United States v. Chevron U.S.A. Inc.

380 F. Supp. 2d 1104 (2005)

UNITED STATES of America, Plaintiff, and
State of Hawaii, Mississippi Commission on Environmental Quality, State of Utah, Bay Area Air Quality Management District, Plaintiff-Intervenors,
v.
CHEVRON U.S.A. INC., Defendant.

No. C 03-4650CRB.

United States District Court, N.D. California.

June 24, 2005.

A. Kent Mayo, Thomas S. Sansonetti, U.S. Department of Justices, Environment & Natural Resources Division, Environmental Defense Sections, Washington, DC, Charles M. O'Connor, Kevin V. Ryan, Office of the U.S. Attorney, San Francisco, CA, for Plaintiff.

William Francis Cooper, Dept. of Attorney General, State of Hawaii, Honolulu, HI, Kelly R. Riley, Mary Jacqueline Easley, Mississippi Department of Environmental Quality, Legal Division, Jackson, MS, Christian Curtis Stephens, Lenore Epstein, Utah Attorney General, Salt Lake City, UT, Adan A. Schwartz, Brian Charles Bunger, BAAQMD/Office District Counsel, San Francisco, CA, for Plaintiff-Intervenors.

J. Michael Rockett, Pillsbury Winthrop Shaw Pittman LLP, San Francisco, CA, for Defendant.

MEMORANDUM AND ORDER ENTERING CONSENT DECREE

BREYER, District Judge.

The United States filed this action alleging violations of various environmental laws by defendant Chevron. Now before the Court is the motion to enter the proposed Consent Decree ("Decree") settling the matter. After carefully considering the positions of all parties and amici, the Court hereby GRANTS the motion.

BACKGROUND

A. The Petroleum Refinery Initiative

The government's approach to this settlement can only be understood in the context of EPA's broader Petroleum Refinery Initiative ("PRI"), which in addition to the Consent Decree before this Court has produced numerous settlements with oil refineries across the nation. Beginning in 1996, EPA identified the need to address widespread environmental law compliance problems in the petroleum refining industry through a nationwide strategy. In developing the strategy, EPA states that the traditional method of enforcing environmental laws — through investigation, identification of violations, further investigation, and finally litigation and/or settlement — was viewed as costly and time-consuming. Therefore, alternative strategies were pursued. In developing such a strategy for petroleum refineries, EPA identified four areas of law that were seen as of the greatest concern: 1) New Source Review and Prevention of Significant Deterioration ("NSR/PSD"); 2) New Source Performance Standards ("NSPS"); 3) Leak Detection and Repair ("LDAR"); and 4) National Emissions Standards for Hazardous Air Pollutants for benzene ("benzene waste NESHAP"). EPA also focused on four types of refinery structures: fluidized catalytic cracking units ("FCCUs"), process heaters and boilers, flares, and sulfur recovery plants ("SRPs").

Under the PRI, EPA identified refineries that were potentially out of compliance by using an industry publication, the Oil & Gas Journal, to determine which refineries had significantly expanded their capacities and were therefore most likely to have engaged in modifications triggering Clean Air Act review. Investigations conducted by EPA based on this and other strategies through early 2000 suggested that over 70% of the targeted refineries had significant non-compliance problems. EPA then identified seven companies seen to be the worst offenders and conducted further investigation of them. Chevron was not among that group, and therefore was not investigated.

EPA's investigations of the targeted refineries identified significant levels of non-compliance in some refinery structures, and moderate levels in others. Specifically, EPA found widespread violations in refinery FCCUs, with as much as three in four units out of compliance. Second Foley Decl. ¶ 50. Somewhat fewer heaters and boilers (less than one-third) were found in violation, although this was still seen as a significant problem requiring action. See id. ¶ 67. Similarly serious levels of non-compliance were also found in refinery sulfur recovery plants and flares. McCoy Decl. ¶¶ 15, 32.

EPA then commenced negotiations with refineries based on what they had learned. During the course of these negotiations, EPA developed a model template for injunctive relief to provide guidance regarding the agency's goals for PRI settlements while also creating consistency across settlements. The provisions of the template were selected based on EPA's understanding of the scope of compliance problems in the industry and list which pollution control strategies EPA aspires to mandate in consent decrees negotiated under the PRI. Because the template was developed through actual negotiations, EPA feels that it strikes the appropriate balance between remedying the perceived scope of non-compliance in the industry and applying settlement criteria that are economically and technologically feasible for the refineries.

B. The Chevron Settlement

During an inspection in 1999 EPA Region 8 identified probable and significant non-compliance with benzene waste NESHAP at Chevron's Salt Lake City refinery. Based on this data alone, EPA initiated settlement negotiations with Chevron pursuant to the PRI model template. At the beginning of the negotiations, EPA told Chevron that it would not engage in any further investigation if it was satisfied that Chevron was pursuing settlement in good faith. Apparently satisfied with Chevron's cooperation, EPA conducted no further investigation.

Chevron agreed to engage in global settlement negotiations in December 2001 and the parties began discussions in March 2002. The negotiators sought to resolve Chevron's potential liability at five refineries located in: Pascagoula, Mississippi; Richmond, California; El Segundo, California; Salt Lake City, Utah; and Kapolei, Hawaii. Joining in the negotiations were environmental agencies from the states of Hawaii, Mississippi, and Utah, as well as the Bay Area Air Quality Management District ("BAAQMD"). The South Coast Air Quality Management District ("SCAQMD"), which has jurisdiction over the El Segundo plant, did not participate.

The parties had differing positions on all of the Consent Decree's key subject areas, including PSD/NSR requirements for each of the major refinery structures. See Mayo Decl. ¶ 23. The timing and method of air pollution reductions, as well as the baseline against which reductions would be measured were all in dispute. See id. The parties worked through all of the terms over the next several months and were able to circulate a draft Consent Decree by June 2003. Negotiations on the final issues continued throughout the summer and the final Decree was lodged before the Court on October 16, 2003, the same day that the United States filed the complaint in this case. The complaint sets forth general allegations of Clean Air Act violations, along with allegations of liability under CERCLA and EPCRA. All the state environmental agencies except SCAQMD later intervened in the case.

The final Consent Decree demonstrates that EPA made significant concessions from the goals it established in the PRI template. EPA found these concessions acceptable because they believed Chevron had a better compliance record than the refineries that were used to create the template. That assumption was based on several factors, chief among them the fact that Chevron had engaged in less capacity expansion in recent years than other refineries.

In November and December of 2003, lengthy and detailed comments mainly criticizing the Clean Air Act provisions of the Decree were submitted by SCAQMD and the amici to this action. The government provided detailed responses rejecting the critiques either as mistaken interpretations of the Decree or as unrealistic given that the Decree reflects a settlement of a still-disputed and not-yet-investigated case. No changes were made to the Consent Decree in response to the comments.

Amici, a group of environmental organizations, unions, and the law firm of Adams Broadwell Joseph & Cardozo have raised similar objections to the current motion to enter the Decree. Essentially, they argue that: 1) this Court lacks jurisdiction because the government did not comply with proper notice procedures; 2) the agreement was not procedurally fair because the government has relied almost exclusively on Chevron in evaluating the violations rather than doing an independent investigation and evaluation; and 3) the agreement is not substantively fair because it releases Chevron from a large universe of liability without forcing the company to significantly reduce pollution. EPA replies that there is jurisdiction, that the decree was the product of months of adversarial negotiations and that it results in significant environmental benefits.

EPA and Chevron now move for this Court's approval of the Consent Decree.

JURISDICTION

Amici lodge two challenges to this Court's subject matter jurisdiction over this matter: (1) that the EPA's failure to give formal written Notice of Violations (NOVs) deprives this Court of subject matter jurisdiction; and (2) that the Consent Decree resolves a broader scope of liability than was pled in the complaint and thus this Court lacks jurisdiction to bless a settlement that covers issues not before it. Both arguments fail, and therefore this Court has jurisdiction.

A. Failure to issue NOVs

Amici argue that the issuance of a formal written Notice of Violation ("NOV") to the violator and the State is a jurisdictional prerequisite to the filing of an enforcement action. This contention is based on Clean Air Act section 113(a), which regulates EPA enforcement:

Whenever, on the basis of any information available to the Administrator, the Administrator finds that any person has violated or is in violation of any requirement or prohibition of an applicable implementation plan or permit, the Administrator shall notify the person and the State in which the plan applies of such finding. At any time after the expiration of 30 days following the date on which such notice of a violation is issued, the Administrator may, without regard to the period of violation (subject to section 2462 of Title 28) —

(A) issue an order requiring such person to comply with the requirements or prohibitions of such plan or permit,

(B) issue an administrative penalty order in accordance with subsection (d) of this section, or

(C) bring a civil action in accordance with subsection (b) of this section.

42 U.S.C. § 7413(a)(1) (emphasis added). It is uncontested that Chevron and the state agencies received actual notice. Therefore, amici's argument is that section 7413 requires EPA to issue a notice that comes in the form of an official written "Notice of Violation," and that other forms of notice are insufficient.

Section 7413 is silent with respect to the form the notice must take. See United States v. BP Exploration & Oil Co., 167 F. Supp. 2d 1045, 1051 (N.D.Ind.2001); United States v. Brotech Corp., No. Civ. A-00-2428, 2000 WL 1368023, at *2 (E.D.Pa. Sept.19, 2000). Therefore, the Court agrees that "[r]ather than formal written notice, actual notice of violations is sufficient." BP Exploration, 167 F.Supp.2d at 1051 citing United States v. B & W Inv. Properties, 38 F.3d 362, 366 (7th Cir.1994). See also Brotech, 2000 WL 1368023, at *2 ("In this context, substance, we believe, is more important than form."). Amici's attempts to distinguish the statements in these cases as dicta fail because that claim directly conflicts with their contention that the notice requirement is jurisdictional.

Amici's other arguments against jurisdiction also fail. First, amici cite EPA's own memorandum titled "Guidance on Complying with Notification Requirements on Section 113(a)(1) and 113(a)(4) of the Clean Air Act" as authority demonstrating that only a formal NOV will suffice for satisfying the notice requirement. The memorandum "recommends that the notification requirements of Section 113(a)(1) be met by the issuance of a written notice of violation (NOV)." First Mayo Decl. Exh. 4 at 2, 4. However, amici do not address why this statement has any jurisdictional implications, since it is on its face a mere recommendation. Indeed, the memorandum states that the notice requirement is satisfied where a source has received "substantial or constructive notice." Id. at 4. It also states that "[t]he notice provisions in Section 113 are general in nature, giving EPA a great deal of latitude," and "the specific procedures recommended [by the memo] are not necessarily compelled by the Act or judicial decisions [and by] recommending specific procedures this guidance is not meant to imply the existence of jurisdictional... limitations on EPA's enforcement authority." Id. at 1-2. This text and the text of the statute therefore do not support a requirement of a specific form of notice as a jurisdictional prerequisite.

The cases cited by amici are also inapposite. See, e.g., United States v. Pan American Grain Mfg., 29 F. Supp. 2d 53, 56 (D.P.R.1998) (holding that EPA may only bring civil enforcement action based on specific violations alleged in NOV); United States v. Louisiana-Pacific Corp., (Louisiana-Pacific II) 682 F. Supp. 1141, 1155 (D.Colo.1988) (same); United States v. Louisiana-Pacific Corp., (Louisiana-Pacific I) 682 F. Supp. 1122 (D.Colo.1987).^[1] None of the cases cited speaks directly to the issue of the form of notice. They instead hold that EPA may only bring a civil enforcement action based on the specific violations alleged in an NOV. Pan American Grain, 29 F.Supp.2d at 56; Louisiana-Pacific II, 682 F.Supp. at 1128. Those cases do not discuss whether notice could come in some other form while still restricting the scope of the enforcement action.

Moreover the cases are also based on the erroneous assumption that the notice requirement was included to allow the violator to come into compliance. The legislative history to the 1990 amendments to section 7413 clarifies that the notice requirement is "intended solely to allow the State an opportunity to take appropriate action." PL 101-549, Senate Rep. No. 101-228, pp. 361-62. It also makes clear that the notice requirement is not intended to be construed in a way that would make EPA enforcement more difficult. Id. In light of such purposes, the reasoning of the Louisiana-Pacific cases and their progeny is put into doubt. There is no reason to assume that failure to provide notice of a specific violation makes the violator immune to suit until such notice has been provided.^[2]

Accordingly, the form of notice does not preclude this Court's jurisdiction over this case.

B. Scope of the Complaint

Amici point out that although the United States' complaint merely alleges that Chevron modified "certain" of the heaters and boilers at its refineries, the Consent Decree releases Chevron from liability for all heater and boiler modifications. Amici reason that to do so is to ask the Court to unlawfully resolve issues not before it.

This contention also fails. A consent decree need only "com[e] within the general scope of the case made by the pleadings" and a federal court is "not necessarily barred from entering a consent decree merely because the consent decree provides broader relief than the court could have awarded after a trial." Local Number 93, Int'l Ass'n of Firefighters v. City of Cleveland, 478 U.S. 501, 525, 106 S. Ct. 3063, 92 L. Ed. 2d 405 (1986) (citations and internal quotations omitted); see also Sierra Club v. Electronic Controls Design, 909 F.2d 1350, 1355 (9th Cir.1990) (holding that the court had the power to enter a consent decree as long as the decree "further[ed] the broad objectives upon which the complaint was based and does not violate the Clean Water Act"). Here, the complaint seeks the broad objectives of remedying alleged widespread violations of several environmental laws. The Consent Decree resolves issues within that general scope and does not violate the Clean Air Act. The Court therefore has jurisdiction.

DISCUSSION

A. Standard of Review

A district court reviews a consent decree to determine whether the decree is "fundamentally fair, adequate and reasonable." United States v. Oregon, 913 F.2d 576, 580 (9th Cir.1990). In addition, while a consent decree "must conform to applicable laws...[it] need not impose all the obligations authorized by law." Id.

The Court's review of the Consent Decree is conducted in light of the public policy favoring settlement. See United States v. Comunidades Unidas Contra La Contaminacion, 204 F.3d 275, 280 (1st Cir.2000). This deference is particularly strong where the decree has been negotiated by the Department of Justice on behalf of an agency like the EPA which is an expert in its field. United States v. Akzo Coatings of Am., Inc., 949 F.2d 1409, 1436 (6th Cir.1991). However, while it is appropriate to pay such deference, the Court must avoid giving a "rubberstamp approval" and instead must conduct an independent evaluation. BP Exploration, 167 F.Supp.2d at 1050 (citations omitted). Where a consent decree impacts public interests, the court has a heightened responsibility if the interests at stake were not represented in the negotiating process. See Oregon, 913 F.2d at 581.

In applying the "fair, adequate and reasonable" standard, courts examine both procedural and substantive fairness. See United States v. Cannons Engineering Corp., 899 F.2d 79, 86 (1st Cir.1990); Environmental Defense v. Leavitt, 329 F. Supp. 2d 55, 70 (D.D.C.2004); State of Ariz. v. Nucor Corp., 825 F. Supp. 1452, 1456 (D.Ariz.1992). Courts typically first examine procedural fairness, and make a determination as to whether the negotiation process was "fair and full of adversarial vigor." United States v. Telluride Co., 849 F. Supp. 1400, 1402 (D.Colo.1994) (citations and internal quotations omitted). If the decree was the product of "good faith, arms-length negotiations," it is "presumptively valid and the objecting party has a heavy burden of demonstrating the decree is unreasonable." United States v. Oregon, 913 F.2d at 581; see also Cannons, 899 F.2d at 87 n. 4. However, where the government has not "pulled the laboring oar" and instead relied heavily on the accused polluter in crafting the settlement, the district court's inquiry is more searching. See Telluride, 849 F.Supp. at 1404; see also United States v. Colorado, 937 F.2d 505, 509 (10th Cir.1991) ("the district court must ensure that the agreement is not ... a product of collusion....").

With respect to substantive fairness, it is not the duty of the court to determine whether "the settlement is one which the court itself might have fashioned, or considers ideal." Cannons, 899 F.2d at 84; see also BP Exploration, 167 F.Supp.2d at 1050 (stating that court should refrain from "substitut[ing] its judgment for that of the parties"). Instead, the "court's approval is nothing more than an amalgam of delicate balancing, gross approximations and rough justice." Oregon, 913 F.2d at 581 (internal quotations omitted). "The court need only be satisfied that the decree represents a reasonable factual and legal determination." Id. (internal quotation omitted).

B. Procedural Fairness

Amici raise several concerns with the fairness of the procedures that led to the final settlement.

1. EPA did not conduct any investigation of Chevron.

Amici's main objection is that the government did not conduct any investigation of Chevron's refineries, and therefore placed themselves in a weakened bargaining position. Amici are correct that the government did essentially no Chevron-specific investigation to make a determination regarding the scope of defendant's non-compliance. Instead, the government extrapolated from what it learned in other PRI settlements. As amici point out, this presents a weak basis for determining Chevron's compliance level, since different refiners almost certainly took different approaches towards the Clean Air Act.^[3] In addition, the government demanded less from Chevron than it did from other refiners based on the assumption that Chevron's violations were not as serious or widespread as those that had been identified as the most serious violators in EPA's initial analysis.

Amici are correct that EPA's failure to conduct an investigation of Chevron's refineries severely limited its ability to engage in informed negotiation. A thorough investigation of Chevron's refineries by EPA was critical to determining Chevron's liability because of the structure of the Clean Air Act provisions involved. PSD, NSR and NSPS emissions standards are only triggered after a refinery builds a new emission source or modifies an existing one. See 42 U.S.C. § 7475(a); 42 U.S.C. § 7502(c)(5); 42 U.S.C. § 7411(b)(1)(B). Chevron's liability under those provisions can therefore only be determined after a complete investigation into all plant activities that may qualify as modifications under the relevant statutes. Because EPA did not conduct any investigation, its understanding of Chevron's liability was based on weakly-supported assumptions.

However, even though it appears that EPA could have been more aggressive in its settlement strategy, the declarations submitted by the government demonstrate that the process of negotiations was non-collusive and therefore procedurally fair. See United States v. Colorado, 937 F.2d 505, 509 (10th Cir.1991) (inquiry into procedural fairness focuses on whether negotiations were non-collusive). Through the declarations of the negotiators involved in the settlement, EPA has demonstrated that there was substantial give-and-take during the months of negotiations leading up to the lodging of the Consent Decree with the Court. It is apparent that a considerable amount of time and resources was invested by both sides in order to come to a settlement that satisfactorily met each party's objectives. These efforts are consistent with an adversarial and non-collusive process.

The Court notes that, were it in a position to apply closer scrutiny to this process, then more questions would be in order. For example, amici point out that EPA could have investigated a small number of Chevron's refinery structures to come to a more accurate conclusion regarding the real level of compliance. If it had done so, EPA would have been in a much stronger position to demand concessions from Chevron in those areas in which Chevron had the greatest potential liability. However, EPA did the opposite by stating at the beginning of negotiations that no investigation whatsoever would be taken. Of course, the only way that this could work as an incentive for Chevron is if Chevron believed that investigation would reveal that its refineries' level of violation was more serious than EPA estimated. It is also apparent from the declarations submitted that EPA entered the negotiations never intending to pursue litigation. As such, EPA was instantly in a weakened bargaining position because it could not seriously use the threat of litigation as an incentive for Chevron to sacrifice more in its positions.

Nonetheless, given the narrow scope of review applicable here, the Court is constrained to show deference to EPA's judgments regarding how to conduct its negotiations, and inquire only into the reasonableness of those choices. See Akzo Coatings, 949 F.2d at 1436. EPA has submitted that through the PRI it seeks to resolve widespread non-compliance in the petroleum refinery industry by entering into global settlements without spending large amounts of agency resources on investigation or litigation. According to EPA this resolves the problems earlier and at lesser expense for both the government and the refiners. See Officers for Justice, 688 F.2d at 625 ("[I]t is the very uncertainty of outcome in litigation and avoidance of wasteful and expensive litigation that induce consensual settlements."). While it is almost certainly true that EPA could have adopted a more aggressive strategy and thereby secured even greater benefits in this settlement, EPA is reasonable in believing that by doing so it may have been deprived of the resources to enter into other settlements against other refiners. Because EPA's justification is reasonable, the negotiations leading up to entry of the Consent Decree were not procedurally unfair.

2. EPA failed to make its own emissions reductions estimates.

Amici also criticize the negotiations as procedurally unfair because EPA did not use its own emissions reductions estimates during the negotiations. EPA does not dispute that it did so. However, this admission is not as problematic as amici claim given that EPA did not rely on these kinds of estimates to gauge their success in the negotiations. Instead, EPA states that the negotiation process focused on attempting to make Chevron agree to technologies, processes and emissions levels outlined in the model template, or something close to those benchmarks. The emissions reductions estimates apparently were primarily used as simplified measurements that could be used for public relations purposes, such as press releases.

EPA's reliance on Chevron's self-measurements is not unreasonable. It is commonplace for environmental regulators to rely on self-reporting and self-monitoring by regulated entities to enforce environmental laws. See, e.g., 42 U.S.C. § 7414 (imposing monitoring and record keeping requirements on sources of air pollution); 33 U.S.C. § 1318 (imposing similar requirements on sources of water pollution). A regulatory target that falsifies such a report is generally subject to civil and criminal penalties. See 42 U.S.C. § 7413. Accordingly, it was not unreasonable for EPA to rely upon Chevron's self-reported emissions measurements to negotiate the Consent Decree. Indeed, Chevron's failure to turn over complete emissions reports was an initial point of contention between the parties. See Second Foley Decl. ¶ 71. It therefore appears that the reports were subject to adversarial scrutiny.

Accordingly, EPA's conduct in this regard was not unreasonable.

C. Substantive Fairness

In evaluating the substantive fairness of a consent decree in an environmental case, it is important for the district court to be fully informed regarding the costs and benefits of the decree. See United States v. Montrose Chemical Corp. of Cal., 50 F.3d 741 (9th Cir.1995) (finding, in a CERCLA action, that the district court abused its discretion by entering the consent decree without having any estimate of the damage done to the environment). One method of accomplishing this end is by determining a benchmark against which the fairness of the Consent Decree may be judged. See id. at 746. This Court has set as an appropriate benchmark the maximum amount of pollution reduction that would have occurred if the government had progressed to trial and prevailed on every claim in the complaint. A reasoned process of evaluation involves comparing such a best-case scenario figure to the actual reductions achieved by the decree, and then "factor[ing] into the equation any reasonable discounts for litigation risks, time savings, and the like that may be justified." Id. at 747 (citation omitted).

EPA has objected to this strategy of evaluation. They argue that Chevron's actual level of non-compliance is likely much lower than what was alleged in the complaint. EPA further claims that reliance on a best-case scenario calculation also obscures the significant litigation risk and resource costs entailed by a failure to settle. They further caution that reliance on gross pollution reduction estimates undervalues refineries which already have good pollution control practices because a dirty plant can reduce gross emissions more cheaply than can a cleaner one.

While these are legitimate concerns, it remains the case that the best-case scenario calculations offer a benchmark that allows a reasoned analysis of the substantive fairness of the Consent Decree. The Consent Decree in this case is complex enough that the Court cannot review its adequacy without translating the relief it secures into some convenient shorthand that is representative — albeit not perfectly — of what it will accomplish. Moreover, by proposing such a best-case scenario benchmark, the Court has not registered any expectation that EPA would have secured anything similar to, or even nearing, that level of emissions reductions. Instead, it is to be expected that the actual relief secured under the Consent Decree will fall short of the best-case scenario. Such a result may be reasonable result of the compromise inherent in any settlement.

1. Pollution Reduction

The government's estimates are summarized below (all numbers are in tons-per-year):

----------------------------------------------------------------------------------------
Estimated        Percent
Estimated        "Best-Case     Accomplished
Reduction Under     Scenario"       by Consent
Unit              Location             Decree         Reduction[4]        Decree
----------------------------------------------------------------------------------------
FCCU — NOx        El Segundo            197             282               70%
-----------------------------------------------------------------------
Hawaii                179             244               73%
-----------------------------------------------------------------------
Pascagoula            108             180               60%
-----------------------------------------------------------------------
Richmond               0               69                0%
-----------------------------------------------------------------------
Salt Lake              68              85               80%
----------------------------------------------------------------------------------------
FCCU — SO2        El Segundo             85              85              100%
-----------------------------------------------------------------------
Hawaii                317             350               90%
-----------------------------------------------------------------------
Pascagoula           1878            1878              100%
-----------------------------------------------------------------------
Richmond             1726            1911               90%
-----------------------------------------------------------------------

Salt Lake             471             482               98%
----------------------------------------------------------------------------------------
Heaters &         El Segundo            224             481               46.6%
Boilers — NOx
-----------------------------------------------------------------------
Hawaii                 0              277                0%
-----------------------------------------------------------------------
Pascagoula           1069            3236               33%
-----------------------------------------------------------------------
Richmond               0              741                0%
-----------------------------------------------------------------------
Salt Lake              0              464                0%
----------------------------------------------------------------------------------------
Heaters &         All                    0                7                0%
Boilers — SO2
----------------------------------------------------------------------------------------
Sulfur Recovery   Consent Decree mandates 100% of what is required under NSPS
Plants            Subpart J.
----------------------------------------------------------------------------------------
Flares            Consent Decree mandates 100% of what is required under NSPS
Subpart J and provides for additional controls not mandated by law.
----------------------------------------------------------------------------------------
Leak Detection    For both LDAR and Benzene NESHAP, EPA emphasizes that the
and Repair        Consent Decree addresses a wide array of non-compliance by requiring
procedures that prevent the escape of harmful emissions. EPA is
Benzene           unable to quantify the amount of emissions reductions that will occur
NESHAP            under the required procedures, but relies on them as a significant
portion of the Consent Decree's accomplishments.
----------------------------------------------------------------------------------------

This summary makes clear that the Consent Decree achieves substantial benefit to the environment when compared to what could have been secured from litigation. The only area in which it appears that EPA could have done substantially better is with respect to emissions from heaters and boilers. However, EPA determined from prior experience that violations in NSR and NSPS regulations in heaters and boilers were not as common as with FCCUs (less than 27% of heaters & boilers according to EPA's expert), and that Chevron was expected to have even fewer incidences of non-compliance. The provisions regarding heaters and boilers for SO2, for which inadequate evidence has been provided, is discussed in more detail below.

It is also notable that most of the benefits secured under the Consent Decree will occur at Chevron's Pascagoula refinery in Mississippi. EPA explains this asymmetry as attributable to the fact that Pascagoula is Chevron's biggest refinery, and was also the refinery that had the least advanced pollution control technology and processes in place. In contrast, the refineries in California, which are estimated to experience the smallest reductions, were already utilizing modern pollution control techniques pursuant to more stringent state and local regulation.

Amici lodge several critiques of the Consent Decree's substantive fairness, which are discussed below.

a. EPA's submissions are incomplete and/or provide an incorrect BACT/LAER estimate.

Amici attack EPA's submissions as incomplete. They first argue that the government incorrectly calculates the benchmark NOx reductions for Chevron's FCCUs because the appropriate benchmark is LAER, regardless of whether that standard would apply in an enforcement action. This is clearly incorrect because LAER only applies in non-attainment areas: those regions of the country that have not met National Ambient Air Quality Standards ("NAAQS"). 42 U.S.C. 7503(a)(2). In areas where NAAQS have been met, a source need only meet BACT in order to obtain a permit following a modification. 42 U.S.C. 7475(a)(4). Therefore, EPA correctly set LAER benchmarks of 5 ppm for Chevron's California refineries, which are both located in nonattainment areas for NOx. Other areas need only have met the BACT standard, which EPA states would be 20 ppm. Therefore, EPA appropriately complied with the Court's request for a benchmark estimate for FCCU NOx.

Amici also claim that EPA produced "no benchmark estimates for PM/PM10 and CO for any emission source, no benchmark estimates for flares and sulfur recovery plants (``SRPs'), and no benchmark estimates for SO2 reductions from heaters and boilers." Amici's Memorandum and Submissions in Response p. 20 (filed March 11, 2004). All but one of these critiques are based on a misreading of the Consent Decree. First, with respect to PM/PM10 and CO the Consent Decree does not resolve Chevron's NSR liability unless Chevron meets a particular standard that EPA claims is similar to BACT.^[5]See Consent Decree ¶¶ 202-03. Because Chevron's compliance with the standard is voluntary,^[6] EPA cannot claim reductions, and therefore comparison to a best-case scenario benchmark is unhelpful. However, if Chevron does choose to implement the restrictions, then it will be released from liability consistent with what is implemented. The Consent Decree does release Chevron from NSPS liability for PM and CO at some refinery units, but this release was granted in exchange for requirements that Chevron adhere to NSPS. The same holds true for flares and sulfur recovery plants, which contrary to amici's contention, are not released from NSR liability. See Consent Decree ¶ 210.

The only legitimate complaint of this type is amici's challenge to EPA's estimates regarding heater and boiler SO2. EPA admits that the Consent Decree will produce no pollution reductions in this category. See Second Foley Decl. ¶ 116. Although certain controls are mandated, those controls are already in place. The required controls are NSPS standards (a standard that results in about 160 ppm SO2 in the treated fuel gas), see id. ¶ 55, even though the Consent Decree releases Chevron from liability from the stricter NSR regulations. See Consent Decree ¶ 201. EPA argues that NSPS is equal to BACT for heater and boiler SO2, and therefore the Consent Decree accomplishes 100% of what could have been achieved through litigation. Amici, however, submit evidence in response that SCAQMD has mandated a stricter standard, which decreases pollution by a factor of four in comparison to the standard under the Consent Decree (40 ppm vs. 160 ppm). In response, the government concedes that the stricter standard may constitute LAER, but would not constitute BACT.

The government is likely incorrect that NSPS is the same as BACT for these units because, as amici point out, the NSPS standards were written in 1974 and technological developments since then have made more restrictive controls more feasible. SCAQMD found that 40 ppm was "technologically and economically feasible" indicating that a level lower than 160 ppm would likely be an appropriate BACT. This component of the Consent Decree, therefore, has not been adequately justified.^[7] However, when viewed in the context of the overall relief of the Consent Decree, particularly the significant reductions in SO2 emissions from FCCUs, this one weakness is not enough to merit rejection of the Consent Decree as a whole.

b. The Consent Decree violates the Clean Air Act because it does not require BACT/LAER at every source.

Amici have throughout these proceedings strenuously argued that the Consent Decree must require BACT or LAER at every refinery unit in order to comply with the Clean Air Act. Although they are correct that EPA's policy guide states that BACT or LAER should be the goal in any settlement of NSR liability, they are wrong in asserting that diversion from the policy constitutes a violation of the Clean Air Act. Because of the structure of the NSR/PSD program, BACT or LAER would only be required under law where there has been a modification, as that term is defined by the Act. In order to prove that Chevron made such modifications without obtaining the proper permit, the government would have to conduct an investigation of all of Chevron's refineries and potentially enter into protracted litigation in order to require that BACT or LAER be installed. Even assuming a successful enforcement action, BACT would have to be determined through a cost/benefit analysis, involving further negotiation. As discussed above, the government has opted to forgo this lengthy process on the belief that it will obtain greater and faster environmental benefits at less cost by settling. Whatever is thought of that strategy, it cannot be said that BACT or LAER is currently required for any of Chevron's refinery units because no modifications have been specifically identified. Therefore, amici are incorrect in asserting that the Consent Decree violates the law.

c. The Decree's pollution control strategies result in negative environmental side-effects.

Amici next claim that some of the pollution control mechanisms required by the Consent Decree produce harmful byproducts that offset the claimed environmental benefits. To the extent it is true that some harmful chemicals are produced by the pollution control technology required, those chemicals are subject to other environmental restrictions and would be regulated accordingly. In any event, amici have not quantified the magnitude of the harmful side-effects, and have not claimed that they are so large as to outweigh the benefits of the pollution control strategies that create them. The Court therefore defers to EPA's expertise in choosing pollution reduction strategies that result in a net gain to the environment.

d. The Pascagoula FCCU SO2 reductions are illusory.

EPA claims that there will be 1,878 tons per year of SO2 reductions at the Pascagoula FCCU. Amici argue that, because Chevron has applied with permitting authorities to modify the Pascagoula FCCU, it must comply with NSPS Subpart J independently of the Consent Decree, and therefore most of the claimed reductions are illusory.

Chevron argues that the alterations to the FCCU will not constitute "modifications" under the Clean Air Act and therefore will not trigger NSPS restrictions. Further, Chevron argues that even if it is required to comply with Subpart J because of the applied-for modifications, the Consent Decree still requires pollution controls that are more strict. The Consent Decree requires Chevron to comply with a FCCU SO2 emission limit of 50 ppmvd @ 0% O2 on a 7-day rolling average basis. See Consent Decree ¶ 16. This is identical to the restriction NSPS Subpart J restriction contained in 40 C.F.R. § 60.104(b)(1). However, if Subpart J applied it would provide Chevron the choice of either complying with this standard or adhering to one of the two other pollution control options contained in the regulation. See 40 C.F.R. § 60.104(b)(2)-(3). According to EPA, the three options are not the same with respect to the emissions produced in practice. See Second Foley Decl. ¶ 29. The first option results in 50 ppm, see 40 C.F.R. § 60.104(b)(1), the second option results in a range of emissions from 200-400 ppmvd @ O% O2, see Second Foley Decl. ¶ 29; 40 C.F.R. § 60.104(b)(2), and the third option allows for a range of 50-500 ppmvd @ 0% O2. See Second Foley Decl. ¶ 29; 40 C.F.R. § 60.104(b)(3). By requiring Chevron to implement the first option, the Consent Decree results in lower emissions than would have occurred if Subpart J applied. Therefore, the claimed SO2 reductions are not illusory.

e. The Richmond FCCU SO2 reductions are illusory.

Amici argue that the Richmond refinery's FCCU has been subject to NSPS Subpart J regulations since 1993, when it got a permit to make a modification to that unit. However, according to amici, the emissions numbers released through this litigation demonstrate that Richmond was not complying with Subpart J.

This critique fails for the same reason as did amici's objections regarding the Pascagoula FCCU. As with Pascagoula, the Richmond refinery's FCCU must comply with the strictest of the three options available to a source under Subpart J. Indeed, Chevron admits that it has been subject to Subpart J and has complied with it since the early 1990s. According to Chevron, it was previously adhering to 40 C.F.R. § 60.104(b)(2). Under the Consent Decree, it will be required to comply with subsection (b)(1), and therefore its emissions will be lowered. Therefore, these claimed reductions are not illusory.

f. The Consent Decree takes too long to achieve its environmental benefits.

The Consent Decree requires Chevron to meet all of its provisions by 2011, although some of the requirements are phased in between lodging and that date so that some benefits will be realized earlier. Amici claims this time period essentially allows Chevron to continue illegally polluting while being shielded from liability. It is, of course, correct that the timing of the relief in the Consent Decree is an important factor of its substantive reasonableness. However, because of the complexity of Clean Air Act litigation, it was reasonable for EPA to conclude that even a due date eight years after the signing of the Consent Decree may create environmental benefits earlier than litigating.

g. Inflated baselines.

The government based its negotiations and its ultimate representations regarding the benefits of the Consent Decree on Chevron's recorded emissions during 2000-2001. All of the Decree's claimed benefits are based on subtracting the estimated future emissions after implementation of the Decree's provisions from that "baseline" time period. Amici argue that the emissions during the 2000-2001 period "appear to be inflated," therefore reducing the claimed benefits of the Consent Decree. The government replies that it used those years as a baseline because they were the years immediately preceding the start of negotiations and newer data was unavailable. Amici have neither provided support nor quantified their contention that the baselines were inflated. As such, this objection is rejected.

h. Chevron is allowed to bank illusory reductions and pollute more in the future.

Relevant air pollution regulations allow a polluter to be credited for pollution reductions that are "voluntary," i.e. over and above what is required by law. See 40 C.F.R. § 52.21(b)(3)(iii). A source can use these credits in subsequent years to comply with regulations where it would otherwise have been emitting pollution in excess of them. The Consent Decree allows Chevron to bank a small portion of the reductions under the Decree in order to comply with Clean Air Act restrictions in the future. See Consent Decree ¶ 105.

Amici argue that, by giving Chevron credit for illusory reductions, the Consent Decree allows Chevron to increase its overall level of pollution by banking illusory credits and using them later. EPA responds that when negotiations were taking place, there was uncertainty as to which reductions would have been required if litigation had taken place. Because Chevron felt that the Decree forces it to comply with standards that are stricter than what is required by law, the parties agreed to give some credits would be awarded in recognition that some of the reductions could be classified as "voluntary."

This is a reasonable resolution of the dispute that complies with applicable law. The Clean Air Act forbids the award of emission reduction credits for reductions otherwise required under the Act. See 42 U.S.C. § 7053(c)(2). Since there has been no adjudication of whether Chevron's reductions are required by law, there is no strict legal prohibition on awarding Chevron credits for some of the reductions under the Consent Decree. Notably, amici only claim that 80 tons per year of NOx reductions are improperly credited. This amount makes up a relatively small amount of the reductions mandated by the Consent Decree and therefore is not enough to scuttle the whole settlement.

i. Trading pollution benefits in Mississippi for reduced air quality in California violates the Clean Air Act.

The government has throughout these proceedings maintained that the settlement should be viewed globally. Amici respond that the structure of the Clean Air Act mandates EPA to regulate air quality on a regional basis, and prohibits transferring benefits from one region to another. Amici are correct. However, while it is true that a disproportionately large portion of the Consent Decree's benefits will occur in Mississippi, this is a product of the fact that the refinery there is the biggest and dirtiest. Further, according to EPA's estimates, the Consent Decree provides substantial environmental benefit in all affected regions. Therefore, the distribution of environmental benefits created by the Consent Decree is not unreasonable or illegal.

2. Fines and penalties

In addition to imposing restrictions on covered emissions, the Consent Decree also releases Chevron of all liability for civil penalties under the relevant environmental laws. EPA did not provide a best-case scenario estimate for penalties because calculation of monetary sanctions involves a complex multi-factor analysis that does not easily reduce to hard number computation. The best estimate EPA could produce is that "the potential maximum penalty could extend into the tens or hundreds of millions of dollars." Jackson Decl., ¶ 26. Amici estimate that Chevron's potential penalty liability is as much as $2 billion. Amici's Opposition to the Motion to Enter the Proposed Consent Decree, p. 27 (filed September 30, 2004). While large, this number may be accurate, since the applicable law calculates penalties on a per-day basis. By contrast, the Consent Decree imposes only $3.5 million in penalties and $4.5 million in Supplemental Environmental Projects ("SEPs").^[8]

EPA states that this level of penalty is consistent with other PRI settlements. They further argue that litigation risk is a major factor in crafting a penalty settlement. For example, courts have split on the statute of limitations for such civil penalties, and application of a limitations period could preclude application of a substantial portion of the penalties potentially owed in this case. Moreover, in practice courts award far less than the statutory maximum.

EPA's failure to investigate Chevron's violations makes evaluation of the reasonableness of the Decree's penalty provisions difficult. However, considering the substantial risk of going to trial, coupled with the important environmental benefits secured by the Consent Decree, this Court does not find that the penalty imposed by the decree is unreasonable.

3. Costs

Chevron estimates that it will spend $235.5 million complying with the Consent Decree. Of course, some or all of these are costs that Chevron would have incurred if it had complied with the law in the first place. But given the uncertainty of the scope of Chevron's violations the substantial costs imposed weigh in favor of approval.

4. Modification provisions

Amici previously voiced concern regarding a provision of the Consent Decree that allowed the parties to modify the Consent Decree without prior approval of the Court. Amici cautioned that such unlimited discretion could render the Court's approval of the Consent Decree meaningless. The parties responded by changing the amendment provisions so as to require the Court's permission for all material modifications. See Consent Decree ¶¶ 107-08, 234.

Nonetheless, amici continue to object that even the new version of the modification provisions allows the parties too much discretion to determine which modifications are material and which are non-material. Under the relevant section of the Decree, material modifications require prior approval from the Court, whereas non-material ones are effective without the Court's assent and need only be reported to the Court on a periodic basis.

The Court is satisfied that the parties have struck an appropriate balance between allowing for judicial review of important modifications and avoiding needless delay of minor changes that do not seriously effect the substance of the settlement approved by the Court. The modification provision states that one example of a non-material modification is an alteration to the Decree's implementation schedule that does not change the ultimate compliance date. Consent Decree ¶ 234. By implication, those modifications that do alter ultimate compliance dates will almost always — if not always — constitute material alterations requiring the Court's approval.

It is also important that all non-material modifications will be filed with the Court. Since the Court retains jurisdiction over the Consent Decree, the Court will have the authority to ensure that the modification provision is not being applied in a manner that contravenes the text of the Decree or undermines the settlement approved by this Court. See Vanguards of Cleveland v. City of Cleveland, 23 F.3d 1013, 1018 (6th Cir.1994) (stating that approving courts "retain jurisdiction over the decree during the term of its existence," can "protect the integrity of the decree with its contempt powers," and has "the inherent power to enforce agreements entered into in settlement of litigation pending before them.").

Amici are also incorrect in asserting that the modification allows the parties to hide any modifications from the view of the public. To the contrary, by requiring that all modifications are filed with the Court, they become a part of the public record and readily accessible on the internet. Further, the amici to this case will be notified of any document filed herein. Amici therefore will have notice and an opportunity to comment on any modification to the Consent Decree.

The modification provisions of the Consent Decree are therefore fair, just and reasonable.

CONCLUSION

While amici have raised several valid critiques of the Consent Decree, when evaluated as a whole the Decree appears to be a reasonable settlement of uncertain liability that saves litigation expenses for the parties. Although disagreeing with the amici's urging that the Consent Decree should be rejected, this Court notes the important role they have played in these proceedings. Because of their comments, the Court has applied much closer scrutiny than would have otherwise been possible. They have drawn the Court's attention to those issues that are the most problematic, while also assisting in the Court in navigating the complicated regulatory landscape on which the Consent Decree sits. Indeed, had the Consent Decree been presented to the Court on a different posture allowing for more searching review, then amici's comments may have resulted in a different outcome.

However, as has been repeated several times above, the standard of review that is applicable here affords the Court very little discretion in evaluating the terms of the Decree. The Court is bound to show substantial deference to EPA and the Department of Justice where, as here, the government has presented evidence that the negotiations leading up to entry of the Decree were adversarial and non-collusive. It also appears that EPA's choice to avoid litigation in favor of a broad-based settlement strategy may have the potential to win greater environmental benefits in the long run. It is not for this Court to say whether that choice is wise, but rather only to review it for its reasonableness. As the Court cannot deem it unreasonable, the motion to enter the proposed Consent Decree is hereby GRANTED.

IT IS SO ORDERED.

NOTES

[1] Amici also cite Garcia v. Cecos Int'l, Inc., 761 F.2d 76 (1st Cir.1985), but that case is clearly off point because it did not involve notice of EPA enforcement under the CAA; instead it involved private plaintiffs suing under a private right of action under a distinct notice provision in RCRA. See id. at 78-80. Moreover, in Garcia, all the parties conceded that there was no actual notice, id. at 78, and therefore the case cannot be read as mandating a specific procedure when there is actual notice.

[2] Amici's claim that section 7413 is also intended to provide public notice, since a well-timed FOIA request would reveal a written NOV, is also spurious. If Congress wanted the NOV requirement to give the public an opportunity for input, it would have required public notice.

[3] Amici further claim that their own investigation revealed that Chevron had substantially more violations than the government estimated. EPA and Chevron vigorously dispute the accuracy of amici's investigation, arguing that it is incomplete and based an incorrect understanding of the facts. Moreover, the parties argue that amici's objections were raised earlier when state environmental agencies were issuing permits to Chevron. Their contentions were rejected in those proceedings.

[4] For the FCCUs and Heaters and Boilers, the complaint alleged violations of New Source Review; therefore the best-case scenario would be BACT or LAER, depending on the location of the refinery. With respect to Sulfur Recovery Plants and Flares, the complaint alleges only violations of New Source Performance Standards. Under those standards, the best-case scenario is determined by NSPS Subpart J. Contrary to amici's protestations, the Consent Decree only releases Chevron from liability for NSPS Subpart J for SRPs; it does not release them from NSR/PSD liability. Consent Decree ¶¶ 201, 210.

[5] The Consent Decree requires Chevron to comply with NSPS Subpart J standards for CO and PM emissions from its FCCUs. See Consent Decree ¶ 23. However, the Consent Decree also provides for releases of NSR/PSD liability if Chevron meets voluntary standards. The voluntary standards are: 1) for PM, an emission limit of 0.5 pound PM per 1000 pounds of coke burned on a 3-hour average basis, Consent Decree ¶ 202; and 2) for CO, an emission limit of 100 ppmvd of CO at 0% 02 on a 365-day rolling average basis. Consent Decree ¶ 203.

[6] There are two exceptions. First, the Richmond refinery is released from NSR liability for CO emitted from its FCCU. See Consent Decree¶ 201. Richmond was given this release because it accepted the 100 ppmvd limit as a mandatory (rather than voluntary) limit. See Consent Decree ¶ 26. Notably, amici claim that 100 ppmvd was BACT for FCCU CO in 1993. See Amici's March 11 brief at 22. Amici have provided no support for their assertion that BACT would be something lower than 100 ppmvd today. However, even if that were true, it would still be the case that the standard imposed by the Consent Decree is substantially similar to BACT and therefore comes close to the best-case scenario.

The second exception is the El Segundo refinery. The Consent Decree provides for a mandatory 0.5 pound PM limit for that refinery's FCCU. See Consent Decree ¶ 22. In exchange, El Segundo receives a release from NSR/PSD liability for that unit's PM emissions. See Consent Decree ¶ 201.

[7] Although EPA attempts to minimize this difference by pointing out that the Salt Lake Refinery's heaters and boilers produce only three tons per year ("tpy") of SO2, EPA fails to mention that other refineries produce much more. For example, heaters and boilers at El Segundo produce 180 tpy SO2, Pascagoula produces 581 tpy and Richmond produces 36 tpy. These quantities are not insignificant.

[8] Supplemental Environmental Projects are a typical component of environmental law penalty assessments. Rather than ordering a violator to pay cash for its violations, EPA will allow the violator to implement an project that is beneficial to the environment.