PUBLISHED

                      UNITED STATES COURT OF APPEALS
                          FOR THE FOURTH CIRCUIT


                                   No. 17-1839


ERGON-WEST VIRGINIA, INCORPORATED,

            Petitioner,

      v.

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY,

            Respondent.


On Petition for Review of Final Agency Action of the United States Environmental
Protection Agency.


Argued: December 7, 2017                                     Decided: July 20, 2018


Before NIEMEYER and AGEE, Circuit Judges, and Paula XINIS, United States District
Judge for the District of Maryland, sitting by designation.


Petition for review granted, final agency action vacated, and remanded for further
proceedings by published opinion. Judge Agee wrote the opinion, in which Judge
Niemeyer and Judge Xinis joined.


ARGUED: Jonathan Grant Hardin, PERKINS COIE LLP, Washington, D.C., for
Petitioner. Patrick Reinhold Jacobi, UNITED STATES DEPARTMENT OF JUSTICE,
Washington, D.C., for Respondent. ON BRIEF: LeAnn M. Johnson, PERKINS COIE
LLP, Washington, D.C., for Petitioner. Jeffrey H. Wood, Acting Assistant Attorney
General, Environmental and Natural Resources Division, UNITED STATES
DEPARTMENT OF JUSTICE, Washington, D.C.; Susan Stahle, UNITED STATES
ENVIRONMENTAL PROTECTION AGENCY, Washington, D.C., for Respondent.
2
AGEE, Circuit Judge:

       Until 2011, Ergon-West Virginia, Inc. enjoyed an exemption as a small refinery

from the Environmental Protection Agency’s renewable fuel standard program, which

requires refineries and other facilities to allocate a certain percentage of their fuel

production to renewable fuels. When Ergon filed for an extension of the small refinery

exemption, the EPA denied its petition on the basis that Ergon’s participation in the

program would not constitute a disproportionate economic hardship. Ergon petitions the

Court for review of the EPA’s denial. Because we conclude that the EPA’s decision was

arbitrary and capricious, we grant Ergon’s petition for review, vacate the EPA’s denial,

and remand for further proceedings.



                                           I.

       We begin with the renewable fuels statute and its history and then turn to the

proceedings in this case.

                                           A.

       With the Energy Policy Act of 2005, Congress added the renewable fuel standard

program (the “RFS Program” or “Program”) as Section 211(o) of the Clean Air Act. See

42 U.S.C. § 7545(o). The statute directs the EPA Administrator to promulgate regulations

“to ensure that transportation fuel sold or introduced into commerce in the United States

(except in noncontiguous States or territories), on an annual average basis, contains at

least the applicable volume of renewable fuel, advanced biofuel, cellulosic biofuel, and



                                           3
biomass-based diesel” 1 required by the Program. 

Renewable fuels,

such as ethanol, are those that are “produced from renewable biomass and that [are] used

to replace or reduce the quantity of fossil fuel present in a transportation fuel.” 

Renewable biomass includes natural materials such as crops, trees, and

animal byproducts. 

The regulations apply “to refineries, blenders,

distributors, and importers.” 

       The applicable volumes of renewable fuel, advanced biofuel, cellulosic biofuel,

and biomass-based diesel that transportation fuels must contain on an industry-wide basis

are found in § 7545(o)(2)(B). For instance, the statute lists the applicable volume of

renewable fuel for 2016 as 22.25 billion gallons. 

To determine

the “applicable percentages” of renewable fuel, advanced biofuel, cellulosic biofuel, and

biomass-based diesel that a facility must use, the EPA first estimates “the volumes of

transportation fuel, biomass-based diesel, and cellulosic biofuel projected to be sold or

introduced into commerce in the United States” the following year. 

The EPA then divides the applicable volume of the particular renewable fuel by the fuel

estimate to arrive at the percentage every refinery must meet and publishes it in the

Federal Register. 

40 C.F.R. § 80.1405. For example, the percentage

of renewable fuel for 2016 was 10.10%. 40 C.F.R. § 80.1405(a)(7)(iv). This


       1
          Although the RFS Program sometimes lists these four fuels separately, it also often
terms them all as “renewable fuel,” perhaps because the definitions of advanced biofuel,
cellulosic biofuel, and biomass-based diesel all refer to those fuels as renewable fuel. See 42
U.S.C. § 7545(o)(1)(B), (D), (E); see also 40 C.F.R. § 80.1401. The U.S. Department of Energy
has called the four fuel categories “‘nested’ standards.” J.A. 19.


                                              4
percentage—or “renewable fuel obligation”—is “applicable to refineries, blenders, and

importers, as appropriate,” 2 and is “expressed in terms of a volume percentage of

transportation fuel sold or introduced into commerce in the United States.” 42 U.S.C.

§ 7545(o)(3)(B)(ii). A refinery will multiply the percentage by the volume of

nonrenewable fuel that it produces or imports to determine its “renewable volume

obligation.” 40 C.F.R. § 80.1407.

       All renewable fuels are identified by a renewable identification number (“RIN”),

which “is a unique number generated to represent a volume of renewable fuel.” 40 C.F.R.

§ 80.1401. An obligated party must “separate” a sufficient number of RINs (i.e., blend

the renewable fuel with nonrenewable fuel) to demonstrate compliance with the Program.

See 

see also 42 U.S.C. § 7545(o)(5) (establishing a credit

program for blending renewable fuels with transportation fuels). If the obligated party

fails to separate the required number of RINs, it can purchase separated RINs from a

party who has separated more RINs than it needs and thereby avoid violating the

Program’s    requirements and        incurring       penalties.   See 40   C.F.R. §§ 80.1428,

80.1460(c)(1); see also 42 U.S.C. § 7545(o)(5)(B) (stating that “[a] person that generates

credits . . . may use the credits, or transfer all or a portion of the credits to another person,

for the purpose of complying with [the Program]”).

       From 2005 until 2011, small refineries—those “for which the average aggregate

daily crude oil throughput for a calendar year . . . does not exceed 75,000 barrels,” 42

       2
        Despite the statute’s instruction, the EPA defines an “obligated party” under the RFS
Program as a refiner or importer and specifically exempts blenders. 40 C.F.R. § 80.1406(a)(1).


                                                 5
U.S.C. § 7545(o)(1)(K); see also 40 C.F.R. § 80.1442—were exempt from the Program,

42 U.S.C. § 7545(o)(9)(A)(i); see also 40 C.F.R. § 80.1441. The statute directed the

Secretary of the Department of Energy to “conduct for the [EPA] Administrator a study

to determine whether compliance with the [Program’s] requirements . . . would impose a

disproportionate       economic      hardship        on    small    refineries.”      42   U.S.C.

§ 7545(o)(9)(A)(ii)(I) (emphasis added). If the DOE determined that a given refinery

would experience disproportionate economic hardship, then the EPA Administrator was

required   to   extend    the     facility’s   exemption     for   at   least   two    years.   

After this first mandatory extension period, the statute provides

that a facility may petition the EPA for extension of the exemption “at any time” due to

disproportionate economic hardship. 

This petition “must specify

the factors that demonstrate a disproportionate economic hardship and must provide a

detailed discussion regarding the hardship the refinery would face in producing

transportation fuel meeting the [Program’s] requirements.” 40 C.F.R. § 80.1441(e)(2)(i).

In evaluating the petition, the EPA—“in consultation with” the DOE—must “consider the

findings of the [DOE’s] study             . . . and other economic factors.” 42 U.S.C.

§ 7545(o)(9)(B)(ii).

                                                B.

       In 2009, the DOE presented the EPA with the Small Refineries Exemption Study

(the “2009 Study”), as required by 42 U.S.C. § 7545(o)(9)(A)(ii)(I). The 2009 Study

recognized that “[o]bligated parties, such as refineries, may fulfill their renewable fuel

requirements through either blending renewable fuels into their products or purchasing

                                                 6
credits from other parties who have exceeded their allocation of renewable fuel

consumption.” J.A. 15. The DOE determined that, “[a]s long as credits are available for

purchase and the market is competitive, small refineries should not be subject to

disproportionate economic hardship from their choice to purchase credits rather than to

generate them.” J.A. 15. 3 The 2009 Study concluded that the general small refinery

exemption should not be extended beyond 2010, based largely on its determination that

“credits are available at a nominal value and compliance volumes have been in excess of

the RFS requirements.” J.A. 16.

       Unsatisfied with this result, Congress directed the DOE to conduct a new, more in-

depth analysis. 4 The DOE released this new study in 2011 (the “2011 Study”). The 2011


       3
        The 2009 Study did recognize, however, that “[i]t [was] too early to project whether
[RIN] markets will continue to be liquid and competitive.” J.A. 27.
       4
           Congress provided this directive in a Senate report, stating the following:
              The January 2009 Small Refineries Exemption Study issued by the
       Department of Energy was intended to determine whether small refineries faced a
       disproportionate economic hardship in meeting Renewable Fuel Standard [RFS]
       requirements beginning in 2011. The Committee understands the study contained
       inadequate small refinery input, did not assess the economic condition of the
       small refining sector, take into account regional factors or accurately project RFS
       compliance costs. Therefore, the Committee does not believe the study is
       complete, nor is the Department able to make the required determination at this
       time. In view of these deficiencies and the importance of the study, the
       Department is directed to reopen and reassess the Small Refineries Exemption
       Study by June 30, 2010. The Department is specifically directed to seek and invite
       comment from small refineries on the RFS exemption hardship question, assess
       RFS compliance impacts on small refinery utilization rates and profitability,
       evaluate the financial health and ability of small refineries to meet RFS
       requirements, study small refinery impacts and regional dynamics by PADD, and
       reassess the accuracy of small refinery compliance costs through the purchase of
       renewable fuel credits. Finally, the Committee notes that the 2009 study does not
       estimate the price of tradable fuel credits, but the Committee is aware that from
       2008 to 2009, price has increased nearly threefold. The Committee expects the
(Continued)
                                                   7
Study defined disproportionate economic hardship as “increased cost of compliance to

the point that the current or future viability of the refinery is impacted.” J.A. 47. The

DOE recognized that:

       [s]mall refineries can suffer disproportionate economic hardship from
       compliance with the RFS program if blending renewable fuel into their
       transportation fuel or purchasing RINs increases their cost of products
       relative to competitors to the point that they are not viable, either due to
       loss of market share or lack of working capital to cover the costs of
       purchasing RINs.
J.A. 47. 5 After conducting a survey of small refineries, the DOE created a scoring matrix

composed of two indices—the “Disproportionate Impact Index” and the “Viability

Index”—to be used to determine whether a small refinery suffers disproportionate

economic hardship:




       Department to undertake an economic review to estimate the actual economic
       impact of the RFS on small refineries on a regional basis.
S. Rep. No. 111-45, at 109 (2009).
       5
          The EPA later disagreed with the DOE’s statement that purchasing RINs may put a
small refinery at an economic disadvantage, stating the following:
       EPA notes that after further review, contrary to statements in [the 2011 Study], it
       has been found that a refinery does not experience disproportionate economic
       hardship simply because it may need to purchase a significant percentage of its
       RINs for compliance from other parties, even though RIN prices have increased
       since the DOE study, because the RIN prices lead to higher sales prices obtained
       for the refineries’ blendstock, resulting in no net cost of compliance for the
       refinery.
J.A. 317.


                                               8
9
J.A. 82, 85.

       In Section 1(a) of the Disproportionate Impact Index, the DOE assesses the small

refinery’s access to capital and credit, primarily through its credit rating. Section 1(b)

considers a refinery’s business lines other than refining and marketing—“in particular

upstream operations such as exploration and development that are less correlated with

refining”—which may insulate the refinery from the volatility of refining margins. J.A.

83. Section 1(c) accounts for the refinery’s geographic location by evaluating how likely

the local market will accept transportation fuels blended with renewable fuels. Although

the category lists subcategories for E10 (a fuel mixture of 10% ethanol and 90%

gasoline), E85 (a fuel mixture of 85% ethanol and 15% gasoline), and biodiesel, the latter

two are “[r]eserved for later evaluation.” J.A. 83. Section 1(d) evaluates a refinery’s

percentage of diesel production in recognition of the fact that “refineries that

disproportionately favor diesel production over gasoline inherently have a more difficult

compliance pathway, as the percentage of renewable fuel available to blend into diesel is

much lower than the 10 percent of ethanol that can be blended into gasoline.” J.A. 83.

                                           10
Section 1(e) contemplates abnormally strict state regulations, such as those states that

“require refiners to sell unblended fuel.” J.A. 83.

       In Section 2(a), the DOE scores the refinery for its relative refining margin—

essentially its refining revenue minus its refining costs, or refining profit—compared to

the three-year industry average. Section 2(b) evaluates the capacity of the refinery to

blend its nonrenewable fuels with renewable fuels, with a lower capacity indicating

greater impairment. Although Section 2(b) has subcategories of ethanol, biodiesel, and

advanced biofuels, only the first is scored, with the others “[r]eserved for later

evaluation.” J.A. 84. Section 2(c) considers whether a refinery operates in a niche

market—such as a refinery that is located in a region “with limited alternative finished

product supply or access to distressed crude oil supply” or one that produces “a specialty

slate of products” like “lube oils, greases, asphalt, etc.” in addition to transportation

fuels—as the refinery’s participation in the niche market may “result in higher than

industry refining margins.” J.A. 84. In Section 2(d), the DOE determines whether the

refinery generates revenue by selling RINs or must purchase RINs in the market. The

2011 Study stated that “[t]his criterion was not utilized in the current assessment due to

lack of consistency among the survey participants.” J.A. 84.

       The Viability Index analyzes “the ability of the refiners to remain competitive and

profitable” while complying with the Program. J.A. 85. In Section 3(a), the DOE

determines the degree to which a facility’s cost of compliance impairs its ability to make

efficiency improvements. Section 3(b) accounts for any events such as a temporary

shutdown that may prevent the facility from fully complying with the RFS Program’s

                                             11
requirements. Section 3(c) evaluates the likelihood that compliance costs will cause a

refinery to shut down.

       The DOE averages the scores in a given index and divides the average by 2. A

refinery is entitled to an exemption if it achieves a score greater than 1 in both indices. To

obtain this score, a facility must earn “a score equivalent to at least four of the eight

metrics for disproportionate impact at the moderate level (5)” and “a positive value for at

least one of the three metrics for” the Viability Index. J.A. 86.

       In 2015, Congress directed the DOE “to recommend to the EPA Administrator a

50 percent waiver of RFS requirements” if a refinery reaches the requisite score on only

one of the two indices. 161 Cong. Rec. H10,105 (daily ed. Dec. 17, 2015). Congress also

stated the following:

       The [DOE] Secretary is reminded that the RFS program may impose a
       disproportionate economic hardship on a small refinery even if the refinery
       makes enough profit to cover the cost of complying with the program.
       Small refinery profitability does not justify a disproportionate regulatory
       burden where Congress has explicitly given EPA authority, in consultation
       with the Secretary, to reduce or eliminate this burden.

December 2016, the EPA issued a memorandum that detailed how it

evaluates small-refinery-exemption petitions. The EPA “considers the findings of the

DOE Small Refinery Study and a variety of economic factors.” J.A. 201. Some of those

factors include “profitability, net income, cash flow and cash balances, gross and net

refining margins, ability to pay for small refinery improvement projects, corporate

structure, debt and other financial obligations, RIN prices, and the cost of compliance


                                             12
through RIN purchases.” J.A. 201. Petitioning facilities include financial information

with their petitions to aid the EPA in its analysis.

                                                  C.

       Ergon owns a refinery in Newell, West Virginia, with a maximum crude oil

capacity of 23,000 barrels per day, well under the small refinery threshold. The facility

primarily produces paraffinic lube oils, and the transportation fuels it produces are

byproducts of that lube oil production. 6 Nearly all (99%) of Ergon’s transportation fuels

are sold within a 170-mile radius of the facility.

       In April 2016, Ergon filed a petition with the EPA for a small refinery exemption

for compliance years 2014, 2015, and 2016. In its petition, Ergon claimed that it is at an

economic disadvantage because, while there was a widespread market for blended

gasoline, there was no such market for blended diesel. Ergon also claimed that its ability

to comply with the Program was limited by its geographic location because customers in

Ergon’s market chose its main competitor’s unblended diesel over Ergon’s blended

diesel. See J.A. 216 (“Because [Ergon] produces diesel at nearly twice the industry

average and biodiesel is blended at lower rates than gasoline, [Ergon] generates fewer

RINs for compliance than a large, vertically integrated refiner like [Ergon’s main

competitor].”). 7




       6
          Approximately two-thirds of the transportation fuel Ergon produces is diesel, with the
other third gasoline.
       7
           Ergon does have the infrastructure to blend ethanol into its gasoline.


                                                  13
       In June 2016, the EPA denied Ergon’s petition for years 2014 and 2015. The DOE

applied the scoring matrix for those years and concluded that Ergon did not achieve the

requisite scores on either the Disproportionate Impact Index or the Viability Index. For

both years, the DOE gave Ergon scores of 0 for Sections 1(a), 1(b), 1(c), 1(e), and 2(a)

through 2(c) in the Disproportionate Impact Index. However, the DOE gave Ergon a

score of 10 for Section 1(d) due to its high level of diesel production. The DOE did not

give Ergon a score for Section 2(d)—the “RINs net revenue or cost” factor—because it

“has not scored this category for any hardship petition evaluations.” J.A. 260. With a total

of ten points, the average across the eight scored sections was 1.25. After dividing that

figure by 2, Ergon’s overall score for the Disproportionate Impact Index was 0.6. The

DOE gave Ergon a score of 0 for each of the three sections of the Viability Index,

resulting in an overall score of 0. The EPA reviewed the DOE’s scoring matrix and

“independently determine[d]” that Ergon “w[ould] not experience ‘disproportionate

economic hardship’ from compliance with the RFS program for 2014 and 2015.” J.A.

262. In short, the EPA “agree[d] with DOE’s determination in reviewing [Ergon’s]

petition that [Ergon’s] 2014 and 2015 RFS compliance costs do not threaten [Ergon’s]

viability.” J.A. 264.

       In August 2016, Ergon withdrew its 2016 petition, informed the EPA that it would

file a revised petition for that year, and asked the EPA to reconsider its decision for




                                            14
2015. 8 In December 2016, Ergon submitted the revised petition for the 2016 compliance

year. In May 2017, the EPA denied Ergon’s 2016 petition for a small refinery exemption.

The EPA again relied on the DOE’s determination (the “DOE’s Report”) that Ergon

achieved the same deficient scores as in the 2014 and 2015 evaluations:




      8
         Ergon did not request reconsideration of the 2014 decision “because the compliance
deadline ha[d] passed.” J.A. 268.


                                            15
16
J.A. 326–27. Unlike the 2014 and 2015 denials, however, the EPA did not focus on

Ergon’s viability specifically in its denial of the 2016 petition.

       Ergon filed a timely petition for review of the EPA’s final agency action regarding

only the 2016 petition. The Court has jurisdiction pursuant to 42 U.S.C. § 7607(b)(1).



                                               II.

       Ergon makes two overarching arguments in its challenge of the EPA’s denial of its

2016 petition. 9 First, Ergon argues that the EPA’s decision was arbitrary, capricious, and

contrary to law because it adopted the DOE’s “error-riddled analysis” of Ergon’s petition.

Opening Br. 23. Second, Ergon contends that the EPA’s conclusion was contrary to law

insofar as it read an extra-statutory “viability requirement” into § 7545(o)(9)(B)’s

“disproportionate economic hardship” determination. While we reject the latter argument,

the former is well-taken. We therefore vacate the EPA’s decision and remand for further

proceedings. 10


       9
            At the outset, we reject the EPA’s assertion that Ergon has waived some of its legal
arguments because it did not raise them during the administrative process. We fail to see how
Ergon could have raised legal arguments addressing the EPA’s alleged errors in denying its
petition before the EPA had actually denied the petition. Nor do we accept the EPA’s contention
that Ergon should have made those legal arguments within its 2016 petition because it knew the
bases for the EPA’s denial of its petition for compliance years 2014 and 2015. See Sinclair Wyo.
Ref. Co. v. U.S. EPA, 

, 992 (10th Cir. 2017) (“Indeed, the [EPA’s] decisions have
no precedential value even for the refiner, since each petition must be resolved on a case-by-case
basis . . . .”).
       10
          This is a case of first impression in this Court. In fact, the EPA’s application of the
small refinery exemption has been challenged only three times, with each of those cases focused
on the EPA’s application of a viability requirement. In Hermes Consol., LLC v. EPA, 

, 574–75, 579–80 (D.C. Cir. 2015), the court rejected a refinery’s challenge to the EPA’s
heavy reliance on the DOE’s analysis of its petition, particularly the Viability Index, but it
(Continued)
                                               17
                                                A.

       This Court must “hold unlawful and set aside agency action, findings, and

conclusions” that are “arbitrary, capricious, an abuse of discretion, or otherwise not in

accordance with law.” 5 U.S.C. § 706(2)(A). Although we accord substantial deference to

an agency’s final action and presume it valid, “the arbitrary-and-capricious standard does

not reduce judicial review to a rubber stamp of agency action.” Friends of Back Bay v.

U.S. Army Corps of Eng’rs, 

, 587 (4th Cir. 2012) (internal quotation marks

omitted). “Agency action is arbitrary and capricious if the agency relies on factors that

Congress did not intend for it to consider, entirely ignores important aspects of the

problem, explains its decision in a manner contrary to the evidence before it, or reaches a

decision that is so implausible that it cannot be ascribed to a difference in view.” United

States v. F/V Alice Amanda, 

, 1085 (4th Cir. 1993).

                                                B.

       Ergon argues that the EPA (1) acted in an arbitrary and capricious manner by

“ignor[ing] important aspects of the problem,” F/V Alice 

, in its




reversed the EPA’s decision due to errors the EPA admitted it made in calculating the refinery’s
net income and net refining margins. In Lion Oil Co. v. EPA, 

, 982–83, 984 (8th
Cir. 2015), the court likewise rejected a refinery’s contention that the EPA erred in relying on the
DOE’s analysis of its petition, with the EPA’s denial again focusing on the Viability Index. And
in Sinclair, the Tenth Circuit found reversible error when the EPA rejected the DOE’s
recommendation that a refinery receive a 50% waiver of the RFS Program requirements because
the EPA had determined that the Program would not affect the refinery’s viability. 

–99. Again, all of these cases involve the EPA’s rejection of the refineries’ petitions
because it concluded that the RFS Program did not threaten their viability. None of the cases
specifically address the Disproportionate Impact Index—primarily the metric at issue here.


                                                18
reliance on the DOE’s analysis of Ergon’s 2016 petition; and (2) acted contrary to law in

determining that Ergon would not receive a waiver because compliance with the RFS

Program would not threaten its viability. We address each argument in turn.

                                             1.

       Ergon makes several arguments attacking the DOE’s conclusions, but we are

limited in our consideration of these arguments. The DOE’s Report itself cannot be

challenged directly in this case. Ergon did not sue the DOE for issuing its

recommendation; 11 rather, it sued the EPA—the action agency—for denying its 2016

waiver petition. Therefore, instead of determining whether the DOE’s Report is arbitrary

and capricious, we may consider only whether the EPA’s reliance on the DOE’s Report

is arbitrary and capricious. See Dow AgroSciences LLC v. Nat’l Marine Fisheries Serv.,

, 266–67 (4th Cir. 2011) (“When a court of appeals reviews the EPA’s

reliance on a [report issued by another agency], it would determine only whether the

EPA’s reliance was arbitrary and capricious.”); City of Tacoma v. FERC, 

, 75

(D.C. Cir. 2006) (“Accordingly, when we are reviewing the decision of an action agency

to rely on [another agency’s report], the focus of our review is quite different than when

we are reviewing a [report] directly. In the former case, the critical question is whether

the action agency’s reliance was arbitrary and capricious, not whether the [report] itself is

somehow flawed.”). While the action agency is not required “to undertake an

independent analysis” of another agency’s conclusions, it may not “blindly adopt [those]

       11
          We express no opinion on whether such a report might be subject to direct judicial
review in a separate action.


                                             19
conclusions.” City of 

. Thus, an action agency’s reliance on a

facially-flawed report is arbitrary and capricious. See 

With these legal principles

in mind, we turn to Ergon’s arguments.

                                              2.

       Ergon first contends that the DOE erred in scoring two factors within the

Disproportionate Impact Index by arbitrarily defining “refining” to Ergon’s detriment. In

Section 1(b) (the “other business lines besides refining and marketing” factor), the DOE

separated Ergon’s refining from its lube oil production, considered the latter as an “other

business line[] besides refining and marketing,” and gave Ergon a score of 0 for this

factor. Then, in Section 2(a) (the “relative refining margin” factor), the DOE treated

Ergon’s lube oil production as refining for purposes of the relative refining margin

measure, again resulting in a score of 0. Both of these decisions negatively impacted

Ergon’s score. Despite this apparent contradiction, however, these arguments go to the

DOE’s scoring methodology and are not apparent on the face of the DOE’s Report.

Therefore, we cannot say that the EPA’s reliance on the DOE’s scoring of these factors

was arbitrary and capricious.

                                              3.

       Ergon next posits that the DOE erred by failing to score Section 1(c) (the “local

market acceptance of renewables” factor), Section 2(b) (the “renewable fuel blending”

factor), and Section 2(d) (the “RINs net revenue or cost” factor). The DOE’s failure to

score these factors is apparent on its face, and Ergon contends that the DOE’s arbitrary

treatment of these sections actively hurt its petition. Because our analysis of Sections 1(c)

                                             20
and 2(b) differs slightly from that of Section 2(d), we provide separate discussions of

these factors below.

                                                 a.

       In Section 1(c), the DOE accords points depending on the acceptance of renewable

fuel in the refinery’s local market. There are three subcategories within this factor: E10,

E85, and biodiesel renewables. The DOE did not score the latter two subcategories at all

in analyzing Ergon’s petition and has apparently never scored those subcategories in any

refinery’s petition. Instead, the DOE gave Ergon’s petition a score of 0 for the local

market’s acceptance of E10, completely disregarding the fact that approximately two-

thirds of Ergon’s transportation fuel production is diesel, which must be mixed with

biodiesel. The DOE treated Section 2(b)—which measures a refinery’s capacity for

blending renewable fuels with nonrenewable fuels—similarly. Although Section 2(b) has

subcategories for ethanol, biodiesel, and advanced biofuel blending, the DOE scored only

the first, ignoring Ergon’s biodiesel blending and giving Ergon a score of 0 for this

category. Had Ergon achieved a score of 10 on either Section 1(c) or Section 2(b), it

would have achieved a score greater than 1 and likely earned a small refinery

exemption. 12



       12
           The DOE’s analysis of Ergon’s 2016 petition resulted in a total score of 0.6 for the
Disproportionate Impact Index and a total score of 0 for the Viability Index. In the
Disproportionate Impact Index, the DOE gave Ergon ten points for Section 1(d) (the “percentage
of diesel production” factor) and no points for the other factors. With eight sections in this index
(not counting the “RINs net revenue or cost” factor, which the DOE did not score and has never
applied to any facility), the average was 1.25. After dividing by 2, the DOE reached the total
score of 0.6. Ergon needed a total of twenty points in this Index to achieve a score greater than 1.
(Continued)
                                                21
       The DOE’s treatment of these two factors—Sections 1(c) and 2(b)—is plainly

arbitrary as it treats unfairly those facilities where diesel makes up a substantial

percentage 13 of their transportation fuel production. For Section 1(c), despite the

widespread acceptance of E10 gasoline, a local market may not readily accept diesel

blended with biodiesel, placing refineries with higher-than-average production of diesel,

like Ergon, at a measurable disadvantage, as the DOE recognized in its 2011 Study. See

J.A. 71 (noting that, “[i]n most states, biodiesel blending is limited because biodiesel

feedstock is expensive and consumer resistance to the blend exists”). Similarly, in

Section 2(b), while a facility may have a high capacity to blend ethanol with its

nonrenewable fuel, it may not have the same capacity to blend biodiesel, so failing to

score this factor again harms those facilities with higher-than-average production of

diesel, like Ergon. These errors are readily apparent on the face of the DOE’s Report as

the index lists “[n]ot available” next to biodiesel in Section 1(c) and “not used” next to

biodiesel blending in Section 2(b). J.A. 326. Because the DOE’s recommendation was

clearly flawed on its face, “a clear error of judgment was made” when the EPA relied

without explanation on the DOE’s Report for its denial of Ergon’s 2016 waiver petition.

Ohio Valley Envtl. Coal. v. Aracoma Coal Co., 

, 192 (4th Cir. 2009). In




Had the DOE scored either Section 1(c) or Section 2(b), Ergon likely would have earned at least
a partial exemption.
       13
          While diesel makes up almost two-thirds of Ergon’s transportation fuel production, the
industry average is for diesel to constitute less than one-third of a given refinery’s transportation
fuel production.


                                                 22
addition, the EPA did not conduct any independent analysis regarding the subject matter

of Sections 1(c) and 2(b).

       Nonetheless, the EPA argues that its consideration of the DOE’s analysis was only

one of several grounds for denying Ergon’s petition. To be sure, the EPA stated in its

denial letter that it “independently review[ed] the information” and “consider[ed] other

economic factors in [its] analysis.” J.A. 327. But the extent to which the EPA relied on

the DOE’s Report and the relative weight of Sections 1(c) and 2(b) are unknown. In its

denial letter to Ergon, the EPA stated, “In determining whether [Ergon] will experience

disproportionate economic hardship, EPA considers whether compliance with its RFS

obligations disproportionately impacts [Ergon]. EPA generally defers to DOE’s

assessment due to DOE’s expertise on the refining industry.” J.A. 327 (emphasis added);

accord J.A. 320 (“EPA considers DOE’s assessment of whether a small refinery will face

disproportionate impacts in complying with its RFS obligations. The DOE analysis

informs EPA’s finding of whether ‘disproportionate economic hardship’ exists and in

turn EPA’s resulting decision about whether to grant or deny a petition for an extension

of the RFS temporary exemption for a small refinery.”). On this record, we cannot

determine whether the EPA would have reached the same conclusion had the DOE

submitted a proper analysis or had the EPA addressed the DOE’s failure to analyze

Sections 1(c) and 2(b). Although the EPA acknowledged that “disproportionate impacts

could disadvantage a refinery relative to the industry average and make compliance with

RFS obligations relatively more burdensome,” it specifically recognized that the “DOE

did not find that [Ergon] demonstrated disproportionate economic and structural impacts”

                                          23
(i.e., did not achieve the requisite score in the Disproportionate Impact Index). J.A. 328.

Although the EPA is statutorily required to consider the DOE’s recommendation, it may

not turn a blind eye to errors and omissions apparent on the face of the report, which

Ergon pointed out and the EPA did not address in any meaningful way. City of 

(“[T]he action agency must not blindly adopt the conclusions of the

consultant agency, citing that agency’s expertise.”). In doing so, the EPA “ignore[d]

important aspects of the problem.” F/V Alice 

.

                                            b.

       Section 2(d) considers whether a facility generates revenue by selling RINs or

suffers costs by purchasing RINs on the market. Like Sections 1(c) and 2(b), the DOE’s

failure to score this section is apparent on the face of the DOE’s Report, and that failure

negatively impacted Ergon’s petition. Unlike Sections 1(c) and 2(b), however, the EPA’s

reliance on the DOE’s Report regarding Section 2(d) was not arbitrary and capricious in

and of itself as the EPA did not rely on that factor in its determination. See J.A. 317

(discussing the 2011 Study in general and stating that “EPA notes that after further

review, contrary to statements [in the 2011 Study], it has been found [in an EPA study]

that a refinery does not experience disproportionate economic hardship simply because it

may need to purchase a significant percentage of its RINs for compliance from other

parties, even though RIN prices have increased since the [2011 Study], because the RIN

prices lead to higher sales prices obtained for the refineries’ blendstock, resulting in no

net cost of compliance for the refinery”). Because the EPA provided a specific response

addressing why it did not consider the 2011 Study’s conclusions concerning RIN

                                            24
prices—thereby implicitly disregarding the scoring of the factor in the DOE’s Report—

the EPA’s reliance on the DOE’s Report as to Section 2(d) was not arbitrary and

capricious.

       Even so, the EPA’s analysis of the effect of RIN prices on Ergon’s refining facility

was arbitrary and capricious on this record because the EPA ignored specific evidence

suggesting that those prices had a negative effect. In that regard, Ergon points out that the

EPA’s analysis of its RIN costs consists of a solitary statement: “EPA acknowledges that

[Ergon] may not be able to satisfy its [Program obligations] exclusively through the

blending of ethanol and biodiesel into its gasoline and diesel; however, the mere fact that

[Ergon] needs to purchase RINs for compliance does not necessarily entitle [Ergon] to an

exemption.” J.A. 330. The EPA cites to an EPA study titled “A Preliminary Assessment

of RIN Market Dynamics, RIN Prices, and Their Effects” for this conclusion. That study,

according to the EPA, merely determined that the refining industry as a whole is not

burdened by rising RIN prices because refineries may pass that cost to purchasers of the

blended fuel. Ergon’s 2016 petition, however, maintains that its refinery cannot pass the

RIN costs on to purchasers because of the local market’s low acceptance of blended

diesel. See 161 Cong. Rec. H10,105 (daily ed. Dec. 17, 2015) (“Since [the 2011 Study],

the dramatic rise in RIN prices has amplified RFS compliance and competitive

disparities, especially where unique regional factors exist, including high diesel demand,

no export access, and limited biodiesel infrastructure and production.”); J.A. 71

(recognizing in the 2011 Study that, “[i]n most states, biodiesel blending is limited

because biodiesel feedstock is expensive and consumer resistance to the blend exists”).

                                             25
Insomuch as the EPA cited generally to an industry-wide study and a nonspecific

nationwide trend to find that RIN prices would not harm Ergon although Ergon provided

specific, contradictory evidence of hardship particular to its refinery due to RIN costs,

the EPA failed to squarely address Ergon’s petition with regards to RIN costs and

“explain[ed] its decision in a manner contrary to the evidence before it.” F/V Alice

. Furthermore, the EPA’s disregard for Ergon’s RIN arguments

appears inconsistent with its statement earlier in the 2016 decision that the EPA considers

“RIN prices[] and the cost of compliance through RIN purchases” in making its

determination. J.A. 327; accord J.A. 201 (explaining in a December 2016 memorandum

that the EPA considers “RIN prices[] and the cost of compliance through RIN purchases”

in evaluating a petition). Consequently, the EPA’s cursory consideration and failure to

address Ergon’s specific evidence regarding RIN costs was an arbitrary and capricious

action. This failure alone warrants granting Ergon’s petition for review.

                                           ****

       Because the EPA relied on the DOE’s facially-deficient recommendation to an

unexplained and unknown degree, and because the EPA failed to properly address

Ergon’s petition with regard to RIN costs, we must vacate the EPA’s decision to deny

Ergon’s 2016 petition as arbitrary and capricious. See Hermes Consol., LLC v. EPA, 787




                                            

, 571 (D.C. Cir. 2015) (“[W]e are unable to conclude that EPA would have

reached the same decision absent its mistakes.”). 14

                                                4.

       Ergon next argues that the EPA read a viability requirement into the definition of

“disproportionate economic hardship” 15 and rejected Ergon’s waiver petition primarily

because compliance with the RFS Program would not threaten Ergon’s viability. 16 Ergon

urges us to adopt the Tenth Circuit’s reasoning in Sinclair Wyoming Refining Co. v. U.S.

EPA, 

(10th Cir. 2017), and hold that the EPA erred in this case by applying

this viability requirement. However, while it does appear that the EPA applied some sort


       14
         At this time, we neither endorse nor find fault with the remaining grounds that the EPA
contends support its denial of Ergon’s petition.
       15
           The parties dispute whether the Court should accord Chevron or Skidmore deference to
the EPA’s interpretation of the term “disproportionate economic hardship” found in 42 U.S.C.
§ 7545(o)(9). Compare Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 

, 843
(1984) (holding that, “if the statute is silent or ambiguous with respect to the specific issue, the
question for the court is whether the agency’s answer is based on a permissible construction of
the statute”), with Skidmore v. Swift & Co., 

, 140 (1944) (holding that a court will
determine how much weight to give an agency’s rulings “depend[ing] upon the thoroughness
evident in its consideration, the validity of its reasoning, its consistency with earlier and later
pronouncements, and all those factors which give it power to persuade, if lacking power to
control”). We do not decide this standard-of-review question because we otherwise find no merit
in Ergon’s argument on the issue raised.
       16
           Couched within the viability argument section of Ergon’s opening brief, Ergon
discusses at some length the DOE’s allegedly erroneous calculation of the industry average
refining margins and comparison of those margins to Ergon’s refining margins. From that
discussion, Ergon presumably argues that the EPA erred in relying on those figures when
assessing Ergon’s 2016 petition. Assuming without deciding that blind reliance on the DOE’s
calculations would constitute per se error, Ergon fails to recognize that the EPA independently
calculated the industry average refining margins. Compare J.A. 323–24 (the EPA’s calculation
of three-year average refining margins for the years 2014 through 2016), with J.A. 326 (the
DOE’s calculation of three-year average refining margins for the years 2013 through 2015).
Because Ergon does not challenge the EPA’s independent calculation, we do not consider
Ergon’s argument.


                                                27
of viability test in its denial of Ergon’s petition for compliance years 2014 and 2015,

there is no indication it used a similar viability requirement in the 2016 petition denial—

the sole decision of the EPA at issue in this case. Compare J.A. 262–65 (2014 & 2015

denial), with J.A. 327–30 (2016 denial); see also J.A. 320 (“In prior decisions, EPA

considered that a small refinery could not show disproportionate economic hardship

without showing an effect on ‘viability,’ but we are changing our approach. While a

showing of a significant impairment of refinery operations may help establish

disproportionate economic hardship, compliance with RFS obligations may impose a

disproportionate economic hardship when it is disproportionately difficult for a refinery

to comply with its RFS obligations—even if the refinery’s operations are not significantly

impaired.”). At most, it appears that the EPA considered viability as only one factor in its

2016 decision, which it was permitted to do. 

(“If long-term

‘viability’ was merely one element the EPA considered in its ‘disproportionate economic

hardship’ analysis, that would be a different story.”). Accordingly, we find no merit to

Ergon’s contention on this issue.



                                            III.

       For these reasons, we grant Ergon’s petition for review, vacate the EPA’s

decision, and remand the case to the EPA for further proceedings consistent with this

opinion.

                                    PETITION FOR REVIEW GRANTED;
                                    FINAL AGENCY ACTION VACATED;
                                    REMANDED FOR FURTHER PROCEEDINGS.

                                            28