Scalia v. Wynnewood Refining

978 F.3d 1175

Opinion

FILED

United States Court of Appeals PUBLISH Tenth Circuit

UNITED STATES COURT OF APPEALS October 27, 2020

Christopher M. Wolpert

FOR THE TENTH CIRCUIT Clerk of Court

_________________________________

EUGENE SCALIA, Secretary of Labor, Petitioner/Cross-Respondent, v. No. 19-9533 WYNNEWOOD REFINING CO., LLC;

Respondent/Cross-Petitioner, and

OCCUPATIONAL SAFETY & HEALTH REVIEW COMMISSION,

Respondent.

------------------------------

CORN REFINERS ASSOCIATION; NATIONAL OILSEED PROCESSORS ASSOCIATION; AMERICAN CHEMISTRY COUNSEL; AMERICAN FUEL & PETROCHEMICAL MANUFACTURERS; AMERICAN PETROLEUM INSTITUTE,

Amici Curiae.

–––––––––––––––––––––––––––––––––––

EUGENE SCALIA, Secretary of Labor, No. 19-9578

Petitioner/Cross-Respondent,

v.

WYNNEWOOD REFINING CO., LLC;

Respondent/Cross-Petitioner, and

OCCUPATIONAL SAFETY & HEALTH REVIEW COMMISSION,

Respondent.

------------------------------

CORN REFINERS ASSOCIATION; NATIONAL OILSEED PROCESSORS ASSOCIATION; AMERICAN CHEMISTRY COUNSEL; AMERICAN FUEL & PETROCHEMICAL MANUFACTURERS; AMERICAN PETROLEUM INSTITUTE,

Amici Curiae.

_________________________________

Appeal from the Occupational Safety & Health Administration (OSHA No. 13-0644)

_________________________________

Eric J. Conn and Micah R. Smith, Conn Maciel Carey LLP, Washington, D.C., for Respondent/Cross-Petitioner.

Ronald J. Gottleib, Senior Trial Attorney, United States Department of Labor, Washington, D.C. (Kate S. O’Scannlain, Solicitor of Labor, Edmund C. Baird, Associate Solicitor of Labor for Occupational Safety and Health, Charles F. James, Counsel for Appellate Litigation, United States Department of Labor, Washington, D.C., with him on the briefs), for Petitioner/Cross-Respondent.*

Michele Schoeppe, American Chemistry Counsel, Washington, D.C.; Johnathan L. Snare and Alana Genderson, Morgan Lewis & Bockius LLP, Washington D.C.; Richard

*

In an Order filed June 18, 2020, we granted the motion of Ronald J. Gottleib to withdraw as counsel.

3

Moskowitz, American Fuel & Petrochemical Manufacturers, Washington, D.C.; and Maryam Hatcher, American Petroleum Institute, Washington, D.C., filed an amicus curiae brief on behalf of Respondent/Cross-Petitioner, for American Chemistry Counsel, American Fuel & Petrochemical Manufacturers, and American Petroleum Institute.

James H. Hulme, Mark S. Dreux, and Alexandra M. Romero, Arent Fox LLP, Washington, D.C., filed an amicus curiae brief on behalf of Respondent/Cross-Petitioner, for the Corn Refiners Association and National Oilseed Processors Association.

_________________________________

Before BACHARACH, MURPHY, and MORITZ, Circuit Judges.

_________________________________

MORITZ, Circuit Judge.

_________________________________

After a boiler exploded at a refinery, the Occupational Safety and Health Administration (OSHA) cited the refinery’s owner, Wynnewood Refining Co., LLC, (Wynnewood or Wynnewood LLC), for violating 29 C.F.R. § 1910.119, which sets forth requirements for the management of highly hazardous chemicals. The Occupational Safety and Health Review Commission (the Commission) upheld the violations. In doing so, it noted that the refinery had previously violated § 1910.119. But it determined that these prior violations occurred before Wynnewood LLC owned the refinery and that they therefore occurred under a different employer. Accordingly, the Commission did not classify the violations as “repeat[] violations” under 29 U.S.C. § 666(a), which permits increased penalties for “employer[s] who willfully or repeatedly violate[]” the regulation.

Wynnewood appeals from the Commission’s order, arguing that § 1910.119 does not apply to the boiler that exploded. Because we find that § 1910.119’s plain text unambiguously applies to the boiler, we affirm the portion of the Commission’s

4

order upholding the violations. Secretary of Labor Eugene Scalia (the Secretary) also appeals from the order, arguing that the Commission erred in failing to characterize the violations as repeat violations. Because we agree with the Commission that Wynnewood LLC is not the same employer as the refinery’s previous owner, we also affirm the portion of the order concluding that these violations were not repeat violations.

Background

Before December 2011, Wynnewood Inc., a subsidiary of Gary-Williams Energy Corporation, owned the refinery. In December 2011, CVR Energy, Inc., (CVR Energy) “acquired all the stock of Gary-Williams Energy Corporation and its subsidiaries, including Wynnewood, Inc.” App. vol. 21, 1145. After the purchase, Wynnewood Inc. became Wynnewood LLC.

In September 2012, one of the refinery’s boilers—the Wickes boiler—

exploded after too much natural gas entered its firebox, which burns gas to heat the boiler. As a result of the explosion, two employees died. Neither party asserts that the Wickes boiler contains highly hazardous chemicals. But it is connected to two other parts of the refinery that do—the fluid catalytic cracking unit (FCCU) and the alkylation unit.

Shortly after the explosion, OSHA began an inspection of the refinery. After investigating, OSHA cited Wynnewood for various violations of the Occupational Safety and Health Act (OSH Act). Relevant here, these citations included violations of § 1910.119, several of which OSHA characterized as repeat violations. An

5

administrative law judge (ALJ) affirmed all but one of these violations. But because OSHA based its repeat-violations finding on violations that occurred under Wynnewood Inc. and not Wynnewood LLC, the ALJ changed the characterization of several of the violations from repeat to serious. The Commission affirmed the ALJ’s decision.

Analysis

Both parties appeal the Commission’s decision. Wynnewood argues that the regulation at issue does not apply to the Wickes boiler and that the Commission therefore erred in affirming these violations. The Secretary argues that the Commission erred by characterizing the violations as serious rather than repeat. I. Application of § 1910.119 to the Wickes Boiler Wynnewood argues that the regulation, which creates a standard for process safety management (PSM) of highly hazardous chemicals, does not apply to the Wickes boiler. The PSM regulation—which the parties also refer to as the PSM standard—sets out “requirements [employers must follow] for preventing or minimizing the consequences of catastrophic releases of toxic, reactive, flammable, or explosive chemicals.” § 1910.119. The regulation applies only to “process[es] which involve[]” a threshold amount of highly hazardous chemicals. § 1910.119(a)(1). And the regulation provides a specific definition of “process”:

Process means any activity involving a highly hazardous chemical including any use, storage, manufacturing, handling, or the on-site movement of such chemicals, or combination of these activities. For purposes of this definition, any group of vessels which are interconnected and separate vessels which are located such that a highly

6

hazardous chemical could be involved in a potential release shall be considered a single process.

§ 1910.119(b).

Below, the Commission determined that the Wickes boiler could be part of a process even though it did not contain any highly hazardous chemicals. The Commission then analyzed the text of § 1910.119(b) and found that the Wickes boiler was part of a process covered by the regulation—in other words, that it was part of a “PSM-covered process” or was “PSM-covered”—for two independent reasons. App. vol. 21, 1150. First, the Commission determined that the Wickes boiler was interconnected with the FCCU and the alkylation unit, both of which are covered by the PSM standard. Second, and alternatively, the Commission determined that the Wickes boiler “was covered by the PSM standard because it was ‘located such that a highly hazardous chemical could be involved in a potential release.’” Id. at 1152 (quoting § 1910.119(b)).

On appeal, Wynnewood argues that (1) the Wickes boiler was not PSM-covered because it did not contain any highly hazardous chemicals, (2) the Wickes boiler was not PSM-covered because of either its interconnection with or location near a process, and (3) we should not defer to the Secretary because his new interpretation deprived Wynnewood of fair notice. And, citing Kisor v. Wilkie, Wynnewood urges us to interpret the regulation by examining its “text, structure, history, and purpose,” even if the text of the regulation is unambiguous. 139 S. Ct. 2400, 2415 (2019).

7

At the outset, we note that Wynnewood’s proposed interpretive method rests on a misunderstanding of Kisor. There, the Supreme Court emphasized that a reviewing court should defer to an agency’s interpretation of its own regulations only where the regulations are “genuinely ambiguous.” Id. at 2414. In doing so, the Court reiterated that a reviewing court cannot conclude that a regulation is genuinely ambiguous until after it thoroughly considers “the text, structure, history, and purpose of a regulation.” Id. at 2415. But Kisor did not indicate that courts must consider these sources before finding a statute unambiguous. Thus, nothing in Kisor changes our longstanding approach to regulatory interpretation: begin with the text of the regulation, and, if the meaning is clear, look no further. See Mitchell v. Comm’r, 775 F.3d 1243, 1249 (10th Cir. 2015); cf. Callahan v. U.S. Dep’t of Health & Human Servs. Through Alex Azar II, 939 F.3d 1251, 1259 n.9, 1262 (11th Cir. 2019) (noting, post-Kisor, that courts need not “consult extra[]textual evidence concerning ‘history’ and ‘purpose’” of regulation where “text is clear”; explaining that “Kisor itself disclaims any groundbreaking”). And, post-Kisor, we have determined on at least three occasions that a regulation was not ambiguous by relying on the text alone. See Sierra Club v. EPA, 964 F.3d 882, 891 (10th Cir. 2020) (citing Kisor and concluding that “regulation is not ambiguous” after analyzing only regulation’s text); In re MDL 2700 Genentech Herceptin (Trastuzumab) Mktg. & Sales Practice Litig., 960 F.3d 1210, 1234 (10th Cir. 2020) (citing Kisor and determining that “the regulatory language resolves the question” of whether regulation is ambiguous); Reyes-Vargas

8

v. Barr, 958 F.3d 1295, 1306 (10th Cir. 2020) (citing Kisor and finding regulation unambiguous based on only “plain language”).

Accordingly, when interpreting the PSM regulation, we “begin by examining the plain language of the text, giving each word its ordinary and customary meaning. If, after engaging in this textual analysis, the meaning of the regulation[] is clear, our analysis is at an end, and we must enforce the regulation[] in accordance with [its] plain meaning.” Mitchell, 775 F.3d at 1249 (internal citations omitted). If the meaning of the text is not plain, then we must look to sources outside of the regulation. Kisor, 139 S. Ct. at 2414. And if after “resort[ing] to all the standard tools of interpretation” we find that the regulation is “genuinely ambiguous,” we may apply Auer deference to OSHA’s interpretation of its regulation. Id. (discussing deference owed to agency’s interpretation of its own regulation under Auer v. Robbins, 519 U.S. 452 (1997)).

A. Whether the Wickes Boiler must Contain Highly Hazardous Chemicals

With that understanding in mind, we return to Wynnewood’s arguments.

Recall that the PSM standard applies only to processes involving a threshold amount of highly hazardous chemicals. § 1910.119(a)(1). And the term “process” is defined in relevant part as “any activity involving a highly hazardous chemical” and includes “any group of vessels which are interconnected and separate vessels which are located such that a highly hazardous chemical could be involved in a potential release.” § 1910.119(b) (emphasis added). Wynnewood first argues that the Wickes

9

boiler could not have been part of a process because it did not contain highly hazardous chemicals. The Secretary agrees that the Wickes boiler does not contain any highly hazardous chemicals, but he argues that “the presence of [a highly hazardous chemical] in a vessel is not a precondition to coverage.” Aplee. Br. 39.

In making its argument, Wynnewood insists that the definition of process in § 1910.119(b) is ambiguous and urges us to rely on extratextual sources. Yet it claims ambiguity by asserting that both it and the Secretary propose two different, permissible interpretations of the text. Notably, Wynnewood never makes a text-based argument for its interpretation of this aspect of the PSM standard. Instead, in its opening brief, Wynnewood bases its interpretation on the regulation’s preamble. And in its reply brief, Wynnewood “concede[s] the standard’s text does not contain [the] explicit statement” that “all interconnected vessels must contain [highly hazardous chemicals] to be considered part of a covered process,” and it insists that “the purpose and structure of the standard imply it, and the history of the standard demand it.” Aplt. Rep. Br. 34. But we do not consider extratextual sources if the text is clear. Mitchell, 775 F.3d at 1249. Thus, we can’t reach Wynnewood’s arguments before first considering whether the text of the regulation is ambiguous.

And so, we turn to the text of the regulation. Recall that the regulation contains two sentences.1 The first sentence of the regulation defines process as “any

1 The dissent suggests that we should ignore the first sentence and “confine”

our analysis to the second sentence. Dissent 9. But the critical question before us is whether the Wickes boiler is part of a PSM-covered process; thus, the first sentence—defining what process means—is clearly relevant to this question. Indeed,

10

activity involving a highly hazardous chemical.” § 1910.119(b). Nothing on the face of this definition requires a vessel to contain a highly hazardous chemical in order to be part of a process. And its plain terms do not suggest that we should construe this definition to apply to such a narrow category of vessels. Instead, its terms suggest the opposite—the comprehensive phrase “any activity involving” captures a wide swath of vessels in that they need only be part of an any activity that involves a highly hazardous chemical. Thus, the definition of process unambiguously includes vessels which do not contain a highly hazardous chemical. And because the text of the regulation is unambiguous, we do not consider Wynnewood’s arguments based on extratextual sources, including the preamble, and we conclude that the Wickes boiler may be considered part of a process even though it did not contain any highly hazardous chemicals. See Mitchell, 775 F.3d at 1249; cf. Peabody Twentymile Mining, LLC v. Sec’y of Labor, 931 F.3d 992, 998 (10th Cir. 2019) (“Here, the limitations that appear in the preamble do not appear in the language of the regulation, and we refuse to engraft those limitations onto the language.”).2

the Commission began its analysis by reciting both sentences. Further, we note that neither of the parties suggest we limit our consideration to only the second sentence. Wynnewood does not meaningfully analyze either sentence because it fails to make a text-based argument when discussing the PSM standard. The Secretary, for his part, includes both sentences in his analysis. Additionally, he emphasizes that the first sentence defines a process as “any activity involving a highly hazardous chemical.” Aplee. Br. 27 (emphasis in original) (quoting § 1910.119). We therefore begin our analysis with the first sentence.

2 The dissent considers extratextual sources like the preamble because it concludes that the second sentence of the regulation “does not clearly say whether a vessel can constitute part of a PSM process even when the vessel itself does not contain any [highly hazardous chemicals].” Dissent 9. But this is not a genuine

11

B. Interconnection to PSM-Covered Units Next, Wynnewood argues that the Commission erred by concluding that the Wickes boiler was part of a process because it was interconnected to the PSM-covered FCCU and alkylation unit. Wynnewood first asserts that a vessel cannot be part of a process simply because it is interconnected to a PSM-covered process. Instead, Wynnewood maintains, an interconnected vessel is not part of a process unless it poses a risk of catastrophic release of highly hazardous chemicals.

With this argument, Wynnewood focuses on the second sentence of the definition of process: “For purposes of this definition, any group of vessels which are interconnected and separate vessels which are located such that a highly hazardous chemical could be involved in a potential release shall be considered a single process.” § 1910.119(b). Wynnewood argues that the requirement that vessels be “located such that a highly hazardous chemical could be involved in the potential release” applies to both “separate vessels” and to “vessels which are interconnected” and that the text of the regulation supports its interpretation. Id. In doing so, Wynnewood invokes the series-qualifier canon: “[W]hen there is a straightforward, parallel construction that involves all nouns or verbs in a series, a prepositive or postpositive modifier normally applies to the entire series.” Aplt. Rep. Br. 32 (quoting Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of

ambiguity. Kisor, 139 S. Ct. at 2415. The plain text of the regulation does not require a vessel to contain a highly hazardous chemical. Accordingly, we will not “wave the ambiguity flag” to supplement an unambiguous text with an additional requirement. Id.

12

Legal Texts 147 (2012)). Wynnewood then asserts that “any group of vessels which are interconnected” and “separate vessels which are located” are two items in a series that are both modified by “such that a highly hazardous chemical could be involved in the potential release.”

But the series-qualifier canon does not support Wynnewood’s reading because, here, there is no “straightforward, parallel construction” of items in a series. See Potts v. Ctr. for Excellence in Higher Educ., Inc., 908 F.3d 610, 615 n.5 (10th Cir. 2018) (explaining that series-qualifier cannon does not apply “when the series is not parallel”). To accept Wynnewood’s construction, we would have to treat the phrase “such that a highly hazardous chemical could be involved in the potential release” as a postpositive modifier. But doing so would require us to consider “separate vessels which are located” as a standalone item in a series, and this phrase cannot grammatically stand on its own as a separate item in the series. Thus, we find that the text of the regulation is unambiguous: the phrase “such that a highly hazardous chemical could be involved in the potential release” applies only to “vessels which are located.” And therefore, the Commission did not err in concluding that the Secretary need not demonstrate that the Wickes boiler posed a risk of catastrophic release of highly hazardous chemicals in order to be part of a process. Rather, he need only prove that the boiler was interconnected with a PSM-covered process.

And here, the Commission found that the Wickes boiler was interconnected to the FCCU and alkylation unit, both of which are “PSM-covered processes by virtue of the flammables contained in each.” App. vol. 21, 1150. In so doing, the

13

Commission noted that “the Wickes boiler is physically connected to both units through [a] pipeline.” Id. Relying on dictionary definitions of “interconnect” and “connect,” the Commission concluded that the “indirect, physical link between the Wickes boiler and the [units] is sufficient for PSM coverage.” Id. at 1151. Although Wynnewood attempts to challenge this conclusion it its reply brief, it fails to do so in its opening brief. That is, in its opening brief, Wynnewood does not argue that the Wickes boiler was not interconnected with the FCCU or alkylation unit or that the Commission relied on the wrong definition of “interconnect.” Instead, Wynnewood’s opening brief focused on the argument discussed earlier in this section—that is, whether interconnected vessels must pose a risk of the release of highly hazardous chemical. Thus, Wynnewood waived the argument that the Wickes boiler was not interconnected with the PSM-covered FCCU and alkylation unit, and we uphold the Commission’s decision on this point. See In re Motor Fuel Temperature Sales Practices Litig., 872 F.3d 1094, 1113 n.5 (10th Cir. 2017) (“[A]rguments raised for the first time in a reply brief are waived.”).

In sum, we conclude that the Wickes boiler can be part of a process as defined in § 1910.119(b) even if it does not contain highly hazardous chemicals. We also conclude that the Secretary was not required to demonstrate that the Wickes boiler posed a risk of releasing a highly hazardous chemical in order to be part of a process through interconnection. And Wynnewood waived its argument that the Wickes boiler was not interconnected with a PSM-covered process. We therefore hold that the text of the PSM regulation supports the Commission’s finding that the Wickes

14

boiler is part of a process covered by the regulation because it is interconnected with the FCCU and the alkylation unit.3 Accordingly, we affirm the portion of the Commission’s order upholding the PSM-standard violation. II. Repeat Violation The Secretary argues that the Commission erred in characterizing several of Wynnewood’s violations as serious instead of repeat. The OSH Act increases penalties for “[a]ny employer who . . . repeatedly violates” standards promulgated under the Act. § 666(a). Repeat liability can apply to successor entities as long as there is “‘substantial continuity’ between the two enterprises.” Sharon & Walter Constr., Inc, 23 BL OSHC 1286, 1295 (No. 00-1402, 2010) (quoting Fall River Dyeing & Finishing Corp. v. NLRB, 482 U.S. 27, 43 (1987)). Here, the Secretary argues that for the purposes of § 666(a), the current iteration of Wynnewood— Wynnewood LLC—is the same employer as Wynnewood Inc., which was the refinery’s legal identity before its 2011 sale to CVR Energy. And therefore, the Secretary continues, the Commission should have considered pre-2011 violations when determining whether to characterize the violations at issue here as serious or repeat.

3 Because we affirm the violations based on our conclusion that the Wickes boiler was interconnected to a PSM-covered process, we need not reach Wynnewood’s argument that the Commission erred in finding that the Wickes boiler was part of a process because it was located “such that a highly hazardous chemical could be involved in the potential release.” § 1910.119(b). And because our holding is based on the text of the regulation rather than deference to the Secretary’s interpretation, we need not reach Wynnewood’s arguments about why we should not defer to the Secretary’s interpretation.

15

Below, the Commission determined that Wynnewood Inc. and Wynnewood LLC were not the same employer and therefore declined to find repeat liability. In reaching this conclusion, the Commission applied the three-factor substantial-continuity test from Sharon & Walter. See Sharon & Walter, 23 BL OSHC at 1295. In that case, the Commission determined that an employer can face repeat liability even if the employer’s legal identity changes between the initial and subsequent violations. Id. In doing so, Sharon & Walter adapted the substantial- continuity test used in labor law. Id. at 1294–95. Under that adapted test, the Commission considers the totality of the circumstances by analyzing three factors: (1) “the nature of the business,” including the “continuity in the type of business, products/services offered[,] and customers served” because “[s]uch continuity . . . typically indicates that the nature of the activities associated with the business and the inherent safety and health considerations are likewise unchanged”; (2) the “jobs and working conditions . . . because of [their] close correlation with particular safety and health hazards”; and (3) the “[c]ontinuity of the personnel,” focusing particularly on “the personnel who specifically control decisions related to safety and health . . . because the decisions of such personnel relate directly to the extent to which the employer complies with the [OSH Act’s] requirements.” Id. at 1295.

Applying that test to the circumstances of the change of ownership between Wynnewood Inc. and Wynnewood LLC, the Commission explained that the issue “came down to the third category[,] given that the refinery’s business, product, jobs, and working conditions were the same under both entities.” App. vol. 21, 1155. It

16

then noted that even though many supervisors at Wynnewood LLC remained the same, new “high-level executives of Wynnewood LLC’s . . . parent company . . . took an increased role in day-to-day operations at the refinery.” Id. at 1155–56. And it found that the Wynnewood LLC supervisors did not control the refinery’s safety policy. Id. Instead, it found that “these managers merely implemented the safety policies set by the previous parent company and then CVR Energy.” Id. at 1156 n.13. The Commission further recognized that the “new management focuse[d] on improving safety, health, and the proper implementation of [the] PSM [standard]” and that it made substantial investments in safety personnel and equipment, resulting in a “safety culture shift.” Id. at 1156. Based on these facts, the Commission concluded that there was not a continuity in personnel and that therefore Wynnewood LLC was not a successor of Wynnewood Inc.

On appeal, the Secretary argues that (1) the Commission legally erred by incorrectly applying the substantial-continuity test and (2) even assuming it correctly applied the test, the Commission nevertheless factually erred by concluding that the third factor, continuity of personnel, favored successor liability. We consider each argument in turn.

A. Application of the Substantial-Continuity Test The parties agree that the substantial-continuity test applies.4 But they disagree on whether the Commission correctly applied that test. The Secretary argues that the

4 Wynnewood alternatively argues that we should apply an alter-ego test, explaining that, in practice, the substantial-continuity test operates like the alter-ego

17

Commission did not correctly apply the substantial-continuity test because it (1) failed to consider the totality of the circumstances and (2) considered irrelevant evidence. As it is a question of law, we review de novo whether the Commission incorrectly applied the substantial-continuity test. Trimmer v. U.S. Dep’t of Labor, 174 F.3d 1098, 1102 (10th Cir. 1999).

As the Secretary explains, the substantial-continuity test considers the totality of the circumstances across the three factors. See Sharon & Walter, Inc., 23 BL OSHC at 1295. The Secretary argues that the Commission displaced the totality-of- the-circumstances analysis by requiring each of the three factors to favor successor liability instead of “evaluat[ing] all of the evidence and explain[ing] why” it weighed the factors as it did. Aplee. Rep. Br. 7 n.2. And he argues that “even if one or more factors does not weigh in favor of successor liability[,] the ‘totality of the circumstances’ can still establish” successor liability. Aplee. Br. 60 (quoting Fall River Dyeing & Finishing Corp., 482 U.S. at 43). Likewise, the dissent contends that we can’t affirm the Commission’s analysis because the Commission “never explained why the third factor would override the other two.” Dissent 20. But the Commission correctly explained the substantial-continuity test, never indicating that the test required continuity across all three factors. Instead, when considering the evidence in light of the three factors, the Commission explained how the change in personnel

test. But in Sharon & Walter, the Secretary made an alter-ego argument to the Commission, and the Commission instead chose to apply the substantial-continuity test. See Sharon & Walter, Inc., 23 BL OSHC at 1296 n.17. We therefore decline to apply the alter-ego test here.

18

directly impacted the safety of the refinery, thus demonstrating why it weighed that factor so heavily. Indeed, the Commission acknowledged that even though the first two factors favored a repeat violation, substantial evidence as to the third factor demonstrated that the “changeover in ownership resulted in changes in management practices, procedures, and culture significant enough to break the chain of liability stemming from Wynnewood Inc.’s previous actions.” App. vol. 21, 1156 n.13.5 We therefore reject the Secretary’s argument that the Commission failed to consider the totality of the circumstances.

Next, the Secretary argues that the Commission deviated from the substantial-

continuity test by considering irrelevant factors.6 First, the Secretary asserts—and the

5 The dissent further argues that the Commission inappropriately considered this change in culture and safety policies. In doing so, the dissent proposes that we limit our analysis of the third factor to determining “whether the employees addressing safety were the same in both the old and new entities.” Dissent 21. But the Commission appropriately criticized this “rote application” of the third factor as “paint[ing] an inaccurate picture of how safety policy was set and how safety decisions were made at the refinery.” App. vol. 21, 1156 n.13. And this criticism is consistent with Sharon & Walter, which explained that the rationale underpinning the third factor was that certain personnel “control decisions related to safety and health”—thus suggesting that this factor permits consideration of whether health and safety decisions remained the same, not simply whether personnel remained the same. 23 BL OSHC at 1295 (emphasis added).

6 In his opening brief, the Secretary suggests that the Commission relied on another type of irrelevant consideration: change in the parent company’s personnel as opposed to Wynnewood’s. However, in his reply brief, he clarifies that he does not argue that “no consideration can be given to changes in management at the corporate parent” but that instead such considerations were inappropriate given the facts of this case. Aplee. Rep. Br. 13. Thus, this argument concerns whether the Commission came to the correct conclusion, not whether the Commission correctly applied the substantial-continuity test. Accordingly, we do not address this argument here, where we determine whether the Commission correctly applied the substantial-continuity test.

19

dissent agrees—that the Commission erroneously relied on the fact that Wynnewood did not change its legal identity to avoid a repeat characterization. And it argues that a company’s motive is not part of the three-factor test articulated in Sharon & Walter. See 23 BL OSHC at 1295. But the Commission did not incorporate Wynnewood’s motive into its analysis of the three factors—instead, the Commission mentioned Wynnewood’s motive only after it analyzed the various factors and concluded that there was no substantial continuity. Because it was “not necessarily involved nor essential to determination of the case,” this statement was dicta. Tuttle v. United States (In re Tuttle), 291 F.3d 1238, 1242 (10th Cir. 2002) (quoting Rohrbaugh v. Celotex Corp., 53 F.3d 1181, 1184 (10th Cir. 1995)). Moreover, the Commission explained in Sharon & Walter that if it were to “to restrict a repeat characterization to the legally identical employing entity, [an employer] could successfully avoid a repeat characterization for future violations by simply changing its legal identity.” 23 BL OSHC at 1293. Thus, a concern about motive animated the Commission’s decision in Sharon & Walter, even if the Commission did not incorporate that concern into the final test adopted in that case. And the Commission’s dicta here merely emphasizes that its conclusion was consistent with that underlying concern; it does not indicate that the Commission misapplied the substantial-continuity test.

The Secretary further argues that the Commission “rest[ed its decision] on the misguided notion that it would be unfair or inequitable to assess heightened penalties against Wynnewood LLC based on safety lapses of its predecessor.” Aplee Br. 69.

20

But the statement the Secretary cites to support this assertion—“[W]e find these are not the appropriate circumstances to affirm a repeat characterization”—doesn’t bear the weight of his argument. Id. (alteration in original) (quoting App. vol. 21, 1156). In the paragraphs prior to this statement, the Commission analyzed the three Sharon & Walter factors, and it preceded this statement with the phrase “[i]n light of the foregoing.” App. vol. 21, 1156. Thus, it is clear that when the Commission referred to the “appropriate circumstances,” it was referring to its analysis of the Sharon & Walter factors. Accordingly, we reject the Secretary’s argument that the Commission relied on irrelevant information. Because we also reject the Secretary’s totality-of- the-circumstances argument, we conclude that the Commission did not legally err by misapplying the substantial-continuity test.

B. Substantial Evidence We now turn to the Secretary’s argument that even if the Commission did not legally err by incorrectly applying the substantial-continuity test, it nevertheless factually erred by failing to find that the third factor, continuity of personnel, favors successor liability. Below, the Commission found that the continuity-of-personnel factor did not support finding substantial continuity. In doing so, it noted that “several of the day-to-day managers were in the same positions under Wynnewood Inc. and Wynnewood LLC.” App. vol. 21, 1155. But it explained that high-level executives at CVR Energy “took an increased role in day-to-day operations at the refinery.” Id. at 1156. It found that these executives “focused on improving safety, health, and the proper implementation of [the PSM standard].” Id. As a result, the

21

refinery’s “safety culture shift[ed].” Id. Based on this analysis, the Commission found that there was no substantial continuity between Wynnewood Inc. and Wynnewood LLC.

In reviewing this finding, we note that the substantial-continuity test “is primarily factual in nature.” Fall River Dyeing & Finishing Corp., 482 U.S. at 43; see also Sharon & Walter, Inc., 23 BL OSHC at 1296 (favorably citing Fall River for this point). We review the Commission’s factual findings “under a substantial[-] evidence standard, which is satisfied if ‘“a reasonable mind” would consider the evidence adequate to support the conclusion reached.’” Jake’s Fireworks Inc. v. Acosta, 893 F.3d 1248, 1257 (10th Cir. 2018) (quoting Universal Const. Co. v. Occupational Safety & Health Rev. Comm’n, 182 F.3d 726, 732 (10th Cir. 1999)). In making this determination, “[w]e do not reweigh the evidence, second-guess the factual inferences drawn therefrom, or substitute our judgment on the credibility of witnesses.” Id. And we “must affirm a Commission determination supported by substantial evidence even though we might reach a contrary result in a proceeding de novo.” CCI, Inc. v. Occupational Safety & Health Rev. Comm’n, 688 F.2d 88, 89 (10th Cir. 1982).

In attempting to show that the Commission’s decision lacked substantial evidence, the Secretary first argues that the Commission inappropriately relied on the change in personnel of Wynnewood’s parent company, CVR Energy. The Secretary suggests that “Sharon & Walter did not place significance on the policy-making authority of executives.” Aplee. Br. 68. Thus, the Secretary reasons, the Commission

22

should not have relied upon the change in executive personnel in concluding there was no continuity of personnel. In doing so, he relies on United Food & Commercial Workers International Union, AFL-CIO, Local 152 v. NLRB, 768 F.2d 1463 (D.C. Cir. 1985). There, the D.C. Circuit found that although a “change in upper level personnel occurred,” the NLRB erred in concluding that there was substantial continuity between two employers. United Food, 768 F.2d at 1473. Specifically, the court found “no substantial evidence to support the [NLRB’s] apparent conclusion that the upper[-]level[-]personnel changes effected any substantial transformation in [workplace] operations.” Id.

Here, by contrast, the Commission found that CVR Energy executives controlled safety-related decisions and, as a “direct result,” “the safety policies at the refinery changed significantly.” App. vol. 21, 1156 n.14. And substantial evidence supports this finding. For example, the operations manager at the time testified that, “as a result of the purchase by CVR Energy,” Wynnewood “went through some pretty drastic changes” related to safety, including hiring two new members of its safety personnel. App. vol. 1, 1714. He also testified that “the corporate management [was] very active, involved every day,” and that CVR Energy permitted “access to capital or to funds” for safety, which were “previous[ly] very difficult” to access. Id. at 1714, 1716. The process-safety manager indicated that CVR Energy “raised the level of the safety programs” and that its “involvement in the safety program included process safety.” Id. at 1625. And a safety specialist testified that CVR Energy executives made the “budget for safety . . . pretty open” and “review[ed,]

23

revis[ed, and] train[ed] on safety policies [and] on safety procedures.” Id. at 2147. Thus, by considering the involvement of CVR Energy executives, the Commission focused on the “[c]ontinuity of the personnel who specifically control decisions related to safety and health,” as required by Sharon & Walter. 23 BL OSHC at 1295.

The Secretary next argues that the Commission ignored the continuity of several refinery supervisors between Wynnewood Inc. and Wynnewood LLC. But the Commission did not ignore the continuity of these supervisors; instead, it found that “safety policy at the refinery was not controlled by these managers” and that they “merely implemented the safety policies set by the previous parent company and then by CVR Energy.” App. vol. 21, 1156 n.14 (emphasis omitted). To be sure, there is evidence that these supervisors were involved in establishing procedures for using the Wickes boiler. But, as noted above, there is also evidence supporting the Commission’s conclusion that the CVR Energy executives controlled the safety- related decisions. And when reviewing the Commission’s decision for substantial evidence, it is not our role to “reweigh the evidence [or] second-guess the factual inferences drawn therefrom.” Jake’s Fireworks Inc., 893 F.3d at 1257. We thus conclude that substantial evidence supports the Commission’s analysis of the involvement of the CVR Energy executives.

Finally, the Secretary argues that we should not consider changes in safety policy because such changes relate to working conditions and the Commission found that the working conditions were the same between the two entities. True, the Commission did indicate that the working conditions under Wynnewood Inc. and

24

Wynnewood LLC were “the same.” App. vol. 21, 1155. But the Commission also emphasized Wynnewood LLC’s safety-related changes at the refinery and, as discussed, those changes were initiated by CVR Energy executives. For example, the Commission credited the “more formalized training programs,” the “$130 million in equipment upgrades,” and the “safety culture shift at the refinery.” Id. at 1157.

We are not convinced that such changes are better classified as changes to working conditions. And even if we accepted this categorization, the Secretary cites no authority indicating that the Commission’s decision to consider the changes under one category rather than another means that we must disregard this evidence on appeal. Cf. Jake’s Fireworks Inc., 893 F.3d at 1257 (explaining that the substantial- evidence standard “is satisfied if ‘“a reasonable mind” would consider the evidence adequate to support the conclusion reached’” (quoting Universal Const. Co., 182 F.3d at 732 )). Moreover, Sharon & Walter emphasized the continuity of “personnel who specifically control decisions related to safety and health” because “safety- related decisions made by managers and supervisors” impact the “type and extent of workplace hazards.” 23 BL OSHC at 1295. Accordingly, substantial continuity depends in part on the continuity of personnel who make decisions that could impact workplace hazards. Id.

Thus, we find that substantial evidence supports the Commission’s finding that there was no substantial continuity between Wynnewood Inc. and Wynnewood LLC. Accordingly, we affirm the Commission’s decision to characterize the violations at issue as serious rather than repeat.

25

Conclusion

Because the Wickes boiler was part of a PSM-covered process, we affirm the PSM-standard violations. And because the Commission correctly applied the substantial-continuity test and substantial evidence supports its finding that Wynnewood Inc. and Wynnewood LLC were not the same employer, we affirm the characterization of violations as serious rather than repeat.

Eugene Scalia, Secretary of Labor v. Wynnewood Refining Co., L.L.C, et al., Nos. 19-9533, 19-9578 BACHARACH, J., dissenting.

In 2012, a boiler at Wynnewood’s refinery exploded, killing two employees. (The boiler is pictured here shortly after the explosion.)

The Secretary of Labor investigated and cited Wynnewood for twelve regulatory violations, eleven involving the boiler. The Secretary characterized five of the twelve violations as “repeat,” triggering heightened penalties.

The Occupational Safety and Health Review Commission upheld the citations involving the boiler. But the Commission rejected the Secretary’s

2

characterization of the violations as “repeat,” reasoning that the previous violations had taken place when a different entity had owned the refinery. Both parties have appealed.

Wynnewood’s appeal focuses on the eleven citations involving the boiler. Though the facts are largely undisputed, the parties disagree about the application of regulations known as the “Process Safety Management” (PSM) standard. The Secretary adopted these regulations to prevent and minimize the impacts of catastrophic chemical releases, imposing stringent requirements for employers to enhance safety. See 29 C.F.R. § 1910.119; R., vol. I, at 1214–15 (testimony by the Secretary’s designated representative). 1 In appealing, Wynnewood argues that the PSM standard doesn’t apply to the boiler. The majority rejects this argument, concluding that the PSM standard applies because the boiler was connected to other vessels containing highly hazardous chemicals (HHCs).

The Secretary’s appeal involves the applicability of heightened penalties for employers who commit repeat violations. When the previous violations were committed, the refinery was owned by a separate legal

1 When the PSM standard doesn’t apply, other regulations may apply.

The employer’s duties may vary based on the applicability of particular regulations. See Teal v. E.I. DuPont de Nemours & Co., 728 F.2d 799, 804 (6th Cir. 1984).

entity. Though the owner’s legal entity changed, the Secretary argues that the new entity is substantially the same as the old one and should be regarded as the same entity. The Commission rejected this argument, concluding that the new entity differs from its predecessor. The majority agrees with the Commission.

I respectfully disagree with the majority in both appeals.

First, I agree with Wynnewood that the Secretary hasn’t shown that the PSM standard applies to the boiler. In deciding the applicability of the PSM standard, we’re confined to the regulatory provision invoked by the Secretary and the Commission. Both relied solely on a provision that unambiguously encompasses only vessels that contain an HHC. But the boiler itself contained no HHC, so I would vacate the eleven violations involving the boiler.

Second, I agree with the Secretary in his appeal regarding the applicability of heightened penalties for repeat violations. For the twelfth citation, the Commission misapplied the legal test for determining whether the entity itself has changed after a change in ownership. So I would remand for the Commission to reconsider the applicability of heightened penalties for the twelfth citation.

I. The Secretary and the Commission relied on a provision that unambiguously applies only to vessels containing HHC.

The Secretary and the Commission relied on a regulation, 29 C.F.R.

§ 1910.119, which regulates vessels constituting part of a covered “process.” We must ask: Does the boiler constitute part of a “process” covered by the PSM standard? The Commission answered “yes,” and the majority agrees. I do not. The Commission relied on a provision in the PSM standard that applied only if the boiler contained at least some HHC. The boiler didn’t, so the PSM standard didn’t apply. 2

A. We apply deferential review of the Secretary’s regulatory interpretation only if the regulation is ambiguous.

When reviewing the Secretary’s regulatory interpretation, we apply deference under the scheme described in Kisor v. Wilkie, ___ U.S. ___, 139 S. Ct. 2400 (2019). Under Kisor, however, courts should defer to an agency’s interpretation only if the regulation is genuinely ambiguous. Kisor, 139 S. Ct. at 2414. In determining whether a regulation is

2 Wynnewood also argues that the PSM standard doesn’t apply to the boiler because

 a risk of catastrophic release must exist for interconnected vessels and

 the Secretary must prove a “reasonable probability” that the vessel could contribute to a catastrophic release.

I would not reach these issues.

ambiguous, we “must exhaust all the ‘traditional tools’ of construction,” including the regulatory text, structure, history, and purpose. Id. at 2415 (quoting Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843 n.9 (1984)).

B. The applicability of the PSM standard turns on two sentences in the regulatory definition of “process.”

The PSM standard defines “process” in two sentences. The first sentence states that “process” is “any activity involving [an HHC] including any use, storage, manufacturing, handling, or the on-site movement of such chemicals, or combination of these activities.” 29 C.F.R. § 1910.119(b). The second sentence explains what constitutes a single process: “For purposes of this definition, any group of vessels which are interconnected and separate vessels which are located such that a highly hazardous chemical could be involved in a potential release shall be considered a single process.” Id.

The two sentences reflect two ways that a vessel can constitute part of a single PSM “process”:

1. if the vessel is used in an “activity” involving an HHC or

2. if a group of vessels is “interconnected” or “located such that [an HHC] could be involved in a potential release.”

Id.

The first sentence focuses on particular activities involving an HHC.

For example, if the boiler had constituted equipment used to store or move

HHCs, a process would exist under the first sentence. 29 C.F.R. § 1910.119(b). And if that process involved enough HHC, the PSM standard would apply. 29 C.F.R. § 1910.119(a)(1). If the first sentence had triggered the PSM standard, the boiler would constitute “process equipment.”

The second sentence focuses on the interconnection with or proximity to other equipment. In my view, the second sentence governs the applicability of the PSM standard to Wynnewood’s boiler.

C. The Secretary and the Commission relied solely on the second sentence of the PSM standard’s definition of “process.”

In considering the boiler part of a PSM process based on interconnection with or proximity to other equipment, the Secretary and Commission relied solely on the second sentence of the definition of “process.” See, e.g., R., vol. XXI, at 454–56 (the Secretary’s Post-Hearing Brief, relying on the second sentence for the applicability of the PSM standard); R., vol. I, at 785 (the Secretary’s statement that his theory of PSM coverage had “always” been interconnection with or proximity to other vessels), 1221 (testimony by the Secretary’s designated representative, agreeing that he had “determined that the boiler was a covered process because it was interconnected”), 1237 (further testimony

7

by the designated representative that his other theory of coverage entailed proximity to a covered process). 3 Though the Secretary and Commission relied solely on the second sentence for the regulatory definition of “process,” the majority relies on the first sentence. According to the majority, Wynnewood frames its arguments around the first sentence. See Majority Op. at 8 (“Wynnewood first argues that the Wickes boiler could not have been part of any such activity . . . .”). But Wynnewood has consistently focused on the second sentence, not the first one. See R., vol. I, at 91 (arguing about the applicability of the PSM standard based on “interconnection” rather than the existence of a “process activity”); Appellant’s Opening Br. 11 (Wynnewood’s summary of its argument about the PSM standard, quoting only the second sentence). Wynnewood’s focus is understandable given the Secretary and Commission’s reliance on the second sentence.

The difference is decisive. The parties agree that the first sentence covers Wynnewood’s fluid catalytic cracking unit and alkylation unit because they contain HHCs. If the boiler had been used in a process with the fluid catalytic cracking unit or alkylation unit, the boiler would be subject to the PSM standard as a PSM “activity” under the first sentence.

3 The designated representative testified that he was the agency’s representative who had decided that the PSM standard applied to the boiler. R., vol. I, at 1105–06.

But the Commission didn’t find that the boiler had been used in a PSM activity. The Commission instead applied the PSM standard based on the boiler’s interconnection with or proximity to two other vessels (the fluid catalytic cracking unit and alkylation unit). Interconnection or proximity of vessels could trigger the PSM standard only under the “process” definition’s second sentence, not the first sentence.

The distinction is evident in the Commission’s order. The administrative law judge had concluded that the boiler was covered as a PSM “activity,” relying heavily on an administrative precedent (Delek Refining, Ltd.) that had focused on whether a piece of equipment was part of a covered activity. Delek Refining, Ltd., 25 BNA OSHC 1365 (No. 08- 1386, 2015), 2015 WL 1957889 at *7, aff’d in relevant part, 845 F.3d 170 (5th Cir. 2016). Delek had examined “the ‘activity’ involving” the covered process, concluding that the equipment was “part of a ‘process’ covered by the PSM standard because [the equipment was] an integral part of the ‘manufacturing, handling [and] on-site movement of [highly hazardous chemicals].” Id. (quoting the first sentence of the “process” definition in 29 C.F.R. § 1910.119(b)). Delek thus rested on the first sentence of the regulatory definition of “process.”

But the Commission rejected the administrative law judge’s reliance on Delek, explaining that even though “Delek did address the PSM standard’s ‘process’ definition, it focused on the first sentence and whether

9

the positive pressurization unit was involved in the ‘manufacturing, handling [and] on-site movement’ of HHCs . . . not whether vessels were interconnected pursuant to the definition’s second sentence.” Commission’s Order at 9 (emphases added). So the Commission focused solely on the second sentence, rejecting the administrative law judge’s reliance on the first sentence. Id. at 9–10.

Given the Commission’s reliance on the second sentence, I would confine our review to the Commission’s reasoning. See Indus. Union Dep’t v. Am. Petroleum Inst., 448 U.S. 607, 631 n.31 (1980) (“[T]he validity of an agency’s determination must be judged on the basis of the agency’s stated reasons for making that determination.”).

D. The text of the second sentence is ambiguous without considering the regulatory preamble.

Wynnewood argues that the second sentence refers only to interconnected and co-located vessels that contain an HHC. To assess that argument, we apply the traditional tools of construction to the text of the regulation. Kisor, 139 S. Ct. at 2415.

Viewed in isolation, the second sentence does not clearly say whether a vessel can constitute part of a PSM process even when the vessel itself does not contain any HHC. The second sentence refers to “any group of vessels,” 29 C.F.R. § 1910.119(b), but the PSM standard doesn’t define “vessels.” So we must consider not only the regulatory text but also the

10

interpretive tools involving the regulatory structure, history, and purpose. See pp. 4–5, above. These interpretive tools include the preamble, which shows that interconnection with or proximity to another vessel can trigger the PSM standard only if both vessels contain HHC.

E. The preamble shows that the second sentence covers vessels only if they contain an HHC.

The parties debate the value of the preamble in interpreting the PSM definition of “process.” Given the apparent ambiguity in this definition, we must consider the regulatory history and purpose. Kisor, 139 S. Ct. at 2415; see pp. 4–5, above. To consider the history and purpose, we may focus on the preamble if it does not override the plain text of a provision:

Some courts and commentators have said that the prologue cannot be invoked when the text is clear. This limitation is reasonable if it means that the prologue cannot give words and phrases of the dispositive text itself a meaning that they cannot bear. But . . . [i]f the prologue is indeed an appropriate guide to meaning, it ought to be considered along with all other factors in determining whether the instrument is clear. The factors undermining its reliability affect its weight, not its relevance.

Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 218 (2012) (footnote omitted; emphasis in original); see Majority Op. at 7–8.

The preamble shows that the Secretary added the second sentence to clarify when multiple vessels containing HHCs would constitute a single process:

[A] new sentence has been added to clarify the fact that interconnected and nearby vessels containing a highly hazardous chemical would be considered part of a single process and the quantities of the chemical would be aggregated to determine if the threshold quantity of the chemical is exceeded.

Process Safety Management of Highly Hazardous Chemicals; Explosives and Blasting Agents, 57 Fed. Reg. 6356, 6372 (Feb. 24, 1992) (emphasis added). The preamble thus focuses on vessels “containing” an HHC. So the Secretary explained in the preamble that the agency would measure the quantity of HHC by combining the HHCs in each vessel:

The boundaries of a “process” would extend to quantities . . . which are interconnected and would include separate vessels located such that . . . an event such as an explosion would affect interconnected and nearby unconnected vessels which contain quantities of the chemical that when added together would exceed the threshold quantity . . . .

57 Fed. Reg. at 6372 (emphasis added). 4 Given this explanation, the second sentence unambiguously applies only when HHCs exist in the interconnected or nearby vessels.

4 The agency also reflected its focus in guidance materials published soon after adoption of the PSM standard. In these materials, the agency addressed the definition of the phrase “aggregate threshold quantities.” The agency explained:

In accordance with the second sentence of the definition of “process,” quantities of a particular hazardous chemical contained in vessels that are interconnected—and in unconnected vessels that may be adversely affected due to an incident at a nearby process—must be combined to determine whether the threshold level of a hazardous chemical has been reached.

F. The Secretary’s designated representative testified that the second sentence applies to an interconnected or nearby vessel only if it contains HHC.

At the hearing before the administrative law judge, the Secretary’s designated representative (Mr. Rick Hartung) testified. He acknowledged that he was the agency representative who had decided that the PSM standard applied to the boiler. See p. 7 n.3, above. In explaining that decision, he admitted that under the second sentence, the PSM standard would apply to the boiler only if it contained HHC:

Q. It says, “For purposes of this definition, any group of vessels which are interconnected and separate vessels which are located such that a highly hazardous chemical could be involved in a potential release shall be considered a single process”; correct?

A. Yes, ma’am.

Q. And since the term “vessel” isn’t defined in the standard, can you tell us what you interpret the term “vessel” to mean?

A. A vessel can be a container. It can also be in some instances a pressure vessel, but for the most part a container.

Q. I’m sorry?

A. For the most part, a container.

Q. Would it need to contain an HHC or flammable?

Process Safety Management of Highly Hazardous Chemicals—Compliance Guidelines and Enforcement Procedures, Occupational Safety & Health Administration (Sept. 28, 1992), https://www.osha.gov/enforcement/ directives/cpl-02-02-045.

A. Yes, I believe so. Yes.

R., vol. I, at 1234 (emphasis added). Given this concession by its designated representative, I would reject the Secretary’s current argument that the second sentence could apply even if the interconnected or nearby vessel contains no HHC.

G. Though the first sentence could apply even when the vessel contains no HHC, the Commission has not relied on the first sentence.

The Secretary argues that other provisions of the PSM standard show “that the presence of a HHC in a vessel is not a precondition to coverage.” Secretary’s Resp. & Opening Br. at 39. The Secretary is right, but only under the “process” definition’s first sentence (when a vessel is used in an activity involving a covered process). See Oral Arg. at 43:58–46:37 (responding to a hypothetical and explaining that vessels without HHCs could be considered process equipment depending on their use); see also Process Safety Management of Highly Hazardous Chemicals—Compliance Guidelines and Enforcement Procedures, Occupational Safety & Health Administration (Sept. 28, 1992), https://www.osha.gov/enforcement/ directives/cpl-02-02-045 (“Furnaces, boilers, heaters, etc., fueled by flammable liquids or gases--regardless of the quantity of the fuel--used in processes that are otherwise covered by the PSM standard . . . are considered part of the process and are covered by the PSM standard.”). Here, though, the Secretary and Commission relied on the “process”

definition’s second sentence, which considers the boiler’s interconnection or proximity to a covered process, not the boiler’s use in a process activity. For coverage under the second sentence, the presence of an HHC is required to trigger the PSM standard.

* * *

Because the boiler did not contain any HHC, I would reverse the Commission’s determination of PSM coverage as to the eleven citations involving the boiler.

II. The Commission committed legal error in applying the test of substantial continuity.

The Secretary characterized five of Wynnewood’s twelve citations as repeat violations, which triggered heightened penalties. See 29 U.S.C. § 666(a) (providing heightened penalties against employers who commit repeated violations). The Secretary based this characterization on violations committed by Wynnewood’s predecessor. But the Commission determined that Wynnewood is not a successor and rejected the Secretary’s characterization of Wynnewood as the same entity cited for previous violations.

The majority concludes that the Commission did not err. I respectfully disagree and would remand for further consideration of heightened penalties for the single “repeat” citation not involving the boiler.

A. We review the Commission’s factual findings for substantial evidence and legal conclusions for conformity with the law.

When reviewing the Commission’s decision, we examine factual determinations under the substantial-evidence standard. 29 U.S.C. § 660(a); see Slingluff v. OSHRC, 425 F.3d 861, 866 (10th Cir. 2005) (“[F]indings of the Commission with respect to questions of fact, if supported by substantial evidence on the record . . . shall be conclusive.” (quoting Interstate Erectors, Inc. v. OSHRC, 74 F.3d 223, 226 (10th Cir. 1996))). We also set aside the Commission’s legal conclusions if they are “not in accordance with law.” 5 U.S.C. § 706(2)(A).

To determine whether a predecessor’s violations may be attributed to a successor, the Commission applies a three-factor test involving “substantial continuity.” See Sharon & Walter Constr., Inc., 23 BNA OSHC 1286 (No. 00-1402, 2010), 2010 WL 4792625 at *9 (focusing on “whether there is ‘substantial continuity’ between the two enterprises” (quoting Fall River Dyeing & Finishing Corp. v. NLRB, 482 U.S. 27, 43 (1987))).

The Secretary argues that the Commission

 improperly treated the absence of one of the three factors as automatically dispositive and

 misapplied that factor.

Responding to this argument, Wynnewood urges application of the substantial-evidence standard on the ground that the test of substantial continuity turns on factual determinations.

Wynnewood has misinterpreted the Secretary’s arguments. The Secretary argues that the Commission misapplied its own precedents, not that the Commission erred in its factual findings. Given the Secretary’s argument, the majority correctly concludes that we apply de novo review. Majority Op. at 17; see Mountain Side Mobile Estates P’ship v. Sec’y of Hous. & Urban Dev., 56 F.3d 1243, 1250 (10th Cir. 1995) (stating that our review is not deferential on legal questions, such as whether the agency applied the correct legal standard).

B. The Commission considers substantial continuity based on three factors.

The Commission applied the “substantial-continuity test” for successorship in Sharon & Walter, adapting the test used in labor-relations cases. 23 BNA OSHC 1286 (No. 00-1402, 2010), 2010 WL 4792625. The substantial-continuity test evaluates “the totality of the circumstances,” focusing “on whether the new [entity] has ‘acquired substantial assets of its predecessor and continued, without interruption or substantial change, the predecessor’s business operations.’” Coastal Derby Ref. Co. v. NLRB, 915 F.2d 1448, 1452 (10th Cir. 1990) (quoting Fall River Dyeing & Finishing Corp. v. NLRB, 482 U.S. 27, 43 (1987)). When applying that

17

standard to violations of the Occupational Safety and Health Act, the Commission considers three factors:

1. the nature of the business, with continuity in the nature of the business indicating a lack of substantive change in the enterprise,

2. the continuity in the jobs and working conditions, which are correlated with safety and health hazards, and

3. the continuity of safety and health personnel.

Sharon & Walter, 23 BNA OSHC 1286, 2010 WL 4792625 at *10.

C. The Commission improperly treated the third factor as automatically dispositive.

As the Secretary argues, the Commission improperly treated the final factor (continuity of personnel) as a required element of substantial continuity.

The Commission found that

 the first two factors of the test support a finding of substantial continuity and

 the third factor cuts against a finding of substantial continuity.

See Commission’s Order at 13 (“[T]he refinery’s business, products, jobs, and working conditions were the same under both entities.”); id. at 15 (“[T]he record does not support the Secretary’s contention that there was . . . continuity in the safety personnel . . . .”). After making these factual findings, the Commission treated the absence of the third factor as automatically dispositive.

The Commission can use discretion in balancing the three factors.

But in exercising that discretion, the Commission treated the absence of the third factor as automatically fatal to the Secretary’s argument on substantial continuity. The Commission erred because the absence of any single factor is not automatically fatal to substantial continuity. See Haw. Carpenters Trust Funds v. Waiola Carpenter Shop, Inc., 823 F.2d 289, 294 (9th Cir. 1987) (stating that the absence of one factor in a similar test does not prevent characterization as the same entity); see also Ind. Elec. Workers Pension Benefit Fund v. Manweb Servs., Inc., 884 F.3d 770, 783 (7th Cir. 2018) (stating that “[t]he presence or absence of any one factor ‘does not compel a particular conclusion’” on successor liability) (quoting Resilient Floor Covering Pension Tr. Fund Bd. of Trs. v. Michael’s Floor Covering, Inc., 801 F.3d 1079, 1091 (9th Cir. 2015)).

The Commission could conclude that one factor outweighed the other two factors. But the Commission didn’t do that. The Commission instead concluded that the failure to satisfy the third factor would necessarily prevent consideration of Wynnewood as a successor: “[W]e agree that the last prong of the three-part ‘substantial continuity’ test . . . was not met in this case—and that [the two companies] are therefore not the same employer for repeat characterization purposes under Sharon & Walter . . . .” Commission’s Order at 15 n.14; see also id. at 15 (“[T]he record does not support the Secretary’s contention that there was sufficient

19

continuity in the safety personnel at the cited entity such that ‘there was a Commission final order against the same employer for a substantially similar violation.’” (emphasis in original) (citation omitted)).

The majority says that the Commission “never indicat[ed] that the test required continuity across all three factors.” Majority Op. at 17. But the Commission expressly found that the employer was not the same because of the Secretary’s failure to satisfy the third factor. In doing so, the Commission treated the third part of the test as a necessary element rather than as a factor.

Wynnewood defends the Commission’s reliance on the third factor, arguing that it is the most important of the three. For this argument, Wynnewood points to Sharon & Walter, explaining that continuity of “control over decision-making in both companies . . . weighs heavily in favor of” finding substantial continuity. 23 BNA OSHC 1286, 2010 WL 4792625 at *10. But Sharon & Walter explained the importance of the third factor in the context of the facts in that case. The third factor “weigh[ed] heavily” there because the same person had owned both the old and new entities. But Sharon & Walter didn’t suggest that the third factor would always outweigh the other factors.

After finding that the third factor cut against a repeat characterization, the Commission needed to explain why that factor outweighed the other two. The majority says that “the Commission

20

explained how the change in personnel directly impacted the safety of the refinery, demonstrating why it weighed that factor so heavily.” Majority Op. at 17–18. But the Commission never explained why the third factor would override the other two. The Commission said only that “the record does not support the Secretary’s contention that there was sufficient continuity in the safety personnel” supporting a repeat violation and “[a]ccordingly, we find that a repeat characterization is not warranted here.” Commission’s Order at 15–16. The Commission’s language shows that it treated the absence of the third factor as automatically fatal to characterization of the violations as “repeat.”

The Commission’s failure to properly weigh the three factors constitutes a legal error.

D. The Commission also misapplied the third factor.

The Commission also improperly applied the third factor.

The Commission framed its analysis of the third factor by considering

 a shift in policies and workplace culture after Wynnewood’s acquisition and

 Wynnewood’s lack of intent to manipulate the successorship doctrine.

But these reasons should not affect consideration of the third factor.

1. A shift in policies and workplace culture does not bear on the third factor.

The Commission relied in part on the change in safety policies and culture at Wynnewood after the acquisition. See Commission’s Order at 14 (“These leadership changes resulted in a safety culture shift at the refinery. In light of the foregoing, we find these are not the appropriate circumstances to affirm a repeat characterization.”) (footnote omitted); id. at 14 n.13 (“[W]e conclude that Wynnewood LLC’s changeover in ownership resulted in changes in management practices, procedures, and culture significant enough to break the chain of liability . . . .”). For the sake of argument, we may assume that the Commission could consider these changes under the test for substantial continuity. Even so, those changes would not affect the third factor, which Wynnewood and the majority characterize as the most important of the three factors.

The third factor doesn’t consider whether the employees of the new entity acted differently than they had in the past. This factor instead focuses on whether the employees addressing safety were the same in both the old and new entities. See Sharon & Walter, 23 BNA OSHC 1286, 2010 WL 4792625 at *10 (examining the continuity of personnel who had “control over decision-making,” but not the continuity of those decisions).

The Commission’s focus undermines the purpose of inquiring into whether the old and new entities are the same. This inquiry matters

22

because once an entity has been cited, the citation “tends to apprise the [entity] of the requirements of the standard and to alert [it] that special attention may be required to prevent future violations of that standard.” Dun-Par Eng’d Form Co. v. Marshall, 676 F.2d 1333, 1337 (10th Cir. 1982). A repeat violation does not depend on a showing of intent and cannot be cured by the entity’s change in policy.

A policy change could conceivably weigh against a finding of willfulness. See Sharon & Walter, 23 BNA OSHC 1286, 2010 WL 4792625 at *5 (“The hallmark of a willful violation is the employer’s state of mind at the time of the violation-an ‘intentional, knowing, or voluntary disregard for the requirements of the Act or . . . plain indifference to employee safety.’” (citation omitted)). But willfulness constitutes a separate basis for heightened penalties. See 29 U.S.C. § 666(a) (“Any employer who willfully or repeatedly violates the requirements . . . may be assessed a civil penalty . . . .”). The Commission improperly conflated the inquiries for substantial continuity and willfulness.

2. The third factor does not address the entity’s possible manipulation.

The Commission also erred by relying on Wynnewood’s lack of manipulation.

In applying the third factor, the Commission noted: “This is not a case where ‘the cited employer has altered its legal identity from that of

23

the predecessor employer . . . [simply to] avoid a repeat characterization.’” Commission’s Order at 14–15 (quoting Sharon & Walter, 23 BNA OSHC 1286, 2010 WL 4792625 at *8) (alterations in original). The Commission also explained that it “s[aw] no rationale here for imposing the OSHA violation history of Wynnewood Inc. upon Wynnewood LLC, a separate legal entity not created to avoid responsibilities under the Act.” Id. at 15 n.14. By focusing on a lack of manipulation, the Commission improperly conflated the substantial-continuity test and the alter-ego test.

The majority recognizes that the substantial-continuity test does not focus on whether a new entity was intended as a sham. See Majority Op. at 19. And Wynnewood admits that the Commission’s formulation of the substantial-continuity test serves as the functional equivalent of the alter- ego test. See Wynnewood’s Resp. & Reply Br. at 10 (“[The Commission’s] substantial continuity test . . . looks for overlap more akin to an alter ego, and especially for evidence of a sham transition . . . .”); Oral Arg. at 50:00–50:10 (“The primary purpose of [the Commission’s] successorship test is to defeat those kinds of efforts by employers to purposefully evade repeat liability by engaging in sham transactions.”). But the tests for alter ego and substantial continuity differ. Given this difference, the

Commission didn’t just use the wrong label; the Commission mistakenly conflated the two tests. 5 The majority discounts the Commission’s discussion of Wynnewood’s motive, calling the discussion “dicta” and assuming that the discussion wasn’t essential to the decision. Majority Op. at 19. How does the majority know that the Commission didn’t rely on its discussion of Wynnewood’s motive? Labelling the discussion “dicta” might affect the discussion’s status as precedent, but doesn’t explain whether the erroneous discussion tainted the Commission’s findings on the third factor.

C. The appropriate remedy is remand.

The Secretary suggests that if we conclude that the Commission had committed legal error, we should conduct our own factual analysis of the substantial-continuity test. But even if we were to agree with the Secretary’s factual analysis, a remand would remain necessary for the Commission to determine whether the cited violations are substantially similar to the prior violations. See Potlatch Corp., 7 BNA OSHC 1061 (No. 16183, 1979), 1979 WL 61360 at *3 (“A violation is repeated . . . if . . . there was a Commission final order against the same employer for a

5 Wynnewood contends that the successorship doctrine should not discourage mobility of capital. Wynnewood’s Resp. & Reply Br. at 24–26. But Wynnewood doesn’t explain how mobility of capital should affect our analysis of the third factor.

substantially similar violation.”). Given the factual nature of this inquiry, I would remand to the Commission to reconsider its application of the substantial-continuity test. III. Conclusion In my view, the PSM standard applies to interconnected or nearby vessels only when each vessel contains an HHC. Because the boiler didn’t contain any HHC, I would vacate the eleven citations involving the boiler. I would also conclude that the Commission improperly applied the substantial-continuity test, so I would remand for the Commission to reconsider heightened penalties for the single repeat violation not involving the boiler. I thus respectfully dissent.