could not be dismissed as trivial merely because no amine absorber had exploded as yet, but not so recent that it was not yet recognized as a hazard. We cannot say that the administrative law judge was wrong to conclude that the danger of explosion was sufficiently well recognized to put a prudent inspector on guard even though no explosion had occurred. There is no “one explosion” rule in OSHA cases comparable to the fabled “one bite” rule of tort liability for injury inflicted by a house pet. The next cluster of issues involves the firefighting training and equipment of Union Oil’s shift fire crew and operating employees. These issues turn primarily on the meaning of OSHA’s regulation concerning “fire brigades.” 29 C.F.R. § 1910.156. The term is defined as follows: “fire brigade (private fire department, industrial fire department) means an organized group of employees who are knowledgeable, trained, and skilled in at least basic fire fighting operations.” 29 C.F.R. § 1910.155(c)(18). The regulation prescribing the training and equipment for fire brigades states that it “contains requirements for the organization, training, and personal protective equipment of fire brigades whenever they are established by an employer,” 29 C.F.R. § 1910.156(a)(1), but that its “personal protective equipment requirements apply only to members of fire brigades performing interior structural fire fighting.” 29 C.F.R. § 1910.156(a)(2). The administrative law judge found that the shift fire crew was part of Union Oil’s fire brigade (neither party contests this finding) but that the operating employees — even those with specific tasks in the event of fire — were not members of the fire brigade, and OSHA contests this finding. He also found that Union Oil had not given the members of the shift fire crew the training required by the fire-brigade regulation. Union Oil contests this finding; but while the evidence concerning the adequacy of the training was in conflict, certainly the administrative law judge committed no clear error in finding for OSHA on this point. Next, however, the administrative law judge held that since the shift fire crew — 1046 — was not charged with fighting interior structural fires, Union Oil was not required to supply the members of that crew with personal protective equipment (fireproof clothing). OSHA disagrees, pointing to a separate regulation which provides that “protective equipment, including personal protective equipment ... shall be provided ... wherever it is necessary by reason of hazards....” 29 C.F.R. § 1910.132(a). OSHA argued that this regulation was applicable to the shift fire crew just as much as to the operating employees — for the administrative law judge deemed the separate regulation applicable to them. But he held that the regulation dealing with fire brigades had been intended to be exclusive of the general regulation, so far as all members of such a brigade were concerned. Another OSHA regulation provides that “if a particular standard is specifically applicable to a condition, practice, means, method, operation, or process, it shall prevail over any different general standard which might otherwise be applicable,” and the administrative law judge relied on this. But the regulation continues: “on the other hand, any standard shall apply according to its terms to any employment and place of employment in any industry, even though particular standards are also prescribed for the industry, ... to the extent that none of such particular standards applies.” 29 C.F. R. § 1910.5(c). The administrative law judge’s interpretation produced a bizarre pattern: Union Oil is not required to furnish the members of its shift fire crew with equipment necessary to protect them from the hazards of fire, because they are members of the company’s fire brigade, but is required to furnish operating employees with such equipment because they are not members of the company’s fire brigade. We think the administrative law judge erred in interpreting the fire-brigade regulation to exclude operating employees expressly charged with firefighting functions although not formally assigned to either fire crew. They are ordered to stand by their posts in the event of fire and to take such measures as they can to extinguish the fire, or prevent it from spreading, until either the daytime or the shift fire crew (depending on the time of day) arrives. In other words, they have firefighting duties, and given the extraordinary hazards posed by fires in a oil refinery they ought to be trained in how to contain a refinery fire. The company did not argue that training in firefighting might distract the operating employees from duties equally or more important to the maintenance of safety and health, or give any other reason why excluding them from the fire brigade would promote the regulation’s objectives. Although not themselves an organized firefighting group, which is part of the definition of a fire brigade, see 29 C.F.R. § 1910.155(c)(18), the operating employees who have express firefighting duties are part of the organized group that consists of them plus the members of the two fire crews. However, it would not be reasonable to infer from the fact that these employees are functionally a part of the company’s fire brigade, and therefore entitled to the training that the fire-brigade regulation requires for members of the fire brigade, that the company does not have to give them personal protective equipment. By the same token, we do not think it was reasonable for the administrative law judge to infer from the fact that members of the fire shift crew are part of the company’s fire brigade that the company is not required to supply them with such equipment. These inferences may seem compelled by the statement in the fire-brigade regulation that “personal protective equipment requirements apply only to members of fire brigades performing interior structural fire fighting.” But the requirements in question appear to be ones designed specifically for interior structural fire fighting. For example, the regulation provides that “after July 1,1985, the employer shall assure that all fire brigade members wear protective clothing meeting the requirements of this paragraph when performing interior structural fire fighting,” 29 C.F.R. § 1910.156(e)(l)(i) (emphasis added), and the paragraph goes on to require water-resistant footwear, and coats and — 1047 — trousers that meet certain requirements for “Protective Clothing for Structural Fire Fighting.” 29 C.F.R. §§ 1910.-156(e)(2)(B)(ii), (e)(3)(B)(ii) (emphasis added). The appropriate clothing for fighting refinery fires may be different. The regulation also states that “the protective clothing requirements do not apply to the protective clothing worn during outside fire fighting operations (brush and forest fires, crash crew operations) or other special fire fighting activities. It is important that the protective clothing to be worn during these types of fire fighting operations reflect the hazards which are expected to be encountered by fire brigade members.” 29 C.F.R. § 1910, App. A to subpt. L, § 1910.156(7)(A). No inference can be drawn, therefore, that in prescribing personal protective equipment suitable for fighting interior structural fires OSHA intended to preclude the application of the general regulation on personal protective equipment (29 C.F.R. § 1910.132(a)) to members of fire brigades engaged in fighting other types of fire. OSHA is entitled to prevail on both points raised in its petition for review. There is no need to discuss Union Oil’s other challenges to the findings by the administrative law judge that the company had violated OSHA regulations (for example by not having an evacuation plan in the event of an uncontrollable fire). These challenges have no merit. But there is merit to its challenge to the finding that the company’s failure to inspect the pressure vessel for hydrogen stress corrosion cracking, and its failure to supply the operating employees charged with specific firefighting duties with fireproof clothing, were willful violations. The failure to inspect was, as our discussion should have made clear, negligent, but a negligent violation of the statute is merely “serious,” see 29 U.S.C. § 666(k), and for a “willful” violation more is necessary. We need not decide how much more — on which see the interesting debate between the majority and dissenting opinions in Ensign-Bickford Co. v. OSHRC, 717 F.2d 1419 (D.C. Cir.1983). There is no evidence that Union Oil was reckless or even grossly negligent. There had never been either a serious accident to an amine absorber by reason of hydrogen stress corrosion cracking, or even a case of hydrogen stress corrosion cracking in the heat-affected zone around the weld of an amine absorber; and while it was careless to suppose that such accidents could not or would not happen, or that the heat-affected zone around a weld was somehow immune from hydrogen stress corrosion cracking, OSHA utterly failed to prove that Union Oil was more than careless. As for the finding that the company was willful in failing to issue protective clothing to operating employees, it was of course not enough merely to show that the company violated the general protective-clothing regulation (29 C.F.R. § 1910.132(a)). The violation would have to be apparent at the time committed. This was not shown. It was not obvious that the general regulation would be extended to operating employees in oil refineries, given that a colorable argument existed that even members of the fire brigades at such refineries are not required to be issued fireproof clothing. A violation is not willful when it is based on a nonfrivolous interpretation of OSHA’s regulations. See Brock v. Morello Bros. Construction, Inc., 809 F.2d 161, 165 (1st Cir.1987); cf. United States v. Dahlstrom, 713 F.2d 1423, 1428 (9th Cir.1983). Totally without merit, however, is the company’s argument that the administrative law judge should have disqualified himself for bias because he ruled against the company on certain discovery matters. Indeed, the argument is frivolous. Bias cannot be inferred from a mere pattern of rulings by a judicial officer, but requires evidence that the officer had it “in” for the party for reasons unrelated to the officer’s view of the law, erroneous as that view might be. See, e.g., Jenkins v. Sterlacci, 849 F.2d 627, 634-35 (D.C.Cir.1988); Liberty Lobby, Inc. v. Dow Jones & Co., 838 F.2d 1287, 1301 (D.C.Cir.1988); United States v. Balistrieri, 779 F.2d 1191, 1202 (7th Cir.1985). Moreover, the rulings of which the company complains were in fact mild sanctions for egregious contumacy in responding to reasonable requests by OSHA for discovery. Union Oil’s opening — 1048 — brief in this court presented a thoroughly misleading picture of the relevant facts by failing even to mention that the rulings it complains about were sanctions for its abuses of discovery. The company’s hard-nosed tactic in making groundless and unfair allegations against the administrative law judge is counterproductive in this court, however successful such tactics may be in other forums. For a frivolous argument, dishonestly presented (and in an oversized brief, to boot, see Fed.R.App.P. 28(g), (h), that we allowed Union Oil to file), we hereby sanction Union Oil under Fed.R.App.P. 38 by directing it to pay OSHA’s attorney’s fees reasonably incurred in this court in defending against Part IX of Union Oil’s brief (“The AU Denied UOC a Fair Trial and Impartial Decision”). OSHA shall submit a verified statement of those fees to the Clerk of this court within fifteen days. OSHA’s petition for review is GRANTED in its entirety; Union Oil Company’s petition for review is GRANTED in part and DENIED in part — with sanctions. The case is returned to the Occupational Safety and Health Review Commission to redetermine, in light of this opinion, the penalties (presumably somewhere in between the $31,000 that OSHA assessed and the $22,-000 that the Commission awarded) that Union Oil must pay.