EXXONMOBIL PIPELINE COMPANY, Petitioner, v. UNITED STATES DEPARTMENT OF TRANSPORTATION; PIPELINE AND HAZARDOUS MATERIALS SAFETY ADMINISTRATION; OFFICE OF PIPELINE SAFETY, Respondents.
ExxonMobil Pipeline Company petitions for review of a Pipeline and Hazardous Materials Safety Administration order following the release of several thousand barrels of crude oil from a pipeline it owned and operated. ExxonMobil specifically challenges Items 1–4, 7, and 8 of the agency's final order. We vacate Items 1–4 and 7 and affirm the agency with regard to Item 8 but remand with an instruction to reevaluate the basis for the penalty associated with this violation.
The 859-mile long Pegasus Pipeline transports crude oil from Patoka, Illinois to Nederland, Texas. In March 2013, the Pegasus Pipeline ruptured, spilling several thousand barrels of oil near Mayflower, Arkansas. The pipeline is owned and operated by ExxonMobil Pipeline Company. In the wake of the oil spill, the Pipeline and Hazardous Materials Safety Administration (the “agency”), an operating administration of the United States Department of Transportation, conducted an investigation. The agency ultimately issued a final order, concluding that ExxonMobil violated several pipeline safety regulations. The agency assessed a $2.6 million civil penalty and ordered ExxonMobil to take certain actions to ensure compliance with those regulations.
A. Regulatory Framework
The Pipeline Safety Act, 49 U.S.C. § 60101 et seq., gives the Secretary of Transportation regulatory and enforcement authority to take actions to protect the public against risks to life and property posed by pipeline transportation and pipeline facilities. The statute provides that the Secretary of Transportation “shall prescribe minimum safety standards for pipeline transportation and for pipeline facilities.” 49 U.S.C. § 60102(a)(2). Pursuant to this authority, the agency has promulgated regulations establishing minimum safety standards. See 49 C.F.R. pts. 190–199.
The Pipeline Safety Act and the agency's integrity management regulations require each pipeline operator to create what is known as an integrity management program (“IMP”) for all pipelines that could affect a high consequence area. High consequence areas include populated areas, areas that are unusually sensitive to environmental damage, or commercially navigable waterways. 49 C.F.R. § 195.452. A pipeline operator's IMP is to be specific to its own pipeline systems. The purpose of developing an IMP is to assist the operator in “address[ing] the risks on each segment” of its pipelines. Id. § 195.452(b)(1).
An IMP must include a written plan to conduct periodic integrity assessments of each of the operator's pipelines and to address any problematic conditions discovered by those assessments. Id. §§ 195.452(b)(3), (f)(2)–(5). The pipeline integrity regulations require operators to “establish an integrity assessment schedule that prioritizes pipeline segments for assessment.” Id. § 195.452(e)(1). This integrity assessment schedule is informed by the pipeline operator's threat identification and risk assessment process. See id. §§ 195.452(e), (j)(5). As part of this process, operators are tasked with evaluating numerous risk factors for each pipeline segment, including, inter alia, the results of the previous integrity assessment; pipe material, manufacturing, and seam type; and leak history. Id. § 195.452(e). The regulations state that the pipeline operator “must consider” these factors in establishing an assessment schedule. Id. § 195.452(e)(1). Based on the results of an operator's risk assessment analysis, the operator must prioritize its pipeline segments for reassessment on five-year intervals. Id. § 195.452(j)(3).
The pipeline integrity regulations also set forth the available methods by which the operator may conduct the periodic integrity assessments. The regulations list three assessment methods available to operators: (1) in-line inspections; (2) hydrostatic pressure testing;1 and (3) external corrosion direct assessment. Id. § 195.452(j)(5). An additional requirement may apply to pipelines constructed of a certain type of pipe known as pre-1970 low-frequency electric resistance welded steel (“LF-ERW”) pipe because this type of pipe is known to have a higher risk of seam failure than other types of pipe due to manufacturing defects. According to the regulations, if—and only if—the LF-ERW pipeline segment is shown to be “susceptible to longitudinal seam failure,” the methods an operator selects to assess that segment “must be capable of assessing seam integrity and of detecting corrosion and deformation anomalies.” Id.
The pipeline integrity regulations are silent as to how operators must determine whether LF-ERW pipe is susceptible to longitudinal seam failure. However, in 2004, the agency commissioned and published a third-party report, referred to here as the Baker Report, which extensively discusses pipeline metallurgy. Section 4 of the Baker Report contains a methodology for determining seam-failure susceptibility. This methodology, illustrated by a decision tree, considers, inter alia, pipe and seam characteristics, in-service and hydrostatic test failures, the cause of those failures, and operating stress levels to determine whether a given segment of LF-ERW pipe is susceptible to seam failure. As outlined in the Baker Report decision tree and as testified to by Dr. John F. Kiefner, one of the authors of the Baker Report, “seam related in-service failures and/or hydrostatic test breaks or leaks by themselves do not indicate that a pipeline is susceptible to seam failure.” Rather, according to the decision tree and the Baker Report's co-author, whenever a seam-related in-service failure or hydrostatic test failure occurs, these failures should be analyzed for two primary causes that would indicate susceptibility to seam failure: pressure-cycle induced fatigue and selective seam corrosion.
In the event that a pipeline operator fails to comply with the federal Pipeline Safety Act or the integrity management regulations, the agency may issue compliance orders and assess civil administrative penalties after notice and a hearing. 49 U.S.C. §§ 60118(b), 60122.
B. ExxonMobil's Application of the Regulations
Under ExxonMobil's IMP plan, its process for analyzing seam failure susceptibility of LF-ERW pipe is based on the methodology outlined in the Baker Report decision tree. ExxonMobil retained the services of Dr. Kiefner to assist it in applying the pipeline integrity regulations and the Baker Report's guidance to its IMP plan. ExxonMobil has conducted a series of periodic integrity assessments on the Pegasus Pipeline, each time applying the framework provided by the Baker Report decision tree. Following each assessment, ExxonMobil concluded that the Pegasus Pipeline segment at issue in this case was not susceptible to longitudinal seam failure and therefore did not warrant prioritization over other pipeline segments for reassessment.
ExxonMobil first evaluated the Pegasus Pipeline's susceptibility to longitudinal seam failure in late 2004 through early 2005. Its evaluation took into account the pipeline's manufacturing history, pipe materials, operating and maintenance history, leak history, as well as the results of prior pressure tests and integrity assessments. Hydrostatic tests performed in 1969 and 1991 revealed several seam failures and a minor in-service seam leak occurred in 1984.2 However, despite these seam failures, ExxonMobil determined that the pipeline was not susceptible to seam failure because the past failures were not caused by either pressure-cycling induced fatigue or selective seam corrosion—the two factors enumerated in the Baker Report decision tree.
A year after the initial evaluation of the pipeline, ExxonMobil conducted a hydrostatic test for the Pegasus Pipeline. The test resulted in eleven seam-related failures, and ExxonMobil replaced the failed joints and hired a third-party expert in metallurgy to conduct an analysis on why the joints failed. Because the analysis did not reveal evidence of either pressure-cycling induced fatigue or selective seam corrosion, ExxonMobil once again concluded that the LF-ERW pipe segments were not susceptible to longitudinal seam failure.
In 2007, ExxonMobil again evaluated the Pegasus Pipeline's susceptibility to longitudinal seam failure. ExxonMobil's evaluation took into account the pipeline's manufacturing history, pipe materials, operating and maintenance history, leak history, the results of prior pressure tests and integrity assessments, as well as the results of subsequent metallurgical analyses. Once again, applying the Baker Report decision tree, ExxonMobil concluded that the line was not susceptible to longitudinal seam failure. Two years later, ExxonMobil again evaluated the Pegasus Pipeline for susceptibility to seam failure and reached the same conclusion.
ExxonMobil performed an in-line inspection in 2010 to assess the integrity of the Pegasus Pipeline with two in-line inspection tools it had deemed appropriate. The following year, ExxonMobil again reevaluated the pipeline's longitudinal seam failure susceptibility determination, taking into account all of the same information as before in addition to the results of the 2010 in-line inspection. Again, ExxonMobil concluded that the pipeline was not susceptible to longitudinal seam failure.
In late 2012 through early 2013, ExxonMobil conducted another inspection of the Pegasus Pipeline with an in-line inspection tool known as a TFI seam/crack tool. It ran the tool through the section of line where the oil release would eventually occur shortly before the release. After the in-line inspection was complete and while the data from the inspection was being processed, the Mayflower oil spill occurred. Even though the third-party vendor processing the inspection results was aware that a seam failure had occurred, the vendor could not identify a defect at the point of rupture.
C. The Agency's Findings
In the wake of the Mayflower release, the agency conducted an investigation and determined that the cause of the release was a manufacturing defect in the seam of the Pegasus Pipeline's LF-ERW pipe. The agency concluded that ExxonMobil's IMP plan had not properly accounted for the risk of longitudinal seam failure and that this was a contributing factor in the Mayflower release. The agency found that ExxonMobil's determination that the pipeline was not susceptible to longitudinal seam failure was erroneous and that ExxonMobil failed to properly assess the pipeline's integrity. The agency also concluded that ExxonMobil's IMP plan was deficient in a number of other respects.
The agency's final order stated that ExxonMobil violated section 195.452(e)(1) “by failing to properly consider the susceptibility of its ERW pipe to seam failure when establishing a continual integrity assessment schedule based on all risk factors on the Pegasus Pipeline.” The agency found that ExxonMobil's conclusion that the relevant portion of the pipeline was not susceptible to seam failure was “flawed” “[g]iven the history of seam-related failures both in-service and during pressure testing of the pipeline.” Specifically, the agency rejected ExxonMobil's position that the Baker Report permitted it to conclude that its pipe was not susceptible to seam failure because the prior seam failures did not exhibit evidence of fatigue or preferential seam corrosion. The agency reasoned that the pipeline's past seam failures—including eleven seam failures during the 2005–2006 hydrostatic test—“strongly suggested the ERW pipe was susceptible to seam failure” and ExxonMobil's conclusion to the contrary was unreasonable.
The agency ultimately cited ExxonMobil for nine separate violations of the regulatory requirements. According to the agency's final order, violations 1 and 4 are based on § 195.452(e)(1), which requires operators to “consider” pipeline risk factors, including seam type and manufacturing information, among other factors. Violation 2 is based on subsection (j)(3), which requires an operator to schedule continual assessments of susceptible pipelines at five-year intervals. Violations 3, 7, 8, and 9 are based on subsection (b)(5), which requires operators to establish and implement an integrity management program. Violation 5 is based on subsection (h)(1), which requires operators to take prompt action to address conditions discovered through an integrity assessment. Finally, violation 6 is based on subsection (h)(2), which requires operators to promptly discover a condition within 180 days of an integrity assessment. The agency assessed a civil penalty in the amount of $2,630,400. The agency also issued a compliance order in which it directed ExxonMobil to take a number of actions to abide by the pipeline integrity regulations.
Now on appeal, ExxonMobil challenges Items 1–4, 7, and 8 of the agency's final order. ExxonMobil does not challenge the violations cited in Items 5, 6, or 9.
II. Standard of Review
The standard of review we apply to the agency's final order is prescribed by the Administrative Procedure Act, 5 U.S.C. § 706. See 49 U.S.C. § 60119(a)(3). Under the APA, we uphold an agency's actions, findings, and conclusions unless we determine them to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A); see also Allred's Produce v. U.S. Dep't of Agric., 178 F.3d 743, 746 (5th Cir. 1999). “Arbitrary and capricious review focuses on whether an agency articulated a rational connection between the facts found and the decision made.” Pension Benefit Guar. Corp. v. Wilson N. Jones Mem'l Hosp., 374 F.3d 362, 366 (5th Cir. 2004). “We must disregard any post hoc rationalizations of the [agency's] action and evaluate it solely on the basis of the agency's stated rationale at the time of its decision.” Luminant Generation Co., L.L.C. v. E.P.A., 675 F.3d 917, 925 (5th Cir. 2012). The party challenging the agency's action bears the burden of establishing that the agency's determination was arbitrary and capricious. La. Pub. Serv. Comm'n v. FERC, 761 F.3d 540, 558 (5th Cir. 2014).
III. Items 1–4 and 7
We first address whether the agency's interpretation of the pipeline integrity regulations should be afforded Auer deference. Because the regulations are unambiguous, we conclude that Auer deference is inappropriate. We next address whether the agency acted in an arbitrary and capricious manner when it found ExxonMobil to be in violation of the regulations under Items 1–4 and 7 of the final order. Because the regulations unambiguously instruct pipeline operators to “consider” certain risk factors, and because the evidence demonstrates that ExxonMobil did carefully consider those factors, we conclude that the agency's decisions pertaining to Items 1–4 and 7 of its final order were arbitrary and capricious.
The central issue before us is what is required of pipeline operators by the term “consider” within the context of § 195.452(e)(1) of the pipeline integrity management regulations.3 As noted, § 195.452(e)(1) of the regulations requires pipeline operators to
establish an integrity assessment schedule that prioritizes pipeline segments for assessment ․ An operator must base the assessment schedule on all risk factors that reflect the risk conditions on the pipeline segment. The factors an operator must consider include, but are not limited to:
(i) Results of the previous integrity assessment, defect type and size that the assessment method can detect, and defect growth rate;
(ii) Pipe size, material, manufacturing information, coating type and condition, and seam type;
(iii) Leak history, repair history and cathodic protection history ․
49 C.F.R. § 195.452(e)(1). If, after considering these risk factors, a pipeline operator determines that a pipeline segment constructed of LF-ERW pipe is susceptible to longitudinal seam failure, the operator must conduct the integrity assessment with a method “capable of assessing seam integrity and of detecting corrosion and deformation anomalies.” Id. § 195.452(j)(5).
ExxonMobil argues that § 195.452(e)(1) is a process-based regulation that requires a pipeline operator to consider factors but does not compel the operator to reach a specific outcome. ExxonMobil contends that it considered all pipeline risk factors under § 195.452(e)(1) and it determined—on four separate occasions—that the Pegasus Pipeline's LF-ERW pipe segments were not susceptible to longitudinal seam failure. Under its interpretation of § 195.452(e)(1), the agency argues that given the history of seam failures to the LF-ERW pipe, it was unreasonable for ExxonMobil to consider the above factors and conclude that the pipeline was not susceptible to longitudinal seam failure. The agency contends that its interpretation of § 195.452(e)(1) is entitled to deference under Auer v. Robbins, 519 U.S. 452 (1997).
We will generally grant Auer deference to an agency's interpretation of its own ambiguous regulation. Delek Refining, Ltd. v. OSHRC, 845 F.3d 170, 175 (5th Cir. 2016) (“[W]e will defer to an agency's reasonable interpretation of its own regulations when the text of the regulation is ambiguous.”); see also Auer, 519 U.S. at 461. “In situations in which the meaning of regulatory language is not free from doubt, the reviewing court should give effect to the agency's interpretation so long as it is reasonable,” and it “sensibly conforms to the purpose and wording of the regulations.” Martin v. OSHRC, 499 U.S. 144, 150–51 (1991) (alterations and internal quotation marks omitted). However, Auer deference does not apply if the petitioner “lacked fair notice” of the agency's interpretation of the regulation that the agency is advancing in the enforcement action. Employer Sols. Staffing Grp. II, L.L.C. v. Office of Chief Admin. Hearing Officer, 833 F.3d 480, 487–88 (5th Cir. 2016); see also Diamond Roofing Co., Inc. v. OSHRC, 528 F.2d 645, 649 (5th Cir. 1976); Christopher v. SmithKline Beecham Corp., 567 U.S. 142, 155 (2012). If the regulation is unambiguous, we will not defer to the agency's interpretation. Christensen v. Harris Country, 529 U.S. 576, 588 (2000); see also Moore v. Hannon Food Serv., Inc., 317 F.3d 489, 495–96 (5th Cir. 2003).
Therefore, to determine whether the agency's interpretation of the term “consider” in the regulations is entitled to deference, we must first determine whether § 195.452(e)(1) is ambiguous. “We interpret regulations in the same manner as statutes, looking first to the regulation's plain language.” Anthony v. United States, 520 F.3d 374, 380 (5th Cir. 2008). Where the language of the regulations is unambiguous, we do not look beyond the plain wording of the regulation to determine its meaning. Copeland v. C.I.R., 290 F.3d 326, 332–33 (5th Cir. 2002). We hold that § 195.452(e)(1) is textually unambiguous and therefore no Auer deference is warranted to the agency's interpretation.
Section 195.452(e)(1) plainly instructs pipeline operators that they “must consider” “all risk factors that reflect the risk conditions on the pipeline segment.” 49 C.F.R. § 195.452(e)(1). The Oxford English Dictionary defines “consider” as follows: “to contemplate mentally, fix the mind upon; to think over, meditate or reflect on, bestow attentive thought upon, give heed to, take note of.” Consider, Oxford English Dictionary (online edition). The Merriam-Webster's Collegiate Dictionary similarly defines consider as “to think about carefully.” Consider, Merriam-Webster's Collegiate Dictionary (Deluxe ed. 1998).
The regulation's requirement to consider certain factors unambiguously requires pipeline operators to carefully undergo an informed decision-making process in good faith, reasonably taking into account all relevant risk factors in reaching a decision. Contrary to the agency's assertion, the term “consider” does not compel a certain outcome, but rather it serves to inform the pipeline operator's careful decision-making process. See J.H. Miles & Co., Inc. v. Brown, 910 F. Supp. 1138, 1156 (E.D. Va. 1995) (explaining that a federal regulation requiring federal fishing officials to “consider” various statutory factors in setting fishing quota recommendations was not a “strict dictate,” but rather officials had “some discretion” in preparing their recommendation); see also Cent. Valley Chrysler-Jeep v. Witherspoon, 456 F. Supp. 2d 1160, 1173 (E.D. Cal. 2006) (“Congress's use of the term ‘consider’ in a statute requires an actor to merely ‘investigate and analyze’ the specified factor, but not necessarily act upon it.”).
Because we conclude that § 195.452(e)(1)'s instruction to “consider” all relevant risk factors unambiguously serves to inform a pipeline operator's careful and deliberate decision-making process rather than to compel a particular outcome, the agency's interpretation of the regulation does not warrant Auer deference. See Exelon Wind 1, L.L.C. v. Nelson, 766 F.3d 380, 399 (5th Cir. 2014) (“[A]n agency is not entitled to deference when it offers up an interpretation of the Regulation that we have already said to be unambiguously foreclosed by the regulatory text.”); Christensen, 529 U.S. at 588 (“Auer deference is warranted only when the language of the regulation is ambiguous.”).
Owing no deference to the agency's interpretation of the regulation, we next address whether ExxonMobil reasonably applied § 195.452(e)(1)'s instruction to “consider” all relevant risk factors in making its pipeline susceptibility determination.4 As discussed, the pipeline integrity regulations plainly instruct pipeline operators that they “must consider” “all risk factors that reflect the risk conditions on the pipeline segment” in creating an assessment schedule. 49 C.F.R. § 195.452(e)(1). By its plain text, this is a process-based requirement that directs pipeline operators to carefully take into account all relevant risk factors when reaching a decision. The regulations do not mandate a particular outcome, but rather prescribe a decision-making process that pipeline operators must undergo in good faith. The record demonstrates that ExxonMobil satisfied its obligation to “consider” various risk factors when it conducted a lengthy, repeated, and in-depth analysis of those risk factors by utilizing the available industry-commissioned guidance of the Baker Report decision tree. We therefore ultimately conclude that ExxonMobil's actions were reasonable and that the agency's decisions pertaining to Items 1–4 and 7 of its final order were arbitrary and capricious.
Contrary to the agency's assertion, nothing in the regulations compels an operator to conclude that a pipeline constructed of LF-ERW pipe is susceptible to longitudinal seam failure when the pipeline has experienced seam failures. If the agency wished to enforce outcome-based requirements instead of the process-based requirements that are currently in place with regards to seam failure susceptibility, the agency could have promulgated regulations to that effect. For example, the regulations could have prescribed that all LF-ERW pipelines are susceptible to seam failure. Alternatively, the regulations could have prescribed that any LF-ERW pipeline that experiences a seam failure during testing or while in service must be deemed seam-failure susceptible. But the process-based regulations currently in place only require that the pipeline operator “consider” various risk factors in making its risk assessment determination, and they ultimately leave it up to the pipeline operator to make the decision on seam failure susceptibility.5
In its final order, the agency found that ExxonMobil violated § 195.452(e) because ExxonMobil allegedly “failed to properly consider the history of seam-related failures and low toughness of the seam.” While the agency suggests that ExxonMobil simply ignored past seam failures, this is contradicted by the evidence. The record clearly demonstrates that ExxonMobil carefully considered the past seam failures. Indeed, even the agency's final order itself acknowledges the various steps ExxonMobil took in considering the pipeline's past seam-related failures:
The failures were analyzed for evidence of pressure cycling induced fatigue and preferential seam corrosion, but neither condition was detected. [ExxonMobil] attributed the failures to mill defects and a lower test temperature, which the Company believed caused the seams to be more brittle. Due to the absence of pressure cycling induced fatigue and preferential seam corrosion, [ExxonMobil] concluded the ERW pipe was not susceptible to seam failure.
Following each seam failure the Pegasus Pipeline experienced, including the seam failures that resulted from the 2005–2006 hydrostatic test, ExxonMobil applied the analysis set forth in the Baker Report decision tree to determine whether the pipeline's LF-ERW pipe was susceptible to seam failure. Applying the Baker Report decision tree was in accordance with the regulations, which while silent as to how operators are to determine seam failure susceptibility, instruct operators to “[f]ollow recognized industry practices.” 49 C.F.R. § 195.452(b)(6). The Baker Report certainly reflects “recognized industry practices.” Not only did the agency publish the Baker Report to its website and incorporate it into its enforcement manual, it has also relied on the Baker Report in prior enforcement decisions.6
To ensure that it was correctly applying the Baker Report and the decision tree, ExxonMobil retained the services of Dr. Kiefner, the Baker Report's co-author who largely developed the methodology that the decision tree represents. Because the pipeline did not reveal evidence of the two factors enumerated in the decision tree as indicating susceptibility to seam failure—namely, pressure-cycling induced fatigue or preferential seam corrosion—ExxonMobil concluded the pipeline was not susceptible to seam failure.7 After determining that the pipeline was not susceptible to longitudinal seam failure, ExxonMobil then considered these results, the pipeline's seam type, and history in developing its integrity reassessment schedules under § 195.452(e)(1). Undergoing this deliberate, considered process is precisely what § 195.452(e) requires.
The agency heavily criticizes ExxonMobil's reliance on the analysis embodied by the Baker Report decision tree. The agency found ExxonMobil to be in violation of the regulations because it determined that it was inappropriate for ExxonMobil to rely on the decision tree to justify ignoring seam failures exposed during testing. The agency contends that the decision tree's methodology is inaccurate because it does not address the pipe's “toughness.” The agency claims that brittle pipe, or pipe with low toughness, will not exhibit the same evidence of fatigue cracking, which is the end point of the decision tree. Therefore, the agency found, ExxonMobil failed to consider “that the absence of fatigue was a result of the low toughness of the pipe.”
We are unpersuaded by the agency's criticism of ExxonMobil's application of the process set forth in the Baker Report decision tree and the agency's interpretation of the Baker Report. Notably, the Baker Report itself specifically directs operators to rely on and apply the decision tree in reaching a seam failure susceptibility determination. Under the heading “Determination of Susceptibility,” the Baker Report states that
[t]he means of determining whether or not the seam of a particular pipeline is susceptible to failure are illustrated in [the decision tree]. ․ [The decision tree] allows one, by supplying appropriate data on a given segment, to determine if a seam-integrity assessment is required based on the federal pipeline integrity management regulations.
The Baker Report unequivocally directs pipeline operators to apply the decision tree, explaining that it is “[t]he means of determining whether or not the seam of particular pipeline is susceptible to seam failure.” To discredit the decision tree's methodology, the agency relies on language found elsewhere in the Baker Report that states that “[i]f a seam-related in-service or hydrostatic test failure has occurred on the segment, the segment is considered susceptible.” However, this argument is unavailing. Notably, yet another passage in the Baker Report states that “[i]f no fatigue-related failures exist, it is reasonable to certify that the pipeline is not susceptible to seam failures in the context of the federal integrity management requirements.” We are not sure what we—or more importantly, pipeline operators—are to make of this conflicting guidance. This causes us to place even more emphasis on the Baker Report's clear pronouncement that the decision tree is the means of determining seam failure susceptibility.
In sum, the record demonstrates that ExxonMobil met its obligation under § 195.452(e)(1)'s instruction to “consider” various risk factors. ExxonMobil concluded that the Pegasus Pipeline was not susceptible to seam failure only after lengthy, repeated, and in-depth consideration of seam failure risk factors and after utilizing the available industry-commissioned guidance of the Baker Report decision tree with the assistance of Dr. Kiefner—a national expert in LF-ERW pipe who co-authored the Baker Report. ExxonMobil's application of the decision tree was in accordance with the guidance of the Baker Report itself. This purposeful, deliberate decision-making process is precisely what § 195.452(e)(1) requires. The agency's finding that ExxonMobil did not “properly consider”8 all of the relevant factors simply because ExxonMobil reached a different determination from the one the agency now, in hindsight, proclaims ExxonMobil should have reached is contradicted by the record and is therefore arbitrary and capricious. Pension Benefit Guar. Corp., 374 F.3d at 366 (“Arbitrary and capricious review focuses on whether an agency articulated a rational connection between the facts found and the decision made.”).
The fact that the Mayflower release occurred, while regrettable, does not necessarily mean that ExxonMobil failed to abide by the pipeline integrity regulations in considering the appropriate risk factors. If it did, then an operator that experiences a seam-related pipeline leak on its pipeline system could never escape liability under pipeline integrity regulations, thus nullifying the regulations and creating a strict-liability regime that Congress has not authorized. See generally 49 U.S.C. § 60101 et seq. The unfortunate fact of the matter is that, despite adherence to safety guidelines and regulations, oil spills still do occur.
Because we conclude that ExxonMobil properly considered the susceptibility of its LF-ERW pipe to seam failure when establishing a continual integrity assessment schedule based on all risk factors on the Pegasus Pipeline, as required by the plain language of § 195.452(e)(1), we vacate Item 1 of the agency's final order. Because Items 2–4 and 7 are premised on the finding that ExxonMobil violated § 195.452(e)(1), we also vacate those Items.
Even assuming arguendo that we were to determine that the regulations governing seam failure susceptibility are ambiguous, the agency's interpretation would still fall short of warranting Auer deference for the additional and independent reason that ExxonMobil lacked fair notice of the interpretation of the regulation that the agency advances in this enforcement action. See Employer Sols., 833 F.3d at 487–89 (holding that “despite the degree of deference potentially owed” to the agency, Auer deference was inappropriate where the petitioner “lacked fair notice” of the agency's interpretation of its regulation).
We have warned that fair notice requires the agency to have “state[d] with ascertainable certainty what is meant by the standards [it] has promulgated.” Diamond Roofing Co., 528 F.2d at 649. This rule requires that agency regulations that “allow monetary penalties against those who violate them ․ must give [a party] fair warning of the conduct it prohibits or requires, and it must provide a reasonably clear standard of culpability to circumscribe the discretion of the enforcing authority and its agents.” Id. (cited by Christopher, 567 U.S. at 156 n.15). The Supreme Court has noted that “[i]n penalty cases, courts will not accord substantial deference to an agency's interpretation of an ambiguous rule in circumstances where the rule did not place the individual or firm on notice that the conduct at issue constituted a violation of a rule.” Christopher, 567 U.S. at 156 n.15 (quoting 1 R. Pierce Administrative Law Treatise § 6.11, at 543 (5th ed. 2010)); see also Employer Sols., 833 F.3d at 488 (“The challenged statute or agency action must ‘give the person of ordinary intelligence a reasonable opportunity to know what is prohibited, so that he may act accordingly.’ ”) (quoting Grayned v. City of Rockford, 408 U.S. 104, 108 (1972)). “In the absence of notice—for example, where the regulation is not sufficiently clear to warn a party about what is expected of it—an agency may not deprive a party of property by imposing civil or criminal liability.” General Elec. Co. v. E.P.A., 53 F.3d 1324, 1328–29 (D.C. Cir. 1995).
Under this analysis, the relevant inquiry is whether the agency's interpretation of the pipeline integrity regulations could have been understood with “ascertainable certainty” by ExxonMobil at the time it engaged in the conduct that allegedly exposed it to this enforcement action. Diamond Roofin g Co., 528 F.2d at 649. As explained by the D.C. Circuit,
we must ask whether the regulated party received, or should have received, notice of the agency's interpretation in the most obvious way of all: by reading the regulations. If, by reviewing the regulations and other public statements issued by the agency, a regulated party acting in good faith would be able to identify, with “ascertainable certainty,” the standards with which the agency expects parties to conform, then the agency has fairly notified a petitioner of the agency's interpretation.
General Elec. Co., 53 F.3d at 1329 (citing Diamond Roofing Co., 528 F.2d at 649). As explained above, the pipeline integrity regulations themselves did not provide ExxonMobil notice that the pipeline's leak history compelled it to label the LF-ERW pipe susceptible to longitudinal seam failure. In fact, references to susceptibility to longitudinal seam failure are surprisingly scarce within the text of § 195.452. Section 195.452(j)(5) requires operators to select an assessment method capable of assessing seam integrity only if the LF-ERW pipe segments are determined to be susceptible to longitudinal seam failure. Critically, however, the regulations are silent as to how operators are to make that determination. Even if this silence creates an ambiguity in the regulations, as the agency asserts, it does not provide ExxonMobil with fair notice that an operator is compelled to deem a pipeline as susceptible to seam failure just because the pipeline is constructed with LF-ERW pipe and has experienced leaks.
Because the regulations themselves are silent as to how operators are to determine seam failure susceptibility, operators are forced to find an extra-regulatory method to make this determination. ExxonMobil turned to the Baker Report. Given the agency's numerous endorsements of the Baker Report methodology, ExxonMobil was entirely justified to rely in good faith on the Baker Report to conduct its seam failure susceptibility analysis. Indeed, to hold otherwise in this enforcement action would constitute unfair surprise and deprive ExxonMobil of the fair notice to which it is entitled. See Diamond Roofing Co., 528 F.2d at 649.
The agency's criticism of ExxonMobil's reliance on the Baker Report decision tree amounts to a post hoc litigation-derived seam-susceptibility standard that deprives ExxonMobil of fair notice. The agency has failed to point to any instance where it has indicated to the industry generally or ExxonMobil specifically that it is inappropriate for a pipeline operator to rely on the Baker Report decision tree. The agency claims that another section of the Baker Report provided ExxonMobil with notice that a pipeline with a history of seam failures should have been deemed susceptible. As noted, this section states that “[i]f a seam-related in-service or hydrostatic test failure has occurred on the segment, the segment is considered susceptible.” While this sentence may contradict the methodology outlined in the decision tree, we are ultimately persuaded that ExxonMobil was justified to rely on the decision tree. The Baker Report explicitly states that the decision tree is “[t]he means” of determining seam failure susceptibility. The importance of the word “the” cannot be overstated. Even though there might be a conflict between the decision tree and other portions of the Baker Report, ExxonMobil was still justified in adhering to the decision tree as the means of determining whether its pipes were susceptible to seam failure.9
Simply put, ExxonMobil lacked any notice that the agency had a specific interpretation of the regulations that departed in any way from the Baker Report decision tree, let alone notice that they could be subject to an enforcement action for strictly adhering to the decision tree. In the absence of some agency instruction provided to the industry that the decision tree is no longer the means for determining seam failure susceptibility, we must conclude the agency's rejection of ExxonMobil's use of the decision tree's methodology violates the principles of fair notice to which ExxonMobil is entitled. Christopher, 567 U.S. at 156 (holding that Auer deference is “unwarranted” where to hold otherwise “would seriously undermine the principle that agencies should provide regulated parties ‘fair warning of the conduct [a regulation] prohibits or requires' ”). Thus, even if we were to determine that the regulations are ambiguous as to the precise methodology a pipeline operator is to use to determine whether pipe is susceptible to seam failure, ExxonMobil lacked fair notice, which alters the deference owed to the agency. Employer Sols., 833 F.3d at 490. The agency's new interpretation of the Baker Report decision tree “does not flow clearly from any authority in existence prior to this action. Thus, Auer ․ [is] inapplicable.” Id.10
In sum, the agency's interpretation of the pipeline integrity regulations does not warrant Auer deference because ExxonMobil lacked fair notice of the interpretation the agency now seeks to enforce in this action. Because the agency's interpretation is not entitled to Auer deference and because ExxonMobil reasonably complied with the guidance found in the Baker Report decision tree, we conclude that the agency acted in an arbitrary and capricious manner when it found that ExxonMobil violated § 195.452(e)(1). Therefore, we vacate Items 1–4 and 7 of the agency's final order on this alternate and additional ground.11
IV. Item 8
In addition to challenging Items 1–4 and 7 of the agency's final order, ExxonMobil also challenges Item 8. Under Item 8, the agency found that ExxonMobil had violated § 195.452(b)(5), which requires a pipeline operator to develop and follow a written integrity management program. 49 C.F.R. § 195.452(b)(5). Because Item 8 is not based on the same grounds as the other challenged violations, and because ExxonMobil has not articulated how the agency acted in an arbitrary or capricious manner in finding that ExxonMobil had violated § 195.452(b)(5) under Item 8, we affirm the agency's finding with respect to this violation.
Under Item 8 of its final order, the agency explained that ExxonMobil's written IMP plan provides for the use of the Threat Identification and Risk Assessment (“TIARA”) program in the risk management process. This program requires ExxonMobil to manually enter information in response to certain questions. One of the questions posed to ExxonMobil was: “Has a ILI crack tool (TFI or UT) been successfully run and have the appropriate repairs been scheduled?” In 2011, ExxonMobil answered “yes” to this question even though no such tool had been run.12 ExxonMobil explained in the administrative hearing that the decision to answer “yes” was based on a belief that it would perform a TFI tool assessment in the near future. The tool run, however was delayed approximately two years. Despite this delay, ExxonMobil never revisited its answer to the question.
In its Final Order, the agency found that
[b]y answering this question in the affirmative, [ExxonMobil] misrepresented the current status of integrity verification on the pipeline. The answer did not accurately reflect the fact that the tool had not been run and no repairs had been scheduled. The issue was then compounded when the tool run became delayed for two years. As a result, [ExxonMobil] failed to adhere to the procedures as written.
Ultimately, the agency concluded that ExxonMobil violated § 195.452(b)(5) “by failing to follow its written procedures for the TIARA program by incorrectly indicating that a TFI tool run had been performed and then failing to correct it when the tool run was delayed.”
ExxonMobil claims in only the most general terms that Item 8 of the agency's final order should be vacated along with the other challenged violations. However, ExxonMobil does not directly address its basis for this argument. Unlike Items 4 and 7, in which the agency expressly relies on Item 1 in finding that ExxonMobil committed a violation, it is not apparent how Item 8 relates to the other contested violations, most importantly those pertaining to § 195.452(e)(1). We therefore conclude that ExxonMobil has forfeited its challenge to Item 8.13 See United States v. Scroggins, 599 F.3d 433, 446–47 (5th Cir. 2010) (holding that an argument is not adequately presented on appeal by a party that fails to identify the relevant legal standards or authority to support its argument).
Even if we were to liberally construe ExxonMobil's argument, its argument is unpersuasive. ExxonMobil appears to argue that even though it misrepresented that it had run an ILI tool assessment which resulted in the TIARA program indicating that there were no identified threats to the pipeline, it went above and beyond what it was required to do and ran a state-of-the-art ILI seam/crack tool anyway in 2012–2013, which did not detect any defects in the pipeline. Therefore, ExxonMobil argues, even if it had accurately answered the question and was required to run an ILI tool as a result of its answer, the ILI tool run would not have detected any defects in the pipeline so it should not be found to have been in violation of the regulations. However, the agency found ExxonMobil to be in violation of § 195.452(b)(5) under Item 8 primarily because ExxonMobil did not follow its written IMP plan when it failed to accurately use the TIARA program in the risk management process, not just because ExxonMobil did not run the appropriate test.
Accordingly, we affirm the agency's finding that ExxonMobil violated § 195.452(b)(5) under Item 8. However, for the reasons explained below, we remand to the agency to reevaluate the penalty levied against ExxonMobil under Item 8.
ExxonMobil challenges the penalties imposed by the agency. We review an agency's penalty determination under the arbitrary and capricious standard. Interamericas Invs., Ltd. v. Bd. of Governors of the Fed. Reserve Sys., 111 F.3d 376, 384 (5th Cir. 1997). First, ExxonMobil argues that the penalties should be capped at $1,000,000 under the Pipeline Safety Act. Second, ExxonMobil argues that the agency erred in increasing the assessed penalty for Item 8 on the basis that it had a “contributory impact” on the Mayflower release. We discuss each of these arguments below.
First, ExxonMobil argues that the total amount of the penalty assessed by the agency must be capped at $1,000,000 under the Pipeline Safety Act, 49 U.S.C. § 60122. We disagree. In 2012, Congress raised the maximum civil penalty that the agency could levy for “a related series of violations” of the pipeline safety regulations from a cap of $1,000,000 to $2,000,000. See 49 U.S.C. § 60122(a). The statute states that “[a] separate violation occurs for each day the violation continues.” 49 U.S.C. § 60122(a)(1) (emphasis added). Therefore, each day a violation under the remaining Items in the final order occurred after 2012 was a separate violation subject to the $2,000,000 penalty maximum. Accordingly, we conclude that the agency did not act in an arbitrary or capricious manner in applying the $2,000,000 penalty maximum.14
ExxonMobil's second argument pertains to Item 8, which we have affirmed above. ExxonMobil argues that the agency erred in increasing the assessed penalty for Item 8 on the ground that it had a “contributory impact” on the Mayflower spill.15 In determining the penalty amount, the agency is directed by statute to consider the following: the nature, circumstances, and gravity of the violation, including adverse impact on the environment; the degree of the violator's culpability, any history of prior violations, and any effect on ability to continue doing business; and the violator's good faith in attempting to comply. 49 U.S.C. § 60122(b). In assessing the $783,300 penalty for Item 8, the agency applied these factors and stated the following in its final order:
With regard to the nature, circumstances and gravity of the violation, including adverse impact on the environment, the Violation Report suggested the violation had the highest level of gravity because the violation was a causal factor in the Mayflower Accident, which was the result of ERW pipe seam failure. In addition, with regard to the degree of culpability and good faith, the Violation Report suggested that [ExxonMobil] had an elevated degree of culpability and that no good faith credit was warranted.
ExxonMobil contends that the violation cited in Item 8 did not contribute to the Mayflower release because even if it had complied with the regulations—i.e. even if it had treated the pipeline as susceptible to seam failure and assessed the integrity of the pipe with a method capable of assessing seam integrity under § 195.452(j)(5)—it would not have discovered the specific defect in the Pegasus Pipeline that caused the release. ExxonMobil bases this argument on the fact that when it ultimately ran such an assessment, the defect was not detected.
We conclude that the agency erred in finding that Item 8 “had the highest level of gravity because the violation was a causal factor in the Mayflower Accident.” While it is true, as discussed above, that ExxonMobil is culpable of misrepresenting that a TFI tool run had been performed, the record does not support the agency's finding that the misrepresentation “was a causal factor in the Mayflower Accident.” When the agency ran the TFI tool in 2012–2013 shortly before the Mayflower release occurred, the tool was unable to identify a defect in the Pegasus Pipeline even though the third-party vendor who was later analyzing the results from the tool run knew that the Mayflower release occurred. It follows that even had ExxonMobil run the TFI tool in a timely manner, the results of the run would not have identified a defect in the pipeline and therefore would not have prevented the oil spill.
The agency contends that the fact that ExxonMobil's integrity assessment tool run did not detect any anomaly at the site of the pipeline's failure “may have” been because ExxonMobil used an inappropriate tool for that assessment. But the agency did not make a finding on this point.16 Indeed, the agency suggested in Item 1 of its final order that the TFI seam/crack tool was appropriate for that assessment.17 Further, the regulations indicate that either an in-line inspection tool or a hydrostatic test are appropriate for assessing LF-ERW pipeline seam integrity. See 49 C.F.R. § 195.452(j)(5)(i)–(ii). For the agency to now come forward with a new rule that this particular in-line inspection tool is not appropriate for assessing seam integrity in these circumstances would implicate fair notice concerns. See Diamond Roofing Co., 528 F.2d at 649.
We conclude that the agency acted contrary to the evidence before it and in an arbitrary and capricious manner when it determined that ExxonMobil's misrepresentation to the TIARA program was a causal factor in the Mayflower release. Accordingly, we remand to the agency to reevaluate what would be an appropriate penalty for Item 8 in light of this determination and in light of our decision vacating the violations associated with Items 1–4 and 7 of the final order. Further, given that we have vacated Items 1–4 and 7, we likewise vacate the penalties associated with those items.
VI. Compliance Order
ExxonMobil challenges the agency's compliance order. The compliance order contains direction pertaining to Items 1, 2, 5, 6, and 8 of the agency's final order. ExxonMobil argues generally that the terms of the compliance order exceed the agency's authority. Under the Pipeline Safety Act, the agency is authorized to “issue orders directing compliance” with a regulation promulgated by the agency and such orders must “state clearly the action a person must take to comply.” 49 U.S.C. § 60118(b).
Because we have vacated Items 1 and 2 of the agency's final order, we similarly vacate Paragraphs 1 and 2 of the compliance order. ExxonMobil also purports to challenge the paragraphs of the compliance order that are related to Items 5, 6, and 8. However, ExxonMobil did not challenge these provisions of the compliance order in the proceedings before the agency. To the extent ExxonMobil now seeks to challenge these provisions of the compliance order, it has forfeited those arguments. See Scroggins, 599 F.3d at 446–47; see also United States v. Chavez-Valencia, 116 F.3d 127, 130 (5th Cir. 1997) (the failure to raise a claim below “constitutes a forfeiture ․ of that right for the purposes of appeal”). Therefore, the paragraphs of the compliance order pertaining to Items 5, 6 and 8 remain in effect.
According to the unambiguous text of the pipeline integrity regulations, pipeline operators are required to “consider” various risk factors when they prioritize pipelines for assessment. This is a process-based requirement that does not mandate a particular outcome, but rather prescribes a careful, informed decision-making process that pipeline operators must undergo in good faith. ExxonMobil complied with this requirement when it determined the Pegasus Pipeline was not susceptible to seam failure by applying the methodology set forth in the Baker Report decision tree and, considering this determination, the pipeline's seam type, and leak history, developed its integrity reassessment schedule under § 195.452(e)(1). We therefore VACATE the agency's final order with respect to Items 1–4 and 7, which are all premised on a finding that ExxonMobil did not properly consider the appropriate factors in developing its integrity reassessment schedule under § 195.452(e)(1), and VACATE the penalties associated with Items 1–4 and 7. We REMAND to the agency with instructions to reconsider the penalty imposed for Item 8 in light of our determination that the Item 8 violation was not a causal factor in the Mayflower release.
I agree with the majority that ExxonMobil lacked sufficient notice of the agency's interpretation of 49 C.F.R. § 195.452(e). Consequently, I agree that Items 1-4 and 7 should be vacated. I also agree that the agency's determination regarding Item 8 should be affirmed, but should be remanded to the agency to reevaluate the basis for the penalty associated with this violation. But the regulation is ambiguous and for that reason, deference is appropriate on that ground. It is well-settled that an agency's interpretation of its own regulation must be given “controlling weight unless it is plainly erroneous or inconsistent with the regulation.” Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945). “[T]his broad deference is all the more warranted when, as here, the regulation concerns ‘a complex and highly technical regulatory program,’ in which the identification and classification of relevant ‘criteria necessarily require significant expertise and entails the exercise of judgment grounded in policy concerns.’ ” Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 412 (1994) (quoting Pauley v. Beth Energy Mines, Inc., 501 U.S. 680, 697 (1991)).
The majority acknowledges that the regulatory program here is highly complex and, at minimum, involves “lengthy, repeated, and in-depth analysis” of risk factors by referencing the Baker Report's elaborate and “conflicting” guidance. Instead of concluding that the agency's interpretation of § 195.452(e) is not plainly erroneous, however, the majority creates its own definition of “consider,” which requires pipeline operators “to carefully undergo an informed decision-making process in good faith, reasonably taking into account all relevant risk factors in reaching a decision.” The majority states that its definition reflects the unambiguous meaning of the regulation. See Christensen v. Harris Cnty., 529 U.S. 576 (2000) (“Auer deference is warranted only when the language of the regulation is ambiguous.”) But its interpretation is not compelled by the regulation's plain language, which tersely states that a pipeline operator “must consider” “all risk factors that reflect the risk conditions on the pipeline segment.” See § 195.452(e)(1). Nor is it compelled by any authority that the majority cites. See, e.g., J.H. Miles & Co., Inc. v. Brown, 910 F. Supp. 1138, 1156 (E.D. Va. 1995) (stating that “consider” “can hardly be read as a strict dictate” but “[t]his does not mean that the Council has carte blanche to ignore plainly relevant information”) (emphasis added). More to the point, the majority's interpretation, however reasonable or well-crafted, cannot supplant the agency's interpretation of its own ambiguous regulation, “unless it is plainly erroneous or inconsistent with the regulation,” which it is not. Seminole Rock, 325 U.S. at 414.
Though the majority is correct that plain language of § 195.452(e) allows pipeline operators some discretion regarding how they choose to comply with the regulation, it is implausible that the agency would enact a regulation that would be so toothless as to be practically unenforceable, and would give regulated parties ultimate power to decide whether they are in compliance. Because the agency's interpretation of its own ambiguous regulation is not “plainly erroneous or inconsistent with the regulation,” I would give it deference. See Seminole Rock, 325 U.S. at 414.
1. A hydrostatic test is performed by subjecting a pipeline to pressures that exceed its maximum operating pressure, thereby identifying the weakest segments of the pipeline.
2. This minor leak is known as a “weep/pinhole leak” and it only released about two gallons of oil before it was repaired.
3. Both parties agree that this is the central issue on appeal. Both parties also agree that the violations cited in Items 1, 2, and 3 of the final order are premised on the agency's finding that ExxonMobil did not properly consider the risk factors as directed by § 195.452(e)(1). The parties disagree, however, as to whether Items 4, 7, and 8 rise and fall with our conclusion regarding § 195.452(e)(1). We conclude that Items 4 and 7 rise and fall based on our interpretation of § 195.452(e)(1) while Item 8 does not.Notably, in explaining its basis for Items 4 and 7, the agency's final order explicitly relies on Item 1—failing to properly consider all risk factors under § 195.452(e)(1) to determine seam failure susceptibility. In other words, the basis for Items 4 and 7 would be undermined if we were to find that ExxonMobil did not violate § 195.452(e)(1) and vacate Item 1. Therefore, we conclude that Items 1–4 and 7 are all premised on whether ExxonMobil properly “considered” the relevant risk factors under § 195.452(e)(1) in determining that the pipeline was not susceptible to longitudinal seam failure.We conclude that Item 8, however, is not premised on the agency's finding that ExxonMobil did not properly consider risk factors as directed by § 195.452(e), but rather is supported by an independent basis, as we discuss below. Our conclusion is consistent with the position ExxonMobil took in its post-hearing briefing before the agency. Even though ExxonMobil argued before the agency that Items 1–4 and 7 were a related series of violations because they all relied on the same assertion by the agency that ExxonMobil failed to consider the Pegasus Pipeline to be susceptible to seam failure, ExxonMobil did not contend that Item 8 was based on this assertion.
4. Ordinarily, even if a regulation is unambiguous such that Auer deference does not apply, the agency's rulings, interpretations, and opinions may still be “entitled to respect” under Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944), “but only to the extent that those interpretations have the power to persuade.” Moore v. Hannon Food Serv., Inc., 317 F.3d 489, 497 (5th Cir. 2003); see also Belt v. EmCare, Inc., 444 F.3d 403, 408 (5th Cir. 2006) (“If the regulation is unambiguous, we may still consider agency interpretation, but only according to its persuasive power.”). We have noted, however, that “[u]ltimately, Skidmore analysis is of limited value in interpreting regulations, given that it stops short of requiring deference and is likely to be invoked only when a court has already found the regulation to be unambiguous.” Moore, 317 F.3d at 498 n.14; cf. United States v. Mead Corp., 533 U.S. 218, 250 (2001) (Scalia, J., dissenting) (“[T]he rule of Skidmore deference is an empty truism and a trifling statement of the obvious: A judge should take into account the well-considered views of expert observers.”). Here, the agency argues only that we should apply Auer deference to its interpretation of the pipeline integrity regulations; it does not offer any alternative argument as to Skidmore deference. When pressed at oral argument as to where we could find the agency's interpretation of the regulations to which the agency argues we should defer, the agency could not point us to a particular interpretation. Therefore, we need not apply Skidmore deference to the agency's interpretation of the regulation—whatever that interpretation might be.
5. Section 195.303(d) of the pipeline safety regulations states that “[a]ll pre-1970 ERW pipe ․ is deemed susceptible to longitudinal seam failures unless an engineering analysis shows otherwise.” 49 C.F.R. § 195.303(d). This regulation was issued in 1998 in a subpart of the pipeline safety regulations entitled “Pressure Testing” and is distinct from and was promulgated prior to the pipeline integrity regulations at issue here. Section 195.303(d) required operators to conduct a one-time hydrostatic pressure test of certain categories of pipelines and allowed operators to elect by December 1998 if they wanted to use risk-based criteria rather than a hydrostatic test. That one-time election was unavailable for pre-1970 LF-ERW pipe. Section 195.303(d) did not have any further applicability after the December 1998 deadline. When the agency promulgated the pipeline integrity regulations at issue in this case, it did not create a similar rebuttable presumption regarding pre-1970 LF-ERW pipe. In its Final Order and Decision on Petition for Reconsideration, the agency specifically cites to § 195.303(d) as support for the position that there is a presumption that all LF-ERW pipe is susceptible to longitudinal seam failure. This presumption is plainly inaccurate because § 195.303(d) is obsolete and is therefore not relevant here. The agency appears to have abandoned this position on appeal, contending that it did not apply a presumption of susceptibility for pre-1970 ERW pipe based on § 195.303(d). In any event, we conclude that the agency cannot base any part of its finding on the now-obsolete § 195.303(d) and that § 195.303(d) did not provide ExxonMobil with fair notice that it was compelled to deem its LF-ERW pipe as susceptible to longitudinal seam failure.
6. In one prior enforcement action, another company argued that there was a lack of industry guidance for how to determine the susceptibility of LF-ERW pipe to seam failures. In re Kinder Morgan Energy Partners, CPF No. 1-2004-5004 (June 26, 2006), at 3. The agency responded that the methodology embodied in the Baker Report is an example of an acceptable means of performing a seam failure susceptibility analysis that is available to the industry. Id.
7. Dr. Kiefner, after reviewing ExxonMobil's inspection and evaluation activities and data, testified that ExxonMobil “correctly followed the guidance described in the Baker Report. This [analysis] would not have resulted in a finding that the failed segment was ‘susceptible to seam failure in the context of Part 195 IMP regulations.’ ” Notably, the agency itself acknowledges that “[t]he evidence supports [ExxonMobil's] assertion that prior seam failures did not exhibit evidence of fatigue.”
8. The agency has recognized that pipeline operators have wide latitude in how they will weigh various risk factors in determining how to prioritize pipeline segments for assessment. See In re Magellan Midstream Partners, L.P., CPF No. 4-2006-5020, at *7 (“Section 195.452(e)(1) lists nine factors that must be considered in establishing a schedule but leaves it up to the operator to determine what other factors need to be considered, how to assign risk scores to each factor and pipe segment, and how to prioritize assessments.”) (emphasis added). Therefore, to say that ExxonMobil did not “properly consider” the appropriate risk factors begs the question of what would constitute proper consideration. To the extent the agency contends that proper consideration mandates a particular outcome, this is not supported by the text of the regulation nor the industry guidance of the Baker Report.
9. As to the agency's contention that applying the decision tree is inappropriate because it allegedly does not address the toughness of the pipe, the Baker Report notes that all LF-ERW pipe “possess bondline regions that are prone to low toughness and brittle-fracture behavior.” Because all LF-ERW pipe is prone to low toughness and brittle cracking, those factors are presumably built into the process for analyzing seam-failure susceptibility represented by the decision tree. To the extent that the agency believes that this is not the case, the agency must “state with ascertainable certainty,” before an enforcement action, that applying the decision tree is inappropriate for this reason. Diamond Roofing Co., 528 F.2d at 649.
10. If the agency now believes there is a proper methodology to determine seam failure susceptibility, it would be entirely appropriate for the agency to promulgate a new regulation to that end moving forward, following notice and comment from all concerned parties. However, the pipeline integrity regulations as currently constituted only broadly instruct pipeline operators to examine whether pipe is susceptible to seam failure and then to factor that consideration into its overall § 195.452(e)(1) risk evaluation.
11. “This circuit follows the rule that alternative holdings are binding precedent and not obiter dictum.” United States v. Potts, 644 F.3d 233, 237 n. 3 (5th Cir. 2011).
12. ExxonMobil does not cite to or contest any of these facts in its brief.
13. Notably, in its brief, ExxonMobil appears to concede that Item 8 is not premised on Item 1 like the other challenged violations are, stating that “[the agency] has alleged one regulatory violation, namely Violation 1 for alleged failure to ‘consider’ certain risk information and conclude that the Pegasus Pipeline was susceptible to seam failure. [The agency] then expanded it to four additional violations (Nos. 2-4 and 7).”
14. ExxonMobil also argues that the agency acted in an arbitrary and capricious manner when it rejected its argument that Items 1–4 and 7 are “a related series of violations” under 49 U.S.C. § 60122(a)(1). However, in light of our decision vacating Items 1–4 and 7, we need not reach this question because the penalties associated with the remaining violations are less than $2,000,000.
15. ExxonMobil also argues that the agency erred with regard to Items 1 and 2 on the same grounds; however, because we have vacated Items 1 and 2, we need not address this argument as to those violations.
16. In its order denying ExxonMobil's Petition for Reconsideration, the agency stated that[a]lthough no anomaly was previously detected at the failure location using a TFI tool, there were questions raised during the proceeding about the appropriateness of using a TFI tool in the first place, given that the types of defects detected by hydrostatic tests in 2005–2006 would not likely be detected with a TFI tool. ․ While the Final Order did not decide if hydrostatic testing would have detected the anomaly that failed, the fact that it was not detected does not negate the contributory impact of the violations.
17. Under Item 1 of its final order, the agency noted that “[i]t was not until 2012–2013 that [ExxonMobil] finally performed an ILI using a TFI seam/crack tool, which is designed to detect certain ERW seam integrity issues.” (emphasis added).
JENNIFER WALKER ELROD, Circuit Judge: