April 2022 Denial of Petitions for
RFS Small Refinery Exemptions:
Appendices
SEPA
United States
Environmental Protection
Agency

-------
April 2022 Denial of Petitions for
RFS Small Refinery Exemptions:
Appendices
United States Environmental Protection Agency
United States
Environmental Protection
^1	Agency
EPA-420-R-22-005A
April 2022

-------
Appendix A - List of Small Refineries and Petitions Covered by the SRE
Denial
Petitioner
Refinery City
Refinery State
Petition Year
2018
Total























































This information has been claimed as
confidential by the affected businesses.









































































































Total


36
36

-------
Appendix B - Comment Summary and Response
This appendix summarizes the comments received and responds to the unique arguments made
therein that were not already addressed in the SRE Denial.1 Similar to the SRE petitions and
supporting documentation considered and addressed in the Proposed Denial, many of the
comments submitted in response to the Proposed Denial raised the same or very similar
arguments, allowing us to group and respond to the arguments once.
In sum, EPA received numerous substantive comments. The parties represented in the
commenters included refineries, biofuel producers, and their respective trade organizations.
Many elected officials, including representatives at the local, state, and federal level, commented
on the interests their constituents have in the SRE provision and RFS program. Many of the
petitioning small refineries submitted their comments under claims of confidentiality and
included refinery-specific data for DOE and EPA to evaluate. To the extent small refineries
raised the general arguments in favor of EPA granting their exemptions, EPA has responded to
those in the SRE Denial and this Appendix B.2 EPA has responded to confidential data and
information by providing confidential, refinery-specific appendices to the submitting refineries.
In all instances, the findings of the SRE Denial apply to all 36 SRE petition denials, regardless of
whether the refinery's comments are further addressed in an individual appendix.
1	"April 2022 Denial of Petitions for RFS Small Refinery Exemptions," EPA-420-R-22-005, April 2022.
2	Throughout this Appendix B, references to Sections I, II, III, IV, V, and VI refer to the corresponding sections in
the SRE Denial, while references to Sections B.I, II, III, and IV refer to the corresponding sections in this Appendix
B.
B-l

-------
I. Procedural Comments and Legal Authority
1. EPA provided an insufficient opportunity for comment on the Proposed Denial.
Comment:
EPA stacked multiple comment periods for RFS-related actions raising numerous economic and
legal issues and posing serious potential consequences for small refineries' compliance into the
same short period. Small refineries did not have sufficient time to prepare adequate comments to
the Proposed Denial due to several comment periods for RFS actions open at the same time.
Response:
EPA's action denying SRE petitions is not a rulemaking, but rather is an adjudication of the SRE
petitions before the Agency and, as such, EPA is not required to provide public notice and an
opportunity for public comment before taking this action. However, EPA chose to provide the
opportunity for public comment to ensure that the Agency had all relevant information available
to it, and that all stakeholders had an opportunity to provide information for EPA's consideration
in making a final decision on the SRE petitions. Further, EPA disagrees that the public comment
period was insufficient because other RFS actions were available for public comment at the same
time. First, small refineries have been on notice regarding the holdings in the RFA opinion since
January 20, 2020. Second, EPA notified the refineries on August 17, 2021, that EPA was
strongly considering applying the holdings from the RFA opinion that remained after the
Supreme Court's decision in HollyFrontier to pending SRE petitions before the Agency. Though
the Proposed Denial was not issued until December 7, 2021, and not published in the Federal
Register until December 14, 2021, small refineries have had access to the RFA opinion since
January 20, 2020, and had actual notice and the opportunity to provide information as early as
August 2021. As discussed in the SRE Denial Executive Summary, stakeholders were notified of
EPA's intent to include the 2018 remanded SRE petitions in the Proposed Denial as early as
January 3, 2022, following the D.C. Circuit's order remanding those petitions just under four
weeks earlier. The fact that other EPA actions in which the same small refineries may have an
interest were also available for public comment is not relevant to the adequacy of the refineries'
opportunity to comment on this action. Moreover, it would be impossible for EPA to ensure that
only one action at a time is open for public comment to avoid stakeholders having to address
more than one proposed action at a time. Further discussion of our reasoning for maintaining the
comment period deadline is available in a January 25, 2022, response letter to a coalition of
small refineries, available in the docket for this action.
Comment:
EPA's contention that small refineries were on notice regarding the substance or importance of
the Proposed Denial from actions and events leading up to it is nonsensical.
B-2

-------
Response:
EPA clearly informed affected refineries by email on August 17, 2021, that the Agency intended
to evaluate SRE petitions following its evaluation of the RFA holdings.3 Therefore, refineries
had actual notice of the factors and analysis that would likely be applied to the pending SRE
petitions, and that EPA would consider relevant information they submitted. Specifically, EPA
stated:
You are receiving this email. . . because EPA has at least one pending small refinery
exemption petition from your small refinery. EPA has received additional
information from certain small refinery exemption petitioners relating to their ability
to recoup their RFS compliance costs in response to the U.S. Court of Appeals for
the Tenth Circuit's January 2020 holding in Renewable Fuels Association v. EPA that
disproportionate economic hardship must be caused by the RFS. In the interest of
equity, EPA wants to be sure that you are aware that EPA is evaluating what this
holding means and that you, too have the opportunity to submit additional
information to support your small refinery exemption petition(s). EPA will consider
all the information you provide in support of your petition when making its decision.
There is no basis to claim that this notice was inadequate simply because it occurred in advance
of the Proposed Denial; in fact, EPA's email provided additional, earlier notice, effectively
extending the time period for providing information to the Agency. In this email, EPA explained
that it was evaluating RIN cost passthrough in the context of the SRE provision due to the RFA
holding on DEH causation. There was no need for EPA to take a definitive position on whether it
intended to apply the RFA holdings—and in fact that was one of the issues on which EPA
expressly requested input—so refineries would be able to comment on the question of whether
the Agency should do so. Therefore, between August 2021 and February 2022, small refineries
had roughly five months' notice and opportunity to comment on EPA's proposed reliance on
these factors.
Comment:
The 35-day comment period provided for submitting comments on the inclusion of the remanded
2018 SRE petitions was insufficient and unlawful for adequate preparation of meaningful
comments.
Response:
As noted above, this action is not a rulemaking and EPA is not required to provide public notice
and an opportunity for public comment before acting on the 2018 SRE petitions. In fact, EPA did
not undertake a notice-and-comment process for the original action on the remanded 2018 SRE
petitions. Moreover, given the substantial overlap of issues between consideration of the
remanded 2018 SRE petitions and the other SRE petitions pending before the Agency, and given
EPA's obligation to issue new decisions on the remanded 2018 SRE petitions by April 7, 2022,
3 A copy of this email is available in the docket for this action.
B-3

-------
pursuant to the D.C. Circuit's order, 35 days was a reasonable amount of time to adjust or
expand the comments already being prepared in response to the Proposed Denial.
Comment:
The comment period was less than 60 days.
Response:
First, a 60-day comment period is not required for adjudications upon which EPA chooses to
request comment, and in fact no opportunity for notice and comment is required at all. In
addition, the Proposed Denial was published on EPA's website on December 7, 2021. The
comment period was extended shortly thereafter to February 7, 2022. Thus, the total time from
public availability to the close of the comment period was 62 days.
Comment:
The time constraints EPA claims prevent it from extending the comment period deadline are
results of the Agency's own doing.
Response:
As explained above, EPA provided a reasonable opportunity for public comment and explained
in its January 25, 2022, letter to a coalition of small refineries its reasons for not extending the
comment period. In addition, there are several reasons for EPA to act on the SRE petitions at this
time. First, EPA is under an order from the Court of Appeals for the D.C. Circuit to issue new
decisions on the remanded 2018 SRE petitions by April 7, 2022. Furthermore, it is important to
provide certainty to both the SRE petitioners and other RFS program participants by deciding the
pending SRE petitions quickly, particularly following the Supreme Court's decision in
HollyFrontier. EPA also intends to take final action to issue a decision on the remaining pending
SRE petitions in the near future. Lastly, EPA also intends to take final action to issue the 2020-
2022 RFS Annual Rule in the near future, and this action directly impacts that final rule.
B-4

-------
2. EPA's Proposed Denial violated the due process rights of the petitioning small
refineries.
Comment:
Small refineries have a property interest in their RFS exemptions that the government cannot
take away without due process.
Response:
Small refineries have no property interest in continued exemptions under the RFS program. As
an initial matter, the Supreme Court has held that "[t]o have a property interest in a benefit, a
person clearly must have more than an abstract need or desire for it. He must have more than a
unilateral expectation of it. He must, instead, have a legitimate claim of entitlement to it." Board
of Regents v. Roth, 408 U.S. 564, 576 (1972). EPA has consistently maintained that small
refineries have no entitlement to an exemption in a given year. There is a presumption of
compliance under the RFS program from which a small refinery may be exempted, only if it can
make a demonstration of DEH caused by the cost of compliance with the RFS program. There is
not, and has never been, a guarantee or promise that a small refinery exempted in one year will
be exempted in the following year. To the contrary, EPA's practice has routinely been to
evaluate SRE petitions based on the circumstances within the petition year. Meaning that, even
in recent years, EPA has denied small refineries' SRE petitions in one year after having granted
an exemption for the same refinery for a prior year, based on the different facts. The small
refinery exemption is further distinguishable from other government benefits recognized by the
Supreme Court as being protected by the Fourteenth Amendment's due process clause in its
purpose. The Court explains that "[i]t is a purpose of the ancient institution of property to protect
those claims upon which people rely in their daily lives, reliance that must not be arbitrarily
undermined." Roth, 408 U.S. at 576. Exemption from otherwise mandatory environmental
standards is not an "ancient institution of property," nor is it analogous to "claims upon which
people rely in their daily lives." EPA does not recognize perpetual exemption as a valid
compliance strategy upon which a business can rely in making strategic decisions. Moreover,
and as explained elsewhere herein, EPA provided ample notice and process—more than was
required under the CAA—and did not violate small refineries' due process rights.
Comment:
EPA violated small refineries' due process rights because the Agency failed to provide adequate
notice of its intention to deny the pending SRE petitions and the remanded 2018 SRE petitions.
Commenters also claim that they reasonably expected that EPA would grant their SRE petitions
since refineries had consistently received exemptions in the past; and that the public notice-and-
comment process does not cure the lack of notice, as the refineries would have had more time to
prepare their SRE petitions than they did to prepare their comments.
B-5

-------
Response:
As noted above, EPA disagrees with commenters that there is a recognized property interest in
receiving an exemption from the RFS program. In addition, EPA provided adequate notice of its
action and, therefore, even if the refineries did have such a property interest, they were not
deprived of due process. First, as explained above, EPA provided ample notice and opportunity
for the refineries and other stakeholders to comment on its Proposed Denial. Second, and as
discussed elsewhere in this Appendix, commenters also have no basis to claim they reasonably
relied on EPA's past actions on other SRE petitions to assume their petitions would be granted,
and EPA disagrees that any such expectation was reasonable. As noted above, EPA has always
evaluated each SRE petition based on information relevant to that petition, and in some cases has
provided relief in one year and denied it in the next.
Finally, the commenters claim that, since the Agency claims the statute is ambiguous, EPA was,
therefore, not compelled to revise its interpretation. The basis for EPA's decision to follow the
holdings of RFA is explained in Section III and depends on the Agency's evaluation of the
statutory text as well as the purpose of the RFS program and of the SRE provision. The fact that
the commenters disagree with EPA's interpretation does not mean that the Agency failed to
provide adequate notice of its action. In fact, commenters had sufficient time and opportunity to
explain in their comments their disagreement with EPA's conclusions, as described in responses
in Section B.I.I.
Moreover, as noted above, small refineries lack a property interest in obtaining an exemption
from the RFS program. And, even if they had such an interest, they would have to show that they
"sustained prejudice as a result of the allegedly insufficient notice." Long v. Board of Governors
of the Federal Reserve System, 117 F.3d 1145, 1158 (10th Cir. 1997). EPA provided small
refineries adequate notice regarding the Agency's specific intent to deny their pending SRE
petitions and solicited comment on the aspects the Agency considers to be the most important
matters of both fact and law. In addition to this comment period, in August 2021, EPA explicitly
requested additional information from small refineries regarding RIN cost passthrough and the
holdings of the RFA opinion. On August 25, 2021, EPA filed a motion for voluntary remand
without vacatur in the D.C. Circuit cases so that EPA could evaluate the impacts of the RFA and
HollyFrontier decisions on its SRE policy and the decisions made on those SRE petitions.4 In
total, small refineries had over five months of notice of what factors EPA believed would be
important in deciding the pending SRE petitions. Small refineries have used that time to provide
comprehensive comments—in meetings and written comments—on the legal and policy issues
raised by the Proposed Denial.
Furthermore, small refineries could have supplemented their SRE petitions at any time during
their pendency at the Agency, and some did submit additional information multiple times over
that period. EPA's use of the notice-and-comment process merely provided small refineries
another opportunity to provide information supporting their SRE petitions and for other RFS
stakeholders to also provide feedback on EPA's implementation of the SRE provision. EPA also
disagrees with the commenters' attempt to equate the time to prepare an SRE petition with the
4 See e.g., EPA's Motion for Voluntary Remand Without Vacatur, Doc. No. 1911606, August 25, 2021, Sinclair
Wyo. Refining Co. v. EPA, No. 19-1196 (consol. with 19-1197) (D.C. Cir.).
B-6

-------
time needed for a sufficient opportunity to comment. The SRE petitions EPA is acting on had
already been prepared and submitted to the Agency for review when EPA notified refineries in
August 2021, and when EPA issued its Proposed Denial. EPA provided the comment period to
ensure that the Agency had before it all relevant information, including any additional
information petitioners wanted EPA to consider before taking its final action.
B-7

-------
3. EPA's Proposed Denial is a veiled retroactive rulemaking with inadequate process
that violates the Administrative Procedure Act.
Comment:
EPA's Proposed Denial is an action that is generally applicable to all small refineries,
characteristic of a legislative rulemaking and not individual adjudications or an interpretive rule,
and otherwise meets the definition of a rule in the Administrative Procedure Act ("APA").
The Proposed Denial proposes to retroactively apply two new interpretations of the SRE statute:
(1) The eligibility provision, and (2) The "disparate economic hardship" provision. If finalized,
the Proposed Denial's eligibility and DEH interpretations would cause either a de facto
regulatory repeal or an amendment of 40 CFR 80.1441 and have devastating economic
consequences on small refineries.
The statutory interpretations EPA proposed are rules, and the process EPA used in adopting
these statutory interpretations (i.e., publishing a notice and request for comment in the Federal
Register) resembles the rulemaking process. Accordingly, this rule cannot be applied
retroactively absent a clear statement of Congress to the contrary, and there is no such
authorization in the CAA. Even if the CAA did authorize retroactive rulemaking of this kind, the
APA prohibits retroactive rulemaking (see Treasure State Res. Indus. Ass 'n v. EPA, 805 F.3d
300, 305 n.l (D.C. Cir. 2015), defining "rule" as a statement of "future effect").
Response:
It is well-settled that "the choice between rulemaking and adjudication lies in the first instance
within the agency's discretion." NLRB v. Bell Aerospace Co., 416 U.S. 267, 294 (1974); see also
SEC v. Chenery Corp., 332 U.S. 194, 203 (1947). It is also well-settled that an agency "is not
precluded from announcing new principles" in an adjudication, see Cassell v. FCC, 154 F.3d
478, 486 (citing to NLRB), and may also address legal issues for the first time. Conference
Group v. FCC, 720 F.3d 957, 965 (2013). Here, EPA is conducting a single adjudication of 36
SRE petitions in reliance on EPA's revised interpretations of the statutory SRE provisions, as
applied to the facts and circumstances of each SRE petition. The SRE Denial is an adjudication
limited to the SRE petitions expressly identified in Section I and Appendix A, redacted under
claims of confidentiality. We are only adjudicating the SRE petitions from small refineries
articulated in this action, not every small refinery participating in the RFS program currently and
in the future (as a rulemaking would necessarily do). If we receive additional SRE petitions in
the future, we will grant or deny them in a subsequent adjudicative action. Furthermore, it is
recognized that agency adjudications may necessarily include statements of policy, and that such
statements do not transform the adjudications into rulemakings. NLRB v. Bell Aerospace, 416
U.S. at 294 (upholding the Board's discretion to forgo rulemaking because "the adjudicative
procedures in this case may also produce the relevant information necessary to mature and fair
consideration of the issues").
Additionally, EPA has not asserted or cited to any of its rulemaking authorities under CAA
sections 206, 211, or 301 to support this action, and this action is clearly not a rulemaking under
B-8

-------
CAA section 307(d) as it is not a promulgation or revision of a regulation under CAA section
211. EPA has relied solely on its authority to adjudicate SRE petitions under CAA section
211(g)(9)(B).
Even if EPA were taking action through a rulemaking, this would not be a retroactive rule. A
retroactive rule "takes away or impairs vested rights acquired under existing law, or creates a
new obligation, imposes a new duty, or attaches a new disability in respect to transactions or
considerations already past." I.N.S. v. St. Cyr, 533 U.S. 289, 321 (2001). The SRE Denial does
not take away or impair small refineries' vested rights, as they have no entitlement to exemptions
from the RFS standards. Additionally, the SRE Denial imposes no new obligations, duties, or
disabilities on small refineries. It merely denies their requests to be excused from compliance
with their existing RFS obligations.
EPA's previous action on the 2018 SRE petitions was remanded to the Agency with an order to
issue new decisions on the SRE petitions at issue. EPA's new decisions on the remanded SRE
petitions are necessary to respond to court direction to address those petitions and are not
retroactive. Rather, they are new actions on past SRE petitions that were sent back to the Agency
by the reviewing court.
The commenters claiming EPA's SRE decisions constitute a "regulatory repeal," or "amendment
of the regulations" are incorrect. The SRE Denial does not make any changes to the RFS
regulations—they remain intact and unchanged. In fact, the SRE Denial is consistent with the
regulations at 40 CFR 80.1441. See 40 CFR 80.1441(e)(2). Those regulations do not speak to
how EPA will interpret the statute or evaluate eligibility to petition or DEH, but simply explain
the process for small refineries to apply for an exemption. Thus, this action is not a "regulatory
repeal" or "amendment." Rather, it is an adjudicatory action to decide SRE petitions, based on
EPA's interpretation of the relevant statutory provision.
B-9

-------
4. If EPA's Proposed Denial is not an improper, retroactive rulemaking, then it is an
unlawful retroactive adjudication causing "manifest injustice."5
Comment:
EPA's denial of the SRE petitions is a retroactive adjudication, and reconsidering those petitions
is especially inappropriate and inequitable. Where an agency imposes a retroactive adjudication,
courts consider "(1) whether the particular case is one of first impression, (2) whether the new
rule represents an abrupt departure from well-established practice or merely attempts to fill a
void in an unsettled area of law, (3) the extent to which the party against whom the new rule is
applied relied on the former rule, (4) the degree of the burden which a retroactive order imposes
on a party, and (5) the statutory interest in applying a new rule despite the reliance of a party on
the old standard." Retail, Wholesale & Department Store Union v. NLRB, 466 F.2d 380, 390
(D.C. Cir. 1972) (describing factors that can result in a manifestly unjust retroactive
adjudication). Even if EPA's new interpretation is permissible, and regardless of whether it is an
adjudication or a rulemaking, retroactive application is impermissible.
Commenters also claim that retroactive application of the new legal rule is unlawful, if a party
has conformed its conduct to a prior legal regime, as small refineries have done: at the time small
refineries were owed decisions (90 days after submitting their SRE petitions), EPA's approach to
SRE petition evaluation included reliance on the 2011 DOE Study, and small refineries had
formulated their petitions accordingly. As such, the Proposed Denial violated small refineries'
settled expectations regarding EPA's SRE petition evaluation process since disproving RIN cost
passthrough was not an eligibility requirement at the time small refineries submitted their
petitions.
EPA is attaching new legal consequences to actions small refineries completed before proposal
of the changed interpretation, and small refineries reasonably relied on the 2018 decision to grant
their SRE petitions and EPA's retroactive revocation results in DEH to small refineries.
Reconsidering the already granted petitions is especially inappropriate and inequitable.
Response:
As explained elsewhere, "the choice between rulemaking and adjudication lies in the first
instance within the agency's discretion." (NLRB, 416 U.S. 267, 194 (1974)). Here, EPA is acting
on SRE petitions through an adjudication, not a rulemaking, and courts do not disfavor
retroactive adjudication but review their validity based on fairness and equity to the affected
party. Cassel v. FCC, 154 F.3d 478, 486 (D.C. Cir. 1998). EPA does not believe that the
decisions issued in the SRE Denial are retroactive; however, even if they are, the final action is
not an impermissible retroactive adjudication as it only clarifies existing law, Aliceville Hydro
Associates v. F.E.R.C., 800 F.2d 1147, 1152 (D.C. Cir. 1986), and does not result in an unfair or
inequitable outcome.
5 Retail, Wholesale & Department Store Union v. NLRB, 466 F.2d 380, 390 (D.C. Cir. 1972) (describing the
considerations that must be weighed when evaluating whether a retroactive adjudication results in manifest
injustice).
B-10

-------
As an initial matter, agency adjudications are generally accepted to be retroactive, though the
retroactivity of adjudications is not limitless. Bowen v. Georgetown University Hosp, 488. U.S.
204, 221(1988), see also AT&T v. F.C.C., 454 F.3d 329, 332 (D.C. Cir. 2006). The assertion that
EPA's Proposed Denial and final action constitute a "substitution of new law for old law that
was reasonably clear," Aliceville, 800 F.2d at 1152), fails to acknowledge the many changes in
SRE petition adjudication that have taken place over the years. These represent different
approaches EPA has taken to DEH evaluation, as the Agency's views and the case law have
evolved. As explained in Section II.D, none of these approaches can be called "old law that was
reasonably clear." Id. And even if they were, EPA's change in approach is more accurately
characterized as a clarification of existing law (i.e., a clarification of what constitutes DEH
within the context of the SRE provisions), or a correction of practice "rectifying] legal mistakes
identified by a federal court." Verizon Telephone Companies v. FCC, 269 F.3d 1098, 1111 (D.C.
Cir. 2001). Accordingly, the SRE Denial is not an impermissible retroactive adjudication.
For the 2018 SRE petitions, EPA is responding to the D.C. Circuit's order remanding EPA's
prior actions and requiring the Agency to issue new decisions on the petitions at issue. Therefore,
EPA has an obligation to act in response to the court decision. EPA is applying the reasoning of
the RFA opinion, which was issued on January 20, 2020. EPA is obligated to take into
consideration changes in the law that occur while it is considering the petitions before it, and
doing so is proper; the petitions are again before the agency. Verizon Telephone Companies, 269
F.3d at 1110-11. Indeed, "the Administrative Procedure Act generally contemplates that when an
agency proceeds by adjudication, it will apply its ruling to the case at hand." Clark-Cowlitz Joint
Operating Agency v. F.E.R.C., 826 F.2d 1074, 1082 (D.C. Cir. 1987).
EPA further disagrees that its action is an impermissible retroactive adjudication based on the
Retail Wholesale factors. Those factors "boil down ... to a question of concerns grounded in
notions of equity and fairness[,]" Clark-Cowlitz at 1082, n.6, and consideration of the factors
demonstrates that EPA's action is not counter to such notions. Here, EPA is applying its new
statutory interpretation via adjudication, which is permissible, and is necessarily applying that
interpretation to the SRE petitions it has before it. EPA's change in position is not an "abrupt
departure" from well-established practice. EPA has taken different approaches to SRE petitions
over the years, as described in Section II.D, and in some cases has granted a refinery's petition in
one year but denied it in the next. In addition, the RFA decision upon which EPA's current
interpretation is based was issued in January 2020, just a few months after the Agency's action
on the 2018 SRE petitions. It was not reasonable for the 2018 petitioners to rely to their
detriment on our exemption grants in light of our past practice, the unsettled case law, and the
January 2020 RFA decision in particular. Additionally, challenges to the 2018 Decision6 were
filed and still pending. Given those circumstances, small refineries should have been aware that
their exemption grants could be reversed, either by the Agency if it concluded that the RFA
decision was correct, or by the courts. Also, EPA has twice extended the 2019 compliance
deadline for small refineries, mitigating any adverse impacts or burdens EPA's change in
interpretation might impose. Lastly, EPA is applying an interpretation that it believes is
consistent with the intent of Congress in adopting the RFS program and in authorizing
exemptions for small refineries.
6 Memorandum: Decision on 2018 Small Refinery Exemption Petitions, August 9, 2019 (memo on file with EPA).
B-ll

-------
EPA also disagrees that its action is disrupting small refineries' settled expectations and
attaching new legal consequences to decisions the refineries made before the Proposed Denial.
As noted above, to the extent small refineries relied on past actions, that reliance was not
reasonable in light of the facts and circumstances. Small refineries were, or reasonably should
have been, aware that EPA's 2018 Decision Memo was challenged by multiple parties
immediately after its issuance. Furthermore, once the RFA opinion was issued in January 2020,
small refineries were additionally put on notice that the same findings relied upon in the 2018
Decision had been found to be impermissible under the CAA. Accordingly, any reliance small
refineries placed on the 2018 Decision was misplaced given the litigation over that decision and
the RFA decision.
Moreover, the legal consequences of EPA's actions are the same regardless of when the denial is
issued (i.e., small refineries remain obligated under the RFS program). Nor are small refineries
prejudiced by having demonstrated compliance while their SRE petitions are still pending, which
was a business decision they chose to make. Furthermore, EPA solicited, and small refineries
submitted, comments in response to the Proposed Denial, providing small refineries an
opportunity to modify and amend the SRE petitions they submitted prior to the Proposed Denial
in order to address EPA's changed SRE policy.
EPA itself provided small refineries with sufficient notice regarding its possible change in
interpretation. EPA proposed to revise its prior interpretation approach to evaluating SRE
petitions on December 7, 2021. However, small refineries should have been aware of the
Agency's consideration of a different interpretation even before the Proposed Denial. EPA
explained in February 2021 that it intended to support the interpretation taken by the Tenth
Circuit in RFA before the Supreme Court. After the HollyFrontier opinion was issued, EPA
solicited information from small refineries directly relevant to the remaining holdings of RFA,
and even highlighted those holdings in its request. Through that request, subsequent requests,
and the Proposed Denial, EPA provided small refineries with the opportunity to supplement their
pending SRE petitions to address the change in EPA's approach, and, in fact, small refineries did
exactly that.
Even if EPA had not provided small refineries with notice and opportunity to supplement their
SRE petitions according to the RFA holdings, the SRE Denial would not likely "trigger[]
retroactivity concerns ''Pine Tree Medical Assocs. v. Secretary of Health and Human Servs., 127
F.3d 118, 121-22 (1st Cir. 1997). In Pine Tree, the First Circuit stated that, "[tjhere is an obvious
difference between rejecting an application because it fails to meet a new regulation governing
the proper format or preparation of applications that was promulgated after that application was
filed, and rejecting an application because the substantive standards for granting the application
on the merits have changed in the period between filing and review." Id. The court explains that
the petitioner "place[d] undue significance on the act of filing an application with an
administrative agency," and that "the mere filing of an application is not the kind of completed
transaction in which a party could fairly expect stability of the relevant laws as of the transaction
date." Id.
B-12

-------
Comment:
EPA's retroactive revocation results in DEH on small refineries.
Response:
As stated above, the legal and policy decisions EPA here adopts are not retroactive. EPA's
previous action on the 2018 SRE petitions was challenged in court, and the court remanded the
actions back to the Agency to issue new decisions. These new decisions do not constitute a
retroactive rulemaking or retroactive adjudications, and are instead made to replace the
challenged actions. Further, small refineries do not experience DEH from compliance with the
RFS program, as explained in the SRE Denial.
Comment:
The Proposed Denial abandons the practice of relying on the 2011 DOE Study and scoring
matrix, which EPA has applied for over ten years. EPA justifies this change by wrongly relying
on DOE's superseded 2009 study. Reliance on the 2009 DOE Study is particularly inappropriate
because EPA in its 2020-2022 RFS Annual Rule Proposal indicated that the RIN market in
illiquid, and reliance on the 2009 DOE Study cannot replace the consultation requirement in the
statute.
Response:
EPA's choice to modify its approach to SRE petition evaluation by moving away from reliance
on the DOE scoring matrix is an appropriate policy decision given the lack of relevant
information provided in the scoring matrix under the RFA causation framework, as explained in
Sections IV.C and D.
The language quoted by the commenter from the 2020-2022 RFS Annual Rule Proposal
regarding the liquidity of the RIN market was mischaracterized by the commenter as describing
the current state of the RIN market. Rather, the language described a hypothetical situation that
could occur if EPA had proposed a different action. Additionally, EPA has satisfied the statutory
requirement to consult with DOE as described in Sections IV.C and B.1.5.
B-13

-------
5. EPA failed to follow the statutory process for deciding SRE petitions.
Comment:
EPA has failed to consult with the DOE and consider the 2011 DOE Study, as required by the
CAA. Commenters also assert that DOE's findings in the 2011 DOE Study have largely proven
correct over time.
Response:
As described in Section IV.C, EPA did consult with DOE through meetings, phone calls, and
written communications. EPA also considered both the 2009 and 2011 DOE Studies and "other
economic factors," and the Agency's consideration is explained in the SRE Denial. EPA is not
bound by any statutory language to a specific form or format for its consultation with DOE, nor
does the statute dictate how EPA should consider the studies or other economic factors. EPA's
consultation and consideration of the 2011 DOE study are consistent with the statutory
requirement. While not legally required, EPA has added a memorandum to the docket for this
action describing the EPA-DOE consultation process. Regarding the assertion that the 2011 DOE
Study has been proven correct over time, as explained in Section IV.C, while DOE was correct
in anticipating the RIN prices could rise in the future, DOE's supposition that this would
advantage fuel blenders has proven not to be true. Furthermore, the 2011 DOE Study did not
anticipate the degree to which those compliance costs would be passed through to refineries in
higher prices for the products they sell.
Comment:
Had EPA issued timely decisions, likely resulting in exemptions for many small refineries under
EPA's prior approach, small refineries would have had the opportunity to purchase RINs at
lower prices than today.
Response:
As explained in the SRE Denial, EPA is acting on the petitions consistent with the holdings of
the RFA decision. While that decision was issued in January 2020, its threshold holding
regarding small refineries' eligibility for an extension of their exemptions was reviewed by the
Supreme Court and reversed in HollyFrontier in July 2021. Had the Court upheld that particular
holding, EPA would not have needed to have consider the other holdings in RFA. Therefore, it
was reasonable for EPA to wait until resolution of that case by the Court before acting on
petitions, in case refineries were no longer eligible. Additionally, the 36 SRE petitions were not
remanded until December 8, 2021, and, by that time, both the RFA and HollyFrontier opinions
had been issued, and EPA must apply those holdings to its current evaluation of SRE petitions.
Finally, small refineries are aware that receiving an exemption in one year does not guarantee an
exemption in the following year, as each year's SRE petition is reviewed separately. Under the
statute, compliance with the RFS program is the default, and small refineries should plan to
comply with their annual obligations until such time as they petition for and receive and
exemption. Moreover, EPA does not agree that the results of the business decisions small
B-14

-------
refineries make regarding the timing of their RIN purchases is a cause of DEH, as explained in
Section IV.D.
Comment:
Where an interagency consultation is required, evidence of such consultation must go beyond a
mere generalized statement that consultation occurred.
Response:
As noted above, the form of consultation is not specified in the statute. Since the Proposed
Denial, EPA has updated Section IV.C to explain the consultation process the Agency used with
DOE. While not legally required, EPA has added a memorandum to the docket for this action
describing the EPA-DOE consultation process.
B-15

-------
6. EPA's Proposed Denial is arbitrary and capricious.
Comment:
EPA is changing its interpretation to effectuate a particular policy outcome, not to correct a legal
error.
Response:
EPA is changing its interpretation to align it with the Tenth Circuit's surviving RFA holdings,
which clarified the meaning of DEH by indicating it must be caused by compliance with the RFS
program and highlighted a failure of EPA in ignoring its finding on RIN cost passthrough when
adjudicating SRE petitions, and which EPA believes is the best interpretation of the statutory
SRE provisions. See Sections II.D and IV.D for more explanation regarding how EPA is
changing its approach in response to the RFA opinion. Additionally, EPA is also basing this
change in approach on the Agency's findings regarding the RIN market and RIN cost
passthrough, consistent with the Tenth Circuit in RFA, as described in Section IV.D.2.d.
Comment:
EPA's Proposed Denial ignores information relevant to assessing whether RFS compliance
would impose DEH on an individual small refinery, as opposed to making a single decision for
numerous SRE petitions. A single decision for all SRE petitions cannot adequately consider the
facts of individual SRE petitions as the CAA requires.
Response:
As an initial matter, this commenter did not specifically identify what information EPA is
allegedly ignoring in its analysis of whether RFS compliance imposes DEH on an individual
small refinery. Without knowing what specific information this commenter is referencing, EPA
cannot respond to this assertion. Regardless, by publishing notice of and requesting comment on
the Proposed Denial, EPA's process was designed to gather all information that small refineries
and other stakeholders considered relevant to deciding SRE petitions. Accordingly, in support of
the SRE Denial, EPA considered and addressed all the substantive information—including the
individual petitions and supplemental information from small refineries—provided by interested
parties to the Agency during the public comment period that those parties considered relevant to
assessing whether RFS compliance imposes DEH on small refineries. After conducting this
review, EPA finds that the petitioning small refineries have not demonstrated that they face
disproportionate RFS compliance costs and, therefore, have not demonstrated DEH warranting
exemption. If this commenter is also asserting there exists other allegedly relevant information
that was not considered, but that information was not provided to the Agency, EPA obviously
could not have considered that information in its analysis unless it was provided to the Agency.
EPA notes that, if a small refinery had provided data and evidence of other economic factors
upon which EPA could determine, after consultation with DOE, that the particular small refinery
had demonstrated that it faced DEH consistent with the criteria described in the SRE Denial and
B-16

-------
contrary to the facts regarding other small refineries, EPA would have issued an exemption to
that small refinery. However, no individual small refinery has made such a showing in the SRE
petitions EPA reviewed in taking this action. As described in more detail in Sections IV.C and D,
EPA has evaluated each individual SRE petition, as well as provided all SRE petitions and
related supplemental materials to DOE as part of the agencies' consultation. Contrary to the
commenters' assertion, the SRE Denial is based on EPA's consultation with DOE on the facts of
individual petition elements, including: the costs of RFS compliance per the memorandum to the
docket regarding DOE-EPA consultation; consideration of individual small refineries' data and
comments as evidenced by EPA's detailed confidential, refinery-specific appendices; and EPA's
response to comments herein and in Section IV.D. Nothing in the SRE Denial contradicts the
fact that, if a refinery demonstrates that it experiences unique DEH in the future, EPA would
issue an SRE to that small refinery.
Comment:
EPA's failure to consider the scoring matrix prepared by DOE contravenes the CAA and belies
EPA's claim to have consulted with DOE. The scores provided by DOE when it applies the
scoring matrix to an SRE petition demonstrate without question whether a small refinery merits
an exemption. Moreover, EPA's unsupported assertion that RIN costs are always passed through
to consumers by all refineries, regardless of market location and situation, ignores the findings in
the 2011 DOE Study on which EPA has based its evaluation of SRE petitions for over a decade.
Response:
As an initial matter, the CAA does not require EPA to use the DOE scoring matrices in its
evaluation of SRE petitions. As the commenter itself acknowledges, the CAA only requires that
EPA, in consultation with DOE, consider the 2011 DOE Study and other economic factors.7 As
described more fully in Section IV.C, EPA expressly considered the 2011 DOE Study and,
importantly, its finding that "[disproportionate economic hardship must encompass two broad
components: a high cost of [RFS] compliance relative to the industry average, and an effect
sufficient to cause a significant impairment of the refinery operations." (emphasis added). EPA
has concluded, consistent with the findings of the 2011 DOE Study and the Tenth Circuit's RFA
decision, that DEH can only occur when the disproportionate impact comes from a high cost of
RFS compliance relative to other refineries. EPA chose not to use the 2011 DOE scoring
matrices because those matrices were designed to differentiate between refineries that would
bear a higher cost of RFS compliance due to an inability to blend biofuels when compared to
refineries that could blend fuels. DOE designed the matrices in this way projecting that"If
certain small refineries must purchase RINs that are far more expensive than those that may be
generated through blending, this will lead to disproportionate economic hardship for those
effected entities." (emphasis added). EPA has, with the benefit of time, experience implementing
the RFS program, and based on the substantial data, contracts, and academic literature provided
to the Agency in the SRE petitions and comments on the SRE Denial, concluded that RFS
compliance costs are the same whether RINs are acquired through blending or by purchasing
RINs.8 With no difference in compliance costs whether a refinery buys RINs or blends
7	CAA section 21 l(o)(9)(B)(ii).
8	See Section IV.D.2.
B-17

-------
renewable fuel to acquire RINs, the evaluation rubric that DOE created to identify small
refineries with limited ability to blend biofuels (i.e., the DOE scoring matrix) has no applicability
to the analysis that EPA is making in this decision.
Comment:
EPA's narrow interpretation of "other economic factors" to allow the Agency to rely exclusively
on a flawed finding of RIN cost passthrough to deny all pending SRE petitions misapplies the
Tenth Circuit's holdings.
Response:
As an initial matter, the commenter misreads EPA's explanation of the RFA opinion in the
Proposed Denial, and misreads the RFA opinion itself, in an attempt to erroneously assert that
EPA can still grant an exemption to a small refinery for hardship caused by something other than
its compliance with the RFS program. EPA strongly disagrees with that assertion. The
commenter also asserts EPA is narrowly and erroneously construing the use of "other economic
factors" to only consider RIN cost passthrough so that EPA may deny the SRE petitions. EPA
also strongly disagrees with that assertion. In Section IV.D, EPA explains how it is following the
statutory provisions in CAA section 21 l(o)(9), as interpreted by the RFA opinion, which requires
the hardship to be caused by compliance with the RFS program. In making that evaluation, and
as further explained in Section IV.D, EPA considers whatever "other economic factors"—which
includes its consideration of the economic principles described as the RIN discount and RIN cost
passthrough—that inform whether a small refinery has demonstrated its hardship is caused by its
RFS compliance.
Moreover, EPA's findings regarding RIN cost passthrough are not flawed but are based on
EPA's analysis of the available information as described throughout the SRE Denial. Where
commenters have presented studies refuting EPA's findings, we have responded to those
comments in Section B.III and in confidential, refinery-specific appendices. This action
harmonizes EPA's findings regarding RIN cost passthrough with the circumstances described in
the SRE petitions and the holdings of the RFA opinion. In relying on EPA's findings regarding
RIN cost passthrough, EPA is also relying on all the data supporting RIN cost passthrough, and
the findings represent months of careful consideration of the information described throughout
this decision and its supporting materials.
Comment:
EPA's interpretation also contravenes the statute because it fails to read "disproportionate" in
context. Small refineries seeking an SRE must demonstrate "disproportionate economic
hardship." But EPA proposes to sever "disproportionate" from that phrase, asserting that small
refineries must demonstrate that their "RFS compliance costs are disproportionate compared to
other refineries' RFS compliance costs." EPA also tries to smuggle in a non-statutory severity
requirement, insisting that any disproportionality must be "of sufficient magnitude to warrant the
exemption." That is not what the statute says. If RFS compliance—on its own or in conjunction
B-18

-------
with "other economic factors"—causes a small refinery to suffer any greater hardship relative to
large refineries, it has suffered DEH, and EPA must grant an exemption.
Response:
Again, as noted in the previous response, this commenter is attempting to read the statute and the
RFA opinion to allow it to obtain an exemption for reasons other than hardship caused by its
RFS compliance. EPA strongly disagrees with that assertion. As explained in Section III, EPA,
DOE, and the Tenth Circuit all share the same understanding of the definition of "DEH" (i.e.,
that DEH must be caused by RFS compliance and that a small refinery's RFS compliance costs
must be higher relative to other refineries). Furthermore, if a small refinery's RFS compliance
costs are higher relative to other refineries, then that higher compliance cost must be significant
enough to constitute "economic hardship," since slightly higher costs may not rise to that level.
Because each obligated party's RFS obligation is determined as a percentage of that party's
gasoline and diesel fuel production, the RFS obligations are, by definition, proportionate across
all obligated parties. Furthermore, in Sections IV.D.2.a and IV.D.3.f, EPA explains how RFS
compliance costs are the same for all obligated parties regardless of a party's chosen compliance
approach (blending or purchasing RINs). This happens because the market prices for
transportation fuel increase to reflect the cost of the RIN, and this increased fuel price allows
obligated parties to recover their RIN costs through the market price of the fuels they produce.
Because the market behaves this way for all parties subject to the RFS program, there is no
disproportionate cost to any party, including small refineries.
Comment:
EPA is basing its Proposed Denial on improper political considerations and the Agency's desired
outcome, not the facts of small refineries' petitions by considering the input of biofuels groups
and others vehemently opposed to any form of relief for small refineries. These outside parties
have no understanding of the CBI provided by small refineries in support of their SRE petitions.
Response:
EPA based the Proposed Denial and the SRE Denial on the extensive information and analysis
presented in those documents and summarized herein and in the supporting materials provided
by the petitioning small refineries before EPA issued the Proposed Denial and by small refineries
and other interested parties during the public comment period. EPA chose to provide public
notice and broadly request comment on its Proposed Denial from all interested parties to ensure
full consideration of all relevant factors. EPA's decisions on SRE petitions have an impact on all
parties participating in the RFS program. As such, EPA believed all parties could provide
meaningful input on all aspects of the Proposed Denial, including EPA's understanding of its
observations in the RIN market. Accordingly, EPA believes the input of all parties was
appropriately considered in the SRE Denial. While EPA acknowledges that other parties are not
able to review, consider, and comment on any materials small refineries claim as CBI, EPA itself
carefully considered that information in the SRE Denial, and applied the statutory criteria after
consideration of all relevant comments.
B-19

-------
Comment:
EPA's Proposed Denial is arbitrary and capricious because it was developed using an unlawful
and opaque process. First, EPA's decision to take public comment on a single decision to deny
multiple SRE petitions submitted by numerous small refineries creates serious procedural
concerns. Congress intended SRE petitions to be adjudicated and decided on a case-by-case
basis, see 42 U.S.C. § 7545(o)(9)(B). In fact—with one exception that is the 2018 SRE
decision—this is how EPA has always conducted the SRE decision-making process. To do
otherwise conflicts with the U.S. Court of Appeals for the Fourth Circuit's decision in Ergon-
West Virginia, Inc. (admonishing EPA that, when assessing the impact of RFS compliance costs
on an individual small refinery, EPA must do more than cite to conclusions about "the refining
industry as a whole" Ergon-W. Va. V. EPA, 896 F.3d 600, 613 (4th Cir. 2018) (emphasis
added).
Second, under the judicial review provisions of the CAA, any refinery whose petition is denied is
entitled to judicial review in the applicable regional circuit. 42 U.S.C. § 7607(b)(1) (petitions for
review of certain enumerated petitions must be filed in the D.C. Circuit while other enumerated
actions and "any other final action of the Administrator under this chapter.. .which is locally or
regionally applicable may be filed only in the United States Court of Appeals for the appropriate
circuit" except for a subset of additional cases that must be filed in the D.C. Circuit because
EPA's otherwise local action "is based on a determination of nationwide scope or effect" that the
Administrator publishes). The process EPA is now using will force dozens of refineries to
challenge the Proposed Denial as a group—without any meaningful opportunity to explain to the
D.C. Circuit why EPA has wrongly denied relief to a given refinery based on information in its
individual exemption petition. If finalized, the Proposed Denial will essentially insulate EPA
from judicial review. When procedural errors such as these are "so serious and related to matters
of such central relevance" to EPA's final action "that there is a substantial likelihood that the
[decision] would have been significantly changed if such errors had not been made," 42 U.S.C. §
7607(d)(8), a court must reverse. Id. § 7607(d)(9)(D); see also 5 U.S.C. § 706(2)(D).
But EPA's procedural failings do not end there. In addition to forcing small refineries into a
public, rulemaking process for what is intended to be a confidential adjudication, EPA has based
many of its conclusions on data and information that it has declined to make public. Although
the commenter appreciates that there may be CBI in many small refineries' submissions, see,
e.g., Proposed Denial at 53, EPA cannot use that as a convenient excuse to obscure the data upon
which it relies. Doing so robs small refineries and other stakeholders of a meaningful opportunity
to comment on and refute the Proposed Denial and will likewise also deny them meaningful
judicial review. To name just one example, EPA states that it found no evidence to support the
lack of passthrough or higher RIN acquisition costs for some small refineries and that this is
"consistent across all the markets it observed." Proposed Denial at 27. Tellingly, EPA does not
say which (or even how many) markets it observed or the type of evidence it sought but could
not find, leaving the commenter with no ability assess this claim or explain why the commenter's
market may be different.
B-20

-------
Response:
First, in reaching this decision, EPA's process was not "unlawful and opaque." As has been
stated above, "the choice between rulemaking and adjudication lies in the first instance within
the agency's discretion." (NLRB, 416 U.S. 267, 294 (1974)), and here EPA chose to decide these
SRE petitions through an adjudication addressing 36 pending SRE petitions. EPA also chose to
employ a public notice-and-comment process to ensure it adjudicated the SRE petitions after
considering all relevant information through a transparent process. In comparison, under the
prior approach, EPA only provided the basis for its decisions to the small refineries themselves
in confidential decision documents, a practice for which EPA has been criticized for its opacity.
Through this public process, EPA has received information from all interested parties and
considered the refinery-specific information submitted. After careful review of the information
submitted in the SRE petitions, petition supplements, and comments on the Proposed Denial,
EPA determined small refineries had not demonstrated DEH because EPA found the cost of RFS
compliance is the same for all obligated parties, including small refineries, as described in
Section IV.D.2.
Second, EPA disagrees with the commenter's assertion that the D.C. Circuit would provide an
inadequate venue in which small refineries could seek judicial review of the SRE Denial. The
venue for judicial review of EPA's actions under the CAA is determined by the statute.9 Further,
there is no reason to believe small refineries would not have a "meaningful opportunity to
explain ... why EPA has wrongly denied relief to a given refinery based on information in its
individual exemption petition," in the D.C. Circuit as in any other circuit court. Indeed, the
commenter fails to explain how the D.C. Circuit would not provide adequate review of this
action.
Importantly, the lack of access to the information that EPA is evaluating in no way diminishes a
small refinery's ability to "explain why the commenter's market may be different." The market
EPA has described is one where the market price of a refinery's products reflects the cost of RFS
compliance (i.e., RIN cost passthrough). Any small refinery wishing to refute that finding for its
local market would do so by providing evidence to the Agency and, if challenging a decision, a
court that the market in which it operates does not behave in this manner. As explained
extensively in EPA's evaluation of economic studies provided by small refineries in Sections
IV.D.2 and B.III, a number of small refineries have attempted to provide such explanations and
EPA has evaluated them. All of this information is part of the administrative record for this
action that the D.C. Circuit would consider in its review of the SRE Denial. The court has rules
and procedures in place to safeguard the claims of CBI while it considers, as does EPA, all the
information presented in making its decision. Thus, neither the venue of the litigation nor claims
of CBI for certain information will in any way impede the parties in obtaining a fair and
impartial consideration of its arguments during judicial review of the SRE Denial.
Lastly, the commenter here provides conflicting assertions, where it first admonishes EPA for an
"unlawful and opaque" process, then states that EPA's decisions must be "confidential
adjudication^]." Thus, the comment both complains of the opacity of EPA's current action while
simultaneously stating that EPA's decisions ought to be less transparent (the whole decision
9 See CAA section 307(b)(1).
B-21

-------
withheld as confidential). Moreover, EPA cannot disclose information that has been submitted to
the Agency under a claim of confidentiality until such time as EPA's Office of General Counsel
makes a final determination that the information is not entitled to confidential treatment.10 Until
such a determination is made, EPA must preserve the information submitted under claims of
confidentiality. To the furthest extent possible, EPA has utilized publicly available information
and aggregated or summarized the confidential information received so that it could be presented
in the Proposed Denial and the SRE Denial.
Comment:
EPA cannot simultaneously claim that the Tenth Circuit's opinion in RFA compelled its change
in interpretation and that the language of the CAA is ambiguous and in need of clarifying
interpretation.
Response:
As explained in Section III, the CAA does not define "disproportionate economic hardship."
Several courts have identified this phrase as an ambiguous term that EPA interprets in
administering the SRE provisions.11 Prior to the RFA decision, EPA had been interpreting and
applying that phrase to allow a small refinery to demonstrate DEH for reasons other than its RFS
compliance. The Tenth Circuit in RFA, however, rejected that interpretation and instead directed
that "disproportionate economic hardship" must be caused by compliance with the RFS
program.12 Accordingly, EPA is adopting the Tenth Circuit's holdings and applying them in the
SRE Denial because the Agency believes this is the correct interpretation of the statutory text
given the purpose of the RFS program and of the exemption.
10	See 40 CFR 2.204.
11	See e.g., Hermes Consolidated, LLC v. EPA, 787 F.3d 568, 574-75 (D.C. Cir. 2015); Sinclair Wyoming Refining
Co. v. EPA, 887 F.3d 986, 996 (10th Cir. 2017) ("The statutory text at issue allows a range of linguistic possibilities
in defining "disproportionate economic hardship.").
12	RFA, 948 F.3d at 1254 ("Granting extensions of exemptions based at least in part on hardships not caused by RFS
compliance was outside the scope of the EPA's statutory authority.").
B-22

-------
7. EPA's statutory interpretation and approach to SRE evaluation is contrary to
congressional intent.
Comment:
EPA's strict interpretation of DEH causation in the Proposed Denial disregards, and is contrary
to, Congress's plainly expressed intent that the DOE survey factors can and do show when a
small refinery is experiencing DEH, as demonstrated in appropriations reports over the years to
both DOE and EPA instructing the agencies on how to implement the SRE provisions.
Response:
The commenter appears to suggest that Congress intended for DOE and EPA to grant relief
whether or not DEH exists. Following the RFA decision, it is clear that in the absence of DEH,
caused by the cost of RFS compliance, EPA has no authority to grant hardship relief. Yet, the
commenter suggests such a conclusion is surplusage and, therefore, EPA must grant relief
regardless of whether DEH does or does not exist. Were that the intent of Congress, Congress
would not have put any condition on issuing SREs, would not have directed EPA and DOE to do
the requisite evaluations, and would not have deemed the exemption "temporary," but rather
would have simply exempted small refineries from the RFS program.
As described in Section II.D, Congressional appropriation committees for DOE and EPA have
offered direction through report language that DOE should apply the scoring matrix in a
particular way, and EPA should grant some form of relief based upon DOE's scoring and
provide an explanation for why EPA chose not to do so. This language has not remained
consistent year-to-year, in some cases recommending for relief only when both parts of the DOE
scoring matrix recommend for relief and in other years when one or the other portion of the
matrix does. Neither EPA nor DOE's current FY2022 appropriations bills or associated report
language contain information directing the agencies regarding these decisions. Furthermore, to
the degree it is appropriate for EPA to consider such earlier report language, the SRE Denial
fully explains the basis for EPA's decision and why it is appropriate under the CAA.
Congress has never opined on EPA's findings regarding RIN cost passthrough, nor, as discussed
in Sections II.D and IV.C, did DOE's 2011 Study make any determination on the veracity of
RIN cost passthrough. EPA has, throughout the Proposed Denial and the SRE Denial, provided
independent studies and its own data analysis supporting an overall finding of RIN cost
passthrough. Nevertheless, in denying the SRE petitions, it is not necessary for EPA to
conclusively demonstrate that RIN costs are always passed through in every market and under all
circumstances. Rather, small refineries requesting relief must demonstrate that they experience
DEH as a result of compliance with the RFS program. EPA invited petitioning small refineries to
submit information demonstrating that they experienced DEH due to the RFS program. EPA has
evaluated all refinery-specific information it has received to determine whether this information
provided evidence of DEH. As detailed throughout the SRE Denial and our response to
comments, we found that no small refinery demonstrated that it experienced DEH due to the RFS
program.
B-23

-------
Moreover, CAA section 21 l(o)(9)(B) does not include specific factors that Congress plainly
expected EPA to consider. As Section IV.C explains, the statute is largely silent on the approach
EPA applies to evaluating SRE petitions. And as some supporting comments have pointed out,
Congress as a whole did not provide instruction to DOE to revisit the 2009 DOE Study. Rather,
it was one committee in the Senate that rejected and called for a reevaluation of the 2009 DOE
Study. One Senate committee does not represent the whole of Congress. Indeed, the text of the
CAA was not amended to instruct DOE to perform an additional analysis and the House of
Representatives' conference report acknowledged the non-binding nature of the Senate
committee's statement: "[t]he conferees ... expect the Department to undertake the [Senate
committee's] requested economic review." H.R. Rep. No. 111-278, at 126 (2009). Thus, the
instructions to DOE and EPA in the various congressional reports do not represent
"congressional intent" regarding the SRE provisions and do not obligate EPA to act in any
particular way, given such language does not modify the statutory provisions.
B-24

-------
8. EPA's statutory interpretation renders the SRE provisions in the CAA surplusage.
Comment:
EPA reads the requirement that it consider "other economic factors" out of the SRE statute,
focusing only on the RIN cost passthrough theory.
Response:
Under the statute, EPA must evaluate an SRE petition to determine whether a small refinery has
demonstrated it experienced DEH caused by compliance with its RFS obligations. The statute
directs EPA, in consultation with DOE, to consider the findings of the 2011 DOE study and
"other economic factors" when making that evaluation. The statute does not require EPA to
consider any particular number or types of economic factors, nor does it require EPA to consider
other circumstances that might affect a small refinery's financial wellbeing once EPA has
determined that a small refinery has not experienced DEH from compliance with its RFS
obligations. As explained in Section IV.D, the RIN cost passthrough analysis and all the
economic data that go into it are part of EPA's consideration of relevant "other economic
factors" in its evaluation of the pending SRE petitions, not a single economic factor as the
commenter asserts. Additionally, in the responses to the studies submitted by small refineries in
Section B.III, and again in the response to the technical comments EPA received in Section
B.IV, EPA further considers "other economic factors" in its evaluation. Taken altogether, EPA
considers all the information small refineries and other interested parties have submitted in
determining that small refineries do not experience DEH caused by RFS compliance given all
refineries are able to recover their costs in the market. The commenter has not provided
information regarding another economic factor that EPA has not considered that clearly
demonstrates that a small refinery experiences DEH caused by its RFS compliance costs. Rather,
the commenter would have EPA consider factors unrelated to RFS compliance, and, as explained
in Sections III and IV, this is not what permissible under the statute.
Comment:
EPA's new interpretation would effectively end SRE relief under the RFS program, in direct
contradiction with the Supreme Court's foundational assumption in HollyFrontier that CAA
section 21 l(o)(9) does not include a sunset provision, and would render meaningless the
Supreme Court's recent opinion in HollyFrontier. EPA's new interpretation is contrary to the
Supreme Court's opinion in HollyFrontier, which instructed that the SRE provisions in the CAA
must be read "fairly, not narrowly."
Response:
In HollyFrontier, the Supreme Court held that the "key phrase at issue before [it]—'A small
refinery may at any time petition the Administrator for an extension of the exemption under
paragraph (A) for the reason of disproportionate economic hardship'—simply does not contain
the continuity requirement the court of appeals supposed. Instead, more naturally, it means
exactly what it says: A small refinery may apply for a hardship extension 'at any time.'"
B-25

-------
HollyFrontier at 2181. The SRE Denial does nothing to prevent small refineries from submitting
petitions "at any time" in the future. Moreover, small refineries are aware of the analysis and
statutory interpretation EPA will apply to their SRE petitions in the future. For these reasons, the
SRE Denial does not render meaningless the Supreme Court's opinion in HollyFrontier. EPA is
not "sunsetting" the exemption provision—refineries may submit exemption petitions in the
future along with their demonstrations of DEH, which EPA will evaluate and act on.
First, the question at issue in HollyFrontier is not the same question before EPA in this action,
and the decision here in no way impacts the holding in HollyFrontier. This action decides the
SRE petitions before the Agency according to the information EPA has reviewed. As the Tenth
Circuit said in RFA, simply because a small refinery may petition for an exemption does not
require EPA to grant the exemption. While EPA is concluding that the information it has
considered, including information submitted by petitioning refineries, does not demonstrate DEH
caused by RFS compliance, it is important to state that the SRE Denial does not prejudge future
SRE petitions or eliminate the possibility of new, different data becoming available in the future
that could support a different conclusion. EPA does not in the SRE Denial judge SRE petitions
that do not yet exist, in the context of future circumstances that do not yet exist; it only decides
the SRE petitions that are currently before us.
B-26

-------
9. EPA's statutory interpretation adds a strict "proximate cause" requirement to the
CAA.
Comment:
The CAA does not require that DEH be caused by, and only by, compliance with the RFS
program. EPA's new causation interpretation also contravenes the text of the statute and evinces
a misunderstanding of the RFS program. EPA asserts that the CAA requires that DEH be caused
by, and only by, RFS compliance, "meaning that a small refinery may not simply experience a
year of poor economic performance or struggle with disadvantageous operational or market
constraints to merit an SRE." But protecting struggling small refineries is precisely what
Congress intended. As EPA has always agreed (until now), Congress "did not constrain the
scope of EPA's [DEH] determination or use language requiring that RFS compliance be the sole
cause of hardship." And as the D.C. Circuit has held, Congress required more than a bare
consideration of compliance costs, "Congress required EPA to consult with DOE and to consider
the findings of the 2011 Study and other economic factors." It is only after doing all three—
consulting with DOE, considering the 2011 Study, and considering other economic factors—that
the statute grants EPA "substantial discretion to decide how to evaluate hardship petitions." The
Proposed Denial improperly eschews the 2011 DOE Study and fails to address the statutorily
required "other economic factors" beyond the cost of compliance.
Response:
As explained in Section IV.D. 1 and in responses to other comments herein, the language of the
CAA requires that DEH be caused by compliance with the RFS program. It was the Tenth
Circuit in RFA that clarified the extent to which EPA may consider DEH to be caused by factors
outside the RFS program, and it determined that such considerations were improper.13
Accordingly, this is the analysis EPA must apply within the Tenth Circuit and in which the
Agency is now applying nationwide. Moreover, as explained in the SRE Denial and in responses
to other comments herein, EPA followed the statutory directive as it evaluated these petitions
and considered "other economic factors" when making its final decision.
13 RFA, 948 F.3d at 1254 ("Granting extensions of exemptions based at least in part on hardships not caused by RFS
compliance was outside the scope of the EPA's statutory authority.").
B-27

-------
10. The RFA opinion lacks any legal force and is not an authority upon which EPA may
rely.
Comment:
The Tenth Circuit vacated the RFA opinion in its entirety after the Supreme Court opined in
HollyFrontier, therefore the Tenth Circuit's opinion lacks legal force and cannot be the basis of
this action.
Response:
EPA disagrees with this assessment of the validity of the holdings within the RFA opinion. On
August 19, 2021, EPA filed a motion for clarification regarding the legal effect of the court's
July 29, 2021, mandate, stating:
EPA wishes to clarify its understanding that the challenged agency orders are remanded
back to EPA without vacatur for further proceedings in accordance with this Court's
January 24, 2020, opinion, as modified by the Supreme Court. Specifically, EPA wishes
to clarify that, pursuant to the mandate: (1) the alternative holdings in the Court's January
24, 2020, opinion not addressed by the Supreme Court remain in effect; and (2) the
orders at issue are remanded to EPA without vacatur. ... If the Court concludes that its
prior orders and mandate do not require further clarification, EPA will proceed in
accordance with its current understanding as reflected in this motion."
EPA's Motion for Clarification of the Court's July 29, 2021, Mandate at 2, RFA, 948 F.3d 1206
(10th Cir. August 19, 2021). On August 26, 2021, the court denied EPA's motion. Order, id.
(10th Cir. August 26, 2021). Accordingly, EPA considers RFA to have legal force and is
proceeding with this understanding, as explained to the court.
Comment:
Even if the Tenth Circuit did not vacate the RFA decision in full, the opinion is only binding
within the Tenth Circuit.
Response:
While the Tenth Circuit's RFA holding is only binding precedent within that court's jurisdiction,
EPA has determined that the RFA decision provides the best reading of the statutory provisions
of CAA section 21 l(o)(9) and is accordingly taking this action for 36 pending SRE petitions.
This is appropriate because EPA has "substantial discretion" for purposes of implementing these
SRE provisions.14 The alternative—to apply RFA only to small refineries within the Tenth
Circuit—would create disparate treatment of those small refineries, which EPA finds would be
14 Hermes, 787 F.3d at 575 ("EPA retains substantial discretion to decide how to evaluate hardship petitions.").
B-28

-------
unworkable and unfair given the national scope of the RFS program and EPA's determination of
what is the best interpretation of the Act.
B-29

-------
11. The CAA requires individual hardship decisions and analysis, not the generalized
approach EPA has taken with the Proposed Denial.
Comment:
EPA's Proposed Denial is unlawful because it applies an improper "one-size-fits-all" analysis
and is not based on an evaluation of the refinery-specific facts raised in each SRE petition.
Response:
As explained in Sections III and IV.D.3 and in responses to other comments herein, EPA
considered the refinery-specific facts in each SRE petition in taking its final action. EPA
provided all of the SRE petitions and supplemental materials to DOE and consulted with DOE
on those submissions. Furthermore, EPA evaluated the design and mechanisms of the RFS
program to assess how small refineries might be impacted, evaluated the fuel and RIN market
data at-large to assess whether actual field data supported the conclusions from that analysis, and
then also evaluated all of the refinery-specific information individual companies provided to
assess whether there was something unique to their circumstances that was not captured by the
broader analysis. This included evaluation of the confidential information provided, as discussed
in the confidential, refinery-specific appendices to this action.
Comment:
The CAA requires individual adjudications of each SRE petition. Individual adjudications
preserve small refineries' access to the Federal Court of Appeal for the Circuit in which they are
located. EPA's single decision for all SRE petitioners forces small refineries into the D.C.
Circuit, insulating EPA's action from judicial review.
Response:
The SRE Denial is not the first instance in which EPA has issued a single decision document
adjudicating multiple SRE petitions.15 Furthermore, EPA has considered the arguments made by
the individual small refineries, many of which were repeated across SRE petitions from other
small refineries. In the SRE Denial, EPA also addresses refinery-specific data in a way that
preserves small refineries' claims of confidentiality. Accordingly, though EPA is deciding
multiple SRE petitions in the SRE Denial, it has considered and evaluated each petition
individually. The comment regarding judicial review of EPA's SRE decisions is addressed in
Section B.I.6.
Comment:
The failure to conduct a case-by-case analysis constitutes a procedural error that warrants
invalidation of the action under CAA section 307(d)(8).
15 2018 Decision (August 9, 2019).
B-30

-------
Response:
As explained in the Proposed Denial and again in the SRE Denial, this action is not a rulemaking
subject to the various statutory and other provisions applicable to a rulemaking; as such, neither the
procedural requirements of CAA section 307(d) (none of which require a case-by-case analysis even
if applicable) nor the procedural standard of review in CAA section 307(d)(8) apply. Instead, it is an
adjudication of 36 pending SRE petitions. As such, EPA has considered and responded to the
arguments (many of them identical) made by the individual small refineries. EPA has also
considered the facts each small refinery submitted purporting to refute or disprove EPA's
analysis of the RIN program and market effects, including RIN cost passthrough and RIN
discount. EPA has addressed the refinery-specific data in a general way to preserve small
refineries' claims of confidentiality. EPA has also issued separate individual responses in
confidential, refinery-specific appendices to certain small refineries that raised unique arguments
to which the Agency could not respond without disclosing confidential information.
Comment:
The Fourth and Tenth Circuits both held that EPA must specifically consider and address each
small refinery's argument that RIN costs are not passed through. Thus, EPA's Proposed Denial is
directly contrary to Fourth Circuit case law and the RFA opinion.
Response:
EPA's final action does consider and address each small refinery's allegation that its RIN costs
cannot be passed through. Small refineries began providing detailed comments to EPA to support
their claims of DEH immediately following the Supreme Court's decision in HollyFrontier (in
which case they were considered by EPA in developing the Proposed Denial) and many took the
opportunity to submit the same or augmented arguments as comments on the Proposed Denial.
EPA addresses and responds to these arguments in Sections IV.D.2 and 3, and throughout this
Appendix in response to specific comments. Section B.III specifically focuses on specific studies
and data submitted by many small refineries in response to the Proposed Denial.
B-31

-------
12. The Proposed Denial does not comply with Executive Order 12898.
Comment:
EPA fails to "identify or address" the fact that small refinery closures (or even reductions in
capacity) caused by the Proposed Denial could have an adverse environmental impact on
environmental justice communities because small refineries have a smaller environmental
footprint and less impact on their surrounding communities than large integrated refineries.
Response:
The commenters provided no analysis to support their assertion that small refineries have less
impact on their surrounding communities than large integrated refineries. Regardless, any small
refinery closures that occur subsequent to this action are not caused by compliance with the RFS
program, as explained in Section IV.D. Furthermore, any adverse environmental impact on
environmental justice communities from these hypothetical small refinery closures is purely
speculative and outside the scope of this action.
B-32

-------
13. The Proposed Denial is contrary to case law on the SRE provisions and RFS
program.
Comment:
EPA's Proposed Denial conflicts with other Tenth Circuit case law in Sinclair, which instructed
EPA to apply a holistic analysis to each small refinery petition, not single factor analysis based
on whether DOE evaluated a small refinery to be financially viable going forward. Here, as then,
EPA is relying on a single factor, the flawed assumption that small refineries can passthrough
their RIN costs.
Response:
EPA disagrees. EPA's interpretation is consistent with the Sinclair precedent by applying a
"holistic" analysis to the pending SRE petitions by considering RIN market observations that
EPA had previously ignored. This analysis accounts for the RFS program's effects on fuel and
RIN pricing for obligated parties and is therefore far from a "single factor" analysis. EPA has
broadly considered all relevant factors of DEH, including the specific arguments raised by small
refineries and the economic studies they provided in their comments to the Proposed Denial. As
described in Section IV.D, EPA has considered and analyzed many factors in reaching the
decision to deny the 36 SRE petitions; this process is the opposite of applying a "single-factor"
assumption to make these decisions.
Comment:
In all prior instances of judicial review over SRE decisions, the courts have never overturned the
framework of the 2011 DOE survey and scoring matrix. This created a reliance interest on the
part of small refineries on this evaluation approach being maintained.
Response:
As explained in Section B.I.4, EPA disagrees that small refineries had any basis to rely on a
particular past Agency action, given the uncertainty in the case law and EPA statements
regarding RIN cost passthrough.
B-33

-------
II EPA's Interpretation on SRE Eligibility is Arbitrary, Capricious, and
Otherwise Contrary to Law
Comment:
The requirement that a small refinery must have received the original statutory exemption early
in the RFS program to petition for an extension of the exemption is contrary to the language of
the CAA. Alternatively, such a requirement is not supported by anything in the CAA. In
addition, EPA's proposed eligibility requirement is contrary to the HollyFrontier opinion.
Response:
EPA disagrees with the commenter's claim that the statute permits a refinery that is not a small
refinery at the time of the original exemption to later become one, either through reducing its
throughput or being newly constructed, and then to receive an "extension" of that original
exemption. The commenter's interpretation is inconsistent with the text of the statute and is not
supported by HollyFrontier. On its face, the interpretation that a refinery must have received the
original statutory exemption under CAA 21 l(o)(9)(A) is consistent with the CAA, which
describes the extension of that exemption in section 21 l(o)(9)(B)(i) as an "extension of the
exemption under paragraph (A)." Contrary to the commenter's assertion, the HollyFrontier court
did not expressly address this question, and in fact, focused on a small refinery's exemption
having "lapsed," which means it would have existed at some time in the past.16 Therefore, the
Tenth Circuit and Supreme Court opinions are consistent with EPA's interpretation that, under
the CAA, a small refinery that held the original blanket exemption is eligible to receive an
extension of that exemption, regardless of whether or not a small refinery's exemption history
following its receipt of the original exemption is continuous. For this reason, EPA announced in
the Proposed Denial that the Agency was considering returning to its original view of eligibility,
and that is the position taken in Sections IV.A.2 and 3.
Comment:
EPA's interpretation regarding eligibility violates small refineries' due process rights.
Response:
Every time EPA has adjudicated an SRE petition, it has done so by applying its then-current
interpretation of a small refinery's eligibility for an exemption under CAA section 21 l(o)(9). In
this case, EPA provided notice to small refineries of its intention to apply its prior interpretation
of small refinery eligibility in the Proposed Denial (i.e., requiring refineries to have obtained the
original blanket exemption) and provided over 60 days of notice and invited public comment.
Additionally, EPA communicated directly with the two refineries that it concluded were
ineligible to petition on these grounds. Accordingly, these refineries had actual notice of EPA's
intent to find them ineligible and had the opportunity to comment on that finding. The refineries
16 HollyFrontier, 141 S.Ct. at 2178 (referring to a "resumption after some interrupting lapse").
B-34

-------
submitted comments, which have been included and addressed here. EPA has responded to other
comments on the subject of due process in Section B.I.2.
B-35

-------
III. Studies Relied on by Small Refineries to Refute RIN Cost Passthrough
1. Dr. Fitzgerald, Texas Tech Study and LSU Study
Comment:
EPA's Proposed Denial relies on a single academic study (Knittel et al., 2017). EPA has ignored
other studies (including Lade & Bushnell, 2019; Li and Stock, 2019; Burkhardt 2019; Pouliot et
al., 2017) that provide strong evidence for imperfect passthrough.
Response:
EPA has not based its RIN passthrough conclusion on a single study, as the commenter alleges.
Rather, EPA has based this conclusion on an abundance of evidence, including economic theory
and empirical studies. The basis for our conclusion is discussed at length in the SRE Denial.
Regarding the studies mentioned by the commenter,17 we do not believe that these studies
provide evidence for imperfect RIN cost passthrough at terminals where obligated parties sell
fuel. Two of the studies (Lade & Bushnell, 2019 and Li and Stock, 2019) focus on "RIN subsidy
pass through" (what in the SRE Denial, EPA calls the RIN "discount") to E85 at retail stations.
EPA does not hold that the RIN discount fully passes through in the prices offered to consumers
at retail stations marketing E85 as EPA previously acknowledged.18 To summarize briefly, EPA
previously found that retail stations selling E85 are rarely in direct competition with other
stations selling E85 and that retail stations likely seek to recover the cost of installing the E85
pumps by marking up the price of E85. Importantly for EPA's analysis here, obligated parties
primarily realize the impacts of RIN cost passthrough and the RIN discount when they sell fuel
at wholesale terminals. Unlike retail stations, most wholesale terminals have a significant
number of position holders all selling fuel in competition with each other from the same
terminals. In that highly competitive setting with posted product prices, the RIN discount is far
more likely to pass through in the prices wholesale fuel sellers receive. Lade in particular notes
that "our finding that pass-through is high in contested markets implies that RIN pass-through at
wholesale terminals is also high in these markets." (emphasis added)
Because the RIN is separated when the fuel is blended and sold as E85 at the wholesale terminal,
it is the wholesale RIN discount that determines the cost for blenders to acquire RINs. The fact
that individual retail stations may markup the discounted E85 wholesale price at retail has no
impact on the cost to obligated parties to acquire RINs as the RIN has been separated and its
17	Gabriel E. Lade & James Bushnell, "Fuel Subsidy Pass-Through and Market Structure: Evidence from the
Renewable Fuel Standard", 6 JAERE 563 (March 22, 2019), available at https://doi.org/10.7910/DVN/AX4LQY:
Jing Li & James H. Stock, "Cost pass-through to higher ethanol blends at the pump: Evidence from Minnesota gas
station data," 93 J. Env. Econ. & Mgmt 1 (2019) available at https://doi.org/.1.0. .1.016/i.ieem.2018.08.003: Jesse
Burkhardt, "The impact of the Renewable Fuel Standard on US Oil refineries," 130 Energy Policy 429 (2019)
available at https://doi.Org/10.1016/i.enpoL2019.03.058: Sebastien Pouliot, Aaron Smith, & James H. Stock, "RIN
Pass-Through at Gasoline Terminals," February 22, 2017, available at
https://scholar.hamird.edu/stock/publications/rin~pass~through~gasoline~terminals.
18	"Denial of Petitions for Rulemaking to Change the RFS Point of Obligation," EPA-420-R-17-008 at 50-51,
November 2017.
B-36

-------
value (discount) realized prior to the sale of the fuel at the retail stations. Hence, these two
studies do not contradict EPA's conclusion that parties that acquire RINs by blending renewable
fuel must discount the renewable fuel by the value of the RIN.
EPA finds the 2019 Burkhardt paper cited in the comments to be largely consistent and
supportive of the conclusions EPA has reached in this action. Notably, the study finds that the
"RIN tax obligation were fully passed through to wholesale gasoline and diesel prices on
average", "that rack level pass-through is generally complete with the largest exception being
firms on the East Coast", and that "full pass-through of RIN costs to nationwide output prices on
average, and no statistical difference between pass-through rates for large and small refineries.
These two findings suggest that exempt refineries that do not bear the burden of the RFS tax
obligation, but enjoy increased output prices, may incur substantial benefits from the policy."19
(emphasis added).
The commenters presumably cite the 2019 Burkhardt paper because of its conclusion that RIN
prices may not pass through in firms operating on the East Coast. The Burkhardt paper itself
(citing Pouliot et al., 2017) suggests the reason for the result on the East Coast:
First, Florida is not on the petroleum pipeline network and second, Atlanta requires a
specific blend of low-sulfur gasoline. These unique properties could lead to more
volatility in the price of blended gasoline, which would lead to lower pass-through rates.
Consistent with the second explanation, I do not find statistically significantly lower
pass-through rates in the ULSD and jet fuel markets in PADD 1.
These explanations from Burkhardt are important for several reasons. First, small refineries
consistently argue that it is diesel fuel in particular where they are unable to pass through the cost
of RVO compliance. Here Burkhardt finds that not only are those costs passed through in diesel
fuel prices on average nationally, but also when evaluated across different geographic regions
and company sizes (i.e., small refineries versus large refineries). Second, to the degree that the
East Coast analysis is skewed due to the unusual market conditions in Florida and Atlanta, those
are two markets primarily served by large refineries (the Colonial Pipeline in the case of Atlanta
and fuel tankers serviced from the major Gulf Coast refineries in the case of Florida). Any
inability to pass through the RVO compliance costs into those markets is unlikely to negatively
impact small refineries. That said, we believe it is more likely that the unusual East Coast market
conditions simply result in more scatter in the data, making it harder to differentiate the impact
of the D6 RIN in particular in those markets during the time period analyzed in the Burkhardt
paper.
The Burkhardt paper analyzed data in the years from 2012-2014 when the total renewable
volume percentage standards ranged from 9.19% to 9.74%. When evaluating E10 prices in
particular during this time period the degree of RIN cost and RIN discount pass through is
particularly hard to measure because the two factors nearly fully offset each other in the price of
19 Burkhardt, Jesse "The impact of the Renewable Fuel Standard on US oil refineries", Energy Policy Volume 130,
July 2019, Pages 429-437
B-37

-------
E10. As described in Section IV.D, EPA expects the price of E10 to be largely determined by the
following equations:
E10 Price = Gasoline Blendstock Price * 90% + (Ethanol Price - D6 RIN Price) * 10%
Gasoline Blendstock Price = Gasoline Price with no RFS Obligation + RIN Costs
RIN Costs = RVO% * Weighted RIN Prices (D6, D4, D3)
Combining the two equations then we can see that the E10 Price would be expected to change
with RIN prices in the following way.
Change in E10 Price = 0.9*9.74%*(Weighted RIN Price) - 0.1 *D6 RIN Price
= 0.0873 *(Weighted RIN Price) -0.1 *D6 RIN Price
Because the weighted RIN price reflects not only the D6 RIN but also the more expensive D4
and D3 RINs, the weighted RIN price is slightly higher than the D6 RIN price such that, in net,
the increase in the gasoline blendstock price due to the 9.74% RVO (the RIN cost passthrough)
is almost exactly offset by the 10% RIN discount from the D6 RIN. In net then, there is almost
zero change in E10 prices when evaluated during this time period. With a near-zero change in
E10 prices with changing RIN prices, it is exceptionally difficult to estimate the impact,
especially in a market with more volatile pricing due to the Florida and Atlanta markets.
Finally, EPA has considered the Pouliot et al. 2017 study that found incomplete RIN passthrough
in PADD 1 and PADD 5. In reviewing this study EPA identified several concerns with the
methodology. First, the study does not appear to account for changes in blending margins over
time. Instead, it appears to attribute any change in the posted price of blended fuels to changes in
the rate of RIN passthrough. Second, and perhaps more importantly, the study does not use
renewable fuel prices actually available at the terminals studied. Instead, the study uses ethanol
prices at the nearest location where spot prices are posted. This is a problem, as the cost of
transporting ethanol even relatively short distances by truck can be significant. To explore the
issues raised in the Pouliot et al. 2017 study further, EPA attempted to use the methods in this
paper, but to add estimated transportation costs for ethanol (and biodiesel) from the markets with
posted prices to the terminals being studied. EPA contracted with Stillwater Associates to
perform this analysis.20 Stillwater ultimately concluded that such an analysis was not feasible for
a variety of reasons, including the unavailability of and inconsistency with the source data,
concerns over extreme price postings and the reliability of the posted prices, and timing
differences between when prices for petroleum and renewable fuels are posted at a major hub
and when these fuels are available at a local market.21 EPA has therefore concluded that the
Pouliot et al. 2017 study provides insufficient evidence to disprove our conclusions on RIN
passthrough.
20	Economic Analysis of Fuel Blending, prepared for the Environmental Protection Agency by Stillwater Associates
LLC, February 9, 2022, pp. 6-7.
21	Id.
B-38

-------
Comment:
An important issue in these data is the treatment of weekend fuel sales. Knittel et al. (2017)
excludes weekends, and RIN price quotes are not available for weekends and major holidays.
Refining and fuel supply is a 24/7/365 industry, which is borne out in the data, with about 28
percent of all transactions logged on Saturday and Sunday.
This has serious implications for the assumption of ratable compliance asserted strongly by EPA.
Dr. Fitzgerald performed three, layered sets of results using the 2019-2021 data. As a control, he
initially excluded weekend fuel sales and analyzed weekday-only sales in the three fuel spreads
across the entire data set, which is the same "methodology used by Knittel et al. (2017) for an
earlier period and endorsed by EPA."
Dr. Fitzgerald then analyzed "each of [the] three fuel spreads using the data that include
weekends. The results indicate that there is less evidence of pass-through when weekend
transactions are included. Notably, the results . . . suggest that the degree of pass-through in the
gasoline market is significantly different" from the results when weekends were excluded. This
"draws into question the relevance of ratable compliance assumptions because trading
arrangements for RINs are substantially less liquid on weekends, and waiting until the next
weekday may expose obligated parties to price risk inherent to compliance. EPA expressly
ignores such a pathway as being 'caused by RFS.'"
Response:
To estimate the rate of RIN cost passthrough in the prices of fuels sold over the weekend, Dr.
Fitzgerald created surrogate RIN prices for Saturdays and Sundays using "a linear interpolation
of missing RIN prices" explaining further that "the difference between Friday and Monday
quotes is split between Saturday and Sunday for a regular weekend." EPA understands this to
mean that if Monday's RIN price was three cents higher than Friday's then Saturday's RIN price
would be estimated to be one cent higher and Sunday's two cents higher such that Monday's
increase occurred evenly over this three-day period. Similarly, if Monday's price was three cents
lower, then Saturday's price would be estimated at one cent lower and Sunday's price at two
cents lower.
Not surprisingly, Dr. Fitzgerald's analysis showed no significant correlation between the
estimated "increase" or "decrease" in RIN prices on Saturday or Sunday because no such RIN
prices exist. Market participants on Saturday or Sunday do not know if RIN prices will rise or
fall on Monday when compared to Friday and so cannot react to what hasn't yet occurred. The
commenter and Dr. Fitzgerald both suggest that this is an issue given EPA's expectation that
refineries will acquire RINs ratably through the year consistent with their production and sale of
refined products. We do not see the lack of pricing information for RINs on Saturday or Sunday
as fundamentally problematic for refineries wishing to acquire RINs ratably with their fuel
production and sales. Such refineries can buy a volume of RINs at Friday's RIN price but at a
volume that reflects Friday, Saturday, and Sunday's sales volumes as Friday's RIN price
information is the information that the market has when it finds the appropriate fuel pricing on
Saturday and Sunday. A similar strategy can be applied to holidays.
B-39

-------
Comment:
EPA's first claim is that the RFS compliance costs are the same for all obligated parties, and thus
no party bears RFS compliance costs that are disproportionate relative to others' costs. We
conclude that this claim is implausible. One economic rationale behind a tradable permit
program, such as RFS, is to achieve the lowest possible compliance cost market-wide. The
economic logic for tradable permit programs is based on the assumption that the costs for each
firm to comply with a regulation differ across firms. Economic theory suggests that this least-
cost outcome will occur at the point at which the marginal compliance cost of each firm is equal
to the tradable credit price (i.e., Renewable Identification Numbers, or RIN price in this context).
However, equalizing marginal compliance costs across firms does not imply that the average
compliance cost per unit of output is the same for all individual firms.
Response:
The primary reason that the LSU study cites in making the claim that compliance costs are not
the same for all obligated parties is the difference in the cost of production of gasoline and diesel
fuel between refineries, due to differences in geography, fuel quality regulations, crude oil costs,
refinery configuration, etc. EPA recognizes that there are significant differences among
refineries, and that these differences affect the cost of production of the petroleum fuels they
produce. However, this does not refute EPA's claims that all obligated parties have the same cost
of RFS compliance. RINs, which obligated parties need to demonstrate compliance, are
generated by renewable fuel producers. RINs are generally separated from renewable fuel by
blenders when renewable fuel is blended with petroleum fuel. Because fuel blenders (whether
they are obligated parties or not) are the source of RINs, the important factor to consider when
evaluating the likelihood that all obligated parties have the same compliance costs is whether the
cost structure for fuel blenders is similar across the industry. Unlike refiners, fuel blender's cost
structure varies very little across the country. The process and cost for blending fuels, whether at
a terminal or refinery rack, is a fairly uniform process.
Further, the cost of obtaining a RIN is not simply the cost of blending renewable fuel with
petroleum fuel, but also the discount the blender must offer on the blended fuel to remain
competitive. Because RIN prices are uniform across the nation, and further because fuel blending
is a competitive market, fuel blenders must discount their fuel blends by the entire value of the
RIN to remain competitive.22 Thus, the cost of acquiring RINs for blenders, whether or not they
are obligated parties, is the same for all parties even though the cost of petroleum fuels and
renewable fuels varies across the U.S. If the cost of acquiring RINs is the same for all parties it
follows that the cost of acquiring RINs will be the same for all obligated parties. In this case the
purpose of a tradable credit program is not necessarily to allow parties with lower blending costs
to blend excess renewable fuels and provide credits to parties with higher blending costs, rather it
is to allow parties already in the business of blending renewable fuels to continue in that business
rather than forcing all refiners to become renewable fuel blenders to meet their RFS obligations.
22 Independently of the RFS program, blenders also charge a blending fee to recover other costs (e.g., capital costs
and operational expenses) and to provide a return on their investments.
B-40

-------
Comment:
Finally, even if there is 100% pass-through of all RIN prices to final product prices, consumers
will respond to the increased price by decreasing the quantity demanded for fuels. This reduction
in demand will negatively impact the refining sector. Both the effect of equilibrium price and
quantity of fuels sold should be considered when assessing the effects of the policy.
Throughout the Proposed Denial EPA focuses exclusively on the impact of RIN costs and other
costs of compliance associated with the RFS Program (such as building biofuel blending
infrastructure). This narrow view arbitrarily overlooks other serious consequences of the RFS
Program. In particular, regardless of whether there is complete RIN cost passthrough, small
refineries bear the burden of decreased demand for their products because of the RFS Program.
Every gallon of biofuel mandated by the RFS on an annual basis is a gallon of gasoline or diesel
that refineries are no longer able to produce. This decrease in product volume is not distributed
equally across refineries. Rather, the highest cost producers see the greatest demand loss, and
small refineries are very often the highest cost producers in their markets. Additionally, even if
the demand reductions were spread evenly across refineries, small refineries generally have
higher fixed costs per gallon and lost volumes can therefore be more impactful on their margins.
EPA does not acknowledge either of these realities in the Proposed Denial.
Response:
As the commenters describe, microeconomic theory states that an increase in prices should lead
to a reduction in demand. Further as one commenter more directly notes, the RFS program itself
by displacing demand for petroleum-based fuels with renewable fuels further reduces the
demand for the gasoline and diesel fuel that refineries produce. The commenters both argue then
that this reduced demand for petroleum-based fuels is itself a form of hardship that EPA should
consider in determining if DEH exists for any particular small refinery.
EPA agrees that microeconomic theory should be reflected in this market and that a reduction in
demand should be expected both due to the marginally higher cost of renewable fuels and further
due to the direct substitution of renewable fuels for petroleum fuels under the RFS program
mandates. However, we do not believe these generalized market outcomes are the kinds of
direct, individual-refinery impacts that Congress intended the SRE provisions to address, and
they do not disproportionately impact small refineries. First, the very purpose of the RFS
program is to displace petroleum fuel with renewable fuel through the RFS mandates set by
Congress. EPA thinks it is unlikely that Congress would then intend to waive the very mandate it
set simply because that mandate was having its intended effect. Instead, we believe the SRE
provisions are intended to address the circumstances where the individual refinery cost of
compliance is the source of the "disproportionate economic hardship," since the statutory
exemption provision refers to DEH "due to compliance" with the RFS program. That cost of
compliance is the cost to acquire the RINs necessary to demonstrate that a refinery has met its
RVOs. As detailed extensively in the SRE Denial, those costs are the same for all refineries and
are passed through to consumers in the prices of gasoline and diesel fuel.
B-41

-------
2. Small Refinery Coalition
Comment:
The data upon which EPA rely are from the 2013 through 2016 time period, before the blend
wall was reached, RVOs exceeded 10.5%, and RIN prices rose to historic highs. Entirely missing
from EPA's analysis are data from the 2017 through 2021 time period. It is arbitrary, capricious,
and methodologically unsound for EPA to rely on pre-2017 data to purportedly determine that all
RIN costs were fully passed through by all parties (including small refineries) during the years of
2019 and 2020.
Response:
There are several factual errors associated with this comment. First, the E10 blendwall was
reached in 2013, which is when D6 RIN prices rose. Second, this comment fails to acknowledge
the EPA analysis provided in Section IV.D.2.d, which provides a RIN market analysis using
more recent data than was available in 2015. Finally, this comment presumes that the structure
and operation of the RFS program is somehow different from one year to the next or as a
function of the RIN prices. Since there have been no meaningful changes to the structure of the
RFS program (or its RIN system) since 2010, there is no reason to believe that this would be the
case and the commenter provided no such basis.
Comment:
None of EPA's studies analyze RIN cost pass-through for diesel fuel. Despite the fact that, as
EPA well knows, small refineries produce disproportionately more diesel than their larger
competitors.
Response:
The commenter is factually incorrect. As discussed in Section IV.D.2.d, as well as in past
analyses supporting evaluating RIN cost passthrough,23 EPA evaluated distillate markets as well.
Furthermore, the form and structure of the RFS program, the RIN system, and compliance
requirements are identical for gasoline and diesel fuel. There is no reason to believe that they
would function differently, and the commenter has not provided any evidence to the contrary. In
addition to EPA, a number of other studies also evaluated RIN cost passthrough in diesel fuel
(e.g., Burkhardt 2019).
Comment:
EPA did not seem to consider that ethanol is often below the price of gasoline. One commenter's
analysis showed that over the past four years, ethanol was discounted below gasoline nearly 60%
of the time. There is no meaningful insight or change in RIN prices on those days. If EPA's
assertion was correct, that the price of the RIN reduces the price of ethanol to meet the "market
23 "Denial of Petitions for Rulemaking to Change the RFS Point of Obligation," EPA-420-R-17-008, November
2017.
B-42

-------
demand" this would imply that if ethanol was below gasoline, Ethanol RINs would quickly move
to 0. Clearly, this is not occurring in the markets. Additionally, a refiner blending ethanol priced
below gasoline would have a RIN cost of zero, while a refiner who had to purchase RINs for
compliance would have a non-zero compliance cost, equal to the D6 RIN.
Response:
As discussed in Section IV.D.2.d, including the examples provided, EPA has in fact considered
the impact of ethanol's cost relative to gasoline in its evaluation. More importantly, the
commenter fails to understand how the RFS program and RIN system function. The commenter
is mistaken that when ethanol is cheaper than gasoline, D6 RIN prices should be near zero. This
should have been, and in fact was the case prior to 2013, before the RFS standards reached the
E10 blendwall. However, the cost of blending ethanol at concentrations above 10% is far greater
than when blending at 10% due to octane value and RVP control costs. Thus, when the RFS
standards require more conventional renewable fuel to be blended than can be met with E10, the
D6 RIN price rises to the point necessary to incentivize the next least expensive source of
renewable fuels use. While this in some cases has meant increasing the use of ethanol through
higher-level blends such as El 5 and E85, in reality it has typically meant that biodiesel and
renewable diesel volumes have increased instead, which is why D6 RIN prices since 2013 have
tended to track with D4 RIN prices. The elevated D6 RIN prices then lower the effective cost of
blending ethanol even as E10, and as explained in Section IV.D.2.d, this value is reflected in the
market pricing of E10.
Comment:
For as long as DOE has been applying the scoring matrix, and EPA has been evaluating petitions
for small refinery hardship relief, DOE has made clear that there was insufficient information to
score metric 2.d, which measures whether a small refinery's RVO is a net cost or a net revenue.
While this metric was not scored in the 2011 Study because of an alleged "lack of consistency"
among the responders to the DOE small refiner survey, DOE expressly noted that "depending
upon the business model of the small refiner, complying with their RVO can either be a net cost
if they purchase all of their RINs or can generate revenue should they be able to actively trade
RINs in the open marketplace."
Now, EPA claims DOE never "assess[ed] in [the 2011 DOE Study] whether their assumptions
that refiners bear different costs for RINs and that they may not be able to pass these costs onto
consumers in the marketplace actually occurred." That is not true. DOE understood that parties
would experience RFS costs differently—that is why DOE included metric 2.d ("RINs net
revenue or cost") in the scoring matrix. It particularly understood that RIN prices would become
untethered from the price of blending after the E10 blend wall was reached—a concept that has
grounded our understanding of the RFS for nearly 10 years but is not mentioned once in the
Proposed Denial. As stated above, DOE predicted that as the RFS mandate increases, RIN-short
parties "will need to purchase RINs and could suffer significant economic hardship."
B-43

-------
Response:
As explained in Section IV.C, the 2011 DOE Study did not consider that blending refineries
would have to discount blended fuel by the price of the RIN; therefore, the projections
envisioned by the 2011 DOE Study have not occurred in practice. Rather, as the 2009 DOE
Study anticipated, the competitive market forces have resulted in the same cost of compliance
whether that cost comes through the purchasing of RINs on the open market or through the
discounting of the price for blended fuel sold by blenders. Moreover, neither the 2009 DOE
Study nor the 2011 DOE Study anticipated the even more significant finding that, without regard
to how refineries experience their RFS compliance costs, the RIN cost passthrough phenomenon
applies—refineries pass those higher costs through to their customers in higher prices for the
refined products they sell.
As part of EPA's evaluation of SRE petitions, EPA requested that petitioning small refineries
provide data on some RFS compliance costs. EPA acknowledges that all of the petitioning
refineries provided estimates of their RIN acquisition costs based upon a standardized
spreadsheet that EPA created in 2013 and shared with petitioners. This spreadsheet is intended to
provide summary information on the petitioner's annual total differential cost of purchased
biofuel (relative to gasoline and diesel), and annual cost of purchased RINs needed for
compliance. DOE has continually declined to evaluate this information as part of its scoring of
metric 2.d RINs Net Revenue or Cost, explaining in 2011 that a "lack of consistency" among the
petitioners made it impossible to score. More recently, DOE has decided not to score this metric
explaining that there is no information available to compare a refinery's RIN cost/revenue with an
industry average obtained from study of refineries' data (rather than a study of national price data) to
determine DEH.
More importantly here, EPA now recognizes that the RIN acquisition costs in this spreadsheet
lack two very important elements that have not been reported or accounted for in the submittals.
First, the cost to acquire RINs for blending is calculated based on the price differential between
the petroleum fuel into which it will be blended and the cost of the renewable fuel. This
simplified approach assumes that the renewable fuel is a direct substitute for the petroleum fuel
and therefore the cost difference should directly reflect the value of the RIN. However, for
ethanol as an example, the lower energy content of ethanol means that ethanol needs to be
discounted about 30% below the price of gasoline for a consumer to choose to buy it for use in a
flex-fuel vehicle (i.e., as E85). Hence, ethanol needs to be discounted below the cost of gasoline.
By not accounting for this additional discount (i.e., by not using the blended fuel price as an
input to the calculation), the spreadsheet underestimates the acquisition cost to blenders to
acquire RINs. In fact, in many cases it suggests that blenders are getting D6 RINs at a negative
cost since ethanol is cheaper than gasoline in many instances, when in fact the actual RIN
acquisition cost to refineries is most definitely significant. Second, the estimate of RIN
acquisitions does not account for the increased revenue that refineries receive in their product
prices as a result of RIN cost pass-through. In fact, we would argue that accounting for that for
each petitioner would indicate no net cost to small refineries due to compliance with the RFS
program.
B-44

-------
Given the spreadsheet was not designed to capture any information regarding RIN cost
passthrough in the price of the unblended transportation fuel nor to capture the RIN discount, the
spreadsheet is not useful in quantitatively comparing RFS compliance costs, which is why DOE
does not use it in scoring metric 2.d and EPA has chosen not to evaluate those estimates in the
SRE Denial.
B-45

-------
IV Non-CBI Comments on EPA's Technical Analysis
1. EPA's analysis of DEH relies on incomplete or selective information.
Comment:
EPA's Proposed Denial fails to acknowledge and address literature showing variation in RIN
cost passthrough based on location and fuels. Although there is general acceptance that there is,
over time, at least some passthrough of RIN costs in most refined product markets in the U.S.,
there is significant dispute in the literature over the extent of passthrough and the level of
variation between locations, over time, and across fuel types. EPA's Proposed Denial entirely
ignores this information.
Even if one accepts the recent EPA analysis as showing passthrough for the examined ULSD
fuel pairs in New York Harbor and the Gulf Coast, those are just two fuel pairs and two markets.
The Knittel, Meiselman, and Stock studies reviewed passthrough in several regions, including
New York Harbor, Gulf Coast, Chicago, and Los Angeles. The results varied significantly, and
some markets did not produce usable results on passthrough levels. While fuel markets are often
thought of as broad, the dynamics of refined product markets that affect passthrough can vary
significantly across North America. For example, only certain markets currently absorb
significant amounts of higher-ethanol blend gasolines or biodiesels, making entire markets short
RINs and therefore net buyers exposed to potential compliance burdens. And as the Burkholder
Report noted, smaller markets may also involve less competition in supply and demand for
certain products, and fully competitive markets are often cited as a prerequisite for full
passthrough.
Response:
EPA has considered a number of other studies, most importantly, all of the studies provided by
the small refineries with their SRE petitions, supplemental submissions, and comments on the
Proposed Denial. While as the commenter notes, some studies fail to find full RIN cost
passthrough in all markets or for all time periods, as described in Section B.III. 1, we find that on
balance these studies provide more evidence in support of the conclusion that RIN costs are
passed through than evidence to suggest they do not. As detailed in Section B.III. 1, in many
cases where they do not directly show RIN cost pass through, we believe there are other factors
that may obscure such observations in the data.
Comment:
EPA notes that many petitioners have claimed that they believe their regional markets have
different passthrough than the major markets dominated by the large integrated refiners.
Proposed Denial at 27. Recognizing this as a significant challenge to their claim of full RIN
passthrough, EPA tried to address variable passthrough, suggesting that passthrough in major
markets leads to passthrough across all markets as prices equilibrate. The Agency stated:
"Through thousands of decisions made by all the market participants each day, the prices
B-46

-------
between the markets equilibrate to the same level, offset by the transportation costs between the
markets." Id. at 30. But EPA offers no proof of this equilibrium.
Response:
As an initial matter, EPA has extended its explanation for the interrelation between linked fuel
markets in Section IV.D.2, noting importantly that many small refineries acting in local markets
directly index the prices for the products they sell to the major coastal markets and posted prices
from those markets. The terms of the contracts ensure that the local market price indexes (rises
and falls) with those major markets. Further, because these local markets face significant
competition, those market prices are the same for all of the market participants. Perhaps most
fundamentally, the reason that EPA believes the markets equilibrate is because there are so many
market actors whose very purpose is to perform arbitrage between these markets (i.e., to exploit
minor price differences and in doing so close those differences). As but one example, fuel
jobbers (operators of fuel tanker trucks) compare the prices at multiple fuel terminals and
refinery truck racks within a region to determine which wholesaler offers the best price for the
load of fuel they will deliver to a particular retail station after considering first their own
transportation costs between the terminals and the retail station. Small variations in price are
enough reason for the jobber to choose a terminal slightly further away, increasing demand at
that terminal and dropping demand at the others. Multiple market actors all serving similar
function provide feedback to the wholesalers at the terminals who, through these actors, must
compete not only with other operators at the same terminal, but also with operators at other
terminals in the region. If as the commenter suggests, there are markets that do not pass through
their costs, those markets will quickly see increased demand while the other markets see a drop
in demand. In other words, competition in the market will act to close that price anomaly and
bring the whole system into equilibrium.
Further, if there are any fuel markets within the U.S. where such competitive dynamics do not
occur (i.e., where the local market is monopolistic or oligopolistic), then the sellers in that
market do have pricing power and can certainly be expected to pass on all of their costs and more
to wholesale consumers. We do not believe the commenter is suggesting that they operate in
such a market.
Comment:
On February 14, 2022, the Proceedings of the National Academy of Sciences published a study
titled Environmental Outcomes of the US Renewable Fuel Standard. This study shows that the
RFS has substantially increased production of corn, and that the increased production and use of
corn for ethanol have significant environmental effects. Among other things, the use of ethanol
for blending with transportation fuel results in more GHGs than the use of transportation fuel
alone.
This conclusion has stark implications for EPA's management of the RFS. One of the goals of
the RFS is the reduction of GHG emissions. And its most recent proposed rule for the 2021 and
2022 RVOs states that the "proposed rule has the potential to reduce GHG emissions." EPA
B-47

-------
cannot adequately meet its obligations or make such statements without considering this new
study.
It would be arbitrary and capricious for EPA to finalize the Proposed Denial without duly
considering this new study. EPA's Proposed Denial is a blanket action that would deny all
pending petitions for small refinery exemptions. The net effect of such a denial would mean that
more ethanol—as the most-frequently-used renewable fuel—would be blended into
transportation fuel. As the study shows, increased use and production of ethanol would likely
result in increased harm to the environment through the higher number of GHG emissions. Thus,
EPA must demonstrate that it has considered this issue before it takes actions that would increase
the amount of ethanol use.
Response:
The comment period on EPA's proposed denial ended on February 7, 2022, one week before this
study was completed and several weeks before the comment was submitted to EPA on March 22,
2022. While the comment may be relevant to the RFS program generally, it is not directly
applicable to EPA's evaluation of small refinery exemptions under the relevant CAA provision
in determining whether the small refinery experiences DEH caused by RFS compliance DEH
and is therefore not addressed in this response to comments.
B-48

-------
2. EPA concedes that there are circumstances under which obligated parties cannot
passthrough their RIN costs.
Comment:
EPA's conclusion that exemption from the RFS standards results in "windfall" profits for small
refineries is incorrect. EPA's implementation of the program has ensured that, even when small
refineries receive hardship relief, they still suffer economic losses through the RFS program. 1)
When EPA issues SREs well past the statutory deadline, small refineries are left scrambling in
the RIN market, buying RINs from and/or selling RINs to unobligated parties that have no
reason to participate in the marketplace other than to profit off of obligated parties. 2) When
small refineries are forced to wait months or, in this case, years past the statutory deadline to
know whether they received hardship relief, EPA's delay affects their ability to make decisions
that could advantage their business.
Response:
As described in Section IV.D.2 and in responses to other comments herein, small refineries were
paid for the cost of RFS compliance through the pricing of the gasoline and diesel in the
marketplace, as those market prices reflected the cost of RINs. Accordingly, any revenue from
RIN sales after an exemption is granted is gratuitous to small refineries' compliance costs.
Nevertheless, EPA is aware of the extenuating circumstances that have resulted in this long
delayed final action, and EPA has already taken reasonable steps to ease the impact of this delay
on small refineries by twice extending the RFS compliance dates for small refineries in 2019,
and all obligated parties for later years. Ultimately, all obligated parties, including small
refineries, have compliance obligations under the RFS program. Those compliance obligations
exist for small refineries until such time as an exemption is granted. We have previously told
small refineries that they should not plan for an exemption but should instead plan to comply
with their obligations, and many have done so.
Comment:
EPA's theory of RIN cost passthrough directly contradicts its longstanding approach to
administering the RFS program and expectation regarding compliance. EPA is now punishing
small refineries for behavior that is not only legal, but that EPA previously approved. Prior to the
Proposed Denial, EPA never required obligated parties to purchase RINs at any particular time.
In fact, EPA considered but abandoned the notion of requiring quarterly RIN retirement
deadlines. Further, EPA has made clear in the past that obligated parties are not required to
obtain RINs at the same time that they produce or import fuel but may, if they choose, simply
purchase the required number of RINs by the end of the compliance period, once their annual
production is known. The uncertainty surrounding EPA's implementation of the RFS Program
for several years now has made it difficult for cash-strapped obligated parties, like small
refineries, to justify spending millions on RINs before they even know their RVO or whether
they will receive an SRE. Without question, EPA's delays in implementing the RFS Program
have caused compliance planning uncertainty. To reach its conclusions in the Proposed Denial,
EPA is taking a fundamentally inconsistent position as to how compliance planning should work.
B-49

-------
In prior statements, EPA claims the RIN market program is designed to provide obligated parties
sufficient time to plan for compliance. Implicit is the recognition that parties do not buy RINs
ratably (and should not have to). For example, in EPA's proposed extension of the 2019-2021
compliance deadlines, the Agency acknowledges "the importance to obligated parties of
planning their compliance for a given calendar year by understanding their obligations for the
years before and after." 86 Fed. Reg. 67,419, 67,422 (Nov. 26, 2021).
Response:
As described in Section IV.D.2.d.i, all obligated parties have the opportunity to match their costs
by buying RINs on a ratable basis and are responsible for making decisions when to buy RINs
and how many RINs to buy at any given time. However, purchasing RINs ratably is not a
requirement, but a compliance flexibility that allows obligated parties to comply with their RFS
obligations without forcing them to blend renewable fuels with their petroleum based
transportation fuels. Indeed, in the absence of the RIN credit program, refineries would have to
directly ensure renewable fuel blending. In such a program design, a small refinery could, under
the annual compliance provisions, choose to delay any renewable fuel blending until the last
month of the year and then attempt to sell exclusively renewable fuel in the last month of the
year at a volume to meet the obligation it accrued through the preceding 11 months. Such an
approach would almost certainly lead to a much higher cost of compliance than would have
occurred had the small refinery worked to demonstrate compliance on an ongoing basis each
month through the year. As alleged by small refinery commenters, EPA would then be
compelled to provide hardship relief due to the higher cost of RFS compliance for the small
refineries that chose such a compliance mechanism. Such an approach, where the business
decisions of the individual companies are made within the regulations but contrary to the purpose
of the program, does not constitute DEH caused by the cost of compliance with the RFS
program, and therefore cannot be a basis for hardship relief. Otherwise, all small refineries could
simply choose such an impossible compliance approach, and then, having made this choice, be
assured of relief from the RFS obligations. Similarly, individual business decisions made by an
obligated party not to ratably accrue RINs as their obligation accrues, but instead to either
purchase RINs in advance or delay RIN purchases until a later date, are business choices that
companies may lawfully make.
Comment:
EPA's theory that all RINs are passed through rests on an assumption divorced from the reality
of the transportation fuels market. To present a simplified picture of how small refineries can
achieve compliance and recover their costs, the Proposed Denial contains zero discussion of how
RIN trading actually works. Instead, EPA claims that "individual business decisions made by an
obligated party not to ratably accrue RINs as the obligation accrues, but instead to either
purchase RINs in advance or delay RIN purchases until a later date, are speculation in the RIN
market, a business activity not required to comply with the RFS program." To characterize small
refinery behavior as "speculation" is disingenuous. EPA's statement reflects a naive
understanding of business behavior. Speculation is defined as "investment in stocks, property, or
other ventures in the hope of gain but with the risk of loss." Small refineries are not entering the
RIN market in an attempt to make profit. They are RIN-short obligated parties required to show
B-50

-------
compliance with the RFS Program and, because they are dependent on the RIN market to do so,
they have no choice but to purchase RINs in the marketplace. If small refineries could avoid the
RIN market altogether, they certainly would. Put simply, small refineries are making decisions
about regulatory compliance, not speculating.
Response:
As discussed in Section IV.D.2.d.i and above in this Section B.IV.2, by purchasing RINs ratably,
all obligated parties have the ability to match their RIN costs with the price they receive when
they sell their fuel (i.e., to pass through their RIN costs). Alternatively, refineries can try to time
their purchases in the RIN market, which may result in greater or lesser RIN costs. Either way,
an obligated party's choices about when to procure RINs represent individual business decisions
rather than RFS compliance requirements. Thus, the costs associated with timing purchases in
the RIN market cannot be considered to represent DEH caused by compliance with the RFS
program.
Comment:
In contrast to small refineries, other parties do speculate in the RIN market, because EPA allows
them to do so. The CAA directed EPA to create a credit trading program in which the credits, or
RINs, could be generated by parties that "over complied" and sold only to parties that needed
them for compliance. Instead, EPA created a program in which any person may participate,
generating credits for blending at any level they choose and selling RINs to anyone for any
purpose. As a result, the RIN market has been captured and is controlled by large integrated
refineries that generate excess RINs, large retailers (who control their own blending but are not
obligated parties), and traders, all of whom are seeking to make a profit in the market.
Response:
As an initial matter, the commenter has failed to provide evidence that allowing any person to
participate in the RIN market has caused harm to small refineries. EPA created the RIN market
to ensure competition and liquidity. Furthermore, this assertion was made previously in the
context of a previous EPA rulemaking, and EPA has since imposed additional reporting and
record keeping requirements to determine whether any obligated party was holding RINs
sufficient to manipulate the RIN market.24 To date, we have not had any party report that they
have exceeded the RIN holding thresholds under the RFS regulations. EPA addresses
commenters' assertion that large retailers and traders disproportionately profit from the RIN
market in Sections IV.D.2.C and d.
24 84 FR 26980 (June 10, 2019).
B-51

-------
3. Small refineries are unable to recover the costs of RFS compliance in the prices of
the fuels they sell.
Comment:
EPA's analysis does not account for the additional cost small refineries bear to buy Q-RINs;
RINs that have been verified by an independent third-party auditor operating under an EPA
approved quality assurance plan or QAP. EPA ignores the higher cost of Q-RINs, which small
refineries would likely be forced to purchase, and rampant fraud in the RIN market.
Response:
This assertion is misleading for several reasons. First, Q-RINs25 merely represent a cost to
renewable fuel producers that is passed through in the cost of RINs in the market, which is then
passed through to consumers as described in Section IV.D.2.d. Furthermore, all obligated parties
have the option to purchase Q-RINs, and the risks of fraud in the RIN market are shared by all
participants in the RFS program. There is no reason to believe that small refineries would have
any greater need to purchase Q-RINs than any other obligated party, and certainly would not be
"forced" to.
Second, the commenter overstates the higher cost of Q-RINs. The majority of Q-RINs are coded
as D3 RINs for cellulosic biofuel, of which roughly 98-99% are Q-RINs. Cellulosic biofuel RINs
are typically the most expensive D-code RIN. When one pools together the costs of all RINs
across the different D-codes, the cost of the Q-RINs will appear on average more expensive.
However, the cost comparison is separated by D-code, for example within the D6 code, the cost
difference between a RIN and a Q-RIN is generally a few cents and the Q-RIN is not always
more expensive. EPA used publicly available data26 to perform this cost comparison and
summarized the data in the following table.
25	Q-RINs are RINs that have been verified by an independent third-party auditor operating under an EPA approved
quality assurance plan or QAP. They are used to demonstrate the authenticity of RINs generated by renewable fuel
producers.
26	EPA has analyzed data available on EPA's website at https://www.epa.gov/fuels~registration~reporting~and~
compliance~help/rin~trades~and~price~information.
B-52

-------
Row Labels
2017
2018
2019
2020
2021 Grand Total
QD3
2.45
1.88
1.43
1.99
2.65
2.03
Q-RIN
2.47
1.90
1.40
2.01
2.72
2.04
Unverified
2.39
1.79
1.52
1.92
2.50
2.01
HD4
0.72
0.53
0.60
0.93
1.37
0.77
Q-RIN
0.73
0.52
0.60
0.89
1.39
0.76
Unverified
0.71
0.54
0.60
0.97
1.35
0.77
BD5
0.82
0.46
0.47
0.86
1.47
0.82
Q-RIN
0.87
0.47
0.51
0.77
1.40
0.82
Unverified
0.79
0.46
0.45
0.92
1.53
0.82
QD6
0.51
0.26
0.42
0.80
1.23
0.55
Q-RIN
0.58
0.24
0.46
0.79
1.19
0.51
Unverified
0.47
0.27
0.40
0.81
1.23
0.56
EPA found that average prices from 2017 through 2021 for Q-RINs compared to unverified
RINs ranged from 3 0/RIN more for D3 RINs to 5 0/RIN less for D6 RINs, while prices for Q-
RINs and unverified RINs were nearly equal for D4 and D5 RINs. Some refineries provided an
analysis that incorrectly compared the average price for all RIN D-codes combined, rather than
comparing individual RIN D-codes, and which consequently showed larger price differentials
between Q-RINs and unverified RINs.
EPA also notes that small refinery commenters do not say what their additional cost would be, or
even if they are certain that they would purchase Q-RINs at all. Small refinery commenters also
say there is "rampant fraud in the RIN market," but the data they provided shows 470 million
fraudulent RINs identified by EPA over a 10-year period, or 47 million RINs per year on
average. During these 10 years, approximately 16.5 billion RINs were retired each year on
average, which means invalid RINs were 0.3% of the total RINs retired. That hardly constitutes
"rampant fraud" as small refineries are claiming.
Comment:
Additionally, the EPA posits that "demand price for the renewable fuel, which is the price the
market is willing to pay for the renewable fuel as a transportation fuel." This is a misconception
and doesn't recognize that E10 is now the standard fuel in the industry. CBOB like ethanol is an
intermediate product in the liquid transportation fuel value chain. Ethanol prices trade in several
markets, including in a very transparent manner on the Chicago Board of Trade. These prices
react to prices of corn, ethanol production, inventories, demand, US, Chinese and South
American crop acreage, yield and production forecasts, weather and many other factors, certainly
independent of RIN prices.
Response:
The commenter appears to be misinterpreting EPA's statement by confusing "renewable fuel"
with "E10." E10 is not the renewable fuel, but rather the fuel blend that results from blending
ethanol with gasoline blendstock (e.g., CBOB). EPA does not otherwise disagree with the
commenter's description of the CBOB and ethanol markets.
B-53

-------
4. EPA's Proposed Denial is based on fundamentally inaccurate assumptions.
Comment:
EPA's theory does not explain the decision made by private companies or identify the
mechanism by which the RIN discount can eliminate all differences in the cost of generating
RINs.
Firms try to maximize profits in part by seeking to minimize their production costs, including the
cost of complying with regulatory requirements. Under the RFS Program, firms have pursued a
variety of different strategies for minimizing their compliance costs. Some have invested
hundreds of millions of dollars to produce biofuel in the belief that they can make money by
increasing revenue or reducing their RFS compliance costs (or both). Others have invested in
pipelines or terminals or other blending infrastructure; some have invested capital to change their
fuel slates, thus enabling them to produce more non-obligated fuel; others have chosen to
increase the amount of fuel they export.
Remarkably, EPA now says that none of these decisions have actually made any difference in
terms of reducing the compliance cost of any refiner relative to the compliance cost of any of its
competitors: "Regardless of the mechanism by which small refineries and other obligated parties
comply with their RFS obligations, the RFS compliance costs are the same for all obligated
parties and thus no party bears RFS compliance costs that are disproportionate relative to others'
costs." Proposed Denial at 1.
Senior executives and Boards of Directors at dozens of refining companies would probably be
surprised to hear this. They believe that their RFS-related decisions, which have resulted in
billions of dollars of expenditures, have given them some economic benefit. EPA insists
otherwise. No matter what these companies do, their RFS compliance costs on a per gallon basis
are all the same.
In the real world, RFS compliance decisions can be understood by the fundamentals of supply
and demand. Refineries have two basic options for meeting their annual RVO obligations. They
can take actions to generate RINs, or they can purchase RINs generated by others. Many
refineries generate as many RINs as they feasibly can and then purchase the additional RINs they
need to meet their annual RFS obligation. Decisions about which actions to take are based on the
price (and expected future price) of RINs. Firms that, because of their circumstances and
ingenuity, can generate RINs at a cost below the market price of RINs will choose to generate
RINs; refineries that do not have this opportunity will purchase them.
RINs are bought and sold in a nationwide competitive market. As with most products, there is an
upward sloping supply curve for RINs. If a refiner would blend biofuels even if the RFS did not
exist, that refiner's cost for the production of RINs is zero. It is generating RINs by doing
something it would have done anyway.
By design, the RFS Program has been increasing the demand for RINs. Higher demand drives
RIN prices higher, which induces more RIN producers into the market—producers whose RIN
B-54

-------
production costs are higher than zero but still below the market price for RINs. According to
economic theory, the market price for RINs will reflect the marginal cost of producing the last
RINs that will sell on the market. Thus, both the price and the total number of RINs sold would
reflect the point at which the upward sloping supply curve intersects the demand curve.
Putting aside for now the RIN discount theory, it would appear that refineries that can generate
RINs at little or no cost but can sell them at the higher market price, can make significant profits
by selling RINs. This is sometimes referred to as "producer surplus" because the market price is
higher than the cost at which they would have been happy to sell their RINs. Because of
producer surplus, some refiners enjoy substantial profits because the money they get from
generating and selling RINs is much more than the cost of producing them. Other refineries,
including some small refineries, must pay enormous amounts to purchase the RINs they need but
cannot produce at a cost below the RIN price. Even so, EPA argues that no refinery is harmed by
the RFS Program because every refinery can pass through 100 percent of its RFS compliance
cost to its customers.
This ignores the fact that for refineries that must pay the market price for a large portion of the
total RINs they need for compliance they are at a significant economic disadvantage compared to
their competitors who are able to generate all the RINs they need at little or no cost. Even if they
are able to passthrough all their costs, they still face DEH under the RFS Program because the
Program confers substantial economic benefits on their competitors.
Under EPA's theory of RIN discount, however, this cannot occur because any profit that any
refinery makes by selling RINs is precisely offset by the amount of the discount that the refinery
must offer in order to sell its fuel. EPA does not provide (or cite to any source that provides) any
data to support this theory. Nor does EPA explain the mechanism by which it works.
To be sure, there is evidence of some RIN discount that reduces the economic advantages that
some refineries would otherwise have compared to others. But nowhere does EPA explain how
this discount can so precisely offset the benefits that some RIN generators have over others
because of their lower costs of production. The market can only discern the marginal cost of the
highest cost RIN producer, as reflected in the RIN price. Thus, the market cannot simply "take
away" the economic advantage enjoyed by relatively lower cost RIN producers, whose
production costs are below the RIN price and cannot be discerned by the market.
EPA concedes that, when it comes to things other than RFS costs, some refineries have
economic advantages over others, for a variety of reasons. The Agency should acknowledge that
there are factors that provide some refineries with advantages over others when it comes to RFS
compliance costs, and that some small refineries experience DEH because of the RFS Program.
Response:
The commenter is correct that a number of companies have made investments and are realizing
returns on those investments to produce renewable fuels, to distribute renewable fuels and to
blend renewable fuels. The distinction EPA makes in accounting for the cost of RFS compliance
is that EPA considers the cost for parties to acquire RINs, not the cost of parties to produce
B-55

-------
renewable fuels. This distinction can be made clear with an extreme example. If a refinery
created a new line of business to produce ethanol from air at zero production costs, the company
could produce ethanol at no cost. Further in the example, the rest of the ethanol industry has a
production cost of $2 per gallon of ethanol, and hence, the market price for that ethanol would be
$2. Lastly for this example, let us assume that RIN prices are $1. The commenter's assessment of
this example would be that this refinery by virtue of its ethanol business has no RFS compliance
costs because it produces renewable fuel at no cost. In EPA's assessment, because the market
price for ethanol is $2, the return on the ethanol that the company gets for its investment (the
same investments the commenter is referring to) is that $2. Whether the company sells the
ethanol into the ethanol market profiting $2 or blends it into E10 and sells it as a blended
product, the company is still profiting the $2. Either way, EPA considers that to be the return on
the ethanol plant investment. In determining this particular refinery's cost to acquire RINs, EPA
would still note that the refinery has to discount the E10 it sells by the value of the RIN, or to sell
its ethanol without the RIN for $1 (i.e., if selling without the RIN, the ethanol has to be sold for
$1 rather than $2), or sell its ethanol with the RIN for $2 and then return to the RIN market to
buy the RIN back for $1. In all three cases, the refinery's cost to acquire the RIN it uses for
compliance with the RFS program is the $1 value of the RIN, whether that $1 value is expressed
in the market price or the RIN discount. In the end, it is the cost for the refinery to acquire the
RIN that determines its RFS cost of compliance.
The same is true for other investments made by parties to blend or distribute conventional or
renewable fuels. Those investments have the potential to earn a return on the investment.27 That
all occurs outside of the cost for parties to acquire RINs for RFS compliance. The RIN may be
providing the demand for ethanol and through it the motivation for the company to invest money
to create a cheaper means to produce ethanol, but in the end the actual RFS compliance costs
become the cost for the company to acquire the RIN itself. Those costs, as described here and
elsewhere, come down to the market price for RINs and the identical market discount for
renewable fuels based on that market price for the RIN.
EPA discusses the data to support RIN discount in Section IV.D.2.d.ii and explains the
mechanisms by which it works in Section IV.D.2.b. Section IV.D.3.f discusses that the cost to
obtain a RIN by blending renewable fuel is not simply the fixed and operating costs for fuel
blending (which are relatively minor), nor is it simply the price difference between renewable
fuel and the petroleum fuel into which it is blended (e.g., the price difference between ethanol
and gasoline or between biodiesel and diesel fuel). Instead, the cost to a blender to obtain a RIN
is the price difference between the volume-weighted cost of the petroleum fuel (e.g., gasoline or
diesel fuel) and the renewable fuel used to produce blended fuel, and the sales price of the
blended fuel (e.g., E10 or B5). The data presented in Section IV.D.2.C demonstrates that the
difference between the cost of the petroleum fuel and the renewable fuel used to produce blended
27 We note that, despite the RFS program requirements, ethanol production has not always been profitable. In many
years the return on investment in ethanol production have been very low. See Irwin, S. "Ethanol Production Profits
in 2021: What a Ride!." farmdoc daily (12): 18, Department of Agricultural and Consumer Economics, University of
Illinois at Urbana-Champaign, February 10, 2022.
B-56

-------
fuel and the sales price of the blended fuel is equal to the market price for the RINs associated
with the blended fuel.28
The finding that there is parity between the cost of obtaining RINs either by blending renewable
fuel or purchasing RINs does not mean that RINs do not provide an incentive for the blending of
renewable fuel. While blending renewable fuel does not result in windfall profits for blenders
(since the revenue from RIN sales is passed through to consumers in a discount on the price for
blended fuel), RIN revenue lowers the effective cost of renewable fuel, allowing blenders to
offer blended fuel containing renewable fuel at lower prices. The examples presented in Section
IV.D.2.C illustrate this point. The incentive for blenders to continue to blend renewable fuel
when there is parity between the cost of obtaining a RIN through blending and the cost to
purchase a RIN is not that the revenue from the sale of the RIN represents a windfall profit, but
rather that the RIN discount allows blended fuel to sell at a lower price relative to unblended fuel
after passing through the revenue of the RIN sales to the consumer.
EPA recognizes that private companies make investments expecting to realize a financial return.
However, just because some obligated parties have chosen to make investments in renewable
fuel production or in pipelines, terminals, or other blending infrastructure, it does not follow that
they have done so in order to reduce their RFS compliance costs. Renewable fuel production can
be, and in the past often has been, a profitable business before considering any impacts of the
RFS program. Similarly, transporting, distributing, and blending both petroleum and renewable
fuels has the potential to return profit to parties that invest in these operations. Many obligated
parties invested in renewable fuel production and fuel distribution well before the RFS program
existed, and many have chosen to divest of these operations after the RFS program was
established. Neither of these actions would make sense if the sole purpose of participating in
these markets was to reduce the cost of RFS compliance.
Comment:
EPA's evaluation of available market data does not support a claim of universal and complete
RIN cost passthrough of RFS compliance costs.
Regarding ULSD and heating oil in New York Harbor, EPA claims that there is "strong
correlation between these data sets" and that "[t]he market price premium for ULSD over that for
heating oil consistently matches the RIN cost (i.e., the cost of purchasing the RINs needed to
meet the RFS obligations." Proposed Denial at 45. However, EPA's analysis only involved
plotting the price spread between these two fuels from 2017-2020 against RIN prices on a time
series graph and a scatter plot, then looking for visual signs of correlation. See Proposed Denial
at 44-46 and Figures IV.D.2.d.i-l and IV.D.2.d.i-2. While these figures suggest that the spread is
correlated with the RIN cost (suggesting some RIN cost passthrough), it is impossible to draw a
definitive conclusion about its extent from a visual inspection alone. Running a simple regression
with the same data relied on by EPA, which included EIA fuel prices and OPIS RIN prices,
shows that the pass-through coefficient is less than one, and far less than one when expanding
the time period back to 2013. Also, it should be noted that observing a relationship on average
28 See SRE Denial Figures IV.D.2.C-2 and 4.
B-57

-------
can still leave room for variability in the relationship, and these variations can lead to different
levels of passthrough for different refiners that are outside of their control.
Response:
As an initial matter, EPA does not assert, nor does it need to establish, "universal and complete
RIN cost passthrough of RFS compliance costs." We recognize that based on these data alone,
definitive conclusions about the degree of RIN cost passthrough in all locations should not be
made. However, the observed correlation between the price spread between these fuels and the
RIN cost strongly suggests RIN cost passthrough, at least in New York Harbor. The way that
fuels are generally priced in the U.S., with local pricing based on the price at a major fuel hub
plus (or minus) transportation costs to or from that hub, strongly suggests that if RIN costs are
passed through in major fuels markets, such as New York Harbor, these costs are passed through
in other markets as well. While the passthrough coefficient is slightly less than one, it is very
close to one (0.94) and likely impacted by observations when the RIN cost was very low. We did
not include data prior to 2017 in our regression analysis because prior to 2017 higher-sulfur
heating oil was sold in many states in the Northeast, and we expect the reported prices reflect this
higher-sulfur heating oil, which is substantially different than ULSD.
Comment:
EPA conducted a similar review for ULSD and jet fuel in the Gulf Coast market. See Proposed
Denial at 45, 47 and Figures IV.D.2.d.i-3 and IV.D.2.d.i-4. Here, EPA notes that "the correlation
between the price difference of ULSD and jet fuel and the RIN cost is not as strong as the
correlation between the price difference of ULSD and heating oil and the RIN cost." Proposed
Denial at 45. EPA admits that this data is less conclusive and only claims there is a "general
relationship" between the spread and RIN costs. Id. at 45. Leaving aside many issues with using
a set of graphs to determine statistical relationship, a claim of full RIN passthrough would
require a finding of a specific relationship, not a general relationship. A general relationship only
suggests that there is some level of passthrough. The specific relationship would be that the price
spread moves 1-to-l with the RIN cost, represented by a coefficient of 1 in a statistical study.
Using the data cited by EPA, a simple regression suggests a coefficient of approximately 0.75.
EPA argues that weaker correlation is expected because of differences in product quality
between ULSD and jet fuel and their differing markets with "distinct supply/demand issues."
EPA admits there is more "noise" in this data, which means there are irregular, possibly random,
variations in the relationship. This noise, which is obvious in EPA's charts, makes it even more
inappropriate to do a simple visual review and claim adequate correlation.
Response:
As with the correlation of the spread between ULSD and heating oil prices, we are not
suggesting that this correlation in isolation demonstrably proves that RIN costs are completely
passed through. Instead, this is one piece of evidence that suggests RIN costs are passed through.
Jet fuel and ULSD are not perfect substitutes, and they have different markets whose demands
can shift semi-independently. Thus, even in a situation with perfect RIN cost passthrough, we
would not expect a one-to-one relationship between the spread between these fuels and the RIN
B-58

-------
cost. Nevertheless, because these fuels have similar properties and relatively similar processing
costs, we would expect there to be a relationship between the price spread between these fuels
and the RIN costs. The observed data demonstrate that this relationship exists and is consistent
with EPA's findings on RIN cost passthrough.
Comment:
EPA also attempted to support its theory of complete and universal RIN cost passthrough by
considering the relationship between RINs and refining margins. Proposed Denial at 30. "EPA
examined the refining margins for three groups of refineries—small refineries, large refineries,
and all refineries—based on available public data (e.g., financial data from publicly traded
companies) and confidential data, including data provided by petitioners. We compared these
refining margins (operating profit per gallon of fuel produced) to the average RIN cost per gallon
(the per gallon cost to acquire the RINs necessary to meet a refinery's RVO)." Id. Again, this
analysis is not convincing. There is simply too much noise in publicly available refining margins
for EPA to draw these conclusions. For example, each refinery has different product slates and
sells into different markets. EPA would need data from many refineries over many time periods
to draw any conclusions. EPA appears to rely on annual data, which is not nearly a large enough
sample to do any statistical analysis. EPA states that it sees no correlation between refining
margins and RIN prices, which would be consistent with full passthrough. However, a visual
inspection of the chart provided by EPA suggests that RIN costs may move in the opposite
direction of margins in most years. See Proposed Denial at 31 & Figure IV.D.2.b-l. Whatever
EPA's sources, this analysis does not allow the Agency (or anyone else) to reach any reliable
conclusions.
Response:
EPA does not believe that the data presented on average refining margins alone are sufficient to
draw conclusions on RIN passthrough or the impact of the RFS program on small refineries.
However, we do believe that these data would identify any consistent and significant impact on
small refineries vs. larger refineries. EPA has received comments stating that parties that blend
renewable fuels acquire RINs for free. This would suggest that in years when RIN prices are
high these parties would see an advantage of >$0.10 per gallon over parties that do not blend
renewable fuels. A competitive advantage of this magnitude should be apparent in a high-level
analysis such as the relationship between RINs and refining margins presented in the SRE
Denial. The fact that a discrepancy between these parties cannot be seen in the data is not
determinative on its own; it is one more piece of evidence EPA considered in reaching our
decision on the pending SRE decisions, along with the rest of the information presented in the
SRE Denial and the responses to comments herein.
Comment:
EPA suggests that its passthrough analysis must be correct because, if not, EPA would expect to
see parties change their business models:
B-59

-------
While some parties dispute EPA's findings on RIN cost passthrough and the RIN
discount, those same parties have not made business decisions over the last decade
that would be logical if RIN cost passthrough and RIN discount were not occurring.
For example, if RIN cost passthrough did not exist, we would expect to see refiners
avoiding RFS obligations by shifting production to non-obligated fuel (e.g., heating
oil, jet fuel) and/or export fuel. We would also expect to see actions to expand or
modify their business models to include additional blending of renewable fuel to reap
the alleged rewards that they claim independent blenders and marketers enjoy.
Proposed Denial at 26.
This is nonsense. Each of those unobligated fuels involves a separate market with unique supply
and demand dynamics and the availability of shifting production to other fuels is refinery-
specific. Some small refineries have explained in their SRE petitions that they are not capable of
avoiding their RFS obligations by shifting production to non-obligated fuels because there is
little to no market for such fuels in their area. Additionally, small refineries have similarly
explained that they cannot change their business model to reduce their RFS obligations by
increasing their exports due to the inability to ship fuels from their refineries to the coasts.
Additionally, even if there was a known benefit to blenders from incomplete passthrough, that
incentive may still be insufficient for a small refiner to build blending capacity in order to obtain
the benefit.
Response:
EPA's assertion that market actors would change their behavior in response to market
opportunities (if they existed) was not meant to suggest that every refinery in the country would
export all of their volume if RIN costs weren't passed through. Rather, if RIN costs are
temporarily not passed through, those market actors that can export will do so until the market
responds by raising the market price to recover the RIN. As EPA has detailed throughout the
SRE Denial and our response to comments herein, economic theory and the data available to
EPA show that market actors behave in a competitive manner and in doing so pass through the
cost of compliance and must reflect the RIN discount in the price of blended fuel they sell. It is
EPA's assessment that it is far more likely that wholesale refined product prices (CBOB and
diesel fuel prices) would rise to cover the cost of RFS compliance if parties that can export
started exporting all of their volume, than the counterfactual that the commenter seems to
suggest, which is that domestic prices would stay static (not accounting for the cost of RFS
compliance) as gasoline and diesel fuel supplies in the U.S. dropped due to increased exports.
With nearly inelastic demand for transportation fuels, we can be very confident that domestic
wholesale prices must rise to cover the cost of the RIN (i.e., RFS compliance) if that is what is
necessary to keep all of the refineries in the Gulf, West, and East Coasts of the U.S. from
exporting all of their fuel.
Comment:
EPA ignores the competitive advantage of refineries that are able to produce renewable fuel by
assuming that the only paths to compliance are by purchasing renewable fuel for blending with
refinery products or by buying RIN credits. There is, in fact, a third option that involves the
B-60

-------
production of renewable fuels. This third option can dramatically lower the cost of RINs as can
be seen in the recent profit margins of renewable diesel (RD) producers such as Valero's
Diamond Green Diesel (DGD) which reported a per barrel EBITDA margin of $2.34 in 2020 and
$2.97 in 1H 2021.
DGD has required roughly $914 million in cumulative capex and utilizes Valero refining sites in
Louisiana and Texas. Marathon and Phillips, similarly, are repurposing existing, obsolete
refining equipment to produce RD. Chevron, meanwhile, has taken extensive advantage of co-
processing at its large diesel hydrotreating units. Lastly, PBF is also planning to use obsolete
refining equipment at its Chalmette, LA refinery to produce RD.
Some small refineries are unable to take advantage of any similar opportunities due to a lack of
scale. They have conducted feasibility studies to examine the production of renewable fuel and
determined it is not economically viable. They are not able to build a facility of this sort of scale,
even if they wanted to invest in renewable diesel, due to a lack of available capital to build a
viable RD production facility and the lack of a site to build it.
Response:
In the context of SREs, EPA only considers the petroleum refining portion of the parent
company's business. Companies that own refineries may also own other businesses, including
businesses that produce renewable fuels, but EPA does not consider the effect of these
businesses when evaluating the cost of compliance with the RFS program. While it is true that a
renewable fuel producer that can produce renewable fuel at a cost below the market price for the
fuel will profit from that ability, EPA considers here that when the renewable producer sells the
fuel at market prices (either as a 100% renewable fuel to other parties for blending or in blended
fuels it sells itself), the profit it earns for that sale is the return on that renewable fuel business.
The cost for RFS compliance is on top of the market price for the renewable fuel because the
very function of the RIN is to discount the renewable fuel below that market price to incentivize
its sale. When a refinery discounts the renewable fuel that it sells based on the "RIN discount"
and retains the RIN it is acquiring that RIN at the price of the discount it must offer. Hence even
if the refinery also produced the renewable fuel, the cost of discounting that fuel to sell it in
compliance with the RFS program is still borne by the refinery.
Production of transportation fuel, whether it is renewable or non-renewable, is a cyclical
business. Companies that own refineries and produce renewable fuels may sometimes produce
renewable fuel and/or non-renewable fuel at a cost less than the market price at which it may be
sold, and sometimes produce renewable fuel and/or non-renewable fuel at a cost greater than the
price at which it may be sold.
Comment:
RIN costs are not fully passed through based on a comparison of the price difference between
gasoline and diesel sold in two different locations (Los Angeles, CA and Tijuana, Mexico) with
RVO cost from July 2018 through late 2021. If RIN prices are included in the price of
B-61

-------
transportation fuel, this price difference should display a high degree of correlation with RVO
cost. However, the price differential does not appear correlated with RVO cost.
Response:
The commenter does not say whether the price difference is based on wholesale or retail prices.
If it is based on retail prices, the comparison makes little sense, since retail gasoline is typically
E10 from which the RIN has already been separated and Mexican gasoline is a non-obligated
fuel sold in a market in a foreign country. There is no reason to expect the price difference
between two different products (one of which, Mexican gasoline, has no RFS obligation, and the
other of which has almost no net RFS compliance costs as the increase in the CBOB price is
offset by the RIN discount on the ethanol portion of the fuel) sold in two different countries to
bear any relationship to RFS compliance costs.29 If the price difference is based on wholesale
prices, the commenter appears to be assuming that the only reason for their calculated price
differentials is RFS compliance costs, which is incorrect. In reality there are many different
factors that affect prices of various products in different markets, including crude oil prices,
current supply and demand of the fuel, projected future supply and demand for the fuel,
inventories of the fuel, and production costs of the fuel. The commenter's pairs do not offer a
direct "apples-to-apples" comparison. They are not the same fuel; one with an RFS obligation
and one without.
29 While E10 sold in the United States can be used to estimate RIN cost passthrough it is very difficult to do so as
the increase in the CBOB price is offset by the RIN discount on the ethanol portion of the fuel. Depending on the
annual RFS percentage standards and the RIN prices at the time, the E10 price may be slightly lower or slightly
higher than it would be without the RFS program (e.g., if it were an export volume in Mexico). To a first order
approximation, the difference is zero.
B-62

-------
5. Small refineries cannot buy their RINs ratably and should not be punished for
employing a completely legal compliance strategy.
Comment:
Contrary to EPA's assertion, many petitioning refineries claim they cannot buy RINs ratably.
First, as small volume refineries, they do not purchase RINs in large enough volumes to
efficiently make purchases on a ratable basis. In contrast to integrated refiners and large
unobligated retailers, small refineries must manage RIN transactions in small blocks, due partly
to the significant price risks from the highly commoditized RIN market and EPA policy that
influence such risk. Further, small refineries do not have the resources to establish a RIN trading
desk to constantly monitor and purchase RINs and, even if they did, the daily RIN demand
would likely fall well below the typical RIN transaction quantities on the market, meaning they
would either have to delay or accelerate RIN purchases by several days, exposing them to market
risk.
Response:
A number of small refineries provided similar comments suggesting that they could not acquire
RINs ratably due to a lack of capital, an inability to afford the RINs, or specific limitations in
their ability to buy RINs in the proper lot sizes without facing a much steeper cost to acquire the
RINs.
Regarding the cost of capital and, more simply in some comments, the ability to afford RINs,
EPA notes that the very concept of ratable RIN purchases means that the acquisition of the RIN
is approximately concurrent with the sale of the fuel. This is different from other costs of
production, such as crude oil, which companies must first purchase and then process before
selling, resulting in a significant carrying cost for the company from the time of the crude oil
purchase until the time of the refined product sale. Here, for RINs, those time sequences can be
directly aligned and there is no need for the company to borrow to purchase the RINs. Rather,
the proceeds from the sale of the fuel can be directly used towards the purchase of the RINs in
ratable proportion to the company's obligation. For this reason, EPA rejects small refinery
arguments regarding the cost of capital and more generally the arguments generally regarding the
ability to afford RINs.
Regarding RIN lot size, EPA contends that small refineries can enter into contracts with various
RIN brokers to purchase RINs on a ratable basis. The contract terms can look very similar (and
quite reasonably might be made to have parallel elements) to the gasoline sales contracts that the
companies enter into with their customers. Specifically, the contract would specify the intent to
purchase a specific volume of RINs per month (e.g., 1.5 million RINs per month) at a price that
is calculated based on the average posted market price for the month. The RIN broker will likely
charge a service fee for such a contract, but we have no reason to believe this fee will be
substantially different from the fee offered to other market participants buying in lots of more
than 1 million RINs nor that such a fee would be more than the cost of a small refinery hiring
staff to execute a series of trades with parties directly to acquire RINs in this manner. Hence, we
do not think a small refinery paying such a fee would face a disproportionately higher cost for
B-63

-------
RIN acquisition than companies that acquire RINs in other ways. Even parties that acquire RINs
through blending have such administrative costs to track and transfer the RINs that the receive
by buying renewable fuel and then separating the RINs when they blend the renewable fuel.
Whether those accounting and administrative functions are done by staff employed by the
company or under contract with service providers, they nevertheless are a cost the blenders face
in order to accumulate the RINs they will use for RFS compliance. We think it unlikely that
these costs are significantly different among the various parties as the actual labor that is being
done and the value that is being added by that labor (i.e., the tracking of the RINs) is very similar
among the various ways that parties may acquire RINs.
Comment:
Some small refineries claim that they have reasonably relied on their previous exemptions when
choosing not to purchase RINs ratably. Citing reliance on the DOE scoring matrix and prior year
exemptions, many petitioning small refineries state that they reasonably believed that they would
receive an SRE from EPA for each of the pending petition years and, therefore, did not make the
significant capital or other investments necessary to comply.
Response:
As an initial matter, EPA does not believe reliance on prior year exemptions is a sound
compliance strategy or justification for not planning for compliance while current SRE petitions
are pending. The requirements of the RFS program are mandatory unless and until EPA grants an
exemption. Small refineries' reliance on prior exemptions is even more unreasonable in light of
the numerous legal challenges to EPA's prior approach to SREs, and EPA's long-standing
findings on RINs costs being passed on in the price of the transportation fuel they sell, as
explained elsewhere in the SRE Denial and herein.
B-64

-------
Appendices C-T - Confidential, Refinery-Specific Comment Summaries and
Responses
[Information Redacted - Claimed as CBI]

-------