Evaluation of RFA/GrE Submission in

            Partial Fulfillment of 40 CFR 79 Tier 1

            and Tier 2 E15 Registration Requirements
&EPA
United States
Environmental Protection
Agency

-------
             Evaluation of RFA/GrE Submission in
            Partial Fulfillment of 40 CFR 79 Tier 1
           and Tier 2 E15 Registration Requirements
                       U.S. Environmental Protection Agency
&EPA
United States
Environmental Protection
Agency
EPA-420-R-12-010
February 2012

-------
 Evaluation of RFA/GrE Submission in Partial Fulfillment of 40 CFR 79 Tier 1 and Tier 2
                         E15 Registration Requirements1
Introduction

      In 2009, Growth Energy and 54 ethanol manufacturers submitted an application
to EPA for a waiver under section 211 (f)(4) of the Clean Air Act (CAA or "the Act").  The
application sought a waiver for gasoline-ethanol blends of up to 15 vol% ethanol (E15)
for all motor vehicles and nonroad vehicles, engines, and equipment.2  On April 21,
2009, EPA published notice of receipt of the application and requested public comment
on all aspects of the application to assist the Administrator in determining whether the
statutory basis for granting the waiver request had been met (74 FR 18228).

      On  October 13, 2010, EPA took two actions on the waiver request based on the
information available at that time ("October Waiver Decision").3 First, it partially
approved Growth Energy's waiver request to allow the introduction of El 5 into
commerce for use in MY2007 and newer light-duty motor vehicles, subject to several
conditions. The October Waiver Decision was based on a determination that E15 will
not cause  or contribute to a failure of MY2007 and newer light-duty motor vehicles to
achieve compliance with the emissions standards to which they were certified under
section 206 of the CAA over their useful lives. Second, the Agency denied the waiver
request for MY2000 and older light-duty motor vehicles, heavy-duty gasoline engines
and vehicles, highway and off-highway motorcycles, and other nonroad engines,
vehicles, and equipment. The Agency also deferred making  a decision  on the waiver
request for MY2001-2006 light-duty motor vehicles to await the results of additional
testing being conducted  by the Department of Energy (DOE). On January 21, 2011,
EPA partially approved Growth Energy's waiver request to allow the introduction of E15
into commerce for use in MY 2001-2006 light-duty motor vehicles after receiving and
analyzing the completed DOE test data ("January Waiver Decision").4  The January
Waiver Decision noted that additional  steps must be completed before  E15 may be
distributed or sold.  The January Decision stated: "These steps include but are not
limited to submission of a complete E15 fuels registration application by the fuel and
fuel additive manufacturers who wish to introduce E15 into commerce,  and EPA review
and approval of the application, under the regulations at 40 CFR Part 79".5

      Section 211  of the Act requires registration of designated  motor fuels and fuel
additives (F/FAs).   Registration requirements at 40 CFR 79 include emissions
speciation testing and a  literature search of the associated emissions (Tier 1  testing)
and animal testing of exposure to emissions for purposes of determining health effects
      The information in the submissions discussed below was evaluated by EPA's Office of Air and
Radiation and Office of Research and Development. This document was prepared utilizing input from
EPA's Office of Air and Radiation, Office of Research and Development and Office of General Counsel.
2 Since E15 has greater than 2.7 wt% oxygen content, E15 needs a waiver under CAA section 211 (f)(4).
3 75 FR 68094 (November 11, 2010).
4 76 FR 4662 (January 26, 2011).
5 76 FR 4663 (January 26, 2011).

-------
(Tier 2 testing).  In certain cases, a small business is exempt from some or all of this
testing.

      On February 18, 2011, October 19, 2011, and February 3, 2012, the Renewable
Fuels Association (RFA) and Growth Energy (GrE) submitted Tier 1 and Tier 2 data and
analyses which would, if approved by EPA, partially satisfy the requirements for a fuel
or fuel additive manufacturer's registration of E15 under 40 CFR Part 79. For any
potential registrant with over $50 million in annual revenue, Tier 1 and Tier 2
requirements must be met before registering E15.  Specifically, RFA/GrE submitted Tier
1 emissions speciation testing and for Tier 2 testing provided an analysis making the
case that Alternative Tier 2 testing previously conducted for E10 and baseline gasoline
should be sufficient for testing that would otherwise be required under the regulations
for E15.  An evaluation of the RFA/GrE submissions for fulfillment of these Tier 1 and
Tier 2 testing requirements is the subject of this document.

      As explained in detail below, based on our evaluation, we conclude that the Tier
1 and Tier 2 data and testing requirements for an E15 registration application would be
fulfilled by the RFA/GrE submissions.  It is important to note, however, that the RFA/GrE
submissions do not constitute a complete application nor do they meet all of the
requirements for a complete application.  Under the Clean Air Act and EPA's fuel
registration regulations, every fuel or fuel additive manufacturer that intends to introduce
E15 into commerce must submit for EPA approval a fuel registration application for E15.
With RFA/GrE's permission, E15 fuel/fuel additive manufacturers may rely on RFA/GrE
submissions for completing the Tier I and Teir 2 portions of their applications. To
submit a complete application, each E15 fuel manufacturer must also submit specified
information about its company. EPA will act on complete applications as they are
received.  An E15 fuel manufacturer may introduce  E15 into commerce only after EPA
has approved its registration application and the fuel manufacturer meets the misfueling
mitigation conditions.
Tier 1 Requirements

      The Tier 1 registration regulations at 40 CFR 79.52 require a characterization of
the emission products which are generated by evaporation and combustion of E15 in a
motor vehicle.  Combustion testing must be conducted with and without aftertreatment
of exhaust emissions.  A literature search for information on the potential toxicological,
environmental, and other public welfare effects is required for emission products, with
the exception that it is not required for those emission products which are the same as
the emission products of baseline gasoline (EO). A test group organized by the
American Petroleum Institute tested EO and E10 and conducted the literature search for
EO's and E10's emission products.

      RFA contracted to have the Tier 1 testing performed by the Southwest Research
Institute (SwRI) in San Antonio, Texas. The evaporative emissions report and the
combustion emissions report for E15 have  been reviewed and would meet the test

-------
requirements required for registration. As noted above, the combustion testing required
testing the engine with and without exhaust aftertreatment (i.e., the catalytic converter).
In all, 21 new emission products were found compared to EO. In the catalyzed exhaust,
only ethanol, 2-methyl hexane, and methacrolein were new. Seventeen of the new
products were in the untreated exhaust.  Ethane, propane, propyne, and ethanol were
new products in the E15 evaporative emissions, although ethane, propane and propyne
were found in the EO combustion emissions. In all, nine new emission products were
found compared to E10.  In the catalyzed exhaust, only 2-methyl hexane was new. All
nine were in the untreated exhaust. Ethane, propane, and propyne were new products
in the E15 evaporative emissions compared to the E10 evaporative emissions, but were
found in the E10 combustion emissions.  RFA/GrE submitted the literature search for all
of these new products on October 19, 2011. It has been reviewed and would meet the
applicable regulatory requirements.

Tier 2 Requirements

      The Tier 2 registration regulations at 40 CFR 79.53 require health effects testing,
and the Alternative Tier 2 testing provisions can be found at 40 C.F.R. § 79.58(c). In an
August 20,  1997 notification, EPA issued Alternative Tier 2 testing requirements for
baseline gasoline and various oxygenated  gasolines including E10.6  This was based
upon EPA's determination that alternative test procedures would yield more useful data,
as described below EPA required alternative testing for baseline gasoline  and various
oxygenated gasoline groups, including E10. The health studies for E10 that were
required under the notification have now been completed and approved. The studies
were undertaken by the Section 211(b) Research Group.

      The  primary difference between the testing for E10 under the Alternative Tier 2
requirements and Standard Tier 2 requirements is  that the Alternative Tier 2 testing
focuses on the health effects of evaporative emissions. Alternative Tier 2  testing for
E10 and other fuel/fuel additives did not include examination of combustion emissions
due to methodological complications caused by carbon monoxide (i.e., carbon
monoxide in the exhaust gases would kill the test animals before any  useful information
could be developed).  In the notification requiring Alternative Tier 2 testing for baseline
gasoline and several oxygenated fuels, including E10, EPA explained the  rationale for
focusing on evaporative emissions and why the combustion emissions studies would
likely not produce useful information. Since then, there have been advances in potential
combustion emission testing techniques that hold promise for effectively addressing the
methodological complications caused by carbon monoxide. Additionally, newer vehicles
have lower carbon monoxide emission rates.  EPA will continue to monitor and assess
these advancements and consider the usefulness of requiring such testing under
additional (Tier 3) testing provisions for any appropriate gasoline test  groups that are or
might become a significant part of the U.S. gasoline market.

      In their Tier 2 submissions, RFA/GrE state that there is no need for Tier 2 testing
of E15 because the emissions species are so  similar to emission species of E10 and
6 The notification can be found at http://www.epa.gov/otaq/consumer/fuels/mtbe/fnlno19a.pdf.

-------
baseline gasoline that the Alternative Tier 2 testing for those fuels should satisfy the Tier
2 health testing requirements for E15. In part, RFA/GrE states:

      Section 211 (e)(3)(C) gives EPA authority to "exempt any person from such
      regulations with respect to a particular fuel or fuel additive upon a finding that any
      additional testing of such fuel or fuel additive would be duplicative of adequate
      existing testing." 42 U.S.C. §7545(e)(3)(C). EPA has interpreted this authority to
      allow fuel or fuel additives "that are similar in composition and usage to those
      already on the market to group with [emphasis added] those similar F/FAs and
      complete the testing with the other F/FAs in their group." 59 Fed. Reg. at 33,051.
      "EPA interprets section 211 (e) and (b) to give it authority to require any
      necessary health or welfare  effects information for F/FAs that are significantly
      different in composition or usage from currently registered products." Id.
      (emphasis added). EPA focuses on "potential emissions-based effects." Id.

      In its submission  regarding Tier 2 testing for E15, RFA/GrE also states that:

      Tier 2 Testing [for E15] is not necessary, as E15 should fall under the approved
      Alternative Testing already completed forE10: EPA's Tier 1 and Tier 2 testing is
      intended to identify compounds for potential regulation under the Act. 59 Fed.
      Reg. at 33,042.  Based on the analysis above, there are no new compounds that
      have not already  been detected, tested, or analyzed under the fuels already
      registered and approved under section 211(b)(2). The attached comparison of
      E10 Tier 1 speciation data and E15 data, shows that E10 Tier 2 health effects
      data are applicable to E15. A review of the compounds for which the Research
      Group conducted health effects testing further supports this determination. The
      12 species that were both found  in E15 and which were tested in GEVC[7]
      constitutes approximately 95% by weight of E15 emissions. This evidence
      indicates that E15 should have similar health effects with registered fuels, and do
      not "pose new or different health risks to the public." Id. at 33,050.

      Although the grouping approach adopted in the regulations and mentioned by
RFA/GrE in its statement above relies heavily on similarity in composition of emissions,
E15 cannot be grouped with E10 for purposes of Tier 1 or Tier 2 testing requirements
under the regulations.8 This is because the regulations were finalized with the
recognition that the  same oxygenate added to fuels at a higher level might produce
different emissions and health effects than when added at a lower level.9 Therefore,
under the testing schemes reflected in the regulations, a fuel with 15 volume percent
(vol%) ethanol would generally require its own Tier 1 and Tier 2 testing.  However, with
regard to Tier 2 health testing, the regulations recognize that, in certain cases,  for one
or more reasons, previous testing might provide results which would be reasonably
7 GEVC refers to vapor condensate of baseline gasoline with ethanol (Gasoline Ethanol Vapor
Condensate) used in the Alternative Tier 2 health effects studies by the Section 211 (b) Research Group,
the organization which conducted the Tier 2 testing on E10.
8 40 CFR § 79.56(e)(4)(ii)(A)(7)(iii).
9 This is discussed in the preamble of the final testing rule found at 59 FR 33058-33061.

-------
comparable to the results of E15 testing were it to be completed ("reasonably
comparable" analysis). The Agency has discretion to decide whether such previous
testing could be substituted for the Tier 2 testing requirements for E15.  This provision
of the regulations is found at 40 CFR 79.53(d) and states that:

      After submission of all information and testing, EPA in its judgment shall
      determine whether previously conducted tests relied upon in the registration
      submission are adequately performed and documented and provide information
      reasonably comparable to that which would be provided by the tests described
      herein. Manufacturers' submissions shall be sufficiently detailed to allow EPA to
      judge the adequacy of protocols, techniques, experimental design, statistical
      analyses, and conclusions. Studies shall be performed using generally accepted
      scientific principles, good laboratory techniques, and the testing guidelines
      specified in these regulations.

      EPA discussed these issues with RFA/GrE, and RFA/GrE submitted an
addendum to its previous submissions providing further support for its position
regarding the reasonably comparable analysis.  EPA has reviewed RFA/GrE's
submissions regarding Tier 2 testing for E15, and  we agree with RFA/GrE's basic
analysis and conclude that replicating E10 Alternative Tier 2 health effects testing for
E15 evaporative emissions would yield reasonably comparable results. This analysis
considers the emissions data submitted by RFA/GrE as well as the emissions and
Alternative Tier 2 testing data from the 211 (b) Research Group that evaluated exposure
to baseline gasoline and GEVC. A summary of our findings is below.

 "Reasonably Comparable" Review

      The Alternative Tier 2 testing program evaluated baseline gasoline vapor
condensate and fuels with six different oxygenates, including ethanol at 10% volume, or
gasoline ethanol vapor condensate (GEVC). The  testing considered subchronic
inhalation exposure and effects on reproduction and development in rats exposed to
baseline gasoline and GEVC at three dose levels  (2,000 mg/m3, 10,000 mg/m3 and
20,000 mg/m3) and a  clean air control.  Health outcomes evaluated for baseline
gasoline and GEVC include genotoxicity, development, neurotoxicity, immunotoxicity,
and reproduction.  Following exposure to GEVC, the following adverse health outcomes
were identified: 1) decreased body weight at the high dose level; 2) neurotoxicity in
male rats at all three exposure levels; and 3) immunotoxicity in male and female mice at
the high dose. These results are explained in more detail in the following sections.

      Prior to discussing the adverse health outcomes in more detail, we first compare
the emissions speciation of the baseline gasoline vapor condensate and the GEVC.
The 211(b) health effects test reports provide speciated data for 18 compounds that
comprised approximately 97 percent of the vapor condensate that the animals were
exposed to. See Table 1, columns A and B. (The  comparison to RFA/GrE data in
columns D-F, related  to the generation of the fuel, is discussed in Section IV.)  Data
were reported on an area-percent basis, and the 18 target hydrocarbons were selected

-------
based on their relative abundance and as being representative of distinct hydrocarbon
classes (e.g.,  paraffins, olefins, aromatics).10 The primary difference between the
baseline gas and GEVC speciation profiles is the presence of ethanol.11 There are
minor differences in the hydrocarbon proportions, but these are most likely impacts
caused by the addition of ethanol and do not indicate any actual quantitative change in
the absolute levels of these hydrocarbon constituents.  Therefore, any differences in
health effects observed between the baseline gasoline and the GEVC fuel  blend can
reasonably be attributed to the presence of ethanol.
10 "Gasoline Vapor Condensate Characterization Final Report.  Study number 167490. Completed March
18, 2009.  Performed for API at ExxonMobil Biomedical Sciences, Inc.
11 The proportional concentrations of 2-methylpropane (isobutane) and 2-methylbutane (isopentane) are
higher in more saturated non-oxygenated hydrocarbon portion of the E10 and E15 (and even EO) used in
the RFA/GrE speciation than they were in GEVC. This issue is discussed later in this document.

                                          8

-------
Table 1. Proportional Composition of Fuel Vapors
COMPOUND
BUTANE1
2-METHYLPROPANE (ISOBUTANE)
ETHANOL
2-METHYLBUTANE (ISOPENTANE)
PENTANE2
TRANS-2-PENTENE
2,3-DIMETHYLBUTANE
2-METHYLPENTANE
3-METHYLPENTANE
HEXANE3
METHYLCYCLOPENTANE4
2,4-DIMETHYLPENTANE
BENZENE
2-METHYLHEXANE
2,3-DIMETHYLPENTANE
3-METHYLHEXANE
2,2,4-TRIMETHYLPENTANE
(ISOOCTANE)
TOLUENE
SUM
A
211 (b) Baseline
Gasoline Vapor
Condensate (%)5
16.1
2.8
NA
34.0
10.8
2.5
3.2
7.6
4.8
2.4
1.3
0.9
2.1
1.3
1.1
1.7
1.2
3.0
96.6
B
21 1(b) Gasoline
Ethanol Vapor
Condensate - E10
(%)6
11.6
2.2
13.3
34.0
10.2
2.1
2.3
5.1
2.9
2.4
1.2
1.0
1.6
1.1
1.1
1.2
1.3
2.4
97.0
C D
RFA/GrE E15 RFA/GrE
(%)7 E10(%)7
6.3 6.4
16.3 16.8
12.4 7.7
50.9 53.8
6.5 6.5
0.0 0.0
0.8 0.9
0.9 1.1
0.4 0.5
0.5 0.6
0.1 0.1
0.1 0.2
0.0 0.0
0.0 0.1
0.0 0.0
0.0 0.1
0.5 0.6
0.0 0.0
95.8 95.4
E
RFA/GrE EO
(%)7
15.2
18.0
0.0
49.9
6.4
0.0
0.8
0.9
0.4
0.5
0.1
0.2
0.0
0.2
0.0
0.1
1.0
0.0
93.9
Bold indicates compounds values that merit further evaluation. Italics indicates the proportion is less in the modern fuel than in the 21 1(b) fuel.

1 - 21 1(b) data were reported as n-butane; RFA/GrE data were reported as butane. We assume these are equivalent categories.
2-211 (b) data were reported as n-pentane; RFA/GrE data were reported as pentane. We assume these are equivalent
categories.
3 - 21 1(b) data were reported as n-hexane; RFA/GrE data were reported as hexane. We assume these are equivalent
categories.
4 - 2,2-Dimethylpentane and methylcyclopentane co-elute. GC peak area split equally between the two compounds in the RFA/GrE data.
5 - These data are from the 21 1(b) Gasoline Vapor Condensate Characterization Report. The values are a calculated average of the three baseline
gasoline vapor condensate (BGVC) fuels analyzed in this study and found in Table 6 of the report: MRD-00-674, MRD-00-695, MRD-03-747. The
values are derived as an area-percent basis.
6 - These data are from the 21 1(b) Gasoline Vapor Condensate Characterization Report, Table 6 (MRD-00-714). The values are derived as an area-
percent basis.
7 - These data are from the Evaporative Emissions Characterization of EO, E10, and E15 in Support of the Fuel and Fuel Additive Registration of
E1 5 prepared by SwRI for RFA/GrE in February 201 1 . They are a subset of information found in Table 5. Average values were used.

-------
I.     Body Weight

      As noted above, in a 1997 notification, EPA issued Alternative Tier 2 testing
requirements for baseline gasoline and various oxygenated gasolines including E10.12
With regard to fuel oxygenates other than MTBE, the testing program was intended to
provide a screening assessment of the potential toxicological effects in test animals of
inhalation exposure to the evaporative emissions of oxygenate gasoline fuel
formulations including baseline gasoline plus ethanol vapor condensates(GEVC). The
assessments were to be conducted in accordance with the relevant provisions of the
Health Effects Test Guideline (870 series) published by EPA's Office of Prevention,
Pesticides and Toxic  Substances (OPPTS). Among the provisions of the guidelines
was that the highest concentration tested should result in toxic effects but not produce
an incidence of fatalities which would prevent a meaningful evaluation (e.g. Health
Effects Test Guidelines OPPTS 870.3465 90-Day Inhalation Toxicity EPA 712-C-98-
204 August 1998). Among the observations from the 90 day inhalation toxicity study of
GEVC were lower mean body weight and lower body weight gain in animals exposed to
20,000 mg/m3, which was the highest concentration tested.13 Decreased body weight
gain was also observed in male rats exposed to 20,000 mg/m3 GEVC during the pre-
mating period in a study of reproductive toxicity. In addition, decreased body weight
gain was observed in pregnant dams exposed to 20,000 mg/m3 GEVC in a study of
developmental toxicity.

      For comparison, animals exposed to baseline gasoline vapor  condensate without
ethanol did not show  statistically significant changes in body weight parameters in rats
exposed for up to 20,000 mg/m3 in a 90-day subchronic inhalation study. However, in a
chronic study including 512 exposure days, male rats exposed to 10,000 and 20,000
mg/m3 had significantly lower body weights than did clean air controls throughout the
study. There were also decreases in body weight gain in FO female  rats and in F1 male
rats exposed to 20,000 mg/m3 baseline vapor condensate in a two-generation
reproduction study. Finally, a significant decrease of mean fetal body weight was
observed in both sexes and all dose groups relative to controls in the study of baseline
gasoline vapor concentrate, although in this study the pup body weights of the air
control group were high relative to  historic controls.

      A reduction of  body weight or lower body weight gain  at 20,000 mg/m3 indicates
that the highest dose  tested achieved an observable toxic effect in accordance with the
requirements of the test rule.  This provision was intended to assure  that the test
program did not miss  any potential adverse outcomes by examining  too low a range of
concentrations.  However, lower body weights, or a reduction in body weight gain, were
observed in several studies following exposure to baseline gasoline vapor condensates
both with and without added ethanol vapors, suggesting the  observed effects on body
weight parameters were not uniquely associated with the ethanol component of the
gasoline vapor mixture. Because weight loss was not uniquely associated with the
12 The notification can be found at http://www.epa.gov/otaq/consumer/fuels/mtbe/fnlno19a.pdf.
13 The studies discussed in these sections are the Alternative Tier 2 test reports performed under the
211 (b) requirements discussed earlier and can be found in the docket at EPA-HQ-OAR-2003-0065.

                                       10

-------
ethanol component of GEVC, this outcome was not considered to be a critical factor in
evaluating the potential for adverse health consequences associated with an increased
exposure to ethanol vapors if changing from E10 to E15.
II.    Neurotoxicity

      In our judgment, the 211 (b) Research Group's Alternative Tier 2 tests14
performed for E10 generally did not indicate any potential neurotoxicity concerns at
environmentally relevant doses. Although the test results were inconclusive on potential
neurotoxicity using GFAP (Glial Fibrillary Acidic Protein, a marker of potential
neurotoxicity) as an indicator, in our judgment the GFAP effects seen in these tests are
not toxicologically significant.  Several factors  mitigate potential concern based on these
results:

   •  Whereas statistically significant increases in GFAP concentrations were found for
      all tested concentrations of E10 in some brain regions of male rats, unsystematic,
      non-monotonic increases were seen in  other brain regions, and no significant
      changes were observed in females.

   •  Tests of brain regions that showed statistically significant changes were those in
      which GFAP values for the control group (exposed to clean air) were low relative
      to the control values from tests of EO and the six other fuel oxygenate blends that
      were evaluated.

   •  The number of significant differences found among the large number of
      statistical tests applied to these data (2 sexes x 9 brain regions x 7 fuels x 3
      comparisons per fuel = 378 tests) may represent the expected number of false
      positives at a 5% statistical significance level.

   •  No significant effects of treatment were observed in neuropathological
      evaluations, brain weight, brain size measurements or in neurobehavioral
      assessments (functional observational battery and motor activity) for animals
      exposed to GEVC.  Had such effects been observed, they could have
      corroborated the GFAP increases.

   •  Whereas others have reported that repeated exposure to ethanol increases
      GFAP in rats,  the internal blood ethanol concentrations required to produce those
      effects were much higher than those likely achieved  in the 211 (b) studies.  For
      example, Udomuksorn et al. (2011)15 reported significant increases in GFAP in
      rat brains after oral doses in the range of 2 - 5 g/kg,  which produced peak blood
14Alternative Tier 2 test reports performed under 211 (b) can be found in the docket at EPA-HQ-OAR-
2003-0065.
15 Udomukorn, W., W. Mukem, E. Kumarnsit, S. Vongvatcharanon, and U. Vongvatcharanon. 2011.
"Effects of alcohol administration during adulthood on parvalbumin and glial fibrillary acidic protein
immunoreactivity in the rat cerebral cortex." Acta Histochemica 113:283-289.

                                        11

-------
      ethanol concentrations (BECs) between 30 and 150 mg/dL In contrast, the
      highest air concentration of E10 used in the 211(b) studies was 20,000 mg/m3,
      for which we can assume that approximately 10% (or 2,000 mg/m3) was ethanol.
      Using a pharmacokinetic model for inhaled ethanol in rats,16 we estimate that the
      peak blood concentrations achieved by exposure to 2,000 mg/m3 ethanol vapor
      for 6 hr would be approximately 0.6 mg/dL. Thus, the BECs achieved in the
      211 (b) inhalation studies are roughly 50 - 250 times lower than those reported to
      increase GFAP after oral exposure.

      Thus, these considerations suggest that the increases in GFAP observed in the
211(b) studies do not indicate potential health concerns at environmentally relevant
doses, and if performed for E15, results would likely be negative.
III.   Immunotoxicity

      As discussed in more detail below, the 211 (b) Alternative Tier 2 tests suggest
immunosuppressive effects attributable to the ethanol in the GEVC.  Therefore, we
consider the potential impact on adverse health outcomes of the additional ethanol in
E15 compared to E10.

      A.    Results of the 211(b) tests

      In the Alternative Tier 2 study (White, 2010)17 of  GEVC, a statistically significant
impairment of immune function was observed in female rats exposed to 20,000 mg/m3.
Immune function was not evaluated in male rats. At the dose of 20,000 mg/m3, the
primary Immunoglobulin M (IgM) antibody response to immunization was decreased by
85%, relative to controls.  This level of immunosuppression is likely to increase
vulnerability to infectious diseases, particularly those caused by common bacteria and
viruses.  The average IgM response appeared to be reduced in animals exposed to
10,000 mg/m3, but the results were not significantly different from clean  air controls.
Exposure to 2,000 mg/m3 GEVC similarly did not cause a statistically significant
impairment of immune function.  No adverse effects on  immune system  function were
observed in the Alternative Tier 2 study of baseline gasoline vapors alone.  The major
difference between the composition of the baseline gasoline vapor condensate and the
GEVC was the ethanol content,  suggesting that immunosuppression was caused either
by the ethanol component or by ethanol plus gasoline vapors in the GEVC mixture.
Gasoline blended with DIPE and ETBE additives also caused immune suppression,
suggesting that other fuel oxygenates in gasoline could produce similar  effects.
16 Pastino GM, Asgharian B, Roberts K, Medinsky MA, Bond JA. (1997). A comparison of physiologically
based pharmacokinetic model predictions and experimental data for inhaled ethanol in male and female
B6C3F1 mice, F344 rats, and humans. Toxicol Appl Pharmacol, 145, 147-57.
  White KL, Immunological evaluation of gasoline ethanol vapor condensate in female Sprague Dawley
rats using the plaque forming cell assay (2nd Audited Draft Report). Submitted to EPA as a component of:
Study No 00-6127 Gasoline ethanol vapor condensate: a 13-week whole-body inhalation toxicity study in
rats with neurotoxicity assessments and 4-week in vivo genotoxicity and immunotoxicity assessments,
(Final Report). 27 January 2010.

                                       12

-------
      B.    E10 and El 5 Comparisons

      Because it appears that the immunosuppressive effects of GEVC exposure may
be attributed to the ethanol component of the mixture or to ethanol plus gasoline vapors,
the implications of increasing ethanol concentration from E10 to E15 are discussed in
detail.  This section focuses on the E10 and E15 evaporative emissions speciation
characterization submitted by RFA/GrE18.  We note that the RFA/GrE's emissions
characterizations used a slightly different gasoline hydrocarbon base fuel stock.
19
      As expected, ethanol content (both absolute mass and percent composition) is
approximately 1.5 times higher in E15 than E10 (Table 2).

      To confirm that ethanol is still the only compound that varies substantially
between the various fuel blends, as was the case in the 211 (b) baseline gasoline and
GEVC fuel blends, we also examined the other speciated constituents in RFA/GrE's
data.  We found four other hydrocarbon constituents were higher on an absolute or
percentage basis with E15 than E10 (Table 2). They include 1) ethane;  2) trans-2-
butene;  3) unidentified C7 compounds; and 4)  unidentified C9-C12+ compounds.  The
reported increases were very small, changing by 0.5 mg or less and 0.1% or less.
These changes are likely to be random fluctuations (i.e.,  "noise") in the data. No
reliable  evidence was presented that vapor components  of evaporative emissions other
than ethanol increased  from E10 to E15 for these fuels.  Even if it were assumed that
the four non-ethanol components reliably increased in E15 evaporative emissions, the
small  magnitude of the  increases makes it unlikely that these changes reflect a credible
concern for adverse health effects.

      In addition to the increases in certain hydrocarbons, there were five compounds
that were present in the E10 evaporative emissions but not in the E15 evaporative
emissions.  There were also 22 compounds that decreased  in the E15 average test
results compared to the E10 average test results.
   Table  2: Emission  component measurements  that were  higher in  RFA/GrE's  evaporative
   emissions speciation for E15 than E10.

Compound
ETHANE
TRANS-2-BUTENE
Absolute Mass (mg)
E10
1.5
0.0
E15
1.7
0.2
Percent Composition (% of total)
E10
0.1
0.0
E15
0.2
0.0
18
  Southwest Research Institute Final Report 03.15812, "Evaporative Emissions Characterization of EO,
E10, and E15 in Support of the Fuel and Fuel Additive Registration of E15", February, 2011, pp. 12-21.
19
  Base fuel stock is the hydrocarbon portion of the gasoline fuel to which the ethanol was added to make
the ethanol gasoline blend.
                                       13

-------
ETHANOL
UNIDENTIFIED C7
UNIDENTIFIED C9-C12+
91.1
0.5
1.0
136.7
1.0
1.1
7.7
0.0
0.1
12.4
0.1
0.1
      C.    Implications of Increased Ethanol in El 5

      As discussed above, ethanol or ethanol combined with gasoline constituents is
the cause of the adverse immunotoxicity result in the 211 (b) GEVC study.  Since the
ethanol component of E15 vapors would be greater than that from the same
concentration of GEVC or E10, it is reasonable to conclude that the immunotoxicity of
exposure to E15 would be greater than that from the same concentration of E10. In
addition, the threshold concentration for producing immune suppression would be
expected to be lower for exposure to E15 than for GEVC or for E10.

      It is possible to approximate the doses of E15 that would be associated with
immune suppression based on the ethanol concentration of the vapors and in
comparison to the results of the 211 (b) Alternative Tier 2 GEVC study.  Since the GEVC
was composed from gasoline plus 10% ethanol, we can approximate the ethanol
concentration of the vapor to be roughly 10% of the total vapor, and likewise we can
approximate the ethanol component  of E15 vapor to be approximately 15% of the total
vapor (Table 3).

Table 3. Approximation of ethanol content of vapors and anticipated results of immunotoxicity testing
given that ethanol concentration
Material
E15
GEVC
E15
GEVC
E15
GEVC
Exposure
concentration
(mg/m3)
20,000
20,000
10,000
10,000
2,000
2,000
Estimated
Ethanol
Concentration
(mg/m3)
3,000
2,000
1,500
1,000
300
200
211(b)Tier2
Immunotoxicity
results

LOAEL

NOAEL

No effect
Presumed
outcome if E15
were tested
based on ethanol
concentration
Effect level

Uncertain

No effect

      Table 3 is ordered from top to bottom based on the estimated concentration of
the ethanol component of the exposure vapor.  For example, if E15 vapor is assumed to
be 15% ethanol then the ethanol concentration of 20,000 mg/m3 would be 3,000 mg/m3
(0.15 x 20,000). The lowest observable adverse effect level (LOAEL) and no
observable adverse effect level (NOAEL) results of the Tier 2 testing for GEVC are
listed from the White 2010 study report. Based on the estimated ethanol concentration
of the vapor, therefore, E15 at 20,000 mg/m3 is anticipated to cause some
immunosuppression since the ethanol concentration of the total vapor is higher than
that of GEVC at 20,000 mg/m3.  Exposure to E15 vapor at 2,000 mg/m3 would be
expected to have no significant effect since the estimated ethanol concentration in the
                                      14

-------
vapor is lower than that from GEVC at 10,000 mg/m3  The likely effect of exposure to
10,000 mg/m3 E15, however, cannot be precisely estimated from these data because
the ethanol component of vapor falls between the LOAEL and NOAEL of the GEVC
studies.

      In summary, these analyses suggest that (a) the ethanol  component of the
emissions is likely contributing to suppressing immune function, (b) the ethanol
component of E15 vapor is higher than that from E10 or GEVC, (c) the immunotoxicity
of El5 vapor therefore would likely be somewhat greater than the same concentration
of E10 vapor were such testing to be conducted, and (d) the LOAEL/NOAEL for E15
causing immune suppression would likely be between 2,000 and 20,000  mg/m3, but
cannot be specified to greater precision from the data available.

      The importance of these differences depends on several factors, such as
whether these levels of exposure are environmentally relevant.  Based on data available
from the 211 (b) exposure study20, ethanol  concentrations in the U.S. appear to be far
below even the  lowest estimated concentrations of ethanol that  may cause
immunotoxicity. For example, the highest  peak concentration of ethanol  recorded in the
211(b) exposure study in Chicago was less than 120 parts per billion by volume (ppbv).
By comparison, the lowest level that could be associated with adverse health effects
observed in the 211 (b) Alternative Tier 2 testing is greater than 1,000 mg/m3 of ethanol
(see Table 3). As a conservative illustration, this NOAEL of 1,000 mg/m  is equivalent
to 530,600 ppbv21,  which is three orders of magnitude above the measured peak
ethanol exposure.  We note that the results discussed here are based only on the health
effects observed in the 211 (b) Alternative Tier 2 testing for GEVC.

      We note  here that the proportional concentrations of 2-methylpropane
(isobutane) and 2-methylbutane (isopentane) are higher in more saturated non-
oxygenated hydrocarbon portion of the E10 and  E15 (and even  EO) used in the
RFA/GrE speciation than they were in GEVC. This is a reasonable consequence of
lower levels of some constituents, specifically aromatics,  which necessarily increases
the proportion of other remaining constituents that constitute the whole mixture. Fuels
utilized in the 211(b) toxicity testing were produced using gasoline specifications more
common in the early 1990s when the testing was designed. These fuels were relatively
high in aromatic hydrocarbons and relatively lower in saturates.  Although the emissions
of E15 utilized for speciation purposes  is higher in isobutane and isopentane than the
GEVC, this is not due to the addition of ethanol, and the levels of these emittants would
vary from gasoline to gasoline around the country and even from batch to batch within a
refinery. Thus, we  do not consider the significance of the increase in these two
compounds in our discussion of immunotoxicity above.
20Zielinski, Barbara and Eric M. Fujita, John C., Sagebiel and David E. Campbell, Desert Research
Institute,Section 211(B) Tier 2 High End Exposure Study of Conventional and Oxygenated Gasoline,
Compiled Final Report, March 25, 2009, EPA Docket Number EPA-HQ-OAR-2003-0065-0673.
21 The equation used to convert the exposure concentration in mg/m3 to ppm is: ethanol molecular weight
* ppm = 24.45 * mg/m3, or ppm = (24.45*1,000) / 46.08. The resulting ppm was multiplied by 1000 to
convert to ppb.

                                       15

-------
IV.   Conclusion

      For a fuel registration application to be complete without Tier 2 test results for the
fuel being registered, the regulations at 40 CFR 79.53(d) require a showing that
replicating E10 Alternative Tier 2 health effects testing for E15 evaporative emissions
would yield reasonably comparable results.  For the reasons mentioned above, we
conclude that, for the immunotoxicity testing, the results for E15 may or may not be
identical to those for E10, but they would be in the same range of exposures and would
be reasonably comparable.  We also conclude that the results for E15 would be
reasonably comparable for the other health effects discussed above.
Future Considerations

      In the future, EPA may consider the potential need for testing, under Tier 3, to
evaluate developmental neurotoxicity. Developmental neurotoxicity is the most
sensitive adverse outcome from ethanol ingestion, but there is a lack of comparable
data following developmental exposure to ethanol by inhalation. Current EPA research
is investigating this gap by evaluating developmental neurotoxicity associated with
inhalation exposure to evaporative emissions from various gasoline/ethanol blends.  At
this point, it is unlikely that EPA would recommend further developmental toxicity
testing, but we reserve final judgment until the completion of our ongoing research.
Similarly, EPA may consider further immunotoxicity testing if environmental exposures
increase substantially. If after the completion of our ongoing research such testing is
considered, it would likely be in the context of Tier 322 requirements for ethanol fuels
generally.

      Additionally, EPA will continue to consider the possibility of a fuller examination
of speciated combustion  emissions and exposures to help refine our understanding of
potential risks associated with the use of E15 or ethanol gasolines generally.  EPA will
consider testing which might include a further examination of speciated combustion
emissions including speciation of particulate matter (PM) and semi-volatile organic
compounds (SVOCs), more robust vehicle testing of emissions (more vehicles, more
tests and broader conditions), and modeling of potential environmental exposures
associated with emissions from selected E15 use scenarios (e.g., near roadways).  EPA
and DOE have ongoing research in these areas that might further inform the need for
additional studies.  For example, EPA is developing new analytical methods for PM and
organic compounds.

      Such testing might also include health effects testing if, in the future, EPA
determines that exposure of test animals to combustion emissions may yield useful
results which would not be compromised by exposure to carbon monoxide emissions.  If
and when such testing is required, EPA would likely also consider such testing for any
other gasoline fuels with  significant market share, including E10.
22 40 CFR 79.54.
                                       16

-------
Conclusion

      Our evaluation therefore concludes that RFA/GrE has submitted data and
analysis that would satisfy the Tier 1 and Tier 2 testing requirements for registration.
                                       17

-------