Fuels Regulatory Streamlining
Amendments
Response to Comments
rnA United States
Environmental Protection
Agency
-------
Fuels Regulatory Streamlining
Amendments
Response to Comments
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
United States
Environmental Protection
Agency
EPA-420-R-24-029
December 2024
-------
Table of Contents
List of Acronyms and Abbreviations
List of Organizations Submitting Comments on the Proposed Rule
1. Fuels Regulatory Streamlining Amendments
1.1 General Comments
1.2 Implementation Dates
2. Sampling and Testing
2.1 Automatic Sampling
2.2 Demonstrating Homogeneity
2.3 Retaining Samples
2.4 PCG
2.5 Precision and Accuracy Demonstration
2.6 Fuel Certification with Domestic Marine Vessels
3. Other Technical Amendments
3.1 Exemptions
3.2 Recordkeeping and Reporting Requirements
3.3 Fuel Certification with Domestic Marine Vessels
3.4 Alternative Test Value Requests
3.5 Incorporation by Reference
3.6 Technical Corrections
4. Beyond the Scope
-------
List of Acronyms and Abbreviations
Numerous acronyms and abbreviations are included in this document. While this may not be an
exhaustive list, to ease the reading of this document and for reference purposes, the following
acronyms and abbreviations are defined here:
BOB
gasoline before oxygenate blending
CBOB
conventional gasoline before oxygenate blending
EPA
U.S. Environmental Protection Agency
ILCP
inter-laboratory crosscheck program
LLOQ
laboratory limit of quantitation
PCG
previously certified gasoline
PLOQ
pooled limit of quantitation
RFS
Renewable Fuel Standard
RVP
Reid Vapor Pressure
SIP
State Implementation Plan
VCSB
voluntary consensus standards body
-------
List of Organizations Submitting Comments on the Proposed Rule
Docket Item
Number3
Commenter or Organization Name
0003
Michael Ravnitzky
0004
Anonymous
0005
Alex Lau
0006
Mira Garcia
0007
Mira Garcia
0008
Clean Fuels Alliance America
0009
ONEOK, Inc.
0010
Sustainable Advanced Biofuel Refiners (SABR) Coalition
0011
bp America, Inc.
0012
American Fuel & Petrochemical Manufacturers (AFPM) and American
Petroleum Institute (API)
0013
Anonymous
0014
Anonymous
0015
Susan Lafferty
0016
Husky Marketing and Supply Company (Cenovus)
a Individual comments from the public (and attachments submitted with comments) submitted to Docket
No. EPA-HQ-OAR-2024-0143 are assigned a unique 4-digit docket number that follows the base docket
number (i.e., XXXX, where "XXXX" represents the unique 4-digit document docket number). For
example, Docket Item No. EPA-HQ-OAR-2024-0143-0005 is presented as 0005 in this table and within
the text of this document.
111
-------
1. Fuels Regulatory Streamlining Amendments
1.1 General Comments
Commenters that provided comment on this topic include but are not limited to: 0003, 0006,
0007, 0009, 0012.
Comment:
Numerous commenters expressed support for the proposed rule.
Response:
We thank the commenters for their support.
-------
1.2 Implementation Dates
Comment:
> 0009 0NEOK
§ 1090.5 Implementation dates.
1090.5(c)(4)(d) The provisions of § 1090.1315(a)(7) through (14) apply beginning March 1,
2025. Fuel manufacturers may optionally apply those new provisions before March 1, 2025.
• Complying with an implementation date for the provisions of § 1090.1315(a)(7) through (14)
beginning March 1, 2025 will create a burden for fuel manufacturers in the midst of the 2024
compliance reporting season. ONEOK encourages EPA to consider allowing parties that hold
in-line blend waivers to comply with the new provisions at § 1090.1315(a)(7) through (14)
by the end of the calendar year 2025.
§ 1090.1800 General provisions.
1090.1800(a) The following parties must arrange for attestation engagement using agreed-upon
procedures as specified in this subpart:
• ONEOK would like to recommend the addition of text clarifying that 2024 compliance year
attestations due June 1, 2025 will be based on the requirements in the CFR prior to the
issuance of a Final Rule.
> 0012 AFPM/API
VII. § 1090.1315 In-line Blending
a. ILB petition and attestation timing requirements
Section II.C. of the preamble to the proposed rule discusses the process for amending in-line
blending (ILB) waivers. It asserts that not all ILB petitions need to be updated to reflect the
updated specifications in § 1090.1315, as most fuel manufacturers already include the proposed
information in their approved petitions. It further suggests that EPA has proposed rule updates
such that "all in-line blending waivers must comply with the proposed new specifications starting
March 1, 2025, with all the proposed new flexibilities being optionally available before that
deadline." This date was then associated with an implied deadline of December 31, 2024, to
submit a revised petition, if necessary, assuming petition updates in accordance with the
Streamlining Technical Amendments Rule would be approved by EPA without delay or questions
or additional information requests from EPA.
According to the proposed regulatory text, March 1, 2025, would be the deadline to submit a
petition, not the deadline for approval of the petition. In § 1090.5(d), EPA proposed an update
that requires compliance with § 1090.1315(a)(7)-(14) by March 1, 2025. However,
2
-------
§1090.1315(a) articulates a requirement to submit a petition requiring certain information, not a
requirement for the petition to be approved by EPA. We believe it is indeed more appropriate to
include a submittal deadline than an approval deadline, even with the additional criteria that EPA
has proposed to add to help expedite the approval process (see updates in § 1090.1315(c)).
Otherwise, the rule would impose requirements that a fuel manufacturer ultimately cannot
control.
EPA's final rule should clearly state whether existing in-line blend waivers need to be
resubmitted to meet any new requirements. Because the updates are meant to "codify . . .
previously approved flexibilities in the regulations" (EPA's description of the changes from
preamble Section II.B), we do not believe it should be necessary to require resubmittal of
previously approved ILB waivers to address new requirements. We recommend that additional
information proposed in § 1090.1315 is only required in petitions that are submitted on or after
the date the rule is promulgated.
Otherwise, all fuel manufacturers with previously approved waivers would be required to submit
updated petitions, even if the only update required would be to add the new statement required
by proposed § 1090.1315(a)(14).
EPA proposes to expand the scope of ILB attestation requirements, and EPA should clarify when
the new provisions become applicable. Similar to allowances made in the prior Streamlining
rulemaking (see § 1090.5(b)(8)(h)), EPA should provide gasoline manufacturers sufficient time
to prepare for any expansion to ILB auditing requirements. If the rule is not promulgated until
late 2024 or mid-2025, batches certified by ILB in those years should not be subject to the new
requirements, and the new requirements would not apply until 2027 when auditing batches
blended in 2026, unless otherwise specified in an approved ILB petition. If the rulemaking is
significantly delayed this date should be pushed out even further.
Recommendation: EPA should update the text of the rule to clarify that resubmissions of existing
ILB petitions are not necessary and clarify ILB attestation applicability requirements. Final dates
should be based on the date a final Streamlining Technical Amendments Rule is promulgated and
petitions submitted after publication of the final rule would be required to include provisions
included in the rule. Additional discussion of effective dates for specific provisions of this rule is
provided in section XVIII below.
Suggested text:
§1090.5. Implementation dates.
(d) The provisions of § 1090.1315(a)(7) through (14) apply beginning March 1, 2025. to
new in-line blend waiver requests and in resubmissions of previously approved in-line
blend waiver requests that are submitted after [insert publication date]. Fuel
manufacturers are not required to resubmit a previously approved in-line blend waiver to
address the requirements of § 1090.1315(a)(7) through (14) or other technical
3
-------
amendments that were promulgated on [insertpublication date], may optionally apply
those new provisions before March 1, 2025.
(e) The provisions of § 1090.1850(b) apply beginning with batches certified in calendar
year 2026 subject to attestation in calendar year 2027.
XVIII. § 1090.5 Implementation dates
EPA initially promulgated the Streamlining Rule (40 CFR part 1090) on December 4, 2020, with
most provisions applicable a short time later (January 1, 2021), unless otherwise excepted. The
original Streamlining Rule included a detailed "Implementation dates" section (§ 1090.5)
outlining requirements for which regulated parties were allowed additional implementation time.
EPA should include additional detail in the proposed revision to § 1090.5 to address
implementation considerations specific to the Fuels Regulatory Streamlining Amendments
proposed rule. Currently, the only addition to that section in the proposed rule updates, pertains
to in4ine blending (ILB) waiver updates.
Although the changes proposed in the rule are largely straightforward technical amendments,
some will still require updates to procedures or other changes that regulated parties will need
time to address after the rule is finalized. Regulated parties will not know with certainty what the
rule requirements will be, or when they will become applicable, until the rule is finalized.
Although regulated parties have begun preparing for the expected changes, several comments are
submitted herein may result clarifications or updates to certain provisions in the proposed rule, as
well as comments from other stakeholders. Accordingly, we request EPA provide lead time
beyond the date of promulgation for regulated parties to implement certain requirements, as
described in more detail below.
When drafting implementation date rule text to address these concerns, please consider the
practical implications of releasing a rule adjacent to the year-end holidays (e.g.,
November/December). Staffing levels are typically relatively low at that time and pulling staff
across an organization for extra hours over the holidays for a rushed implementation effort is not
justified by the benefits of implementing these technical amendments. Implementing the original
Streamlining Rule on January 1, 2021, under similar holiday timing circumstances was important
due to the significant rule changes, including updates to product quality and averaging, banking,
and trading requirements. Those changes made sense to apply at the start of a new calendar year,
but those unique qualities are not present in this technical amendment rulemaking. Furthermore,
there have been fewer discussion drafts and no workshops for this rule to help affected parties to
prepare in advance when comparing this proposal to the 2020 Streamlining Rule.
Recommendation: The table below includes a summary of proposed rule updates that will
require significant lead time to implement.
4
-------
• If not included in the below table, the proposed change should become effective upon
promulgation of the rule.
• If included in the below table, additional time is needed to implement all or part of the
revisions to each section, or rule revisions are requested that would eliminate the need for
additional time (see details in the notes/special considerations column).
• In the column indicating how much time is needed:
o Where it says "> 1 calendar quarter," please draft the rule update to make the change
effective on the last day of the calendar quarter following the publication date of the
final rule. For example, if the final rule is published on December 1, 2024, the change
would be effective by the end of the first quarter of 2025, so by March 31, 2025.
o Where it says "> 1 calendar half," please draft the rule updates to make the change
effective on the last day of the semiannual calendar period following the publication
date of the final rule. For example, if the final rule is published on December 1, 2024,
the change would be effective by the end of the first half of 2025, so by June 30,
2025.
o This approach is recommended to avoid having unusual mid-month compliance
deadlines, inspired by similar language used by EPA in recent revisions to the
Renewable Fuel Standard (RFS) in 40 CFR part 80.
o Where it says "significant," this indicates the change may trigger long lead time
capital projects or other substantial changes versus current compliance requirements,
which we do not think is EPA's intent with the Fuels Regulatory Streamlining
Amendments rule. Edits to the proposed rule language as suggested in other sections
of the letter may clarify this such that additional time is not needed.
Rule section
Need
time?
How
much
additional
time?
Refer
to
letter
section
Notes/special considerations:
§1090.5
Implementation dates.
This is where exceptions for any/all of the below need to be included.
See below summary and more detailed comments throughout this
letter on each section.
§ 1090.95
Incorporation by
reference.
Yes
> 1
calendar
quarter
XVII
Need sufficient time to update
corporate and site-specific standard
operating procedures (SOPs) to
match updated versions of the
ASTM test methods and standard
practices that will become
incorporated by reference.
Additionally, some updates may be
needed to blend monitoring and
control systems to match the
updated version of ASTM D4177.
5
-------
Rule section
Need
time?
How
much
additional
time?
Refer
to
letter
section
Notes/special considerations:
§ 1090.610
Exemptions for
temporary research,
development, and
testing.
Maybe
(see
note to
right)
> 1
calendar
half
I
If EPA accepts the edits
recommended in section I of this
comment letter, no additional time
needed. Otherwise, sufficient lead
time will be needed to implement
new requirements for R&D fuel
additive exemptions.
§ 1090.905 Reports
by gasoline
manufacturers.
Yes
> 1
calendar
half
(depending
on nature
of change,
see note to
right)
IV
Sufficient lead time is needed for
any updates to batch reporting
forms. If the only change is to add a
volatility category for Midwest
states that opted out of the 1 psi
waiver, updates would presumably
not apply until such batches are
blended. If the updates are more
expansive than that, regulated
parties would need additional time
to track additional criteria for each
batch, including updates to software
systems, product codes, etc. If that
is the case, regulated parties need at
least 1 calendar half to start tracking
additional criteria for each batch.
§ 1090.1315 In-line
blending.
Yes
> 1
calendar
half
(depending
on nature
of change,
see note to
right)
Vila
In summary, clarifications are
needed to confirm previously
approved ILB waivers do not have
to be resubmitted by a particular
date, but that all waivers submitted
or updated after the rule is
promulgated must address the new
criteria. However, if it is EPA's
intent that all previously approved
petitions are resubmitted by a
particular date to address the
technical amendments in the rule,
regulated parties will need sufficient
time to make the updates, work out
any preliminary questions with
EPA, and secure the necessary
internal approvals before
resubmitting to EPA.
6
-------
Rule section
Need
time?
How
much
additional
time?
Refer
to
letter
section
Notes/special considerations:
§ 1090.1320 Adding
blendstock to PCG.
Specifically, §
1090.1320(a)(l)(i)(A).'
Maybe
Significant
Vila
If EPA accepts edits to § 1090.1320
(a)(l)(i)(A), recommended in
section VII of this comment letter,
no additional time needed.
Otherwise, significant lead time
needed.
§ 1090.1337
Demonstrating
homogeneity.
Specifically, §
1090.1337 (f)(1).
Maybe
> 1
calendar
quarter
X.d
If EPA accepts edits to §
1090.1337(f)(1), recommended in
section X of this comment letter, no
additional time needed. If finalized
as proposed, regulated parties need
sufficient time to update procedures
for homogeneity testing when
results are below PLOQ, LLOQ, or
valid range.
§ 1090.1337
Demonstrating
homogeneity.
Specifically, §
1090.1337(d).
Yes
> 1
calendar
quarter
X.b
Need sufficient time to update
procedures for sampling/testing
GTAB and/or TGP for homogeneity.
§ 1090.1345 Retaining
samples.
Maybe
> 1
calendar
half
XI
If EPA accepts the edits that are
recommended in section XI of this
comment letter, no additional time
needed. Otherwise, significant lead
time needed to address retain
procedures/storage limitations.
§ 1090.1375 Quality
control procedures.
Yes
> 1
calendar
half
XII
Sufficient time would be needed to
implement the significant changes
that are proposed to SQC
requirements (e.g., time to update
procedures, source check standards,
update training, etc.).
1 In contrast, the revisions to § 1090.1320(a)(l)(i)(D) would not require additional time to implement (clarifications
on reporting presumed or alternate test values for PCG should be effective upon promulgation).
7
-------
Rule section
Need
time?
How
much
additional
time?
Refer
to
letter
section
Notes/special considerations:
§ 1090.1605
Importation by
marine vessel.
Maybe
> 1
calendar
quarter
XIII
If EPA accepts the edits that are
recommended in section XIII of this
comment letter, no additional time
needed. Otherwise, regulated parties
would need sufficient time to update
procedures for certifying imports of
RBOB (e.g., ensure preparation of
hand blends with ethanol on each
individual compartment rather than
just on a manual composite across
compartments).
§ 1090.1800 General
provisions.
Yes
See note to
right
Vila
& XV
In summary:
• Please clarify that the 2025
attestation of batches blended in
2024 will be based on the
existing version of the
regulations.
• Likewise, assuming some
provisions of this rule will not
become applicable until mid-
2025 or later, please clarify that
the 2026 attestation of batches
blended in 2025 may be based on
compliance with either version of
the rule, as applicable.
• Furthermore, where the
attestation procedures themselves
are being expanded in scope (i.e.,
ILB requirements in §
1090.1850(b)), EPA should
confirm the new provisions apply
beginning with batches blended
in 2026 that are attested in 2027
(unless otherwise specified in an
approved blend waiver).
§ 1090.1850
Procedures related to
in- line blending
waivers.
Yes
See note to
right
XV
See section XV of this letter and the
above comment on § 1090.1800 for
timing of the expanded scope for
ILB attestation.
8
-------
Response:
We generally acknowledge regulated parties' need for more time to implement the new
regulatory amendments because they will require adjustments to their business practices in order
to come into compliance with them. While many of the amendments are minor, provide
additional flexibility, or require minimal lead time to implement, the commenters did identify
several provisions that they believe will require at least three or more months of lead time.
Rather than taking a piecemeal approach in which different provisions would take effect at
different times, we are providing regulated parties with at least six months to adjust their
business practices by making the effective date of this action July 1, 2025, with two exceptions
discussed below. This approach addresses commenters' concerns about needing more time for
certain new requirements and should result in minimal disruption to existing business practices.
For ease of reference, a table has been included below to reflect the applicable effective dates for
the sections identified by AFPM/API.
First, fuel manufacturers with existing in-line blending waivers must have an approved in-line
blending waiver that complies with the new requirements for in-line blending waivers in 40 CFR
1090.1315(a)(7) through (13) no later than January 1, 2026. The fuel manufacturer must comply
with the associated attestation engagement requirements for in-line blending waivers in 40 CFR
1090.1850(b) starting with either the 2025 or 2026 compliance period, whichever is the earliest
compliance period for which the fuel manufacturer has an approved in-line blending waiver that
complies with all the provisions of § 1090.1315(a)(7) through (13).
As noted above, we revised the compliance deadline for fuel manufacturers to update their
existing in-line blending waivers to comply with the requirements of 40 CFR 1090.1315(a)(7)
through (13) from March 1, 2025, to January 1, 2026. We have also revised 40 CFR 1090.5(d) to
clarify that fuel manufacturers must submit their updated in-line blending waivers and have them
approved by January 1, 2026. Now that the regulation includes a provision for automatic
approval of requests to update in-line blending waivers, fuel manufacturers are able to control
the schedule for complying with the deadline, even though the deadline includes getting EPA
approval. It is only logical that fuel manufacturers should have their updated in-line blending
waivers approved by the deadline. This is also necessary as a matter of practicality because some
of the new specifications are subject to the attestation provisions. In sum, we would want
attestation reports for the 2026 compliance period to include only systems that consistently and
reliably comply with the full set of specifications that we are adopting in this final rule.
With regards to the comments about the process and timing for updating in-line blending
waivers, we first note that Section II. C of the proposal preamble stated that all in-line blending
waivers would need to comply with the proposed new specifications (89 FR 70051-52). We
appreciate the commenter highlighting that the added specifications in 40 CFR 1090.1315 are
drafted in the context of a requirement to submit information. This contrasts with our statement
in the proposal preamble that fuel manufacturers should submit their requests for updated
waivers at least 60 days ahead of the requirement to comply with the new provisions based on an
approved waiver. 89 FR 70051-52.
9
-------
It was not our intent to require resubmissions of unchanging ILB waivers and we have clarified
that intent in the final rule. As described in Section 2.1 of this document, we are amending the
relevant regulation to clarify that fuel manufacturers must request approval for an updated in-line
blending waiver in any case where the in-line blending process changes such that the previously
approved waiver becomes incorrect or incomplete. This would include consideration of all the
items identified in 40 CFR 1090.1315(a)(7) through (13). We are also requiring the addition of
the statement specified in 40 CFR 1090.1315(a)(14), but only if the fuel manufacturer requests
approval of an updated waiver for other reasons. Per 40 CFR 1090.1315(c)(2), such statement
does not need to be included in the waiver until the fuel manufacturer otherwise updates their in-
line blending waiver under 40 CFR 1090.1315(c)(1).
Second, the new gasoline batch reporting requirement to indicate whether the 1-psi waiver
applies in 40 CFR 1090.905(c)(l)(viii)(A)(2), (c)(2)(viii)(A)(2), and (c)(8)(vii)(A)(2) begins
with the 2026 compliance period and, therefore, it will not be reflected on batch reports until
March 31, 2027. We believe this implementation date is appropriate for these provisions because
they will take effect as part of a new compliance period, which will make the reporting,
recordkeeping, and annual attestation engagement audits more straightforward to implement. It
also provides the same if not more time than was requested by commenters.
Finally, we also note that annual attestation engagements for the 2024 compliance period will not
cover any amendments included in this action as none of the finalized changes will be in effect
during the 2024 compliance period. Furthermore, to avoid any potential confusion, we have also
intentionally set the July 1, 2025, effective date of this action—including the changes to the
attestation engagement requirements—to occur after the 2024 attestation engagement reporting
deadline of June 1, 2025.
For clarity, we specifically address the issues raised in AFPM/API's comments in the following
table:
Section
Additional Time
Requested?
EPA Response
§ 1090.95 Incorporation
by reference.
Yes.
> 1 calendar quarter
Changes will take effect July 1, 2025.
§ 1090.610 Exemptions
for temporary research,
development, and testing.
Maybe.
> 1 calendar half
EPA accepted suggested revisions. Changes
will take effect July 1, 2025.
§ 1090.905 Reports by
gasoline manufacturers.
Yes.
> 1 calendar half
Changes will take effect on January 1, 2026.
§ 1090.1315 In-line
blending.
Yes.
> 1 calendar half
Changes will take effect on January 1, 2026.
§ 1090.1320(a)(l)(i)(A)
Adding blendstock to
PCG.
Maybe.
Significant
EPA accepted suggested revisions. Changes
will take effect July 1, 2025.
§ 1090.1337(f)(1)
Demonstrating
homogeneity.
Maybe.
> 1 calendar quarter
EPA did not accept suggested revisions.
Changes will take effect July 1, 2025.
10
-------
Section
Additional Time
Requested?
EPA Response
§ 1090.1337(d)
Demonstrating
homogeneity.
Yes.
> 1 calendar quarter
Changes will take effect July 1, 2025.
§ 1090.1345 Retaining
samples.
Maybe.
> 1 calendar half
EPA accepted suggested revisions. Changes
will take effect July 1, 2025.
§ 1090.1375 Quality
control procedures.
Yes.
> 1 calendar half
Changes will take effect July 1, 2025.
§ 1090.1605 Importation
by marine vessel.
Maybe.
> 1 calendar quarter
EPA accepted suggested revisions. Changes
will take effect July 1, 2025.
§ 1090.1800 General
provisions.
Yes.
Changes will take effect July 1, 2025 (i.e.,
after the 2024 attestation engagement
reporting deadline of June 1, 2025) and will
apply for attestation engagements for 2025
batches.
§ 1090.1850 Procedures
related to in-line
blending waivers.
Yes.
Changes will take effect July 1, 2025.
Language has been added to 1090.5 to
clarify that these provisions do not apply
until 2026 batches unless the fuel
manufacturer complies with new in-line
blending waiver requirements early (i.e.,
during the 2025 compliance period).
11
-------
2. Sampling and Testing
2.1 Automatic Sampling
Comment:
> 0011 bp America
III. Automatic Sampling
EPA proposes at Section 1090.1335(c)(5) that, where automatic sampling returns one or more
non-compliant samples (other than composite samples), the volume of noncompliant fuel is the
volume in between valid, compliant samples, bp understands that EPA's intention is to identify
with specificity the noncompliant portion of an in-line blended batch in light of available
certification samples and additional spot sampling. See Preamble, 89 Fed. Reg. at 70,050.
However, EPA suggests that it will limit the evidence it will consider in determining the
noncompliant volume to those samples and exclude any "nonstandard tests." Id. bp suggests that
EPA reserve flexibility to evaluate all credible evidence that can demonstrate the compliant
volume of a batch where a single spot sample may be non-complaint. Such evidence could
include additional mixing of the batch with PCG prior to leaving the facility, representative
online analyzer data, or other indication that the batch or a portion of the batch met fuel
specifications.
> 0012 AFPM/API
IX. § 1090.1335 Collecting, preparing, and testing samples
b. Determining the volume of a noncompliant batch
When determining the volume of a noncompliant batch, the Associations believe EPA should
allow flexibility to evaluate all credible evidence demonstrating when the batch's characteristics
returned to compliant, including whether additional mixing with PCG prior to leaving the facility
meant that no non-compliant material left at all.
Recommendation: The Associations recommend adding additional wording to capture alternative
volumes that are approved by EPA.
Suggested Text:
§ 1090.1335 Collecting, preparing, and testing samples.
* * *
12
-------
(5) If the composite sample fails to meet any applicable per-gallon standard, the entire
batch volume is noncompliant. If one or more separate samples fail to meet any
applicable per gallon standard, the volume of noncompliant fuel is the volume starting
with the last valid passing result before the failing result (or the start of the batch), up to
the first valid passing result after the failing result (or the end of the batch), or alternative
volume approved by EPA.
Response:
We disagree with the comment. The penalty provisions described at 40 CFR 1090.1710(g) allow
for a flexible approach relying on all credible evidence to properly assess the volume of
noncompliant fuel. These established provisions are preferred over adding a redundant clause in
40 CFR 1090.1335(c)(5) for alternative volumes. That regulatory provision allows EPA to
establish alternative values when assessing noncompliance based on any relevant information,
including consideration of alternative values proposed by the entity failing to meet the
requirements. We consider alternative test values to replace the presumed values in 40 CFR
1090.1710(g) to be equivalent to assessing alternative volumes of noncompliant fuel under 40
CFR 1090.1335(c)(5). From that perspective, fuel manufacturers can make an argument for a
smaller volume of noncompliant fuel based on tests that are different than valid tests for
certifying fuel, if those tests make a clear showing that a batch of fuel met standards at a point in
time. We are therefore finalizing as proposed the requirement in 40 CFR 1090.1335(c)(5) to
identify noncompliant fuel volumes based on a "valid passing result."
Comment:
> 0009 ONEOK
§1090.1315 In-line blending.
1090.1315 (a)(8) Describe any circumstances in which it is not possible to achieve a sampling
frequency corresponding to a margin of error of 0.01. Also describe how you will adjust target
values to account for the greater measurement variability. For example, if the greater margin of
error corresponds to a 2 percent increase in measurement variability, adjust measured values of
all parameters subject to per-gallon and average standards upward by 2 percent.
• The proposed language appears to limit automatic sampling frequency with a margin of error
of 0.01. ONEOK requests clarification and alignment with 1090.1335(c)(3) to stipulate that
the sampler should be configured to sample on a frequency of 9,604 grabs, a margin of error
of 0.01, or sampling frequency that does not exceed 20 seconds throughout the batch.
§ 1090.1335 Collecting, preparing, and testing samples.
1090.1335(c)(3) Set a sampling frequency based on collecting 9,604 grabs or a smaller number
of grabs that achieves a margin of error of 0.01 or less as specified in Section 19.1.3 of ASTM
13
-------
D4177; however, less frequent sampling is acceptable as long as the interval between samples
does not exceed 20 seconds throughout the batch. Keep records to show that the sampling
frequency meets specifications.
• ONEOK requests that 1090.1335(c)(3) be clarified and aligned with 1090.1315(a)(8) to
stipulate that the sampler should be configured to sample on a frequency of 9,604 grabs, a
margin of error of 0.01, or sampling frequency that does not exceed 20 seconds throughout
the batch.
> 0012 AFPM/API
VII. § 1090.1315 In-line Blending
b. Clarify ILB flexibilities
The example provided in proposed § 1090.1315(a)(8) seems to add confusion.
Recommendation: The Associations recommend removing the example in § 1090.1315(a)(8), as
it adds unnecessary confusion. The EPA should collaborate with the location in their petition to
include any target adjustment. As an alternative, if EPA disagrees with removing the example in
§ 1090.1315(a)(8), it is recommended to change "adjust measured values" to "adjust target
values" in the example. The adjustment should be made to the target value, not the measured
value from the testing.
Suggested Text:
§ 1090.1315 In-line blending.
* * *
(8) Describe any circumstances in which it is not possible to achieve a sampling
frequency corresponding to a margin of error of 0.01. Also describe how you will adjust
target values to account for the greater measurement variability. For example, if the
greater margin of error corresponds to a 2 percent increase in measurement variability,
adjust measured values of all parameters subject to per gallon and average standards
upward by 2 percent.
Alternate Recommendation:
§ 1090.1315 In-line blending.
14
-------
* * *
(8) Describe any circumstances in which it is not possible to achieve a sampling
frequency corresponding to a margin of error of 0.01. Also describe how you will adjust
target values to account for the greater measurement variability. For example, if the
greater margin of error corresponds to a 2 percent increase in measurement variability,
adjust measured target values of all parameters subject to per-gallon and average
standards upward downward by 2 percent.
IX. § 1090.1335 Collecting, preparing, and testing samples
a. Default sampling frequency
The proposed wording revisions to § 1090.1335(c)(3) lists less frequent sampling as not
exceeding 20 seconds, when the intent is to allow another option for sampling frequency as long
as the sampling frequency does not exceed 20 seconds.
Recommendation: The Associations recommend modifying the wording in § 1090.1335(c)(3), as
a sampling frequency that does not exceed 20 seconds is not less frequent sampling. We believe
that the recommended wording changes below would better capture the intent.
Suggested Text:
§ 1090.1335 Collecting, preparing, and testing samples.
* * *
(ti-3) The default sampling frequency must follow the recommended approach of at least
9,601 samples to represent a batch. L Set a sampling frequency based on collecting 9,604
grabs or a smaller number of grabs that achieves a margin of error of 0.01 or less as
specified in Section 19.1.3 of ASTM D4177; however, loss frequent sampling is
acceptable as long as the interval between samples or does not exceed a sampling interval
of 20 seconds throughout the batch. Keep records to show that the sampling frequency
meets specifications.
Response:
The proposed amendments to the redesignated 40 CFR 1090.1335(c)(3) were intended to
identify a maximum margin of error of 0.01 as the fundamental criterion to qualify the sampling
frequency for automatic sampling systems. The secondary reference to 9,604 grabs was intended
15
-------
to allow for meeting the 0.01 margin of error specification without relying on calculations.
Retaining the reference to a minimum sampling interval of 20 seconds throughout the batch was
intended to be a permissible sampling frequency as an exception to the requirement for meeting a
margin of error of 0.01. The commenters' suggested approach is to bypass those distinctions and
simply list three alternative methods of demonstrating an appropriate sampling frequency. In
response to this comment, we have modified the proposed amendment to include the suggested
simpler approach for the sampling frequency requirements.
The proposed new provision in 40 CFR 1090.1315(a)(8) was drafted to align with the proposed
amendment to the redesignated 40 CFR 1090.1335(c)(3), with the reference to the primary
requirement to set a sampling frequency with a margin of error of 0.01 or less. ONEOK suggests
including all three options for meeting sampling frequency requirements in the regulation at 40
CFR 1090.1315(a)(8). Since we are changing the description in 40 CFR 1090.1335(c)(3) to
identify three alternative means of meeting sampling frequency requirements, we are revising the
proposed text in 40 CFR 1090.1315(a)(8) to simply cross reference 40 CFR 1090.1335(c)(3)
instead of repeating the specific requirements.
We disagree, however, with AFPM/API's comment on proposed 40 CFR 1090.1315(a)(8)
suggesting that the proposed paragraph was confusing. The example identified in the proposed
regulatory text was intended to clarify a specific case of accounting for measurement variability
in a compliance demonstration when a fuel manufacturer was not able to meet sampling
frequency requirements. Removing the example would require us to resolve most issues on a
case-by-case basis. While we are keeping the specific numerical reference in the example, we
have modified the regulation to incorporate the commenter's suggested alternative
recommendation to describe a downward change in target values rather than adjusting measured
values upward. We view this change as acceptable because the new provision is intended to
guide fuel manufacturers in preparing their contingency plans in anticipation of a sampling
frequency that will fall short of sampling frequency requirements. We do not need fuel
manufacturers to adjust measured values as part of their compliance demonstration. Rather, we
are satisfied with adjusting process controls to dial in target values that create a greater
compliance margin to account for the greater measurement variability resulting from not meeting
sampling frequency requirements. We note that this change to the proposed amendment aligns
with Section II.B.l of the proposal preamble (89 FR 70050).
Comment:
> 0009 ONEOK
§ 1090.1850 Procedures related to in-line blending waivers.
1090.1850(b)(1) Obtain test data, including head, middle, and tail results, for each batch
produced under the gasoline manufacturer's in-line blending waiver.
• The provisions at 1090.1850(b)(l)-(6) require the auditor to preform reviews of each batch
produced under the gasoline manufacturer's in-line blending waiver, which will be
impractical for the auditor and gasoline manufacturer alike. ONEOK would suggest
16
-------
amending the language at 1090.1850(b)(1) as follows to align with common audit
procedures:
• 1090.1850 (b)(1) Obtain test data, including head, middle, and tail results, for the batches
selected as part of the representative sample under §1090.1810 of each batch produced under
the gasoline manufacturer's in-line blending waiver. Perform the procedures at
1090.1850(b)(2)-(6) upon the representative sample of batches.
> 0012 AFPM/API
XV. § 1090.1850 Procedures related to in-line blending waivers
The Associations believe that there may be an error or oversight in the proposed amendments to
§ 1090.1850, as the intent of this section is for attestation, which involves auditing of selected
batches. We believe that the wording in proposed § 1090.1850(b) needs modification to clarify
that the listed items are for the selected batches that are part of the attestation. Without the
recommended edits below, it can be interpreted that every batch is required to have all of the
attestation actions listed in the section.
Recommendation: The Associations recommend this be reworded as a selection of batches to be
consistent with attestation requirements listed in § 1090.1805. Exceptions should be included in
an individual ILB petition. We recommend deletion of paragraphs (5) and (6), because this
would be covered in the individual ILB petition.
Suggested Text:
§ 1090.1850 Procedures related to in-line blending waivers.
(b) For the selected batches:
(1) Obtain test data, including head, middle, and tail results, for each batch produced
under the gasoline manufacturer's in-line blending waiver.
(2) Review the alternative sampling plan to meet requirements to test head, middle, and
tail samples for small batches under § 1090.1315(a)(9).
(3) Report as a finding any instance where only a single sample was taken for a small
batch involving more than 8 hours of blending or more than 1 million gallons of fuel.
(4) Report as a finding any instance where two samples were unevenly distributed for a
small batch or where only two samples were taken for a small batch involving more than
16 hours of blending or up to 2 million gallons of fuel.
(5) Determine and report the percentage of in line blending batches where the gasoline
manufacturer failed to perform the required head, middle, and tail samples due to
17
-------
unforeseen circumstances. Report as a finding if this percentage is greater than 10 porcont
of in line blending batches for the calendar year.
(6) Dotormino and report each instance where a contingency plan for alternative sampling
was utilized under § 1090.1315(a)(12).
Response:
We disagree that there was an error in the proposed attestation requirements related to in-line
blending waivers. The proposed changes to the attestation requirements in 40 CFR 1090.1850(b)
were intended to account for the proposed reduced-sampling option for refiners in 40 CFR
1090.1315. This option could apply to any batch and up to 10 percent of batches produced during
the compliance period. Auditors must therefore determine whether the number of batches using
the reduced-sampling option exceeds 10 percent of total in-line blending batches, which is not
possible by auditing only a representative sample of batches. We anticipate that auditors will
filter data from all in-line blended batches to complete the new attestation requirements in 40
CFR 1090.1850(b) by, for example, excluding batches less than 8 hours or less than 1 million
gallons, and excluding batches less than 16 hours or less than 2 million gallons. We are therefore
finalizing 40 CFR 1090.1850(b) as proposed.
Comment:
> 0012 AFPM/API
VII. § 1090.1315 In-line Blending
c. ILB contingency plans and redundant sampling
Proposed § 1090.1315(a)(ll) would add language related to a redundant system to collect a
second composite sample; however, the intent of the clause "where possible" is not clear.
Recommendation: The Associations recommend changing "Where possible" to "For example".
Suggested Text:
§ 1090.1315 In-line blending.
* * *
(11) Describe contingency plans for alternative sampling and testing in cases involving
failure of the automatic compositor or other essential equipment. Where possible For
example, include collecting a second composite sample with a redundant system.
18
-------
Response:
In response to the comment, we have revised the proposed amendment to identify redundant
systems for collecting composite samples as an example of an appropriate contingency plan. The
proposed provision in 40 CFR 1090.1315(a)( 11) was intended to steer fuel manufacturers toward
preferentially setting up their systems to include redundant measurements to prepare for a
possible failure of automatic compositors or other essential equipment. Considering the specific
change suggested by the commenter, we believe the suggested wording appropriately identifies
redundant measurement as one of the possible approaches that fuel manufacturers could take to
prepare for equipment failure, without stating a preference for redundant measurement. We
continue to believe that relying on redundant measurements is a good way to prepare for
unforeseen circumstances, but we will consider any appropriate plan to demonstrate compliance
with standards if there is an equipment failure that prevents the fuel manufacturer from reporting
measurements from the conventional measurements from the sample collected from the
compositor.
Comment:
> 0012 AFPM/API
VII. § 1090.1315 In-line Blending
d. ILB contingency plans and failing to meet a per-gallon standard
Proposed § 1090.1315(a)(12) introduces language regarding contingency plans for alternative
sampling demonstrations if an automatic sampling test fails to meet a per-gallon standard. The
wording appears to be an example, but clarity is needed. Additionally, it should be consistent
with § 1090.1310(b) for certifying a batch, so we recommend referencing that section instead of
the fuel manufacturing facility gate.
Recommendation: The Associations recommend specifying that the proposed language is an
example, by adding "For example" should be added as a lead-in for the second sentence.
The Associations recommend a modification to the wording in the second sentence to point to §
1090.1310(b) when referring to certifying the batch in the tank.
The Associations recommend removing the last sentence, as having an additional automatic
sampler as the material leaves a holding tank would be a practice that is not common. It could
cause confusion throughout the industry if EPA were to retain the proposed language.
Suggested Text:
§ 1090.1315 In-line blending.
19
-------
* * *
(12) Describe any contingency plans for an alternative sampling demonstration if an
automatic sampling test result fails to meet a per-gallon standard. For example, the Your
plan may include certifying the batch based on manual sampling in a tank if you collect
the whole batch in the tank before it loaves the fuel manufacturing facility gate and the
batch is certified before release as defined in § 1090.1310(b). Similarly, as long as the
fuel remains at the facility, you may certify the batch based on secondary automatic
sampling as fuel comes out of a holding tank that you use to collect the fuel that failed to
moot a per gallon standard.
Response:
The proposed 40 CFR 1090.1315(a)(12) starts with the principle of allowing for a contingency
plan to address a failing test result. The scenarios described in the proposal are examples
intended to illustrate that principle. We intended for the examples to clarify the types of plans we
expect to see and approve as part of an in-line blending waiver. We have therefore further
modified the proposed amendment to use the commenter's suggested wording to name the
scenarios as examples. This change aligns with the description in Section II.B.3 of the proposal
preamble (89 FR 70051).
The second proposed scenario, while perhaps less common, is nevertheless a reasonable
approach that we want to highlight as a matter of clarifying the range of available solutions that
we expect to see and approve. We are revising the wording in the proposed amendment to name
this second scenario as another example. Commenter's suggestion would result in removing
content that is relevant to other companies with different facility configurations and who
otherwise would want to exercise the additional option.
The commenter suggested referencing 40 CFR 1090.1310(b) instead of naming the fuel
manufacturing facility gate to consistently identify how to meet the requirement to certify fuel
before shipment. The regulation at 40 CFR 1090.1310(b) describes a requirement to certify fuel
before it leaves a facility, which is not meaningfully different than naming the fuel
manufacturing facility gate as the point at which fuel may not pass without certification. The
commenter might have intended for the reference to 40 CFR 1090.1310(b) to include the
proposed content related to loading marine vessels in 40 CFR 1090.1310(f). We do not anticipate
a scenario of remediating fuel in a barge4oading situation that does not involve taking steps at
the facility to correct the problem. As a result, we are finalizing the language as proposed to
allow fuel manufacturers to describe their contingency plans for remediation based on steps they
can take at the facility to demonstrate that a batch of fuel meets all the standards that apply.
20
-------
Comment:
> 0012 AFPM/API
VII. § 1090.1315 In-line Blending
e. Updating ILB waivers for material changes
Proposed § 1090.1315(c)(1) lists changing analyzer hardware or programming as an example of
a material change that appears to require submitting an updated waiver request to EPA. It is
possible for analyzers to be replaced over time, due to obsolescence, improved technology, or
changing to another manufacturer's analyzer. The Associations do not believe changing the
analyzer or programming should necessitate that a new request be submitted.
Recommendation: The Associations recommend removal of the wording about changing the
analyzer hardware or analyzer programming.
Replacement of an analyzer should not constitute a material change that would require a
resubmittal of an ILB petition.
Likewise, a change in analyzer programming should not require a resubmittal of an ILB petition.
Is it expected that each time an analyzer undergoes calibration there is a requirement to resubmit
an ILB petition, as analyzer calibration can be considered a change in the programming?
If analyzer replacement is information that EPA would like to capture, this should be captured as
part of the attestation section.
Suggested Text:
§ 1090.1315 In-line blending.
* * *
(1) You must submit an updated in lino blending waiver request to EPA 60 days before
making any material change to your in-line blending process. Examples of material
changes include changing analyzer hardware or programming, changing the location of
the an analyzer if this would include a change in the sample tap location, changing the
piping configuration, changing the mixing control hardware or programming logic,
changing sample compositors or compositor settings, or expanding fuel blending
capacity. Changing the name of the company or business unit is an example of a change
that is not material.
21
-------
Response:
The proposal included a change in the process for amending approved waivers. Specifically, the
final rule revises the process for manufacturers requesting changes to existing waivers to deem
those requests to be approved 60 days after EPA acknowledges receiving the request. With that
change in the process, the only question about the documentation requirements is how much
effort fuel manufacturers should invest to keep EPA's records aligned with their current practices.
From that perspective, we have a strong interest in having fuel manufacturers send us updated
information to keep our records consistent with ongoing blending practices, including
information on analyzer hardware. As such, we are finalizing that requirement as proposed.
We are aware that programming changes might range from minor software updates to
substantially recalibrating analyzer functions. Toward that end, the instruction for submitting an
updated waiver request applies for programming changes only if the changes being made are part
of the documentation that was included with the approved waiver. Conversely, if any
documentation related to analyzer programming from the approved waiver becomes incorrect or
obsolete, the fuel manufacturer would need to submit an updated waiver request. Again, the
operative principle is that EPA's records should be consistent with the facility and processes
covered by the waiver. To clarify that principle, we are revising 40 CFR 1090.1315(c)(1) to
include a clarifying statement that material changes generally include anything that causes the
previously approved waiver to be incorrect or incomplete. The examples listed in the regulation
can therefore be understood to be things that fuel manufacturers might need to treat as material
changes.
In response to the comment that "the location of the analyzer" is a nonspecific criterion for
updating waiver requests, we added text to 40 CFR 1090.1315(c)(1) to clarify this requirement
because we agree that the requirement to update a waiver request would apply only if the
analyzer's location changes with respect to drawing samples of blended fuel from the in-line
blending operation.
We note that attestation to address any material changes to in-line blending equipment and
procedures was already required under 40 CFR 1090.1850. As described above, we consider
replacement of an analyzer to be material if that change causes the approved waiver to no longer
accurately represent the fuel manufacturer's in-line blending equipment and procedures. Routine
maintenance and calibration of analyzers would generally not require updating a facility's in-line
blending waiver as long as the fuel manufacturer did not modify the analyzer to be inconsistent
with the documentation reflected in the approved waiver.
22
-------
2.2 Demonstrating Homogeneity
Comment:
> 0012 AFPM/API
X. § 1090.1337 Demonstrating homogeneity.
a. Unsafe conditions limiting ability to sample tank for homogeneity
We appreciate the inclusion of proposed language in § 1090.1337(a)(5) permitting an alternative
to homogeneity testing during inclement weather, however there may be other equally unsafe
situations that could prevent normal sampling operations.
Recommendation: Expand the conditions that may prevent safe sampling operation beyond
inclement weather to include other conditions that may be deemed unsafe. This would be
inclusive for any number of typically short-term unsafe conditions that are not weather related.
In both the case of inclement weather and unsafe conditions, EPA approval of mixing procedure
would be required.
Suggested text:
§ 1090.1337 Demonstrating homogeneity.
* * *
(5) Your tank configuration depends on roof sampling for homogeneity demonstration,
but inclement weather or other conditions that are unsafe prevent collecting roof samples
and EPA has already approved a plan for a mixing procedure to ensure a homogeneous
batch for your specific tank configuration. EPA approval of the mixing procedure will
include consideration of product type, fill level, and other relevant parameters for specific
tank configurations and batch characteristics. Keep records to document EPA approval of
the mixing procedure, your actions to follow the approved mixing procedure, and the
forcing woathor ovont condition.
Response:
We disagree with the commenter. The commenter's suggested change would, for example, allow
exceptions for any number of unsafe conditions. We have a limited ability to monitor these
judgments and are concerned that an open-ended provision allowing for the simplified
homogeneity demonstration would increase the risk that companies would take advantage of the
provision more broadly than we intend. We are unable to find and establish objective criteria for
what should be considered unsafe conditions. In contrast, documenting an exception for
homogeneity testing based on inclement weather allows for compliance oversight by
23
-------
investigating local weather conditions at the time of sampling. We are therefore finalizing the
provision for inclement weather as proposed.
Comment:
> 0012 AFPM/API
X. § 1090.1337 Demonstrating homogeneity.
b. GTAB homogeneity demonstration is unnecessary
A general requirement for demonstrating homogeneity of gasoline treated as blendstock (GTAB)
is unnecessary unless testing for parameters to report, and the existing and proposed rules
already require demonstrating homogeneity when appropriate to account for impacts to benzene
and sulfur credit balances. If GTAB is used as a blendstock in a compliance by addition blend,
the requirements in § 1090.1320(a)(2) specify demonstrating GTAB homogeneity. Alternatively,
if GTAB is used as a blendstock in a compliance by subtraction blend, the impact on benzene
and sulfur balances is accounted for by the PCG and newly created gasoline batch properties, and
thus, requirements in § 1090.1320(a)(2) do not call for blendstock homogeneity. If GTAB is to
be used directly as a certified gasoline, the proposed requirements in § 1090.1605(b)(1) allow for
either assuming individual compartment homogeneity or demonstrating homogeneity across
compartments.
Recommendation: We recommend striking GTAB from the proposed amendments to the
requirements for demonstrating homogeneity in § 1090.1337(d).
Suggested text:
§ 1090.1337 Demonstrating homogeneity.
(d) Demonstrate homogeneity for gasoline, GTAB, and TGP using two of the procedures
specified in this paragraph (d) with each sample. For summer gasoline, the homogeneity
demonstration must include RVP measurement.
> 0013 Anonymous
I am submitting a comment related to the requirement added in this proposal for requiring
homogeneity testing on GTAB imports under 1090.1337(d). The proposal seems to clarify that
GTAB imports are subject to a homogeneity testing requirement, however, there is no reporting
of GTAB parameters under the regulations. The reporting for GTAB consists of volume only, as
the GTAB quality is eventually reflected in the gasoline that is produced using the GTAB at the
24
-------
production facility. Since the GTAB parameters are not reported or used to determine
compliance, the requirement to demonstrate homogeneity is overly costly and burdensome.
> 0015 Susan Lafferty
EPA is proposing to require homogeneity testing for GTAB in § 1090.1337(d), but GTAB
parameters are not sampled and tested as stated in § 1090.1615(d)(1). GTAB is treated as a
blendstock, so homogeneity requirements apply when certifying the final blend. The proposed
language appears to require homogeneity testing prior to certification as a final blend ie, upon
import of GTAB, and therefore should not be adopted.
Response:
We agree with commenters that we should not require a homogeneity demonstration for GTAB
and in response to these comments have removed the proposed requirement from the final rule.
Comment:
> 0012 AFPM/API
X. § 1090.1337 Demonstrating homogeneity.
c. ASTM D4052 precision statement
We appreciate the practical flexibility offered in proposed § 1090.1337(f)(3) to use ASTM
D4052 to "calculate the homogeneity criteria based on the reproducibility of the test method at
the limit of the valid range for testing, even if measured results extend beyond the valid range,"
We would also highlight there is an apparent inconsistency between what was provided in the
preamble to the proposed rule versus what is included in the regulatory text on this topic. We
affirm that the desired flexibility for calculating the reproducibility while assessing homogeneity
is consistent with what is included in the proposed regulatory text.
We also appreciate the December 31, 2026, date offered for API gravity greater than 66. ASTM
D4052 is a method that is used extensively by the fuels industry for measuring density and API
gravity. The industry is well down the path of developing a new precision statement above 66
API gravity and expect to be finished prior to that date. We would ask in the final rule preamble
or response to comments that EPA acknowledge that, should the ILS and ASTM D4052 revision
be delayed, EPA would consider a further extension. Also, since the version of ASTM D4052
that is incorporated by reference in § 1090.1337(d) in the final rule will not reflect the updated
precision statement, regulatory text should be included to permit use of the updated
reproducibility when testing API gravity greater than 66. Without this, parties would have to
revert to the reproducibility of the version of ASTM D4052 that was incorporated by reference or
EPA would have to incorporate the newer version in a subsequent rulemaking prior to December
31, 2026.
25
-------
A concern has also been raised that ASTM D4052 does not have a precision statement for the
API gravity range of 45-51. Renewable diesel typically has an API gravity in this range and
could pose an increasing concern as batches of renewable diesel increase. ASTM has not started
on updating the precision in this range.
Recommendation: Include regulatory text that allows for the use of the updated D4052
reproducibility for API gravity greater than 66 after December 31, 2026. Also, include a
secondary date that permits the use of the ASTM D4052 precision statement for the 29-45 API
for distillates, base stocks, and lubricating oils between 45-51.
Suggested text:
§ 1090.1337 Demonstrating homogeneity.
* * *
(3) If using ASTM D4052 (incorporated by reference in § 1090.95) for measuring density
or API gravity to demonstrate homogeneity through December 31, 2026, you may
calculate the homogeneity criterion based on the reproducibility of the test method at the
limit of the valid range for testing, even if measured results extend beyond the valid
range. This is extended to December 31, 2027 for distillate fuel that has an API gravity
between 45-51. After December 31, 2026, you may utilize the reproducibility from a
subsequent version of D4052 for measuring density or API gravity greater than 66 API.
Response:
We do not have authority under the Clean Air Act to adopt a regulatory provision that allows for
the use of future versions of ASTM D4052. However, we recognize that the changes under
development to address the range issues for ASTM D4052 are intended to allow for evaluating
homogeneity based on API gravity for the fuels being produced today, consistent with existing
industry practice. As a result, we are revising the proposed amendment to include a provision
allowing companies to use an updated, future version of ASTM D4052 if it addresses the
problem by properly identifying reproducibility at measured values that are not included in
ASTM D4052-22, subject to EPA approval. EPA's review of such a request would include
consideration of any additional changes to the test method that may be objectionable. If we
approve the use of a future version of ASTM D4052, we would plan to pursue a corresponding
change to the regulation at the next opportunity.
We appreciate that work is going forward to update ASTM D4052 to determine reproducibility
for API gravity above 66°, but we also recognize the need for the regulation to include a
contingency plan to allow for continued use of the interim provision if that work is delayed. We
are therefore revising the proposed amendment to include an additional provision allowing EPA
26
-------
to approve continued use of the already allowed interim provision beyond December 31, 2026, if
ASTM D4052 has not been updated by that time.
For both the approval to use the updated version of ASTM D4052 and for any approved
extension of the deadline, we would expect a request to come from a company or industry
association. We would then evaluate the request and make a decision that would apply broadly
for any party affected by the decision.
We believe that the regulatory provision as proposed, and amended as described above, fully
addresses the concerns raised about testing with renewable diesel fuel or other products with API
gravity between 45-51°. We note that the proposal described a process for identifying
reproducibility without naming a specific range of API gravity values. The additional provisions
described above for the final rule also apply independent of API gravity values. This means that
companies may rely on the interim provision for products with API gravity between 45-51°
without EPA involvement through 2026. They may continue to rely on the interim provision after
December 31, 2026, subject to EPA approval. Once ASTM D4052 has been updated with
reproducibility for API gravity between 45-51°, we should be able to evaluate a request to use
that updated method for density measurements to make the needed homogeneity demonstrations.
Comment:
> 0012 AFPM/API
X. § 1090.1337 Demonstrating homogeneity.
d. Homogeneity determinations where test results are below the PLOQ, LLOQ,
or valid range limit of the test method
In the proposed rule, EPA provided additional criteria for how test results below the applicable
pooled limit of quantification (PLOQ), lower limit of quantification (LLOQ), or valid range limit
should be managed with respect to homogeneity determinations. These new requirements
effectively increase the stringency of the homogeneity requirements and unfairly penalize fuel
manufacturers with low sulfur and benzene concentrations in their products, without providing
significant improvements to certification data quality or to actual product quality. This proposed
change unnecessarily disincentivizes the production of low sulfur/benzene fuel, contrary to the
emission reduction goals of 40 CFR part 1090.
As proposed, a gasoline manufacturer with very low sulfur and/or benzene may be required to do
additional testing (either to test for additional parameters to demonstrate homogeneity, or to do
full certifications on each tank layer to certify as a non-homogeneous tank, or to further mix a
tank that is already on-spec and sufficiently mixed), while a gasoline manufacturer with
relatively high concentrations of sulfur and/or benzene may be able to certify a tank with a less
rigorous sampling and testing regime and with less rigorous final product mixing practices.
If sulfur or benzene homogeneity test results (e.g., results from testing upper, middle, or lower
tank samples for sulfur or benzene) are below these limits, consider: (1) there is no evidence the
27
-------
batch is out of compliance with any per-gallon limit (to the contrary, if it is sulfur, there are
multiple data points indicating the batch complies with the per-gallon limits), and (2) average
compliance is likely also to be reported at or near the PLOQ, LLOQ, or valid range limit
(presuming the average sample test result is likely also to be very low if upper/middle/lower
results are all that low).
Furthermore, EPA is confirming in proposed section § 1090.1337(b) that homogeneity samples
are certification samples. EPA has previously acknowledged in other sections of the rule that
homogeneity test results are still considered valid test results for certification testing and per-
gallon compliance without making exceptions for test results that are below the PLOQ, LLOQ,
or valid range limits. The same approach with respect to LLOQ or PLOQ should be applied in
both contexts.
Finally, consider that, practically speaking, only sulfur and benzene are in question here. The
other parameters that may be used to demonstrate homogeneity (i.e., RVP and API/density) do
not come back below PLOQs/LLOQs/method limits. As such, in the summer, as required by §
1090.1337(d), at least one of the two parameters used to demonstrate homogeneity for gasoline
batches will not establish homogeneity via rounding up to the PLOQ/LLOQ/method limit
(because at least one must be RVP).
Recommendation: Rather than disincentivize the production of low sulfur/benzene fuels, which
is contrary to the emission reduction goals of the rule, and introduce inconsistency with the
treatment of other certification results, EPA should continue to allow fuel manufacturers to use
sulfur and benzene test results that are below the PLOQ, LLOQ, or valid range limits, consistent
with ASTM requirements for managing test results in those ranges (i.e., by rounding the
measured result up to the lower limit).
Suggested text:
§ 1090.1337 Demonstrating homogeneity.
* * *
(1) Do not use tost results for a given tost method for a parameter to demonstrate
homogeneity if multiple If a measured values are is at or below the test method's PLOQ,
LLOQ, or valid range, you may use the PLOQ, LLOQ, or valid range, as applicable,
when using that test result to evaluate homogeneity. Alternatively, you ¥eu may instead
use a different test method as allowed under this subpart to get test results with the same
parameter or with a different parameter.
28
-------
Response:
The homogeneity demonstration assesses whether a tank of fuel is well mixed. If a tank of fuel is
not well mixed, individual samples taken at different points might have very different measured
values for one or more parameters. EPA's decision to trust a single tap sample to represent a tank
of fuel depends on establishing that proper mixing causes the fuel to have consistent properties
throughout the tank volume. This led EPA to allow for certifying a tank of fuel based on worst-
case test results in cases where the tank did not meet homogeneity requirements.
As noted by the commenter, we adopted the requirement to demonstrate gasoline homogeneity
based on two parameters largely because there is no in-depth study establishing how much
mixing is associated with a certain level of variability in measured values for all the different
parameters. In each case, a valid result for the homogeneity demonstration is based on
consistency of measured values relative to the reproducibility of the test method at the measured
levels for each parameter. If benzene or sulfur levels are so low that they are below PLOQ,
LLOQ, or the valid range for the test method, there would be no expectation that measured
values would vary substantially, even with a highly stratified fuel volume. It is also self-evident
that parameter values below PLOQ, LLOQ, or the valid range of the test method do not provide
a useful indicator for differentiating measured values across the fuel volume. As a result, the
suggested approach of allowing homogeneity demonstration based on very low measured
benzene or sulfur levels would contribute little or nothing to add assurance that the fuel volume
is well mixed for certifying with respect to other parameters. Section II.E.2 of the proposal
preamble described these concerns as the motivation for the proposed changes (89 FR 70053).
The commenter does not address any of these concerns.
The commenter expresses a concern about increasing stringency, which would involve a change
to require a greater degree of mixing. Steering labs away from using parameter measurements
with very low quantities will not increase the level of mixing required. As a result, we do not
agree that the proposed approach for demonstrating homogeneity with very low measured values
would increase the stringency of the homogeneity specifications.
We also disagree with the claim that the proposed change would penalize fuel manufacturers
with low sulfur or benzene concentrations. As noted, homogeneity testing must include RVP
measurements for summer gasoline, and RVP testing is also available for gasoline otherwise. We
are revising the proposed amendments to facilitate homogeneity measurements with density, as
described in the response to those comments. Also, if a batch of gasoline has either benzene or
sulfur levels that are above the disqualifying low levels, those measurements are also available
for making the demonstration. Homogeneity testing may involve any two of the four candidate
parameter measurements, so in our view an emphasis on using measurements with the lowest
levels of benzene or sulfur, would perhaps be to avoid a rigorous assessment of mixing.
While measurements at or below the specified lower limits do not bring into question the
compliance with per-gallon standards, measuring very low levels of one parameter provides no
helpful assessment of mixing, which can compromise the compliance demonstration for other
parameters. Relying on measurements below PLOQ, LLOQ, or the valid range of the test method
29
-------
would not add value toward assessing the level of mixing and would therefore effectively be a
circumvention of the requirement to establish homogeneity based on two parameters.
We are finalizing the proposed homogeneity provisions related to very low measured values
without modification.
30
-------
2.3 Retaining Samples
Comment:
> 0012 AFPM/API
XI. § 1090.1345 Retaining Samples
The Associations believe the wording in proposed § 1090.1345(a)(6) could be misinterpreted as
requiring additional storage capacity for an increased number of retains in order to ensure there
is always an untested sample.
Recommendation: The Associations recommend a modification of the wording to §
1090.1345(a)(6) that still captures the intent of having a retained sample that is representative of
the batch.
Suggested Text:
§ 1090.1345 Retaining samples.
* * *
(6) If you have only a single sample for testing, keep that sample after testing is
complete. If you collect multiple samples from a single batch, keep an untested sample or
other sample that represents the batch, or you croato a hand blond, select a representative
sample as follows:
Response:
Section II.F of the proposal preamble described a simple approach to keep a single untested
sample for each batch (89 FR 70053). We believe that the commenter's approach is appropriate
in all cases except for summer gasoline. We recognize that, except for summer gasoline,
retaining either tested or untested samples could allow for later testing to properly characterize
the batch. Summer gasoline is subject to RVP testing, which requires that the retained sample be
untested for a valid test result. In other words, keeping a tested sample of summer gasoline
would prevent any repeat testing for RVP because the RVP test methods require measurements
with untested samples. We have therefore revised the proposed amendment to require a single
retained sample of untested summer gasoline that represents the batch, and in other cases they
must retain any (tested or untested) sample that represents the batch. This approach relies on the
principle that demonstrating homogeneity allows for any sample from the batch to be used for
measurements to establish test values to characterize the batch. Untested samples allow for
testing summer gasoline for RVP, and any representative sample provides a useful starting point
for measuring any other parameters.
31
-------
2.4 PCG
Comment:
> 0009 0NEOK
§ 1090.1320 Adding blendstock to PCG.
1090.1320(b)(5); 1090.1320(b)(6); 1090.1320(b)(5) If the quality assurance testing under
paragraph (b)(4) of this section shows that certified butane or certified pentane fails to meet one
or more of the standards specified in §1090.250 or §1090.255, the certified butane or pentane
received from that distributor during that testing period is deemed to be in violation of the
relevant per-gallon standard. Any later shipment of certified butane or certified pentane received
from that distributor will also be deemed to be in violation of the relevant per-gallon standard
unless another quality assurance test is conducted demonstrating that certified butane or certified
pentane received from that distributor meets the standards specified in §1090.250 or §1090.255.
1090.1320(b)(6) If certified butane or certified pentane is deemed to be in violation under
paragraph (b)(5) of this section, the certified butane or certified pentane blender must calculate
its compliance obligations using paragraph (a)(2) of this section using the test results from the
quality assurance program and obtain any necessary sulfur or benzene credits.
• As certified butane and certified pentane are excluded from ABT calculations, the inclusion
of certified butane and certified pentane blendstock volumes in the blender's compliance
calculations does not align with the concept of certified butane and certified pentane. The
logistic complexity of assigning quality assurance program test results to blended volumes
under 1090.1320(b)(6) will be problematic when a single certified butane or certified pentane
distributor supplies multiple facilities within an organization. A single oversight sample
which fails to meet the per-gallon standards at 1090.250 or 1090.255 should at most be
assigned to the volume of product tested under the single failed sample. Additionally, with
respect to benzene, clarity is needed to identify the intent of 1090.1320(b)(6) as currently
written. Should the blender use the 1090.250 standard of 0.03 to calculate compliance values
for a failed certified butane oversight sample, rather than the benzene average standard of
0.62 at 1090.210? ONEOK would suggest alternative approach of stepped consequence to
failed oversight samples.
• E.g., If the quality assurance testing under paragraph (b)(4) of this section shows that
certified butane or certified pentane fails to meet one or more of the per-gallon standards
specified in §1090.250 or §1090.255, the certified butane or pentane blender must obtain an
immediate additional quality assurance sample and test to meet the standards at §1090.250 or
§1090.255. The certified butane or certified pentane blender may also validate with a new
sample from the certified butane or certified pentane distributor. If the results of the
immediate additional sample meet the per-gallon standards at §1090.250 or §1090.255, no
further action is required.
32
-------
> 0012 AFPM/API
VIII. § 1090.1320 Adding blendstock to PCG
b. Sampling oversight failures
The Associations recommend EPA modify § 1090.1320 to address failures of certified butane and
certified pentane quality assurance. As certified butane and certified pentane are excluded from
fuels averaging, banking, and trading calculations, the inclusion of certified butane and certified
pentane blendstock volumes in the blender's compliance calculations does not align with the
concept of certified butane and certified pentane. The logistic complexity of assigning quality
assurance program test results to blended volumes under proposed § 1090.1320(b)(6) will be
problematic when a single certified butane or certified pentane distributor supplies multiple
facilities within an organization. A single oversight sample which fails to meet the per-gallon
standards at § 1090.250 or § 1090.255 should at most be assigned to the volume of product
tested under the single failed sample. Additionally, with respect to benzene, clarity is needed to
identify the intent of § 1090.1320(b)(6) as proposed. Should the blender use the § 1090.250
standard of 0.03 to calculate compliance values for a failed certified butane oversight sample,
rather than the benzene average standard of 0.62 at § 1090.210?
Recommendation: Due to the complexity discussed above, we suggest an alternative approach of
a stepped consequence to failed oversight samples with an allowance similar to that given in §
1090.1315(a)(10) for in-line blending missed head, middle, and tail samples.
Suggested text:
§ 1090.1320 Adding blendstock to PCG.
(b) * * *
(5) If the quality assurance testing under paragraph (b)(4) of this section shows that
certified butane or certified pentane fails to meet one or more of the per-gallon standards
specified in § 1090.250 or § 1090.255, the certified butane or certified pentane received
from that distributor during that testing period is deemed to be in violation of the relevant
per gallon standard. Any later shipment of certified butane or certified pentane received
from that distributor will also be deemed to be in violation of the relevant per gallon
standard unless another quality assurance tost is conducted demonstrating that certified
butane or certified pentane received from that distributor meets the standards specified in
§ 1090.250 or § 1090.255. blender must obtain an immediate additional quality assurance
sample and test to meet the standards at § 1090.250 or § 1090.255. If the results of the
immediate additional sample meet the per-gallon standards at § 1090.250 or § 1090.255,
no further action is required. If more than 10% of quality assurance samples from a
distributor fail to meet the standards at § 1090.250 or § 1090.255 in a given compliance
year, once this percent of failed oversight tests has been reached, the blender must
33
-------
consider the certified butane or certified pentane supply uncertified and test under §
1090.1320(a)(1) or § 1090.1320(a)(2).
(6) If certified butane or certified pontano is doomed to bo in violation under paragraph
(b)(5) of this section, the certified butane or certified pentane blender must calculate its
compliance obligations using paragraph (a)(2) of this section using the test results from
the quality assurance program and obtain any necessary sulfur or benzene credits.
Response:
These new regulatory provisions are intended to provide clarity to the regulated community
regarding what happens when a quality assurance test conducted under 40 CFR 1090.1320(b)(4)
shows that certified butane or certified pentane received by the blender does not meet an
applicable standard in 40 CFR 1090.250 or 1090.255, respectively. If the quality assurance test
conducted by the blender shows that the certified butane or certified pentane does not meet an
applicable standard, the certified butane or certified pentane from that distributor during that
testing period is deemed in violation of the applicable per-gallon standard unless another quality
assurance test is conducted during that testing period that shows compliance (in which case the
remaining volumes from that distributor during the testing period would be considered in
compliance with the applicable per-gallon standard).
The commenters recommended that if a quality assurance test shows noncompliance, the blender
be required to immediately re-test the certified butane or certified pentane and that the certified
butane or certified pentane would only be deemed in noncompliance if the re-test also showed
noncompliance or if more than 10 percent of the quality assurance samples from a single
distributor fail to meet the applicable standard. We do not believe that multiple failed sampling
results are required and maintain that a single quality assurance test result that shows that the
certified butane or certified pentane does not meet an applicable standard is sufficient for that
fuel to be deemed in noncompliance. The regulations already allow the blender to re-test the
same shipment or a later shipment of certified butane or certified pentane from that distributor,
and if the re-test shows compliance with the applicable standards, the re-tested batch and all
subsequent volumes from that distributor during that testing period would be deemed in
compliance with the standards. Further, the regulations require that the blender test the first
shipment of product it receives from each distributor during each testing period, so if the blender
opts to re-test certified butane or certified pentane from a distributor after receiving a failing
result from the first batch received and the re-test shows compliance, only the initial volume of
certified butane or certified pentane from that distributor during that testing period would be
deemed to be in noncompliance.
The commenters also seek clarity on consequences that result from a failed quality assurance test
result, including the applicable benzene standard the noncompliant certified butane or certified
pentane is subject to and how to handle a situation where there is a failed quality assurance test
result for a distributor that supplies multiple facilities within an organization. Certified butane
and certified pentane are subject to per-gallon standards in 40 CFR 1090.250 and 1090.255,
34
-------
respectively, and this fuel would be deemed in violation of the applicable per-gallon standard if
the results of a quality assurance test demonstrate that the certified butane or certified pentane do
not meet these standards. For purposes of averaging, banking, and trading, the fuel deemed to be
in violation of a per-gallon standard becomes subject to the benzene standard in 40 CFR
1090.210.
Regarding how to handle a situation where there is a failed quality assurance test result for a
distributor that supplies multiple facilities within an organization, the regulations have been
updated to specify that only the certified butane or certified pentane received by the butane
blending facility or pentane blending facility where the failing quality assurance test occurred
would be deemed in violation of the applicable per-gallon standard.
Finally, per one commenter's request, we have revised the proposed amendment to state that, if
certified butane or certified pentane is deemed in violation of a per-gallon standard, the blender
may calculate their compliance obligations using either compliance by subtraction under 40 CFR
1090.1320(a)(1) or compliance by addition under 40 CFR 1090.1320(a)(2).
Comment:
> 0011 bp America
II. Sampling and Determining Homogeneity of Commingled PCG
As API's comment notes, bp understands that EPA's proposed language at Section
1090.1320(a)(l)(i)(A) requires parties to demonstrate homogeneity on a consolidated batch of
PCG where "multiple batches of PCG" are commingled together prior to creating a new final
batch through compliance by subtraction. As API notes, it is often operationally infeasible and
could adversely impact gasoline supply to require manufacturers to segregate each PCG batch
and separately confirm homogeneity; we do not understand the proposed regulatory language to
require such sampling procedures. Rather, a gasoline manufacturer can confirm homogeneity
across a single volume comprised of multiple PCGs.
In addition, given limited tankage at many terminals, it is imperative operationally that EPA
allow pipeline sampling for confirming homogeneity. As described in API's comment, such
pipeline sampling should allow manufacturers to take one to three spot samples depending on the
volume of PCG, consistent with proposals EPA has made in this rulemaking for in-line blending.
EPA's proposed amendment at Section 1090.1320(a)(l)(i)(A) cross-references the existing
homogeneity rules at Section 1090.1337, which themselves reference the tap sampling
techniques appropriate for tanks described in Section 1090.1335(b)(2). But EPA currently
recognizes at Section 1090.1335(b)(3) that tap sampling may be impractical in some situations
and directs parties to ASTM D4057, which provides for manual pipeline spot sampling, bp
recommends that EPA expressly state that manufacturers may rely on pipeline sampling for PCG
sampling under Section 1090.1320 as described in API's comment.
35
-------
> 0012 AFPM/API
VIII. § 1090.1320 Adding blendstock to PCG
a. PCG homogeneity
The proposed requirement in § 1090.1320(a)(l)(i)(A) to demonstrate homogeneity of each batch
of previously certified gasoline (PCG) used in a compliance by subtraction blend is unclear as to
whether homogeneity is required for each of the "multiple batches of PCG" or just for the total,
consolidated volume of PCG. The Associations read this proposed amendment as requiring
homogeneity for the consolidated PCG batches, and not on an individual batch basis. Requiring
homogeneity on an individual PCG batch basis would be unworkable in regions where pipelines
commingle shipments and would require additional tankage all along the distribution system to
isolate individual batches of PCG. Additionally, the proposed requirement is not amenable to the
current practice of sampling PCG shipments on fungible pipelines and would similarly require
additional and largely unavailable tankage.
Recommendation: We request that EPA clarify that PCG being used for compliance by
subtraction blending and comprised of multiple batches of PCG does not require a homogeneity
demonstration on each of the batches of PCGs used, and we request that EPA adopt a pipeline
sampling methodology allowing manufacturers to take one to three spot samples depending on
volume as suggested in § 1090.1315(a)(9) for in-line blending or confirm compliance of all
samples and use the lowest benzene and lowest sulfur measured.
Suggested text:
§ 1090.1320 Adding blendstock to PCG.
* * *
* * *
* * *
(A) Sample and test the sulfur content, benzene content, and oxygenate content of each
batch of PCG using the procedures in § 1090.1350. Demonstrate homogeneity as
specified in § 1090.1337 if blending involves multiple batches of PCG, or if a single
batch of PCG was certified without demonstrating homogeneity under § 1090.1337(a)(1).
if the batch of PCG being used for blending is comprised of multiple batches or if a
single batch of PCG was certified without demonstrating homogeneity under §
1090.1337(a)(6), with the following exceptions:
(1) A PCG batch comprised of a single batch or a PCG batch comprised of multiple
batches that are well mixed do not require homogeneity testing.
36
-------
(2) The blending manufacturer does not need to test PCG for oxygenate content if they
can demonstrate that the PCG does not contain oxygenates as specified in paragraph
(a)(l)(i)(C) of this section or § 1090.1310(e)(1).
(3) For PCG sampled during pipeline transfer sample as specified in § 1090.1335(c) or
take spot samples of the received volume as follows. Demonstrate homogeneity
following § 1090.1337(d) and § 1090.1337(f) except that RVP measurement is not
required for summer gasoline:
(7) If volume is less than one million gallons, a single spot sample may be utilized to
represent the volume.
(77) If volume is greater than one million gallons but less than 2 million gallons, two
evenly distributed spot samples are required.
(777) For volumes greater than two million gallons, three evenly distributed spot samples
are required.
(7'v) Composite the samples and test and report the benzene and sulfur of the composite.
(v) The homogeneity testing requirement does not apply if each sample is tested for
benzene and sulfur and the lowest test result of each parameter is reported.
Response:
In response to comments, we have revised the proposed amendment to clarify that the
homogeneity demonstration applies for the consolidated batch if the blending process involves
multiple batches of PCG.
The comment included a specific suggestion to waive the homogeneity testing requirement for
blending with multiple batches that are well mixed. This approach is circular, since we rely on
homogeneity testing to determine that fuel is well mixed. As a result, we are not including this
change in the regulations.
We recognize the validity of using multiple pipeline transfer samples as suggested by the
commenter and have included this provision in 40 CFR 1090.1320(a)(l)(i). We have revised the
proposed amendment to say that homogeneity testing is not required for volumes less than one
million gallons, rather than implying that outcome by saying that only a single spot sample is
required for volumes less than one million gallons.
We have also added text in the final rule to clarify how fuel manufacturers would demonstrate
compliance if fuel does not meet homogeneity requirements, for either tank or pipeline
configurations. Specifically, they would use 40 CFR 1090.1337(a)(4), which allows for using
worst-case test values instead of demonstrating homogeneity. In the case of compliance by
37
-------
subtraction, it is important to note that the samples with the lowest measured values of benzene,
sulfur, and oxygenate are worst-case test values.
Comment:
> 0011 bp America
I. Determining PCG Characteristics for "Wrapper Batches"
EPA's proposed amendments include two key changes to Section 1090.1320, which sets forth
sampling and testing requirements for gasoline manufacturers that add blendstock to PCG to
produce a new batch of gasoline. First, where manufacturers must exclude the sulfur and benzene
content of the PCG through compliance by subtraction, EPA proposes that they must
"[demonstrate homogeneity" for the PCG "as specified in § 1090.1337 if blending involves
multiple batches of PCG, or if a single batch of PCG was certified without demonstrating
homogeneity under § 1090.1337(a)(4)." Proposed Regulatory Text, 40 C.F.R. §
1090.1320(a)(l)(i)(A), 89 Fed. Reg. at 70,079. EPA additionally proposes to clarify in a new
provision at Section 1090.1320(a)(l)(i)(D) that, if manufacturers are "unable to measure a PCG
parameter," they "must comply using either the presumed value for the PCG volume or an EPA-
approved alternative value as described in § 1090.1710(g)." Proposed Regulatory Text, 40 C.F.R.
§ 1090.1320(a)(l)(i)(D), 89 Fed. Reg. at 70,079.
Both proposed changes to the regulations at Section 1090.1320 are implicated where gasoline
manufacturers send and receive PCG and blendstock shipments on pipelines for further blending
and certification downstream. Specifically, to meet gasoline demand as efficiently and cost-
effectively as possible, gasoline manufacturers sometimes ship batches of blendstock on a
pipeline to be blended with PCG at the final destination, tested, and then certified using either
compliance by addition or compliance by subtraction. For purposes of preserving gasoline
quality on a fungible pipeline system, such batches must be "wrapped" in PCG, meaning a
certain volume of PCG both precedes and follows the blendstock volume. For example, a batch
might initially be comprised of 12,500 barrels of PCG followed by 100,000 barrels of blendstock
followed by 12,500 barrels of PCG. Such batches are commonly referred to as "wrapper
batches."
Sampling and testing these PCG volumes preceding and following the blendstock in accordance
with the current regulations and the proposed amendments poses substantial operational
challenges. For instance, there may be multiple junctures along the pipeline where the wrapper
batch shipment exits the pipeline into breakout tankage and is temporarily stored in separate
PCG and blendstock tankage. In such instances, the pipeline conservatively includes some
volume of PCG from the blendstock interface in the blendstock tankage to ensure the quality of
the remaining PCG is maintained. That causes some volume of PCG to mix with the blendstock
shipment, while the remaining PCG is routed to breakout tankage where it mixes with other
PCG. That mixed PCG from the breakout tankage is then used again to "wrap" the front and
back of the blendstock shipment to move it further down the pipeline. This process may repeat
multiple times along a wrapper batch's journey on a pipeline to its final destination. Figure 1
illustrates the process.
38
-------
Figure 1. Example of Wrapper Batch Movement into Breakout Tankage
12.5kb 12.5kb
As an example, the initial wrapper batch is comprised of 12,500 barrels of PCG followed by
100,000 barrels of blendstock followed by 12,500 barrels of PCG. However, by the end of the
pipeline, the batch is comprised of a total of 110,000 barrels of a PCG/blendstock mixture
(represented by the dark green and light green portions in Figure 1), with PCG from the breakout
tanks on both ends (represented by the yellow portions in Figure 1).
a. Challenges Posed by Current and Proposed Regulations
The current regulations at Section 1090.1320 require fuel manufacturers to sample and test PCG
prior to blending with blendstock into new, certified batches, and the proposed regulations would
require two relevant additional testing burdens: (1) gasoline manufacturers must demonstrate
homogeneity of "multiple" commingled batches of PCG; and (2) if PCG cannot be sampled in
accordance with the regulations, presumed values of zero for sulfur and benzene must be used
unless EPA approves an alternative value. Moreover, under the Renewable Fuel Standard
regulations, gasoline manufacturers must exclude the full PCG volume from renewable volume
obligation calculations to avoid double counting the volume. See 40 C.F.R. § 80.1407(c).
For the PCG wrappers that are commingled with other fuel in a fungible pipeline and in breakout
tanks along the course of the pipeline, gasoline manufacturers cannot separately sample the PCG
or determine its homogeneity. Although the PCG and blendstock shipment can be separately
sampled and tested at their point of origin, it is not possible to subsequently sample in
accordance with the current and proposed Section 1090.1320 requirements the portion of the
PCG that is mixed with the blendstock as part of the shipment process.
Further, as EPA's proposed amendments describe, manufacturers are obligated to fall back on
presumed values under Section 1090.1710(g) to characterize PCG for purposes of Part 1090
reporting when they are unable to sample the fuel. However, that approach impacts the sulfur and
benzene credit markets with unnecessary purchases—especially where it must be employed by
default every time those parties ship blendstock on a fungible pipeline.2 It is also not necessary
to ensure the integrity of the sulfur and benzene characteristics of the gasoline pool because
2 bp appreciates EPA's recognition at proposed Section 1090.1320(a)(l)(i)(D) that using presumed values under
Section 1090.1710(g) is a valid pathway for compliance.
39
-------
using the presumed values at Section 1090.1710(g) does not reflect the characteristics of the
certified gasoline.
b. Proposed Solution
bp recommends that EPA provide manufacturers with an alternative method to characterize the
commingled PCG/blendstock portion of the wrapper batches. Specifically, bp proposes that EPA
allow fuel manufacturers to mathematically determine the sulfur and benzene characteristics of
the blended PCG wrapper batches based on the change in the blendstock's properties between its
point of origin and its ultimate destination. Such a calculation merely requires sampling the
blendstock prior to shipment as well as at the end of the pipeline, which is operationally feasible
to do. With an understanding of the properties of the blendstock before and after shipment, it is
possible to ascertain mathematically the characteristics of the PCG and use those characteristics
in compliance reporting for compliance-by-subtraction under Section 1090.1320.
For example: A batch of blendstock illustrated in Figure 1 is tested at the point of origin with
sulfur and benzene values of 7 ppm and 0.46 volume percent, respectively. Such blendstock
might be shipped for downstream blending because it does not meet other commercial gasoline
specifications due to low octane, higher RVP, or other parameters. The PCG wrappers described
in Figure 1 would then be included before and after the blendstock as it is shipped on a fungible
pipeline. At its intended destination, the resulting mixture of blendstock and PCG is sampled for
sulfur and benzene values of 9 ppm and 0.48 volume percent. The volume of PCG mixed into the
blendstock in the course of its shipment can be determined because the overall volume of the
blendstock has increased by 10,000 barrels.
A simple equation can calculate the sulfur and benzene characteristics of the 10,000 barrels of
added PCG: (Value of Sulfur or Benzene at Destination x Volume at Destination — Value at
Origin x Volume at Origin) / (Volume at Destination — Volume at Origin). Applying the
equation to this hypothetical, the sulfur and benzene of the added PCG would be 29 ppm and
0.68 volume percent.
The suggested text of the new regulatory provision could state:
§1090.1320 Adding blendstock to PCG.
(1) Compliance by subtraction.
(vii) Where PCG is used as a buffer in shipping blendstock on a pipeline and a portion of
PCG commingles with the blendstock, determine the volume of added PCG based on
difference in the blendstock's volume between the point it entered the pipeline system
and its intended destination. The sulfur content and benzene content of the PCG are
determined as follows:
40
-------
Spcg = (Sd x Vd - So x Vo) / (Vd - Vo)
Bzpcg = (Bzd x Vd - Bzo x Vo) / (Vd - Vo)
Where:
Spcg = Sulfur characteristics of PCG mixed into the blendstock
Bzpcg = Benzene characteristics of PCG mixed into the blendstock
S = Sulfur, ppm
Bz = Benzene, volume %
V = Blendstock volume
D = At Destination terminal
O = At Origin terminal
The lack of such an alternative method of compliance has material consequences for gasoline
manufacturers that must otherwise use unrealistic presumed values for the PCG at substantial
potential financial burden. This path also avoids gasoline manufacturers petitioning EPA for
approval of alternative values for PCG where such values can be readily and reliably calculated
and verified as part of the annual attestation process.
Response:
The regulations currently include a provision for the scenario of being unable to measure a PCG
parameter by allowing for compliance using either the presumed value for the PCG volume or an
EPA-approved alternative value as described in 40 CFR 1090.1710(g). The commenter's
discussion of wrapper batches fits well within the existing provisions to allow for compliance
based on an EPA-approved alternative value under 40 CFR 1090.1710(g). We are therefore not
making further changes to the regulation at this time in response to this comment.
41
-------
2.5 Precision and Accuracy Demonstration
Comment:
> 0005 Alex Lau
If the 'between-lab bias' component is the dominant contributor towards test method
reproducibility (R), Z or Z' may remain consistently on one side of zero
In other words, when between-lab bias is the dominant contributing factor towards method R, Z
and/or Z' remaining on one side of zero is not an 'unusual' expectation
Response:
We thank the commenter for their feedback. As discussed in the proposed preamble, we may
consider future amendments that account for bias in establishing pass criteria for quality testing
under 40 CFR 1090.1375.
Comment:
> 0009 ONEOK
§ 1090.1375 Quality control procedures.
1090.1375(d) Demonstrating accuracy by participating in crosscheck programs. You may meet
accuracy requirements under paragraph (c) of this section by comparing your results to the
accepted reference value in an inter-laboratory crosscheck program sponsored by ASTM
International or another VCSB at least three times per year (two times per year for RVP), subject
to the following provisions:
• Requirement seems to limit an individual/company from utilizing its own ILCP if it has more
than 16 participating labs and requires the use of a 3rd party subscription that might be more
inclusive than what is needed. The below should be struck -
Demonstrating accuracy by participating in crosscheck programs. You may meet
accuracy requirements under paragraph (c) of this section by comparing your results to
the accepted reference value in an inter-laboratory crosscheck program sponsored by
ASTM International or another VCSB at least three times per year (two times per year for
RVP), subject to the following provisions:
> 0012 AFPM/API
XII. § 1090.1375 Quality control procedures.
a. § 1090.1375(d)(1) Proposed accuracy demonstration using ILCP
42
-------
In general, we believe the new proposed language complicates the option to demonstrate
accuracy via Inter-Laboratory Cross Check Program (ILCP) and adds additional liability to the
point that laboratories may choose to discontinue ILCP participation. One example is that the
proposal will likely cause instruments to fail accuracy demonstration for reasons that are not
valid and that are outside of the control of the lab. These scenarios include:
• ILCP samples that have lost integrity during shipment (loss or light ends and/or
empty/broken bottles).
• ILCP samples lab that were not properly blended/mixed.
• ILCP samples which were received late without time to test and report results prior to the
deadline.
• ILCP samples submitted but not enough labs participated to develop an ARV.
Because of the added risks above, along with the proposed additional tracking/documentation
requirements and tight timing restrictions, laboratories may shift to accuracy demonstration using
check standards. Furthermore, the proposed language may cause laboratories to stop
participation in ILCPs like ASTM RFG and ULSD completely. Even if a laboratory chooses by
regulation to demonstrate accuracy by running check standards, laboratories have concerns that
EPA will hold them accountable for failing results from an ILCP even if participation in the ILCP
is for non-regulatory reasons. Participating in ILCPs is an industry best practice and laboratories
receive the benefits of allowing sites to monitor performance as compared to peers and to
identify opportunities for improvement. EPA's proposal will cause additional liability for ILCP
participation and added complications. This may result in a reduction of ILCP participation
representing a backsliding on routine industry review and reducing the ability to identify
performance issues which would have otherwise been identified.
The EPA should permit an Accepted Reference Value (ARV) and the necessary statistic
information to be developed by any ILCP program if it meets the requirements of ASTM D6299.
Several of our member companies operate multiple laboratories and conduct internal ILCPs.
These companies should be granted the ability to develop their own ARV program that is
independent of ASTM International or a voluntary consensus standards body (VCSB) provided
they comply with the same standard that ASTM International or a VCSB would utilize.
Furthermore, ASTM D6299 does not establish that the ILCP must be done within different
companies. ASTM D6299 6.2.2 addresses this with the requirement that the data set be generated
from an "interlaboratory" testing. By definition, interlaboratory only means across multiple labs.
Restricting the ILCP to ASTM International and VCSB would increase costs for these companies
and ultimately discouraging the use of an ILCP for instrument accuracy demonstrations.
New requirements have been proposed for determining whether a laboratory's results in an ILCP
are valid for demonstrating compliance with accuracy requirements. We agree that the proposed
§ 1090.1375(d)(l)(i) and (ii) are appropriate for establishing a robust ARV for the ILCP and
criteria for the laboratory's result relative to the accuracy of the referee method, respectively.
However, the newly proposed § 1090.1375(d)(l)(iii) and (iv) that establish criteria for
performance relative to a methods check standard, and z-score are not appropriate, overly
restrictive, and establishes unclear requirements. We request that § 1090.1375(d)(l)(iii) and (iv)
43
-------
not be included in the future regulation to maintain suitable criteria for demonstrating accuracy
for method-defined VCSB procedures between § 1090.1375 paragraphs (c) and (d).
§ 1090.1375(d)(l)(iii) requires that the ARV is less than the method-defined limit for check
standard accuracy to be valid. The specific method-defined check standard accuracy limits differ
and can be more restrictive than the EPA referee method. For example, accuracy for D5769 at
0.75*Rfor 0.60 vol% is 0.118% vol, whereas check standard requirements for D3606B in
method is 0.090 vol%, based on R of benzene/SQRT(2). The intent of the ongoing accuracy
determination is to show that your test method accuracy is consistent with the EPA referee (D-
5769) by comparing to 0.75*R. Attempting to manage multiple accuracy limits for a given ILCP
sample will be challenging to comply, including confusion on which limits to apply. Similarly,
some test methods have various check standards allowed/required. For example, ASTM D5191
identifies n-pentane, pentane/toluene, neohexane as reference standards. Despite being identified
as reference standards it is unclear if these compounds are to be treated as check standards in the
context of §1090.1375(d)(l)(iii). It would also not be clear which limits to compare to relative to
ASTM D5191-20 published R. Finally, these accuracy limits are specific to the reference
standard used and not applicable to gasoline matrix fuels.
§ 1090.1375(d)(l)(iv) proposes requirements that for an ILCP result to be used to demonstrate
ongoing accuracy, the result must fall within 2-z-scores. The 2-z-score designation is a function
of the standard deviation of the ILCP data set for the test method and not an indication of
validating accuracy to the referee test method (0.75*R) tolerance. Because Z-score is calculated
as (lab result - data set mean)/st. dev of data set, for ILCP samples that exhibit a low standard
deviation relative to the published R of the EPA referee, a lab result can be well within accuracy
limits defined under § 1090.1375(c)(2) but be deemed "invalid" for meeting ongoing accuracy
via an ILCP. As an example, ASTM RFG2402 for D3606B had a data set standard deviation of
0.025 vol% for a robust ARV of 0.581 vol%. 2 z-score data requires compliant results to fall
within 0.581 vol% +/- 0.05 vol%. In comparison, the 0.75*R accuracy limit for a robust ARV
check standard for the R of the EPA referee D-5769 is 0.581 vol% +/- 0.117 vol%, factoring in
35 lab results used in the PTP The 2 z-score requirement makes accuracy tolerance criteria
greater than two times more restrictive than the 40 CFR part 1090 check standard accuracy
requirements outlined in § 1090.1375(c)(2). Additionally, ASTM D-7372 (Interpreting PTP Data)
indicates in Section 6 that a PTP z-score between 2z and 3z is a warning limit that should be
investigated from a continuous improvement standpoint, not a disqualification of accuracy
qualification.
Recommendation: Permit ILCPs that meet ASTM D6299 to be used for accuracy qualification
and do not finalize the additional criteria related to method check standard accuracy and results
being outside of 2 z-scores.
Suggested text:
§ 1090.1375 Quality control procedures.
44
-------
(c) Accuracy demonstration. For absolute fuel parameters (VCSB and non-VCSB) and
for method-defined fuel parameters using non-VCSB methods, you must show that you
meet accuracy criteria as specified in this paragraph (c). For method-defined VCSB
procedures, you may meet accuracy requirements as specified in this paragraph (c) or by
comparing your results to the accepted reference value in an inter-laboratory crosscheck
program as specified in paragraph (d) of this section. Voluntary participation in an inter-
laboratory crosscheck program, including results from that program, will not be
considered for accuracy demonstration if accuracy criteria is already met as specified in
this paragraph (c).
(d) Demonstrating accuracy by participating in crosscheck programs. You may meet
accuracy requirements under paragraph (c) of this section by comparing your results to
the accepted reference value in an inter-laboratory crosscheck program sponsored by
ASTM International or another VCSB that meets the specifications outlined in ASTM
D6299 at least three times per year (two times per year for RVP), subject to the following
provisions:
* * *
(iii) The difference between the tost result and the ARV is greater than the method
defined limit for check standard accuracy, if applicable.
(iv) The measured value lies outside of two Z scores.
Response:
Regarding the comment recommending removal of the requirement for inter-laboratory
crosscheck programs to be sponsored by ASTM International or another VCSB and instead to
allow for conducting test programs across multiple laboratories to avoid increasing testing costs,
the existing regulation already includes a requirement for inter-laboratory crosscheck programs
to be sponsored by a VCSB in 40 CFR 1090.1375(c). As a result, we disagree that the VCSB
requirement will increase testing costs. The proposed change was intended to require that
companies make an affirmative decision from such a test program that they meet accuracy
requirements, rather than meeting quality assurance requirements only by participating. Implicit
in the existing regulations is an expectation that all responsible parties would respond to a failing
result from an inter-laboratory crosscheck program by identifying and correcting any problems.
However, experience has shown that the existing regulation has two shortcomings. First,
companies may apply different judgments about what should be considered a failing result.
Second, we have also found that companies in some cases do not take expected measures to
correct problems that may have contributed to a failing result. Both outcomes could unfairly
reward laboratories that may exclude results that they do not like using the discretion allowed
under the existing regulation. The proposed change to require an affirmative determination for
the accuracy demonstration, with specific failure criteria, provides the accountability that is
integral to participating in a VCSB-sponsored test program. We are therefore not removing the
45
-------
existing or proposed requirement for inter-laboratory crosscheck programs to be sponsored by
ASTM International or another VCSB.
Regarding the comment suggesting a revision that would allow for any crosscheck program that
is suitable for making the accuracy demonstration, we have also left unchanged the provision
allowing companies to make their accuracy demonstration based on in-house testing by
comparing to a check standard instead of participating in an inter-laboratory crosscheck program.
It is possible that companies will be more likely to choose the former option to demonstrate
accuracy. We do not believe any decreased reliance on inter-laboratory crosscheck programs will
significantly affect the prevailing practice of relying on these test programs to achieve a wide
range of benefits for validating measurement procedures across laboratories. Longstanding
experience with past test programs indicates that the particular scenarios highlighted by one
commenter as problematic complexities are in fact exceptional cases that could also be cited as
arguments even to disallow the current requirement to demonstrate accuracy by participating in
inter-laboratory crosscheck programs.
The regulation at 40 CFR 1090.1375 is set up to allow companies to make accuracy
demonstrations based on check standards, and 40 CFR 1090.1375(c)(1) states that a successful
result qualifies a test instrument for 130 days. There is also the requirement in 40 CFR
1090.1300(d) to apply good laboratory practices in all cases. We would expect any company
with a problematic result from an inter-laboratory crosscheck program to quickly and thoroughly
assess their methods, equipment, and calibrations to ensure that they can properly characterize
fuel parameters for demonstrating compliance with the regulation. As such, relying on the
successful testing with the check standard and disregarding the information from the inter-
laboratory crosscheck program would not be consistent with good laboratory practice. As a
result, if a company fails to meet the accuracy criteria for testing in an inter-laboratory
crosscheck program under 40 CFR 1090.1375(d) less than 130 days after successful testing with
a check standard, the company would need to follow the specified steps to correct the problem.
We recognize that this arrangement could create a disincentive for participating in inter-
laboratory crosscheck programs in addition to testing with check standards. At the same time, we
are unable to support a conclusion that would prioritize broader participation in inter-laboratory
crosscheck programs to the detriment of maintaining a high expectation of testing quality for
demonstrating compliance with the regulation.
Consistent with the commenters' recommendation, we are revising the proposed amendment by
removing the criterion proposed at 40 CFR 1090.1375(d)(l)(iii) to require the ARV to be less
than the method-defined limit for check standard accuracy. We agree that synthetic standards are
inherently different from fuel samples and that applying the check standard accuracy limits from
ASTM methods to results from a crosscheck program would impose undue restrictions, given the
performance difference between synthetic reference standards and gasoline matrix fuels.
As also noted in the preamble, we acknowledge that the proposed criterion referencing a 2 Z-
score threshold is overly restrictive and we have removed this criterion from the final rule. While
we maintain that laboratories whose results fall outside of 2 Z-scores should interpret such data
as a signal to initiate an investigation for potential improvements, such results should not
automatically constitute a failure to demonstrate accuracy. In light of these comments, we are
46
-------
adopting an alternative criterion to more effectively align with the parameters of the crosscheck
programs. Specifically, the alternative criterion instead stipulates that results must fall within the
three-sigma range of the data from the relevant ILCP. This adjustment provides a clear standard
for evaluating results, without being overly restrictive.
Comment:
> 0009 ONEOK
§ 1090.1375 Quality control procedures.
1090.1375(d)(2) If your results from the crosscheck program are not valid under paragraph (a)(1)
of this section, perform a root cause analysis and document your findings and the steps you take
to correct the problem. You continue to meet accuracy requirements under this section for the
affected parameter only if you correct the problem and demonstrate compliance with the
accuracy requirements of this section within 35 days after learning of a failure under paragraph
(d)(1) of this section. The compliance demonstration may be based on in-house testing using a
check standard qualified by a third party, or on testing in the next crosscheck program.
• ONEOK would like to recommend removing the requirement to use a check standard
"qualified by a third party". A company may be able to qualify testing internally if the
internal processes and procedures meet the qualifications for a crosscheck program. If the
concern is bias, this could be attested to and reviewed on a routine basis.
> 0012 AFPM/API
XII. § 1090.1375 Quality control procedures.
b. Grace period for subsequent ILCP result
There is concern with the 35-day period proposed for resolution of an accuracy failure. Since
EPA is proposing to allow for the accuracy to continue to be maintained by a satisfactory result
on the next cross check sample, the period should include time until the subsequent cross check
report is received even if that is after the 35-day period.
In the last sentence of proposed § 1090.1375(d)(2), there is reference to check standard
"qualified by a third party." This term lacks definition and does not align with the § 1090.1375(c)
which requires check standard as meeting specifications outlined in ASTM D6299. The language
in proposed § 1090.1375(d)(2) should be changed to align to § 1090.1375(c).
Recommendation: Replace "qualified third party" to any check standard developed in accordance
with ASTM D6299. Also extend the grace period to allow for receipt of subsequent ILCP result
if is beyond 35 days.
Suggested text:
47
-------
§ 1090.1375 Quality control procedures.
(d) * * *
(2) If your results from the crosscheck program are not valid under paragraph (a)(1) of
this section, perform a root cause analysis and document your findings and the steps you
take to correct the problem. You continue to meet accuracy requirements under this
section for the affected parameter only if you correct the problem and demonstrate
compliance with the accuracy requirements of this section within 35 days after learning
of a failure under paragraph (d)(1) of this section. The compliance demonstration may be
based on in-house testing using a check standard qualified by a third party meeting
specifications outlined in D6299, or on testing in the next crosscheck program. A valid
accuracy demonstration based on testing from the next crosscheck program received
more than 35 days after learning of a failure is acceptable.
Response:
We are not removing the requirement for third parties to qualify the check standard as suggested
by the commenters. As noted above, companies can meet requirements for accuracy
demonstrations based on either periodic testing that compares measured values to a check
standard or a positive result from participating in an inter-laboratory crosscheck program. The
proposed provision at 40 CFR 1090.1375(d)(2) describes the testing and demonstration
requirements that apply after a company has results from a crosscheck program that are not valid
based on the pass-fail criteria for such testing. Because the testing is needed to confirm that the
company has corrected a known problem, it is important to create a higher level of confirmation
to ensure that the test results appropriately allow for concluding that the affected test instruments
have once again demonstrated that they meet accuracy requirements to support ongoing testing.
That higher level of confirmation includes a requirement for testing based on a check standard
that has been qualified by a third party. This requirement creates an important additional level of
assurance that the test results properly support a conclusion that the test instruments meet
accuracy requirements.
While the final regulation includes the third-party test requirement as proposed, we are aware
that an additional option for making a demonstration to re-qualify a test instrument would be to
do testing as part of a non-VCSB correlation program administered by a third party. We are
therefore amending the regulation to include this as an additional option as a remedial approach
following a failing result from an inter-laboratory crosscheck program.
In addition, we note that companies can re-establish their test instrument after a failing result by
participating in the next inter-laboratory crosscheck program, which would similarly depend on
comparing test results to a check standard established outside of the company's laboratory.
48
-------
Experience with crosscheck programs suggests that once samples are collected, there is a routine
and consistent process for conducting tests and reporting results. We recognize that collecting
samples at the beginning may cause some delay for the start of testing. However, to account for
the logistics of participating in the next crosscheck program and to allow for any inefficiencies in
the crosscheck program, we are increasing the grace period after learning of a failing result from
the proposed 35 days to 45 days. Also, given that any delay in results from the crosscheck
program will likely occur in the earliest stages, companies needing timely results would likely
learn about any risk of not meeting the 45-day deadline early enough to allow for in-house
testing with a check standard to qualify affected test instruments.
Comment:
> 0012 AFPM/API
XII. § 1090.1375 Quality control procedures.
c. Failure to meet precision or accuracy requirements.
EPA has proposed new criteria for when presumed values should apply when there is a failure to
meet accuracy requirements. While implied in the proposed text, we recommend that it be
explicit that the presumed values only apply to certification measurements.
Recommendation: Specifically limit the presumed values when accuracy is not demonstrated to
parameter certification measurements.
Suggested text:
§ 1090.1375 Quality control procedures.
(e) Failure to meet precision or accuracy requirements. The presumed values specified in
§ 1090.1710(g) apply for parameter certification measurements with instruments failing
to meet precision or accuracy requirements under this section. If you fail to meet the
deadlines for resolving crosscheck-related issues under paragraph (d)(2) of this section,
the presumed values apply to parameter certification measurements for the instrument
that failed to meet precision or accuracy requirements starting at the point of learning of a
failure under paragraph (d)(1) of this section.
> 0014 Anonymous
Under the Quality Control Procedures section, EPA has proposed to add subpart (e) specifically
requiring importers/producers to use the presumed values (essentially the penalty properties of
339 ppm sulfur and/or 1.68% benzene, and llppm RVP) of section 1710(g) if the laboratory fails
49
-------
to meet precision or accuracy requirements. This additional language should not be added.
Importers and Blending Manufacturers conduct sampling and testing using an independent third-
party laboratory with internal controls, but controls that the importer/producer does not have
direct control/insight, internal expertise, nor timely insight into the individual lab equipment
being used, but they are acting in good faith by having reputable laboratories perform these
duties on their behalf. Outlining a specific requirement to use presumed values is not justifiable
when an SQC failure would still result in fairly accurate result and other tests may be available to
help demonstrate the actual values of the fuel (such as load COAs or tank results after an import,
etc). Outside of refineries that maintain their own controls on testing, all other parties are using
independent third-party inspectors to certify/test their products and do not have the inhouse
expertise nor the visibility on how those labs maintain their testing equipment. It is extreme to
force a company acting in good faith that their independent lab is operating according to industry
norms to penalize them in the same way as if no testing/certification was even done. They should
not have the same recourse of having to use the presumed value as someone who didn't test a
batch at all and was grossly negligent. This new requirement would cause an absurdly high credit
requirement (in the event of SQC failure), resulting in millions of dollars in costs to obtain
proper credits from the presumed valued being used on batches that have real-world results
showing they are nowhere near these limits, as well as result in other fines from EPA for having
used penalty properties and not meeting per-gallon standards. This punishment is extreme and
does not align with the intent of the regulation since it would falsely inflate environmental harm
to the public that does not exist.
> 0015 Susan Lafferty
EPA is proposing changes to § 1090.1375 by adding new (d). This proposed language appears to
be directed to refineries that own their own testing labs and will know about results from
crosscheck programs. Importers and blenders do not own their own labs but work with 3rd party
labs, and therefore do not have insight into the results of crosscheck programs. Therefore, it is
quite onerous to impose the presumed values on the 3rd party lab customers when the 3rd party
lab has failed to correct identified issues within 35 days of learning of a failure. An importer is
unlikely to know that an issue was flagged, or play in role in the correction of the issue, but in
the meantime its entire cargo(s) is being implicated in a punitive manner with high economic
stakes that is unlikely to match the offense. EPA should not adopt this provision to apply to
importers and blenders. If timely correction is a problem that needs addressing, EPA should work
with the labs accordingly.
Response:
We have revised the wording of 40 CFR 1090.1375(e) to explicitly tie the noncompliance and
presumed penalty parameters to testing used for certification and clarify that the presumed
penalty applies only with respect to the specific test instruments that fail to meet precision or
accuracy requirements.
The commenters focus on companies that rely on third-party laboratories to meet testing
requirements being vulnerable to compliance determinations, processes, and penalties for actions
that are outside their control. We disagree with their concerns. First, it is important to maintain a
50
-------
level playing field by holding in-house and third-party laboratories to the same standard for
certifying fuels. Setting a lower bar would disadvantage companies that rely on in-house testing
and could create a market distortion by creating an incentive for relying on third-party testing.
Second, the consequence of applying presumed penalty parameters does not apply immediately
upon finding that a laboratory has a failing result for accuracy demonstrations. Rather, the
presumed fuel parameters would apply only if the laboratory failed to correct the problems
identified by the root cause analysis and repeat testing with valid test results within the specified
timeframe which is 45 days. On the other hand, if corrective action is not taken to remedy the
failing result within the specified timeframe, the presumed fuel parameters would apply relative
to certification for parameter measurements with the test instruments failing to meet precision or
accuracy requirements starting on the date the laboratory received the first failing report from the
crosscheck program.
Third, anyone hiring an outside laboratory to do certification testing has the responsibility as due
diligence to select a testing partner that will reliably produce high-quality results, including
practices to respond to any adverse test results in routine tests to qualify test instruments relative
to precision and accuracy demonstrations under 40 CFR 1090.1375.
Fourth, it may be appropriate as a matter of contingency planning for anyone working with an
outside laboratory to retain the services of a second laboratory for further assurance of
uninterrupted service. From that perspective, companies that rely on third-party laboratories may
have an advantage over companies that perform their own in-house testing.
Finally, the presumed fuel parameters at 40 CFR 1090.1710(g) include explicit provisions that
allow EPA to set alternative penalty parameters based on any relevant information. As explained
in the preamble, failing to meet precision or accuracy requirements indicates that test instruments
are not suitable for generating valid test results for certification.
The factors identified by Commenter 0014 are the kind of information we would expect to
consider in determining whether available data, even compromised by an adverse finding relative
to the accuracy demonstration, would support a finding that alternative test values would be
appropriate for assessing penalties under 40 CFR 1090.1710(g).
51
-------
2.6 Fuel Certification with Domestic Marine Vessels
Comment:
> 0012 AFPM/API
VI. § 1090.1310 Testing to Demonstrate Compliance with the Standards
a. Flexibility for co-located terminals
The proposed technical amendments do not include co-located terminals as an extension of the
refinery.
Recommendation: It is recommended to add language in § 1090.1310(f) that would allow for the
15- mile distance to be captured for both the marine vessels and shore tanks in a common
section.
Suggested Text:
§ 1090.1310 Testing to demonstrate compliance with standards.
(f) Refiners and blending manufacturers may meet the testing requirements of paragraph
(b) of this section by loading gasoline or diesel fuel onto a marine vessel or in a tank,
subject to the following conditions:
(1) The marine vessel remains within 15 miles of the fuel manufacturing facility after
loading.
(2) Each vessel compartment is sampled for meeting certification testing requirements as
specified in § 1090.1605(b)(1).
(3) No additional loading occurs after sampling is complete.6
(4) The tank is located within 15 miles, or as approved by EPA, of the fuel manufacturing
gate.
(4)(5) The refiner or blending manufacturer ensures that the fuel meets all applicable per-
gallon standards before the fuel leaves the area specified in paragraph (f)(1) of this
section.
6 Edits to this provision are recommended below in section VI.b.
52
-------
Response:
This comment is beyond the scope of this rulemaking. We did not propose to include co-located
terminals in the proposed provisions for carving out an exemption from the certification
requirements in 40 CFR 1090.1310(b). At proposal, we explained that comments received on
topics or on portions of the regulations that are not the subject of the proposal this action would
be considered as beyond the scope of this rulemaking. 89 FR 70049.
Comment:
> 0012 AFPM/API
VI. § 1090.1310 Testing to Demonstrate Compliance with the Standards
b. Loading restriction should not apply to compartments not involved in the
batch certification
The intent of proposed § 1090.1310(f)(3) "No additional loading occurs after sampling is
complete." is unclear.
Recommendation: The Associations' primary recommendation is to delete § 1090.1310(f)(3), as
it is understood that no additional blending occurs into the marine vessel or tank until
certification is complete. As an alternative, if EPA disagrees with deleting § 1090.1310(f)(3), the
Associations recommend a modification for clarity.
Suggested Text:
§ 1090.1310 Testing to demonstrate compliance with standards.
* * *
(3) No additional loading occurs after sampling is complete.
Alternate Recommendation:
§ 1090.1310 Testing to demonstrate compliance with standards.
(f) * * * Refiners and blending manufacturers may meet the testing requirements of paragraph
(b) of this section by loading gasoline or diesel fuel onto a marine vessel, subject to the following
conditions:
53
-------
(3) No additional loading blending volume into a vessel compartment or tank occurs after
sampling until certification is complete.
Response:
Consistent with the commenter's alternative recommendation, we are revising 40 CFR
1090.1310(f)(3) to disallow adding blending volumes before sampling and certification are
complete. It continues to be appropriate to also disallow additional loading before sampling and
certification are complete because any additional loading would make the blended fuel
unrepresentative of the batch as measured for certification. The fuel manufacturer is still
accountable for any noncompliance and maintains responsibility for addressing any
noncompliance before the fuel is discharged. The fuel manufacturer would need to return to port
to bring fuel back to the fuel manufacturing facility if an issue is identified with the fuel after it
has left the dock. Once the fuel certification is confirmed, the vessel can move to another dock
for additional loading in other compartments.
These restrictions apply for fuel loaded onto a marine vessel, so there is no need or benefit to
reference vessel compartments or tanks when describing the restrictions. We do not want to
introduce the possibility of interpreting the restriction to apply differently for certain
compartments.
54
-------
3. Other Technical Amendments
3.1 Exemptions
Comment:
> 0012 AFPM/API
I. § 1090.610 Exemptions for temporary research, development, and testing
EPA has proposed expanding the temporary research, development, and testing exemption
section to include fuel additive and regulated blendstocks. Research and development (R&D) on
fuel additives occurs regularly as part of developing detergents and other performance enhancing
additives. Fuel additives that are not offered for sale should not be subject to the same R&D
requirements as fuels or regulated blendstocks under § 1090.610. Furthermore, 40 CFR
79.4(b)(2) exempts additives designated as "research, development, or test status" from
registration requirements provided that they are not "offered for commercial sale to the public."
Requiring what would essentially be ongoing EPA approval for any research, development and
testing would stifle innovation.
Recommendation: Limit the scope of this section to fuels and regulated blendstocks, not fuel
additives. If fuel additives are kept within this section, then an exemption should be maintained
for R&D of fuel additives.
Suggested text:
§ 1090.610 Exemptions for temporary research, development, and testing.
* * *
(3) Any person who performs research, development, and testing of fuel additives
consistent with the conditions of 40 CFR 79.4(b)(2) is exempt from the standards
specified in this part.
Response:
We have added language to 40 CFR 1090.610 to clarify that a fuel additive exempt under 40
CFR 79.4(b)(2) is also exempt from the standards in part 1090.
Comment:
> 0012 AFPM/API
II. § 1090.625 Exemptions for California gasoline and diesel fuel
55
-------
The regulation currently requires California (CA) gasoline and diesel to be segregated from other
fuels. The proposal adds the word "completely" (i.e., must be completely segregated). Complete
segregation is not feasible as pipelines transporting CA gasoline and diesel also transport
gasoline and diesel that are designated as EPA gasoline and diesel. There are also common piping
systems in refineries and terminals that would not permit complete segregation between C A and
non-CA grades.
EPA also did not provide an adequate justification for this change in the preamble. Adding the
word "completely" could erroneously imply a new standard for segregation in the regulations
and we question the necessity given that C A fuel specifications are more stringent than those of
the EPA.
Recommendation: Maintain the current regulatory language within this section relative to
segregation of C A gasoline and diesel.
Suggested Text:
§ 1090.625 Exemptions for California gasoline and diesel fuel.
(b) * * *
(2) Designated California gasoline or diesel fuel must be completely segregated from fuel
that is not California gasoline or diesel fuel at all points in the distribution system.
Response:
We do not believe that there is a meaningful difference between the requirement that exempt fuel
be "segregated" versus "completely segregated." Because some sections in 40 CFR part 1090,
subpart G, required that exempt fuel be "completely segregated" while other sections required
that exempt fuel merely be "segregated," we proposed to add the word "completely" before all
instances of "segregated" in order to harmonize the segregation requirement language throughout
the subpart. In response to this comment, however, EPA is removing all instances of "completely
segregated" and replacing them with "segregated" in order to create a uniform standard for all
exempt fuel with a segregation requirement in subpart G. EPA is not imposing a more stringent
exemption requirement or a higher standard for an exempt fuel that is required to be "completely
segregated.". For example, for purposes of complying with the segregation requirement in 40
CFR 1090.625(b)(2), we expect that designated California gasoline or diesel fuel will not be
commingled in the same tank as non-California gasoline or diesel fuel.
56
-------
3.2 Recordkeeping and Reporting Requirements
Comment:
> 0009 ONEOK
§ 1090.905 Reports by gasoline manufacturers.
1090.905(c)(l)(viii)(A); 1090.905(c)(2)(viii)(A); 1090.905(c)(8)(vii)(A) These sections have
added a new requirement to report whether the ethanol 1.0 psi waiver under §1090.215(b)
applies to a given production batch.
• The determination of whether an ethanol 1.0 psi waiver applies is dependent solely on the
distribution point. Products can be redesignated downstream of the production point leading
to error in the gathered information. E.g., summer RFG can be produced and marked with
"no ethanol 1.0 psi waiver" then be sold in a conventional or SIP area that does have a
waiver. This is a common process across many lower RVP product gasolines due to limited
tankage to address all the boutique fuels across a distribution system. Additional clarity is
needed on how the EPA will require this information to be reported as this is not part of the
current reporting process. ONEOK would like to suggest adding language to make clear that
this requirement pertains to whether the ethanol 1.0 psi waiver applies at the time of fuel
manufacture.
> 0012 AFPM/API
IV. § 1090.905 Reports by gasoline manufacturers
EPA has proposed an additional reporting requirement to indicate whether the ethanol 1.0 psi
waiver under § 1090.215(b) applies. We believe adding a separate requirement to report whether
the 1.0 psi waiver is applicable is not necessary as the waiver applicability is already captured
within the reporting of gasoline volatility standards in the STR0200 Gasoline, Gasoline
Blendstock and Diesel Batch Report. It would be more appropriate to add a new reporting code
within the STR0200 report's field 15 for summer batches to accommodate low-RVP conventional
blendstock for oxygenate blending (CBOB) that is required to meet the Federal 9.0 psi standard
inclusive of ethanol. Whether the ethanol 1.0 psi waiver is prohibited would then be indicated by
selecting "Federal 7.4 psi Standard", "Federal SIP", or a new a "Federal 9.0 psi Standard w/o
ethanol 1.0 psi waiver". Each of these volatility selections prohibit the waiver under §
1090.215(b) by regulation and therefore would not require an additional reporting column.
This approach allows for communication of the intent of the gasoline manufacturer of whether
the ethanol 1.0 psi waiver is expected to apply at the time of certification. It does not, however,
provide any specific assurance whether parties downstream may employ the ethanol 1.0 psi
waiver if the product is distributed to a location with a less restrictive vapor pressure standard
and is appropriately redesignated.
57
-------
Recommendation: Delete the proposed regulatory requirement to report whether the ethanol 1.0
psi waiver under § 1090.215(b) applies and indicate in the preamble of the final rule that
"Instructions for STR0200: Gasoline, Gasoline Blendstock, and Diesel Batch Report" will be
updated at least 6 months prior to March 31, 2026, to accommodate compliance year 2025
reporting.
Suggested text:
§ 1090.905 Reports by gasoline manufacturers.
* * *
* * *
(viii) * * *
(A) The applicable RVP standard, as specified in § 1090.215., and whether the ethanol 1.0
psi waiver under § 1090.215(b) applies.
The above suggested revisions are also applicable to § 1090.905(c)(2)(viii)(A) and (c)(8)(vii)(A).
Instructions for STR0200: Gasoline, Gasoline Blendstock, and Diesel Batch Report
Field No
Field Name
Units
Field Formats, Codes & Special Instructions
15
Gasoline
Volatility
Standards
AA; Character. Identify the designation under the gasoline
volatility standards found at 1090.215
Summer Gasoline
VI = Federal 7.8 psi standard
V2 = Federal 9.0 psi standard
V3= Federal 9.0 psi standard w/o ethanol 1.0 psi
waiver
V4=Federal SIP
V5= Federal 7.4 psi standard
EX=Exempt from RVP control
Response:
We are finalizing as proposed the amendment to require the reporting of whether the ethanol 1.0
psi waiver under 40 CFR 1090.215(b) applies to a batch of gasoline because we believe it is
58
-------
necessary to distinguish whether a batch of summer gasoline is eligible for the 1.0 psi waiver.
This change is in response to the actions taken by EPA to grant several governors' request to
remove the 1.0 psi waiver in eight states.3 Without such a change, EPA would be unable to
determine which RVP standard applied to the batch based on information currently supplied in
the annual batch reports. We also believe it is important to include this modification in the
reporting regulations to provide regulatory certainty for regulated parties. However, we
appreciate the commenter's request for adequate time to prepare reports for compliance, and we
generally agree with the commenter's suggested approach to update the gasoline batch reports.
As discussed in RTC Section 1.2, in order to allow industry sufficient time to adjust to the
revised reporting form, we are not requiring this information to be reported until the 2026
compliance period and, therefore, it will not be reflected on batch reports until March 31, 2027.
We believe this addresses the commenter's request for more time. We are already in the process
of making changes to the STR0200 form to modify field 15 to incorporate a new code for
reporting volatility standards as suggested by the commenter. We intend to make this amendment
during the current ICR update and the issue was flagged as part of the comment process for the
ICR update.4
Comment:
> 0012 AFPM/API
V. § 1090.1205 Recordkeeping requirements for all regulated parties
EPA proposes to require collecting the name and title of the person collecting samples and
performing testing. Maintaining records of the person's title conducting sampling and testing is
not necessary and does not provide any practical benefit as titles are not standardized across
companies. Irrespective of the person's title, they are acting as a representative of the regulated
party.
Recommendation: Do not proceed with adding the requirement to maintain record of the title of
the person conducting sampling and testing.
Suggested text:
§ 1090.1205 Recordkeeping requirements for all regulated parties.
* * *
(2) The name and title of the person who collected the sample and the person who
performed the testing.
3 89 FR 14760 (February 29, 2024).
4 88 FR 34497 (May 30^ 2023).
59
-------
Response:
In response to the comment, we have removed the proposed requirement to provide the title of
the person conducting the sampling and testing.
60
-------
3.3 Fuel Certification with Domestic Marine Vessels
Comment:
> 0012 AFPM/API
XIII. § 1090.1605 Importation by marine vessel
We appreciate the added flexibility suggested by the proposed amendments to § 1090.1605(b)(1).
However, the proposal is not clear whether a composite with an ethanol hand blend is allowed.
For a multi-compartment vessel (many have 12), a very large number of individual compartment
hand blends may be necessary. Further, the location of the text in § 1090.1605(b)(l)(iii) appears
incorrect.
Recommendation: We request clarity that a hand blend of the vessel-volumetric composite is
allowed, and we suggest embedding the language of § 1090.1605(b)(l)(iii) into §
1090.1605(b)( 1 )(ii).
Suggested text:
§ 1090.1605 Importation by marine vessel.
(b) * * *
(1) The importer must sample each compartment of the vessel and use one of the
following methods to meet testing requirements:
(i) Treat each compartment as a separate batch. Each individual compartment is deemed
to meet the homogeneity requirements in § 1090.1337.
(ii) Except as specified in paragraph (b)(l)(iii) of this section, Measure the neat RVP of a
sample collected from each compartment for summer gasoline. For all other testing, G
combine samples from separate compartments into a single, vessel-volumetric composite
sample using the procedures in Section 9.2.4 of ASTM D4057 (incorporated by reference
in § 1090.95). Test results from the composite sample are valid only after if single
samples are collected from each affected compartment and together meet the
homogeneity is demonstrated for all samples as specified requirements in § 1090.1337.
(iii) Measure the RVP of a sample collected from each compartment for summer gasoline.
61
-------
Response:
Importing gasoline blendstock with a marine vessel puts the importer in the same position as
domestic gasoline refiners producing BOB to which oxygenate is added downstream. We address
the specific testing requirements for that circumstance in 40 CFR 1090.1310(c)(1) and (2). Since
domestic gasoline refiners and importers are engaging in the same regulated activity (i.e., the
certification of fuels for purposes of meeting EPA fuel quality requirements), they should be
subject to the same testing requirements. We have therefore revised 40 CFR 1090.1605(b)(l)(ii)
to reference 40 CFR 1090.1310(c)(1) and (2) instead of repeating those same testing
specifications.
As suggested by the commenter, we have also revised 40 CFR 1090.1605(b)(l)(ii) to include a
description of RVP testing requirements before describing all other testing requirements.
62
-------
3.4 Alternative Test Value Requests
Comment:
> 0012 AFPM/API
XIV. § 1090.1710 Penalties
The proposed amendment to § 1090.1710(g) requires that alternative test values must be
submitted to EPA within 30 days of discovering a sampling requirement failure. Alternative test
value requests are commonly done concurrently with eDisclosures following EPA's Audit Policy
so that the submittals are consistent. The alternative test values request benefits from the
investigation/corrective actions required by the eDisclosure process and should have a time
requirement consistent with EPA's Audit Policy, https://www.epa.gov/sites/default/files/2021-
02/documents/epaauditpolicyprogramfaqs2021.pdf
Recommendation: We recommend editing the timing for requesting alternative test values in §
1090.1710(g) to provide an option to submit by the Audit Policy Compliance Certification
deadline. This equates to 81 days from discovery, or potentially later if EPA provides an
extension under the Audit Policy.
Suggested text:
§ 1090.1710 Penalties.
(g) The presumed fuel parameter values of fuel parameters in this paragraphs (g) (i)
through (6) of this section apply for cases in which any person fails to comply with the
sampling or testing requirements and must be reported, unless EPA, in its sole discretion,
approves a different value. Any person requesting the use of alternative test values must
submit their request to EPA as specified in § 1090.10 within 30 days of discovering
failure to comply with sampling and testing requirements or concurrently with a
Compliance Certification timely submitted in compliance with EPA's Audit Policy,
whichever is later. EPA may consider any relevant information to determine whether a
different value is appropriate.
Response:
We agree with the commenter's suggestion and have revised the language in 40 CFR 1090.1710
accordingly. As a reminder, submitting a request for an alternative test value is a separate process
than self-disclosing a potential violation in eDisclosure. Additional information regarding how to
submit alternative test value requests and self-disclosures can be found here:
63
-------
https://www.epa.gov/fuels-registration-reporting-and-compliance-help/2023-enviroflash-
announcements-about-epa-fuel#020923.
We also note that information on EPA's audit policy can be found here:
https://www.epa.gov/compliance/epas-audit-policy.
64
-------
3.5 Incorporation by Reference
Comment:
> 0012 AFPM/API
XVII. § 1090.95 Incorporation by reference
We appreciate EPA's commitment to periodically review and update the materials that are
incorporated by reference, to ensure that the most up-to-date standards and reference materials
are cited in the regulations. We are aware of items that have been updated since the issuance of
the proposed rule, so we have suggested edits to those standards below. Additionally, there may
be references that will receive updates after our submission of comments, but prior to EPA
issuing a final rule. For those items, we would invite further discussion with EPA to ensure that
agency staff are aware of all revised materials to ensure that the most recent versions are
incorporated into the regulations.
Recommendation: We recommend the following revisions to the ASTM International (ASTM)
references proposed for § 1090.95(c), to incorporate the latest versions of these ASTM methods
and specifications.
Suggested text:
§ 1090.95 Incorporation by reference.
* * *
(3) ASTM D975 20a24a, Standard Specification for Diesel Fuel, approved Juno 1, 2020
May August 1, 2024 ("ASTM D975"); IBR approved for § 1090.80.
(7) ASTM D2163-23ell/l (Reapproved 2019), Standard Test Method for Determination
of Hydrocarbons in Liquefied Petroleum (LP) Gases and Propane/Propene Mixtures by
Gas Chromatography, approved May 1, 2019 March April 1, 2023 ("ASTM D2163");
IBR approved for § 1090.1350(b).
(8) ASTM D2622 162124, Standard Test Method for Sulfur in Petroleum Products by
Wavelength Dispersive X-ray Fluorescence Spectrometry, approved January 1, 2016
Docombor 1, 2021 July 1, 2024 ("ASTM D2622"); IBR approved for §§ 1090.1350(b),
1090.1360(d), 1090.1365(b), and 1090.1375(c).
65
-------
Response:
We have reviewed the changes in the newer versions of the reference standards and agree that the
suggested changes are appropriate. We have accordingly updated the test methods in the final
regulations.
Comment:
> 0008 CFAA
Clean Fuels supports EPA's decision to include the updated ASTM standards: ASTM D6751-24,
Standard Specification for Biodiesel Fuel Blend Stock (B100) for Middle Distillate Fuels,
approved March 1, 2024, and ASTM D975-24, Standard Specification for Diesel Fuel, approved
May 1, 2024.
However, we encourage EPA to create a system that allows for the adoption of these and related
ASTM standards to be updated more frequently. For example, there are already plans to update
the ASTM standards for biodiesel for other applications (home heating oil, railroads, on/off road
diesel fuels) for blends up to B100 in the next several years.
As the biodiesel industry is doing the technical work for use of up to B100 in all markets, our
efforts to constantly update the biodiesel standards at ASTM have not only been the key to
increasing biodiesel volumes, it has been vital to ensure customer confidence and OEM support.
To address the constant technical work that is being conducted, we suggest that EPA update the
specification for Biodiesel Fuel Blend Stock (B100) for Middle Distillate Fuels and Standard
Specification for Diesel Fuel to "ASTM D6751-24 or the most recent version" and "ASTM
D975-24 or most recent version" respectfully.
However, if this is not judicious, we alternatively suggest that EPA adopt ASTM D6751-24 and
ASTM D975-24 and then plan to update these ASTM standards again every 3 years, or
whenever major changes occur in the ASTM standards as a result in changes in fuel processing
or engine and aftertreatment technology to ensure that there is not another great lag between
improved ASTM standards and those adopted by EPA.
> 0010 SABR
The Proposed Rule would incorporate by reference "ASTM D6751-24, Standard Specification
for Biodiesel Fuel Blend Stock (B100) for Middle Distillate Fuels, approved March 1, 2024 "5
5 89 Fed. Reg. at 70,061. Currently, Part 1090 incorporates by reference "ASTM D6751-20, Standard Specification
for Biodiesel Fuel Blend Stock (B100) for Middle Distillate Fuels, approved January 1, 2020 ('ASTMD6751'); IBR
approved for §1090.1350(b)." 40 C.F.R. §1090.95(c)(36). This is an earlier revision to the standard than that
incorporated into the Part 80 regulations.
66
-------
The proposed rule states that "[t]his updated standard describes the characteristics of biodiesel."
SABR Coalition supports the Proposed Rule's proposed revision to update the standard.6
However, SABR Coalition also submits these comments to urge EPA to ensure that it
consistently revise its references to ASTM D6751 in Part 1090 and in the Renewable Fuel
Standard program to ensure consistency or revise the definition of biodiesel in the Renewable
Fuel Standard program's regulations as follows: (1) revise the definition of biodiesel to
incorporate by reference the "most recent version of ASTM D6751" instead of incorporating by
reference a specific version of the standard; (2) provide notice of enforcement discretion that
provides that any biodiesel producer that is complying with the most recent version of ASTM
D6751 may properly generate RINs under the Renewable Fuel Standard program, even if EPA
has not yet updated the ASTM standard in §80.12; or (3) revise the definition of biodiesel to
remove the reference to ASTM D6751 and replace it with reference to biodiesel registered under
Part 79 of EPA's regulations. At a minimum, EPA should also revise §80.12 to incorporate by
reference the most recent ASTM D6751 standard as it is proposing for Part 1090. It should also
set up a process to update the standard on a more regular basis to avoid inconsistent requirements
and confusion in the marketplace. If needed, EPA should make this change through a direct final
rule, as we do not believe simply updating the ASTM reference would be controversial.
Response:
We do not have authority under the Clean Air Act to adopt a regulatory provision that allows for
updating referenced materials or standards either by any sort of automated process or by direct
final rule. The Office of Federal Register's Incorporation by Reference (IBR) Handbook states
that "[ajgencies must follow the approval process in order to properly incorporate material by
reference into the CFR."7 Furthermore, the IBR process only incorporates a specific version of a
publication;8 any subsequent update of that publication must go through the IBR process again in
order to reference the updated material in the CFR.9 But we strive to update the incorporation by
reference of materials in Part 1090 on a periodic basis.
Note that the reference to ASTM D6751-24 for 40 CFR part 1090 is limited to carving out an
exception for testing to demonstrate compliance with cetane or aromatics standards. If the scope
of ASTM D6751 changes in the future to include a wider range of fuels that should be excluded
from that testing, we would amend the regulation accordingly.
Comments related to the incorporation by reference of ASTM D6751 under the Renewable Fuel
Standard (RFS) program, which is contained within 40 CFR part 80, are beyond the scope of this
rulemaking.
6 89 Fed. Reg. at 70,061.
7 Office of Federal Register, "Incorporation by Reference Handbook," June 2023 Edition (ed. rev. 8/28/2023) ("IBR
Handbook"), p. i. https://www.archives. gov/files/federal-register/write/handbook/ibr.pdf.
8 "The Director only approves a specific version of a publication to a specific CFR section." IBR Handbook, p. 18.
9 "When you decide to use a publication, you must request new approval for: a. Using a new, revised, or updated
version of a publication;" IBR Handbook, p. 18.
67
-------
3.6 Technical Corrections
Comment:
> 0009 ONEOK
§ 1090.700 Compliance with average standards.
• There are multiple instances where the benzene calculation is incorrect. In each instance the
volume percent needs to be converted to Benzene volume by dividing by 100. Calculations in
current rule equate to 0.0062 vs 0.62 as proposed. This needs to be addressed in each instance
where there is reference to 1090.210 (a) for use in an equation for compliance with the
benzene annual average. There is also a presumed value of 0.68 vol% (PBV) that is in the
oxy correction that needs to be addressed at 1090.700(b)(4).
> 0012 AFPM/API
III. § 1090.700 Compliance with average standards
The proposed rule omits a missing factor on benzene equations. EPA has proposed modifications
to equations used in determining compliance with the benzene annual average and benzene
deficits from downstream blendstock for oxygenate blending (BOB) recertification. In each
instance the value 0.0062 is replaced with a reference to benzene in volume percent (i.e.,
benzene average standard, per § 1090.210(a), in volume percent). Expressing the value in
volume percent would be the value 0.62, not 0.0062. This represents a material change which we
believe was unintentional.
Recommendation: In each of the following instances divide the benzene average standard
expressed in volume percent by 100. This correction is needed in §§ 1090.700(b)( 1 )(ii),
1090.700(b)(2), 1090.700(b)(2)(ii), and 1090.740(b)(4).
Suggested Text:
§ 1090.700 Compliance with average standards.
(b) * * *
* * *
(ii) The total amount of benzene produced is determined as follows:
68
-------
Where:
Bztot.y = Total amount of benzene produced during compliance period y, in benzene
gallons.
Vi = Volume of gasoline produced or imported in batch i, in gallons.
Bzi = Benzene content of batch i, in volume percent.
i = Individual batch of gasoline produced or imported during the compliance period,
n = Number of batches of gasoline produced or imported during the compliance period.
Response:
We have corrected the errors.
Comment:
> 0012 AFPM/API
XVI. Typographical errors
The following items are apparent typographical errors that the Associations recommend EPA
correct in the final regulations:
Suggested Text:
§ 1090.1300 General provisions.
* * * *
* * * *
(ii) You must meet requirements for documenting and supporting exclusion of data points
as specified in § 1090.1375(a)(5) (4).
§ 1090.1337 Demonstrating homogeneity.
69
-------
(b) Any test to establish homogeneity is considered a certification test relative to a per-
gallon standard for a given parameter if the test result is the worst-case value from all
testing performed for the batch. Report the highest measured value as specified in 8
1090.1335(e)(3) (1).
Response:
We have corrected the errors.
70
-------
4. Beyond the Scope
Comment:
Several commenters (including but not limited to: 0003, 0010, 0016) addressed numerous
additional topics, including but not limited to the following:
• Scope and applicability of 40 CFR part 1090.
• Allowing for the use of online analyzers for the measurement of RVP.
• Changes to provisions of the RFS program.
Response:
These comments are all beyond the scope of this rulemaking. At proposal, we explained that
comments received on topics or on portions of the regulations that are not the subject of the
proposal this action would be considered as beyond the scope of this rulemaking. 89 FR 70049.
We also explained that we were not reopening unchanged, republished provisions for comment.
89 FR 70061. These topics are not further addressed in this document.
71
------- |